diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,40227 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.40009755077866427,
+  "eval_steps": 500,
+  "global_step": 5742,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 6.967912761732224e-05,
+      "grad_norm": 2.733745568803013,
+      "learning_rate": 6.999999924522766e-07,
+      "loss": 1.7192,
+      "step": 1
+    },
+    {
+      "epoch": 0.00013935825523464447,
+      "grad_norm": 2.709369185413448,
+      "learning_rate": 6.99999969809107e-07,
+      "loss": 1.6019,
+      "step": 2
+    },
+    {
+      "epoch": 0.0002090373828519667,
+      "grad_norm": 2.9135830924208594,
+      "learning_rate": 6.999999320704921e-07,
+      "loss": 1.8484,
+      "step": 3
+    },
+    {
+      "epoch": 0.00027871651046928894,
+      "grad_norm": 2.777419413695461,
+      "learning_rate": 6.999998792364337e-07,
+      "loss": 1.756,
+      "step": 4
+    },
+    {
+      "epoch": 0.00034839563808661115,
+      "grad_norm": 2.8075005508831694,
+      "learning_rate": 6.999998113069345e-07,
+      "loss": 1.7913,
+      "step": 5
+    },
+    {
+      "epoch": 0.0004180747657039334,
+      "grad_norm": 2.668293350418484,
+      "learning_rate": 6.999997282819976e-07,
+      "loss": 1.6321,
+      "step": 6
+    },
+    {
+      "epoch": 0.0004877538933212556,
+      "grad_norm": 2.4745384488107187,
+      "learning_rate": 6.999996301616271e-07,
+      "loss": 1.5594,
+      "step": 7
+    },
+    {
+      "epoch": 0.0005574330209385779,
+      "grad_norm": 2.3430588976825826,
+      "learning_rate": 6.999995169458277e-07,
+      "loss": 1.6613,
+      "step": 8
+    },
+    {
+      "epoch": 0.0006271121485559001,
+      "grad_norm": 2.2357126724663092,
+      "learning_rate": 6.999993886346046e-07,
+      "loss": 1.676,
+      "step": 9
+    },
+    {
+      "epoch": 0.0006967912761732223,
+      "grad_norm": 2.342047279317775,
+      "learning_rate": 6.999992452279643e-07,
+      "loss": 1.6192,
+      "step": 10
+    },
+    {
+      "epoch": 0.0007664704037905445,
+      "grad_norm": 2.2540685801049944,
+      "learning_rate": 6.999990867259133e-07,
+      "loss": 1.559,
+      "step": 11
+    },
+    {
+      "epoch": 0.0008361495314078668,
+      "grad_norm": 1.971114382158942,
+      "learning_rate": 6.999989131284595e-07,
+      "loss": 1.6581,
+      "step": 12
+    },
+    {
+      "epoch": 0.000905828659025189,
+      "grad_norm": 1.5959396288528505,
+      "learning_rate": 6.999987244356111e-07,
+      "loss": 1.721,
+      "step": 13
+    },
+    {
+      "epoch": 0.0009755077866425112,
+      "grad_norm": 1.4800088775047004,
+      "learning_rate": 6.999985206473771e-07,
+      "loss": 1.7666,
+      "step": 14
+    },
+    {
+      "epoch": 0.0010451869142598335,
+      "grad_norm": 1.3687402013136123,
+      "learning_rate": 6.999983017637673e-07,
+      "loss": 1.726,
+      "step": 15
+    },
+    {
+      "epoch": 0.0011148660418771558,
+      "grad_norm": 1.391725060939141,
+      "learning_rate": 6.999980677847923e-07,
+      "loss": 1.7181,
+      "step": 16
+    },
+    {
+      "epoch": 0.0011845451694944779,
+      "grad_norm": 1.1938617951720005,
+      "learning_rate": 6.999978187104631e-07,
+      "loss": 1.5963,
+      "step": 17
+    },
+    {
+      "epoch": 0.0012542242971118002,
+      "grad_norm": 1.2751779681441469,
+      "learning_rate": 6.999975545407918e-07,
+      "loss": 1.6692,
+      "step": 18
+    },
+    {
+      "epoch": 0.0013239034247291223,
+      "grad_norm": 1.2381900634264318,
+      "learning_rate": 6.99997275275791e-07,
+      "loss": 1.7631,
+      "step": 19
+    },
+    {
+      "epoch": 0.0013935825523464446,
+      "grad_norm": 1.153546411353916,
+      "learning_rate": 6.999969809154741e-07,
+      "loss": 1.7649,
+      "step": 20
+    },
+    {
+      "epoch": 0.001463261679963767,
+      "grad_norm": 1.1646630042222352,
+      "learning_rate": 6.999966714598552e-07,
+      "loss": 1.7358,
+      "step": 21
+    },
+    {
+      "epoch": 0.001532940807581089,
+      "grad_norm": 1.1023651577627083,
+      "learning_rate": 6.99996346908949e-07,
+      "loss": 1.67,
+      "step": 22
+    },
+    {
+      "epoch": 0.0016026199351984113,
+      "grad_norm": 1.1875711007742904,
+      "learning_rate": 6.999960072627713e-07,
+      "loss": 1.622,
+      "step": 23
+    },
+    {
+      "epoch": 0.0016722990628157337,
+      "grad_norm": 1.0365697887615943,
+      "learning_rate": 6.999956525213383e-07,
+      "loss": 1.7234,
+      "step": 24
+    },
+    {
+      "epoch": 0.0017419781904330558,
+      "grad_norm": 0.9973919994515652,
+      "learning_rate": 6.99995282684667e-07,
+      "loss": 1.7004,
+      "step": 25
+    },
+    {
+      "epoch": 0.001811657318050378,
+      "grad_norm": 1.3266756191624058,
+      "learning_rate": 6.999948977527749e-07,
+      "loss": 1.7692,
+      "step": 26
+    },
+    {
+      "epoch": 0.0018813364456677002,
+      "grad_norm": 1.4466381604576397,
+      "learning_rate": 6.999944977256808e-07,
+      "loss": 1.4938,
+      "step": 27
+    },
+    {
+      "epoch": 0.0019510155732850225,
+      "grad_norm": 1.5268786523739337,
+      "learning_rate": 6.999940826034036e-07,
+      "loss": 1.7042,
+      "step": 28
+    },
+    {
+      "epoch": 0.002020694700902345,
+      "grad_norm": 1.4725142311197557,
+      "learning_rate": 6.999936523859633e-07,
+      "loss": 1.4254,
+      "step": 29
+    },
+    {
+      "epoch": 0.002090373828519667,
+      "grad_norm": 1.4663543510668822,
+      "learning_rate": 6.999932070733806e-07,
+      "loss": 1.5621,
+      "step": 30
+    },
+    {
+      "epoch": 0.002160052956136989,
+      "grad_norm": 1.5525614196605468,
+      "learning_rate": 6.999927466656765e-07,
+      "loss": 1.7332,
+      "step": 31
+    },
+    {
+      "epoch": 0.0022297320837543115,
+      "grad_norm": 1.3577154717206668,
+      "learning_rate": 6.999922711628735e-07,
+      "loss": 1.5684,
+      "step": 32
+    },
+    {
+      "epoch": 0.0022994112113716336,
+      "grad_norm": 1.2464378205015698,
+      "learning_rate": 6.999917805649943e-07,
+      "loss": 1.7046,
+      "step": 33
+    },
+    {
+      "epoch": 0.0023690903389889557,
+      "grad_norm": 1.2571057892933462,
+      "learning_rate": 6.999912748720621e-07,
+      "loss": 1.6985,
+      "step": 34
+    },
+    {
+      "epoch": 0.0024387694666062783,
+      "grad_norm": 1.046761844185471,
+      "learning_rate": 6.999907540841015e-07,
+      "loss": 1.5903,
+      "step": 35
+    },
+    {
+      "epoch": 0.0025084485942236004,
+      "grad_norm": 1.0190382659116664,
+      "learning_rate": 6.999902182011371e-07,
+      "loss": 1.6446,
+      "step": 36
+    },
+    {
+      "epoch": 0.0025781277218409225,
+      "grad_norm": 0.9420347898121777,
+      "learning_rate": 6.999896672231951e-07,
+      "loss": 1.6862,
+      "step": 37
+    },
+    {
+      "epoch": 0.0026478068494582446,
+      "grad_norm": 0.9594587432273478,
+      "learning_rate": 6.999891011503013e-07,
+      "loss": 1.6502,
+      "step": 38
+    },
+    {
+      "epoch": 0.002717485977075567,
+      "grad_norm": 0.8779244551322024,
+      "learning_rate": 6.999885199824832e-07,
+      "loss": 1.5531,
+      "step": 39
+    },
+    {
+      "epoch": 0.002787165104692889,
+      "grad_norm": 0.8480799062449409,
+      "learning_rate": 6.999879237197686e-07,
+      "loss": 1.4879,
+      "step": 40
+    },
+    {
+      "epoch": 0.0028568442323102113,
+      "grad_norm": 0.9497551105971518,
+      "learning_rate": 6.99987312362186e-07,
+      "loss": 1.6479,
+      "step": 41
+    },
+    {
+      "epoch": 0.002926523359927534,
+      "grad_norm": 0.8618279518712127,
+      "learning_rate": 6.999866859097648e-07,
+      "loss": 1.662,
+      "step": 42
+    },
+    {
+      "epoch": 0.002996202487544856,
+      "grad_norm": 0.9018906795561105,
+      "learning_rate": 6.999860443625348e-07,
+      "loss": 1.5774,
+      "step": 43
+    },
+    {
+      "epoch": 0.003065881615162178,
+      "grad_norm": 0.8612830023344303,
+      "learning_rate": 6.999853877205271e-07,
+      "loss": 1.8141,
+      "step": 44
+    },
+    {
+      "epoch": 0.0031355607427795006,
+      "grad_norm": 0.9007192891107818,
+      "learning_rate": 6.999847159837729e-07,
+      "loss": 1.6209,
+      "step": 45
+    },
+    {
+      "epoch": 0.0032052398703968227,
+      "grad_norm": 0.8585620295584652,
+      "learning_rate": 6.999840291523043e-07,
+      "loss": 1.6896,
+      "step": 46
+    },
+    {
+      "epoch": 0.0032749189980141448,
+      "grad_norm": 0.8582642386199358,
+      "learning_rate": 6.999833272261546e-07,
+      "loss": 1.5751,
+      "step": 47
+    },
+    {
+      "epoch": 0.0033445981256314673,
+      "grad_norm": 0.9273580515998116,
+      "learning_rate": 6.999826102053571e-07,
+      "loss": 1.5397,
+      "step": 48
+    },
+    {
+      "epoch": 0.0034142772532487894,
+      "grad_norm": 0.8614901819193356,
+      "learning_rate": 6.999818780899462e-07,
+      "loss": 1.5651,
+      "step": 49
+    },
+    {
+      "epoch": 0.0034839563808661115,
+      "grad_norm": 0.8055350569759604,
+      "learning_rate": 6.999811308799572e-07,
+      "loss": 1.5396,
+      "step": 50
+    },
+    {
+      "epoch": 0.0035536355084834336,
+      "grad_norm": 0.7667301976321873,
+      "learning_rate": 6.999803685754257e-07,
+      "loss": 1.5635,
+      "step": 51
+    },
+    {
+      "epoch": 0.003623314636100756,
+      "grad_norm": 0.7558223853302638,
+      "learning_rate": 6.999795911763883e-07,
+      "loss": 1.4957,
+      "step": 52
+    },
+    {
+      "epoch": 0.0036929937637180782,
+      "grad_norm": 0.8569578181282904,
+      "learning_rate": 6.999787986828822e-07,
+      "loss": 1.5791,
+      "step": 53
+    },
+    {
+      "epoch": 0.0037626728913354003,
+      "grad_norm": 0.8084428013047424,
+      "learning_rate": 6.999779910949456e-07,
+      "loss": 1.6145,
+      "step": 54
+    },
+    {
+      "epoch": 0.003832352018952723,
+      "grad_norm": 0.7559081128636418,
+      "learning_rate": 6.999771684126168e-07,
+      "loss": 1.5544,
+      "step": 55
+    },
+    {
+      "epoch": 0.003902031146570045,
+      "grad_norm": 0.7689467380996695,
+      "learning_rate": 6.999763306359357e-07,
+      "loss": 1.5876,
+      "step": 56
+    },
+    {
+      "epoch": 0.003971710274187367,
+      "grad_norm": 0.7544375614009099,
+      "learning_rate": 6.99975477764942e-07,
+      "loss": 1.5481,
+      "step": 57
+    },
+    {
+      "epoch": 0.00404138940180469,
+      "grad_norm": 0.7831679946606233,
+      "learning_rate": 6.999746097996769e-07,
+      "loss": 1.5945,
+      "step": 58
+    },
+    {
+      "epoch": 0.004111068529422011,
+      "grad_norm": 0.750603996125084,
+      "learning_rate": 6.999737267401817e-07,
+      "loss": 1.5258,
+      "step": 59
+    },
+    {
+      "epoch": 0.004180747657039334,
+      "grad_norm": 0.7953623833277211,
+      "learning_rate": 6.999728285864989e-07,
+      "loss": 1.5848,
+      "step": 60
+    },
+    {
+      "epoch": 0.004250426784656656,
+      "grad_norm": 0.7574855755742249,
+      "learning_rate": 6.999719153386717e-07,
+      "loss": 1.621,
+      "step": 61
+    },
+    {
+      "epoch": 0.004320105912273978,
+      "grad_norm": 0.776253899297405,
+      "learning_rate": 6.999709869967437e-07,
+      "loss": 1.6088,
+      "step": 62
+    },
+    {
+      "epoch": 0.0043897850398913005,
+      "grad_norm": 0.7610571494782742,
+      "learning_rate": 6.999700435607591e-07,
+      "loss": 1.5922,
+      "step": 63
+    },
+    {
+      "epoch": 0.004459464167508623,
+      "grad_norm": 0.8152423242847946,
+      "learning_rate": 6.999690850307637e-07,
+      "loss": 1.6002,
+      "step": 64
+    },
+    {
+      "epoch": 0.004529143295125945,
+      "grad_norm": 0.7968474238058425,
+      "learning_rate": 6.99968111406803e-07,
+      "loss": 1.5156,
+      "step": 65
+    },
+    {
+      "epoch": 0.004598822422743267,
+      "grad_norm": 0.7974893416155076,
+      "learning_rate": 6.999671226889238e-07,
+      "loss": 1.655,
+      "step": 66
+    },
+    {
+      "epoch": 0.00466850155036059,
+      "grad_norm": 0.7912729979364881,
+      "learning_rate": 6.999661188771734e-07,
+      "loss": 1.628,
+      "step": 67
+    },
+    {
+      "epoch": 0.0047381806779779115,
+      "grad_norm": 0.7550934267500832,
+      "learning_rate": 6.999650999716002e-07,
+      "loss": 1.7023,
+      "step": 68
+    },
+    {
+      "epoch": 0.004807859805595234,
+      "grad_norm": 0.7295311010941107,
+      "learning_rate": 6.999640659722525e-07,
+      "loss": 1.589,
+      "step": 69
+    },
+    {
+      "epoch": 0.0048775389332125565,
+      "grad_norm": 0.7758262875672367,
+      "learning_rate": 6.999630168791802e-07,
+      "loss": 1.6251,
+      "step": 70
+    },
+    {
+      "epoch": 0.004947218060829878,
+      "grad_norm": 0.7449174445401683,
+      "learning_rate": 6.999619526924338e-07,
+      "loss": 1.5023,
+      "step": 71
+    },
+    {
+      "epoch": 0.005016897188447201,
+      "grad_norm": 0.7692068872313949,
+      "learning_rate": 6.999608734120638e-07,
+      "loss": 1.6567,
+      "step": 72
+    },
+    {
+      "epoch": 0.005086576316064523,
+      "grad_norm": 0.8275918260872214,
+      "learning_rate": 6.999597790381221e-07,
+      "loss": 1.708,
+      "step": 73
+    },
+    {
+      "epoch": 0.005156255443681845,
+      "grad_norm": 0.7872033802821833,
+      "learning_rate": 6.999586695706612e-07,
+      "loss": 1.7163,
+      "step": 74
+    },
+    {
+      "epoch": 0.0052259345712991675,
+      "grad_norm": 0.8159200519159217,
+      "learning_rate": 6.999575450097343e-07,
+      "loss": 1.5663,
+      "step": 75
+    },
+    {
+      "epoch": 0.005295613698916489,
+      "grad_norm": 0.7956432710560783,
+      "learning_rate": 6.999564053553953e-07,
+      "loss": 1.697,
+      "step": 76
+    },
+    {
+      "epoch": 0.005365292826533812,
+      "grad_norm": 0.7130908379784697,
+      "learning_rate": 6.999552506076987e-07,
+      "loss": 1.6055,
+      "step": 77
+    },
+    {
+      "epoch": 0.005434971954151134,
+      "grad_norm": 0.7274551108988991,
+      "learning_rate": 6.999540807667e-07,
+      "loss": 1.5876,
+      "step": 78
+    },
+    {
+      "epoch": 0.005504651081768456,
+      "grad_norm": 0.7646927508527166,
+      "learning_rate": 6.99952895832455e-07,
+      "loss": 1.6219,
+      "step": 79
+    },
+    {
+      "epoch": 0.005574330209385778,
+      "grad_norm": 0.8844817783804978,
+      "learning_rate": 6.999516958050207e-07,
+      "loss": 1.5852,
+      "step": 80
+    },
+    {
+      "epoch": 0.005644009337003101,
+      "grad_norm": 0.8046092626958843,
+      "learning_rate": 6.999504806844547e-07,
+      "loss": 1.6941,
+      "step": 81
+    },
+    {
+      "epoch": 0.005713688464620423,
+      "grad_norm": 0.7565003415350436,
+      "learning_rate": 6.999492504708149e-07,
+      "loss": 1.5299,
+      "step": 82
+    },
+    {
+      "epoch": 0.005783367592237745,
+      "grad_norm": 0.7491176510626603,
+      "learning_rate": 6.999480051641605e-07,
+      "loss": 1.4925,
+      "step": 83
+    },
+    {
+      "epoch": 0.005853046719855068,
+      "grad_norm": 0.7518065408047131,
+      "learning_rate": 6.999467447645511e-07,
+      "loss": 1.6606,
+      "step": 84
+    },
+    {
+      "epoch": 0.005922725847472389,
+      "grad_norm": 0.7208224360353647,
+      "learning_rate": 6.999454692720472e-07,
+      "loss": 1.5146,
+      "step": 85
+    },
+    {
+      "epoch": 0.005992404975089712,
+      "grad_norm": 0.7845222532902206,
+      "learning_rate": 6.999441786867097e-07,
+      "loss": 1.6499,
+      "step": 86
+    },
+    {
+      "epoch": 0.006062084102707034,
+      "grad_norm": 0.7179978138866828,
+      "learning_rate": 6.999428730086007e-07,
+      "loss": 1.6328,
+      "step": 87
+    },
+    {
+      "epoch": 0.006131763230324356,
+      "grad_norm": 0.7211806613439429,
+      "learning_rate": 6.999415522377827e-07,
+      "loss": 1.5726,
+      "step": 88
+    },
+    {
+      "epoch": 0.006201442357941679,
+      "grad_norm": 0.7546953013822226,
+      "learning_rate": 6.999402163743188e-07,
+      "loss": 1.5601,
+      "step": 89
+    },
+    {
+      "epoch": 0.006271121485559001,
+      "grad_norm": 0.7566779977768506,
+      "learning_rate": 6.999388654182733e-07,
+      "loss": 1.684,
+      "step": 90
+    },
+    {
+      "epoch": 0.006340800613176323,
+      "grad_norm": 0.6969488028265275,
+      "learning_rate": 6.999374993697108e-07,
+      "loss": 1.5598,
+      "step": 91
+    },
+    {
+      "epoch": 0.006410479740793645,
+      "grad_norm": 0.705981536567612,
+      "learning_rate": 6.999361182286967e-07,
+      "loss": 1.5545,
+      "step": 92
+    },
+    {
+      "epoch": 0.006480158868410968,
+      "grad_norm": 0.8466488562144117,
+      "learning_rate": 6.999347219952973e-07,
+      "loss": 1.6598,
+      "step": 93
+    },
+    {
+      "epoch": 0.0065498379960282896,
+      "grad_norm": 0.7800529271957665,
+      "learning_rate": 6.999333106695795e-07,
+      "loss": 1.6096,
+      "step": 94
+    },
+    {
+      "epoch": 0.006619517123645612,
+      "grad_norm": 0.7752456360469325,
+      "learning_rate": 6.999318842516109e-07,
+      "loss": 1.6597,
+      "step": 95
+    },
+    {
+      "epoch": 0.006689196251262935,
+      "grad_norm": 0.6961730716580421,
+      "learning_rate": 6.999304427414599e-07,
+      "loss": 1.4991,
+      "step": 96
+    },
+    {
+      "epoch": 0.006758875378880256,
+      "grad_norm": 0.7111901307005418,
+      "learning_rate": 6.999289861391954e-07,
+      "loss": 1.5348,
+      "step": 97
+    },
+    {
+      "epoch": 0.006828554506497579,
+      "grad_norm": 0.7729231425029259,
+      "learning_rate": 6.999275144448875e-07,
+      "loss": 1.7086,
+      "step": 98
+    },
+    {
+      "epoch": 0.0068982336341149005,
+      "grad_norm": 0.7207807898538051,
+      "learning_rate": 6.999260276586065e-07,
+      "loss": 1.4942,
+      "step": 99
+    },
+    {
+      "epoch": 0.006967912761732223,
+      "grad_norm": 0.8209832140800927,
+      "learning_rate": 6.999245257804236e-07,
+      "loss": 1.6462,
+      "step": 100
+    },
+    {
+      "epoch": 0.0070375918893495456,
+      "grad_norm": 0.7372508955817842,
+      "learning_rate": 6.999230088104111e-07,
+      "loss": 1.6425,
+      "step": 101
+    },
+    {
+      "epoch": 0.007107271016966867,
+      "grad_norm": 0.7103080276487075,
+      "learning_rate": 6.999214767486415e-07,
+      "loss": 1.5064,
+      "step": 102
+    },
+    {
+      "epoch": 0.00717695014458419,
+      "grad_norm": 0.7453563805733722,
+      "learning_rate": 6.999199295951881e-07,
+      "loss": 1.5735,
+      "step": 103
+    },
+    {
+      "epoch": 0.007246629272201512,
+      "grad_norm": 0.7316234271209868,
+      "learning_rate": 6.999183673501252e-07,
+      "loss": 1.487,
+      "step": 104
+    },
+    {
+      "epoch": 0.007316308399818834,
+      "grad_norm": 0.8306495639187648,
+      "learning_rate": 6.999167900135276e-07,
+      "loss": 1.6901,
+      "step": 105
+    },
+    {
+      "epoch": 0.0073859875274361565,
+      "grad_norm": 0.6877525548972586,
+      "learning_rate": 6.99915197585471e-07,
+      "loss": 1.5244,
+      "step": 106
+    },
+    {
+      "epoch": 0.007455666655053479,
+      "grad_norm": 0.7545823598451985,
+      "learning_rate": 6.999135900660315e-07,
+      "loss": 1.6525,
+      "step": 107
+    },
+    {
+      "epoch": 0.007525345782670801,
+      "grad_norm": 0.7557325144444581,
+      "learning_rate": 6.999119674552864e-07,
+      "loss": 1.5528,
+      "step": 108
+    },
+    {
+      "epoch": 0.007595024910288123,
+      "grad_norm": 0.7363162596156012,
+      "learning_rate": 6.999103297533131e-07,
+      "loss": 1.6752,
+      "step": 109
+    },
+    {
+      "epoch": 0.007664704037905446,
+      "grad_norm": 0.7779785261187538,
+      "learning_rate": 6.999086769601905e-07,
+      "loss": 1.6054,
+      "step": 110
+    },
+    {
+      "epoch": 0.007734383165522767,
+      "grad_norm": 0.7375695814517242,
+      "learning_rate": 6.999070090759976e-07,
+      "loss": 1.6362,
+      "step": 111
+    },
+    {
+      "epoch": 0.00780406229314009,
+      "grad_norm": 0.7295801652020935,
+      "learning_rate": 6.999053261008142e-07,
+      "loss": 1.6278,
+      "step": 112
+    },
+    {
+      "epoch": 0.007873741420757412,
+      "grad_norm": 0.7564600054233208,
+      "learning_rate": 6.999036280347211e-07,
+      "loss": 1.5602,
+      "step": 113
+    },
+    {
+      "epoch": 0.007943420548374734,
+      "grad_norm": 0.7325116000746935,
+      "learning_rate": 6.999019148777999e-07,
+      "loss": 1.5854,
+      "step": 114
+    },
+    {
+      "epoch": 0.008013099675992057,
+      "grad_norm": 0.750333027438248,
+      "learning_rate": 6.999001866301322e-07,
+      "loss": 1.6378,
+      "step": 115
+    },
+    {
+      "epoch": 0.00808277880360938,
+      "grad_norm": 0.7077391988654039,
+      "learning_rate": 6.998984432918011e-07,
+      "loss": 1.4496,
+      "step": 116
+    },
+    {
+      "epoch": 0.008152457931226702,
+      "grad_norm": 0.7061194584771281,
+      "learning_rate": 6.998966848628901e-07,
+      "loss": 1.7211,
+      "step": 117
+    },
+    {
+      "epoch": 0.008222137058844023,
+      "grad_norm": 0.753741884650152,
+      "learning_rate": 6.998949113434836e-07,
+      "loss": 1.5725,
+      "step": 118
+    },
+    {
+      "epoch": 0.008291816186461345,
+      "grad_norm": 0.8776126975121025,
+      "learning_rate": 6.998931227336665e-07,
+      "loss": 1.4603,
+      "step": 119
+    },
+    {
+      "epoch": 0.008361495314078668,
+      "grad_norm": 0.7457924178802692,
+      "learning_rate": 6.998913190335243e-07,
+      "loss": 1.5817,
+      "step": 120
+    },
+    {
+      "epoch": 0.00843117444169599,
+      "grad_norm": 0.6662869164399065,
+      "learning_rate": 6.998895002431439e-07,
+      "loss": 1.4134,
+      "step": 121
+    },
+    {
+      "epoch": 0.008500853569313313,
+      "grad_norm": 0.730929392364289,
+      "learning_rate": 6.998876663626121e-07,
+      "loss": 1.6551,
+      "step": 122
+    },
+    {
+      "epoch": 0.008570532696930635,
+      "grad_norm": 0.7485428242937736,
+      "learning_rate": 6.998858173920168e-07,
+      "loss": 1.5917,
+      "step": 123
+    },
+    {
+      "epoch": 0.008640211824547956,
+      "grad_norm": 0.7041960816293819,
+      "learning_rate": 6.998839533314466e-07,
+      "loss": 1.5663,
+      "step": 124
+    },
+    {
+      "epoch": 0.008709890952165279,
+      "grad_norm": 0.7065757407197867,
+      "learning_rate": 6.998820741809911e-07,
+      "loss": 1.5738,
+      "step": 125
+    },
+    {
+      "epoch": 0.008779570079782601,
+      "grad_norm": 0.7167544228047786,
+      "learning_rate": 6.998801799407402e-07,
+      "loss": 1.4507,
+      "step": 126
+    },
+    {
+      "epoch": 0.008849249207399924,
+      "grad_norm": 0.7373858233625152,
+      "learning_rate": 6.998782706107846e-07,
+      "loss": 1.5638,
+      "step": 127
+    },
+    {
+      "epoch": 0.008918928335017246,
+      "grad_norm": 0.762790306505906,
+      "learning_rate": 6.998763461912157e-07,
+      "loss": 1.5871,
+      "step": 128
+    },
+    {
+      "epoch": 0.008988607462634567,
+      "grad_norm": 0.7997672459056562,
+      "learning_rate": 6.998744066821259e-07,
+      "loss": 1.7292,
+      "step": 129
+    },
+    {
+      "epoch": 0.00905828659025189,
+      "grad_norm": 0.6949799061752019,
+      "learning_rate": 6.998724520836082e-07,
+      "loss": 1.557,
+      "step": 130
+    },
+    {
+      "epoch": 0.009127965717869212,
+      "grad_norm": 0.7429111002352855,
+      "learning_rate": 6.998704823957561e-07,
+      "loss": 1.6075,
+      "step": 131
+    },
+    {
+      "epoch": 0.009197644845486535,
+      "grad_norm": 0.7247660368087409,
+      "learning_rate": 6.99868497618664e-07,
+      "loss": 1.4907,
+      "step": 132
+    },
+    {
+      "epoch": 0.009267323973103857,
+      "grad_norm": 0.753497515434336,
+      "learning_rate": 6.998664977524273e-07,
+      "loss": 1.5622,
+      "step": 133
+    },
+    {
+      "epoch": 0.00933700310072118,
+      "grad_norm": 0.7525401766307482,
+      "learning_rate": 6.998644827971414e-07,
+      "loss": 1.6525,
+      "step": 134
+    },
+    {
+      "epoch": 0.0094066822283385,
+      "grad_norm": 0.7719829875157536,
+      "learning_rate": 6.998624527529032e-07,
+      "loss": 1.7565,
+      "step": 135
+    },
+    {
+      "epoch": 0.009476361355955823,
+      "grad_norm": 0.7780799099616043,
+      "learning_rate": 6.998604076198099e-07,
+      "loss": 1.5999,
+      "step": 136
+    },
+    {
+      "epoch": 0.009546040483573145,
+      "grad_norm": 0.6788851566259799,
+      "learning_rate": 6.998583473979593e-07,
+      "loss": 1.5832,
+      "step": 137
+    },
+    {
+      "epoch": 0.009615719611190468,
+      "grad_norm": 0.6759649759261548,
+      "learning_rate": 6.998562720874505e-07,
+      "loss": 1.5318,
+      "step": 138
+    },
+    {
+      "epoch": 0.00968539873880779,
+      "grad_norm": 0.722411480856436,
+      "learning_rate": 6.998541816883826e-07,
+      "loss": 1.4956,
+      "step": 139
+    },
+    {
+      "epoch": 0.009755077866425113,
+      "grad_norm": 0.7963536281427879,
+      "learning_rate": 6.99852076200856e-07,
+      "loss": 1.6532,
+      "step": 140
+    },
+    {
+      "epoch": 0.009824756994042434,
+      "grad_norm": 0.711429048225027,
+      "learning_rate": 6.998499556249715e-07,
+      "loss": 1.5737,
+      "step": 141
+    },
+    {
+      "epoch": 0.009894436121659756,
+      "grad_norm": 0.6897849482508333,
+      "learning_rate": 6.998478199608309e-07,
+      "loss": 1.5071,
+      "step": 142
+    },
+    {
+      "epoch": 0.009964115249277079,
+      "grad_norm": 0.699945633650181,
+      "learning_rate": 6.998456692085361e-07,
+      "loss": 1.5943,
+      "step": 143
+    },
+    {
+      "epoch": 0.010033794376894401,
+      "grad_norm": 0.6689453841253,
+      "learning_rate": 6.998435033681907e-07,
+      "loss": 1.4587,
+      "step": 144
+    },
+    {
+      "epoch": 0.010103473504511724,
+      "grad_norm": 0.7133832784569326,
+      "learning_rate": 6.998413224398982e-07,
+      "loss": 1.5861,
+      "step": 145
+    },
+    {
+      "epoch": 0.010173152632129047,
+      "grad_norm": 0.7036410895242622,
+      "learning_rate": 6.998391264237631e-07,
+      "loss": 1.6346,
+      "step": 146
+    },
+    {
+      "epoch": 0.010242831759746367,
+      "grad_norm": 0.6886627438173859,
+      "learning_rate": 6.998369153198908e-07,
+      "loss": 1.5589,
+      "step": 147
+    },
+    {
+      "epoch": 0.01031251088736369,
+      "grad_norm": 0.7367558477965669,
+      "learning_rate": 6.998346891283872e-07,
+      "loss": 1.6064,
+      "step": 148
+    },
+    {
+      "epoch": 0.010382190014981012,
+      "grad_norm": 0.8105693813585202,
+      "learning_rate": 6.998324478493588e-07,
+      "loss": 1.6741,
+      "step": 149
+    },
+    {
+      "epoch": 0.010451869142598335,
+      "grad_norm": 0.7110523737644423,
+      "learning_rate": 6.998301914829133e-07,
+      "loss": 1.5858,
+      "step": 150
+    },
+    {
+      "epoch": 0.010521548270215658,
+      "grad_norm": 0.7292303967140454,
+      "learning_rate": 6.998279200291587e-07,
+      "loss": 1.644,
+      "step": 151
+    },
+    {
+      "epoch": 0.010591227397832978,
+      "grad_norm": 0.7220512349531498,
+      "learning_rate": 6.998256334882037e-07,
+      "loss": 1.5486,
+      "step": 152
+    },
+    {
+      "epoch": 0.0106609065254503,
+      "grad_norm": 0.7698867140395331,
+      "learning_rate": 6.998233318601581e-07,
+      "loss": 1.5194,
+      "step": 153
+    },
+    {
+      "epoch": 0.010730585653067623,
+      "grad_norm": 0.9732965207520899,
+      "learning_rate": 6.998210151451322e-07,
+      "loss": 1.5542,
+      "step": 154
+    },
+    {
+      "epoch": 0.010800264780684946,
+      "grad_norm": 0.7118245740667292,
+      "learning_rate": 6.998186833432369e-07,
+      "loss": 1.5825,
+      "step": 155
+    },
+    {
+      "epoch": 0.010869943908302268,
+      "grad_norm": 0.6767247887601149,
+      "learning_rate": 6.998163364545839e-07,
+      "loss": 1.5421,
+      "step": 156
+    },
+    {
+      "epoch": 0.010939623035919591,
+      "grad_norm": 0.7463837156055579,
+      "learning_rate": 6.998139744792858e-07,
+      "loss": 1.3482,
+      "step": 157
+    },
+    {
+      "epoch": 0.011009302163536912,
+      "grad_norm": 0.7023208534797809,
+      "learning_rate": 6.998115974174558e-07,
+      "loss": 1.5365,
+      "step": 158
+    },
+    {
+      "epoch": 0.011078981291154234,
+      "grad_norm": 0.7779897717499841,
+      "learning_rate": 6.998092052692076e-07,
+      "loss": 1.5526,
+      "step": 159
+    },
+    {
+      "epoch": 0.011148660418771557,
+      "grad_norm": 0.7022624765789008,
+      "learning_rate": 6.998067980346561e-07,
+      "loss": 1.5957,
+      "step": 160
+    },
+    {
+      "epoch": 0.01121833954638888,
+      "grad_norm": 0.7690238675537974,
+      "learning_rate": 6.998043757139166e-07,
+      "loss": 1.6412,
+      "step": 161
+    },
+    {
+      "epoch": 0.011288018674006202,
+      "grad_norm": 0.6977898185845065,
+      "learning_rate": 6.99801938307105e-07,
+      "loss": 1.5129,
+      "step": 162
+    },
+    {
+      "epoch": 0.011357697801623524,
+      "grad_norm": 0.7360024484783224,
+      "learning_rate": 6.997994858143384e-07,
+      "loss": 1.6246,
+      "step": 163
+    },
+    {
+      "epoch": 0.011427376929240845,
+      "grad_norm": 0.743352598561228,
+      "learning_rate": 6.997970182357341e-07,
+      "loss": 1.6589,
+      "step": 164
+    },
+    {
+      "epoch": 0.011497056056858168,
+      "grad_norm": 0.7438090842992685,
+      "learning_rate": 6.997945355714104e-07,
+      "loss": 1.7993,
+      "step": 165
+    },
+    {
+      "epoch": 0.01156673518447549,
+      "grad_norm": 0.7806976756154644,
+      "learning_rate": 6.997920378214862e-07,
+      "loss": 1.6904,
+      "step": 166
+    },
+    {
+      "epoch": 0.011636414312092813,
+      "grad_norm": 0.7617691034349805,
+      "learning_rate": 6.997895249860815e-07,
+      "loss": 1.5514,
+      "step": 167
+    },
+    {
+      "epoch": 0.011706093439710135,
+      "grad_norm": 0.6897868904950596,
+      "learning_rate": 6.997869970653164e-07,
+      "loss": 1.7041,
+      "step": 168
+    },
+    {
+      "epoch": 0.011775772567327458,
+      "grad_norm": 0.7171263670275106,
+      "learning_rate": 6.997844540593121e-07,
+      "loss": 1.5569,
+      "step": 169
+    },
+    {
+      "epoch": 0.011845451694944779,
+      "grad_norm": 0.7349273462264111,
+      "learning_rate": 6.997818959681906e-07,
+      "loss": 1.6214,
+      "step": 170
+    },
+    {
+      "epoch": 0.011915130822562101,
+      "grad_norm": 0.7595114670832553,
+      "learning_rate": 6.997793227920744e-07,
+      "loss": 1.7286,
+      "step": 171
+    },
+    {
+      "epoch": 0.011984809950179424,
+      "grad_norm": 0.6743455616725484,
+      "learning_rate": 6.997767345310868e-07,
+      "loss": 1.5967,
+      "step": 172
+    },
+    {
+      "epoch": 0.012054489077796746,
+      "grad_norm": 0.8442688339279371,
+      "learning_rate": 6.997741311853519e-07,
+      "loss": 1.5291,
+      "step": 173
+    },
+    {
+      "epoch": 0.012124168205414069,
+      "grad_norm": 0.7135683838950063,
+      "learning_rate": 6.997715127549944e-07,
+      "loss": 1.547,
+      "step": 174
+    },
+    {
+      "epoch": 0.01219384733303139,
+      "grad_norm": 0.7680553601030344,
+      "learning_rate": 6.997688792401398e-07,
+      "loss": 1.5167,
+      "step": 175
+    },
+    {
+      "epoch": 0.012263526460648712,
+      "grad_norm": 0.7351716147785469,
+      "learning_rate": 6.997662306409142e-07,
+      "loss": 1.5149,
+      "step": 176
+    },
+    {
+      "epoch": 0.012333205588266035,
+      "grad_norm": 0.708247558633638,
+      "learning_rate": 6.997635669574448e-07,
+      "loss": 1.5914,
+      "step": 177
+    },
+    {
+      "epoch": 0.012402884715883357,
+      "grad_norm": 0.7447006944945493,
+      "learning_rate": 6.997608881898589e-07,
+      "loss": 1.6293,
+      "step": 178
+    },
+    {
+      "epoch": 0.01247256384350068,
+      "grad_norm": 0.7344799477527223,
+      "learning_rate": 6.997581943382852e-07,
+      "loss": 1.6402,
+      "step": 179
+    },
+    {
+      "epoch": 0.012542242971118002,
+      "grad_norm": 0.7278240397379574,
+      "learning_rate": 6.997554854028525e-07,
+      "loss": 1.5459,
+      "step": 180
+    },
+    {
+      "epoch": 0.012611922098735323,
+      "grad_norm": 0.7020532255212869,
+      "learning_rate": 6.997527613836908e-07,
+      "loss": 1.5233,
+      "step": 181
+    },
+    {
+      "epoch": 0.012681601226352646,
+      "grad_norm": 0.7588235184758088,
+      "learning_rate": 6.997500222809307e-07,
+      "loss": 1.5568,
+      "step": 182
+    },
+    {
+      "epoch": 0.012751280353969968,
+      "grad_norm": 0.7530180397421503,
+      "learning_rate": 6.997472680947033e-07,
+      "loss": 1.5685,
+      "step": 183
+    },
+    {
+      "epoch": 0.01282095948158729,
+      "grad_norm": 0.7180796924001129,
+      "learning_rate": 6.997444988251405e-07,
+      "loss": 1.6382,
+      "step": 184
+    },
+    {
+      "epoch": 0.012890638609204613,
+      "grad_norm": 0.7664151954290458,
+      "learning_rate": 6.997417144723754e-07,
+      "loss": 1.5858,
+      "step": 185
+    },
+    {
+      "epoch": 0.012960317736821936,
+      "grad_norm": 0.7437188630824249,
+      "learning_rate": 6.99738915036541e-07,
+      "loss": 1.6035,
+      "step": 186
+    },
+    {
+      "epoch": 0.013029996864439257,
+      "grad_norm": 0.7201391130892155,
+      "learning_rate": 6.997361005177718e-07,
+      "loss": 1.5728,
+      "step": 187
+    },
+    {
+      "epoch": 0.013099675992056579,
+      "grad_norm": 0.7259717563800355,
+      "learning_rate": 6.997332709162024e-07,
+      "loss": 1.6235,
+      "step": 188
+    },
+    {
+      "epoch": 0.013169355119673902,
+      "grad_norm": 0.8047726194305217,
+      "learning_rate": 6.997304262319686e-07,
+      "loss": 1.7083,
+      "step": 189
+    },
+    {
+      "epoch": 0.013239034247291224,
+      "grad_norm": 0.7286276274616162,
+      "learning_rate": 6.997275664652066e-07,
+      "loss": 1.65,
+      "step": 190
+    },
+    {
+      "epoch": 0.013308713374908547,
+      "grad_norm": 0.7179888240618036,
+      "learning_rate": 6.997246916160535e-07,
+      "loss": 1.4977,
+      "step": 191
+    },
+    {
+      "epoch": 0.01337839250252587,
+      "grad_norm": 0.6454186584546411,
+      "learning_rate": 6.99721801684647e-07,
+      "loss": 1.4368,
+      "step": 192
+    },
+    {
+      "epoch": 0.01344807163014319,
+      "grad_norm": 0.8428174000154314,
+      "learning_rate": 6.997188966711259e-07,
+      "loss": 1.6111,
+      "step": 193
+    },
+    {
+      "epoch": 0.013517750757760513,
+      "grad_norm": 0.7142931494445387,
+      "learning_rate": 6.997159765756289e-07,
+      "loss": 1.4782,
+      "step": 194
+    },
+    {
+      "epoch": 0.013587429885377835,
+      "grad_norm": 0.7013011170224142,
+      "learning_rate": 6.997130413982963e-07,
+      "loss": 1.6277,
+      "step": 195
+    },
+    {
+      "epoch": 0.013657109012995158,
+      "grad_norm": 0.7323477085758628,
+      "learning_rate": 6.997100911392687e-07,
+      "loss": 1.6999,
+      "step": 196
+    },
+    {
+      "epoch": 0.01372678814061248,
+      "grad_norm": 0.6782232596228719,
+      "learning_rate": 6.997071257986873e-07,
+      "loss": 1.5167,
+      "step": 197
+    },
+    {
+      "epoch": 0.013796467268229801,
+      "grad_norm": 0.7661525942281635,
+      "learning_rate": 6.997041453766945e-07,
+      "loss": 1.5548,
+      "step": 198
+    },
+    {
+      "epoch": 0.013866146395847124,
+      "grad_norm": 0.7078911668543628,
+      "learning_rate": 6.997011498734329e-07,
+      "loss": 1.6429,
+      "step": 199
+    },
+    {
+      "epoch": 0.013935825523464446,
+      "grad_norm": 0.7303745750427693,
+      "learning_rate": 6.996981392890463e-07,
+      "loss": 1.5804,
+      "step": 200
+    },
+    {
+      "epoch": 0.014005504651081769,
+      "grad_norm": 0.6831734889505485,
+      "learning_rate": 6.996951136236786e-07,
+      "loss": 1.4779,
+      "step": 201
+    },
+    {
+      "epoch": 0.014075183778699091,
+      "grad_norm": 0.7347205886402175,
+      "learning_rate": 6.99692072877475e-07,
+      "loss": 1.5836,
+      "step": 202
+    },
+    {
+      "epoch": 0.014144862906316414,
+      "grad_norm": 0.7096195389620719,
+      "learning_rate": 6.996890170505813e-07,
+      "loss": 1.6197,
+      "step": 203
+    },
+    {
+      "epoch": 0.014214542033933734,
+      "grad_norm": 0.7562441052582111,
+      "learning_rate": 6.996859461431439e-07,
+      "loss": 1.7908,
+      "step": 204
+    },
+    {
+      "epoch": 0.014284221161551057,
+      "grad_norm": 0.7334066931518938,
+      "learning_rate": 6.996828601553098e-07,
+      "loss": 1.4389,
+      "step": 205
+    },
+    {
+      "epoch": 0.01435390028916838,
+      "grad_norm": 0.6840997752191323,
+      "learning_rate": 6.99679759087227e-07,
+      "loss": 1.5409,
+      "step": 206
+    },
+    {
+      "epoch": 0.014423579416785702,
+      "grad_norm": 0.7665481621203577,
+      "learning_rate": 6.996766429390443e-07,
+      "loss": 1.5991,
+      "step": 207
+    },
+    {
+      "epoch": 0.014493258544403025,
+      "grad_norm": 0.7817725118434202,
+      "learning_rate": 6.996735117109106e-07,
+      "loss": 1.5223,
+      "step": 208
+    },
+    {
+      "epoch": 0.014562937672020347,
+      "grad_norm": 0.7283839189183391,
+      "learning_rate": 6.996703654029764e-07,
+      "loss": 1.6684,
+      "step": 209
+    },
+    {
+      "epoch": 0.014632616799637668,
+      "grad_norm": 0.7032582333066247,
+      "learning_rate": 6.996672040153923e-07,
+      "loss": 1.5299,
+      "step": 210
+    },
+    {
+      "epoch": 0.01470229592725499,
+      "grad_norm": 0.8077693130167044,
+      "learning_rate": 6.996640275483096e-07,
+      "loss": 1.5585,
+      "step": 211
+    },
+    {
+      "epoch": 0.014771975054872313,
+      "grad_norm": 0.7057816760896902,
+      "learning_rate": 6.996608360018808e-07,
+      "loss": 1.6727,
+      "step": 212
+    },
+    {
+      "epoch": 0.014841654182489636,
+      "grad_norm": 0.7312558829705393,
+      "learning_rate": 6.996576293762589e-07,
+      "loss": 1.5828,
+      "step": 213
+    },
+    {
+      "epoch": 0.014911333310106958,
+      "grad_norm": 0.6892094634096153,
+      "learning_rate": 6.996544076715972e-07,
+      "loss": 1.5521,
+      "step": 214
+    },
+    {
+      "epoch": 0.014981012437724279,
+      "grad_norm": 0.7622336534357208,
+      "learning_rate": 6.996511708880504e-07,
+      "loss": 1.5927,
+      "step": 215
+    },
+    {
+      "epoch": 0.015050691565341601,
+      "grad_norm": 0.8615964785895364,
+      "learning_rate": 6.996479190257735e-07,
+      "loss": 1.495,
+      "step": 216
+    },
+    {
+      "epoch": 0.015120370692958924,
+      "grad_norm": 0.7461775817348781,
+      "learning_rate": 6.996446520849225e-07,
+      "loss": 1.6432,
+      "step": 217
+    },
+    {
+      "epoch": 0.015190049820576246,
+      "grad_norm": 0.702229537181909,
+      "learning_rate": 6.996413700656536e-07,
+      "loss": 1.5812,
+      "step": 218
+    },
+    {
+      "epoch": 0.015259728948193569,
+      "grad_norm": 0.7563790834731114,
+      "learning_rate": 6.996380729681244e-07,
+      "loss": 1.4811,
+      "step": 219
+    },
+    {
+      "epoch": 0.015329408075810892,
+      "grad_norm": 0.7242920709990179,
+      "learning_rate": 6.99634760792493e-07,
+      "loss": 1.4832,
+      "step": 220
+    },
+    {
+      "epoch": 0.015399087203428212,
+      "grad_norm": 0.7170579173498212,
+      "learning_rate": 6.996314335389177e-07,
+      "loss": 1.5585,
+      "step": 221
+    },
+    {
+      "epoch": 0.015468766331045535,
+      "grad_norm": 0.7477520743417225,
+      "learning_rate": 6.996280912075582e-07,
+      "loss": 1.6152,
+      "step": 222
+    },
+    {
+      "epoch": 0.015538445458662857,
+      "grad_norm": 0.7365668010013201,
+      "learning_rate": 6.996247337985746e-07,
+      "loss": 1.6056,
+      "step": 223
+    },
+    {
+      "epoch": 0.01560812458628018,
+      "grad_norm": 0.7407770775087316,
+      "learning_rate": 6.99621361312128e-07,
+      "loss": 1.6089,
+      "step": 224
+    },
+    {
+      "epoch": 0.0156778037138975,
+      "grad_norm": 0.7501162465150802,
+      "learning_rate": 6.996179737483797e-07,
+      "loss": 1.5875,
+      "step": 225
+    },
+    {
+      "epoch": 0.015747482841514823,
+      "grad_norm": 0.703304864155775,
+      "learning_rate": 6.996145711074923e-07,
+      "loss": 1.4987,
+      "step": 226
+    },
+    {
+      "epoch": 0.015817161969132146,
+      "grad_norm": 0.7613996009018047,
+      "learning_rate": 6.996111533896286e-07,
+      "loss": 1.517,
+      "step": 227
+    },
+    {
+      "epoch": 0.01588684109674947,
+      "grad_norm": 0.7295854207507647,
+      "learning_rate": 6.996077205949528e-07,
+      "loss": 1.5285,
+      "step": 228
+    },
+    {
+      "epoch": 0.01595652022436679,
+      "grad_norm": 0.7074237961088794,
+      "learning_rate": 6.99604272723629e-07,
+      "loss": 1.46,
+      "step": 229
+    },
+    {
+      "epoch": 0.016026199351984113,
+      "grad_norm": 0.7560396455970501,
+      "learning_rate": 6.996008097758227e-07,
+      "loss": 1.505,
+      "step": 230
+    },
+    {
+      "epoch": 0.016095878479601436,
+      "grad_norm": 0.7333373513299797,
+      "learning_rate": 6.995973317516995e-07,
+      "loss": 1.5871,
+      "step": 231
+    },
+    {
+      "epoch": 0.01616555760721876,
+      "grad_norm": 0.640181179700889,
+      "learning_rate": 6.995938386514265e-07,
+      "loss": 1.5224,
+      "step": 232
+    },
+    {
+      "epoch": 0.01623523673483608,
+      "grad_norm": 0.7101909631345903,
+      "learning_rate": 6.995903304751709e-07,
+      "loss": 1.6038,
+      "step": 233
+    },
+    {
+      "epoch": 0.016304915862453404,
+      "grad_norm": 0.7454843789810528,
+      "learning_rate": 6.995868072231007e-07,
+      "loss": 1.627,
+      "step": 234
+    },
+    {
+      "epoch": 0.016374594990070723,
+      "grad_norm": 0.7024591060628032,
+      "learning_rate": 6.99583268895385e-07,
+      "loss": 1.5807,
+      "step": 235
+    },
+    {
+      "epoch": 0.016444274117688045,
+      "grad_norm": 0.8070089400837613,
+      "learning_rate": 6.995797154921931e-07,
+      "loss": 1.5066,
+      "step": 236
+    },
+    {
+      "epoch": 0.016513953245305368,
+      "grad_norm": 0.7292465008262952,
+      "learning_rate": 6.995761470136955e-07,
+      "loss": 1.5627,
+      "step": 237
+    },
+    {
+      "epoch": 0.01658363237292269,
+      "grad_norm": 0.7678168219350536,
+      "learning_rate": 6.995725634600631e-07,
+      "loss": 1.6508,
+      "step": 238
+    },
+    {
+      "epoch": 0.016653311500540013,
+      "grad_norm": 0.7171811611510969,
+      "learning_rate": 6.995689648314677e-07,
+      "loss": 1.6866,
+      "step": 239
+    },
+    {
+      "epoch": 0.016722990628157335,
+      "grad_norm": 0.7466351681320509,
+      "learning_rate": 6.995653511280816e-07,
+      "loss": 1.6049,
+      "step": 240
+    },
+    {
+      "epoch": 0.016792669755774658,
+      "grad_norm": 0.7266320402780408,
+      "learning_rate": 6.995617223500782e-07,
+      "loss": 1.5321,
+      "step": 241
+    },
+    {
+      "epoch": 0.01686234888339198,
+      "grad_norm": 0.7063647038715658,
+      "learning_rate": 6.995580784976312e-07,
+      "loss": 1.5047,
+      "step": 242
+    },
+    {
+      "epoch": 0.016932028011009303,
+      "grad_norm": 0.7085394609516711,
+      "learning_rate": 6.995544195709153e-07,
+      "loss": 1.7102,
+      "step": 243
+    },
+    {
+      "epoch": 0.017001707138626625,
+      "grad_norm": 0.7046716168804864,
+      "learning_rate": 6.995507455701059e-07,
+      "loss": 1.4328,
+      "step": 244
+    },
+    {
+      "epoch": 0.017071386266243948,
+      "grad_norm": 0.72750562924709,
+      "learning_rate": 6.995470564953791e-07,
+      "loss": 1.5067,
+      "step": 245
+    },
+    {
+      "epoch": 0.01714106539386127,
+      "grad_norm": 0.6583564224352184,
+      "learning_rate": 6.995433523469114e-07,
+      "loss": 1.5941,
+      "step": 246
+    },
+    {
+      "epoch": 0.01721074452147859,
+      "grad_norm": 0.6829271380484534,
+      "learning_rate": 6.995396331248806e-07,
+      "loss": 1.5829,
+      "step": 247
+    },
+    {
+      "epoch": 0.017280423649095912,
+      "grad_norm": 0.7501970674237022,
+      "learning_rate": 6.99535898829465e-07,
+      "loss": 1.5741,
+      "step": 248
+    },
+    {
+      "epoch": 0.017350102776713235,
+      "grad_norm": 0.718589643719869,
+      "learning_rate": 6.995321494608432e-07,
+      "loss": 1.6546,
+      "step": 249
+    },
+    {
+      "epoch": 0.017419781904330557,
+      "grad_norm": 0.7018977730704544,
+      "learning_rate": 6.995283850191951e-07,
+      "loss": 1.6441,
+      "step": 250
+    },
+    {
+      "epoch": 0.01748946103194788,
+      "grad_norm": 0.7116814108818308,
+      "learning_rate": 6.99524605504701e-07,
+      "loss": 1.6247,
+      "step": 251
+    },
+    {
+      "epoch": 0.017559140159565202,
+      "grad_norm": 0.7417559961150321,
+      "learning_rate": 6.995208109175422e-07,
+      "loss": 1.6553,
+      "step": 252
+    },
+    {
+      "epoch": 0.017628819287182525,
+      "grad_norm": 0.7195473670622379,
+      "learning_rate": 6.995170012579004e-07,
+      "loss": 1.4587,
+      "step": 253
+    },
+    {
+      "epoch": 0.017698498414799847,
+      "grad_norm": 0.6889531952135118,
+      "learning_rate": 6.995131765259583e-07,
+      "loss": 1.6015,
+      "step": 254
+    },
+    {
+      "epoch": 0.01776817754241717,
+      "grad_norm": 0.6800146240575404,
+      "learning_rate": 6.995093367218991e-07,
+      "loss": 1.542,
+      "step": 255
+    },
+    {
+      "epoch": 0.017837856670034492,
+      "grad_norm": 0.706603053149328,
+      "learning_rate": 6.995054818459067e-07,
+      "loss": 1.516,
+      "step": 256
+    },
+    {
+      "epoch": 0.017907535797651815,
+      "grad_norm": 0.7258769121019596,
+      "learning_rate": 6.99501611898166e-07,
+      "loss": 1.5032,
+      "step": 257
+    },
+    {
+      "epoch": 0.017977214925269134,
+      "grad_norm": 0.7475511019702241,
+      "learning_rate": 6.994977268788624e-07,
+      "loss": 1.5867,
+      "step": 258
+    },
+    {
+      "epoch": 0.018046894052886456,
+      "grad_norm": 0.74299733714245,
+      "learning_rate": 6.99493826788182e-07,
+      "loss": 1.6865,
+      "step": 259
+    },
+    {
+      "epoch": 0.01811657318050378,
+      "grad_norm": 0.7066253885889029,
+      "learning_rate": 6.99489911626312e-07,
+      "loss": 1.6019,
+      "step": 260
+    },
+    {
+      "epoch": 0.0181862523081211,
+      "grad_norm": 0.8182502015789678,
+      "learning_rate": 6.994859813934395e-07,
+      "loss": 1.4875,
+      "step": 261
+    },
+    {
+      "epoch": 0.018255931435738424,
+      "grad_norm": 0.7488901193748735,
+      "learning_rate": 6.994820360897534e-07,
+      "loss": 1.5313,
+      "step": 262
+    },
+    {
+      "epoch": 0.018325610563355747,
+      "grad_norm": 0.7298797429645452,
+      "learning_rate": 6.994780757154423e-07,
+      "loss": 1.5985,
+      "step": 263
+    },
+    {
+      "epoch": 0.01839528969097307,
+      "grad_norm": 0.7494430286408182,
+      "learning_rate": 6.994741002706963e-07,
+      "loss": 1.6039,
+      "step": 264
+    },
+    {
+      "epoch": 0.01846496881859039,
+      "grad_norm": 0.749305407018932,
+      "learning_rate": 6.994701097557058e-07,
+      "loss": 1.6647,
+      "step": 265
+    },
+    {
+      "epoch": 0.018534647946207714,
+      "grad_norm": 0.7393237097632813,
+      "learning_rate": 6.99466104170662e-07,
+      "loss": 1.6921,
+      "step": 266
+    },
+    {
+      "epoch": 0.018604327073825037,
+      "grad_norm": 0.7052812834008583,
+      "learning_rate": 6.99462083515757e-07,
+      "loss": 1.4998,
+      "step": 267
+    },
+    {
+      "epoch": 0.01867400620144236,
+      "grad_norm": 0.7656850091837507,
+      "learning_rate": 6.994580477911834e-07,
+      "loss": 1.551,
+      "step": 268
+    },
+    {
+      "epoch": 0.018743685329059682,
+      "grad_norm": 0.7262479768296017,
+      "learning_rate": 6.994539969971345e-07,
+      "loss": 1.6432,
+      "step": 269
+    },
+    {
+      "epoch": 0.018813364456677,
+      "grad_norm": 0.6957600485001357,
+      "learning_rate": 6.994499311338046e-07,
+      "loss": 1.6025,
+      "step": 270
+    },
+    {
+      "epoch": 0.018883043584294323,
+      "grad_norm": 0.6905054997272053,
+      "learning_rate": 6.994458502013883e-07,
+      "loss": 1.4301,
+      "step": 271
+    },
+    {
+      "epoch": 0.018952722711911646,
+      "grad_norm": 0.7179445614004244,
+      "learning_rate": 6.994417542000813e-07,
+      "loss": 1.5864,
+      "step": 272
+    },
+    {
+      "epoch": 0.01902240183952897,
+      "grad_norm": 0.7072984855963963,
+      "learning_rate": 6.994376431300801e-07,
+      "loss": 1.6053,
+      "step": 273
+    },
+    {
+      "epoch": 0.01909208096714629,
+      "grad_norm": 0.7570107373874823,
+      "learning_rate": 6.994335169915813e-07,
+      "loss": 1.6383,
+      "step": 274
+    },
+    {
+      "epoch": 0.019161760094763614,
+      "grad_norm": 0.7274334966890454,
+      "learning_rate": 6.99429375784783e-07,
+      "loss": 1.625,
+      "step": 275
+    },
+    {
+      "epoch": 0.019231439222380936,
+      "grad_norm": 0.7563658771609096,
+      "learning_rate": 6.994252195098834e-07,
+      "loss": 1.4685,
+      "step": 276
+    },
+    {
+      "epoch": 0.01930111834999826,
+      "grad_norm": 0.6811578449534527,
+      "learning_rate": 6.994210481670817e-07,
+      "loss": 1.5123,
+      "step": 277
+    },
+    {
+      "epoch": 0.01937079747761558,
+      "grad_norm": 0.6786727534901147,
+      "learning_rate": 6.994168617565782e-07,
+      "loss": 1.4686,
+      "step": 278
+    },
+    {
+      "epoch": 0.019440476605232904,
+      "grad_norm": 0.7655364222476935,
+      "learning_rate": 6.994126602785729e-07,
+      "loss": 1.4576,
+      "step": 279
+    },
+    {
+      "epoch": 0.019510155732850226,
+      "grad_norm": 0.6821324813628526,
+      "learning_rate": 6.994084437332676e-07,
+      "loss": 1.5197,
+      "step": 280
+    },
+    {
+      "epoch": 0.019579834860467545,
+      "grad_norm": 0.7262970704517422,
+      "learning_rate": 6.99404212120864e-07,
+      "loss": 1.4957,
+      "step": 281
+    },
+    {
+      "epoch": 0.019649513988084868,
+      "grad_norm": 0.743676021736133,
+      "learning_rate": 6.993999654415654e-07,
+      "loss": 1.6085,
+      "step": 282
+    },
+    {
+      "epoch": 0.01971919311570219,
+      "grad_norm": 0.6378550410731294,
+      "learning_rate": 6.993957036955749e-07,
+      "loss": 1.602,
+      "step": 283
+    },
+    {
+      "epoch": 0.019788872243319513,
+      "grad_norm": 0.7246033604464897,
+      "learning_rate": 6.993914268830967e-07,
+      "loss": 1.5291,
+      "step": 284
+    },
+    {
+      "epoch": 0.019858551370936835,
+      "grad_norm": 0.6880276199996247,
+      "learning_rate": 6.99387135004336e-07,
+      "loss": 1.4173,
+      "step": 285
+    },
+    {
+      "epoch": 0.019928230498554158,
+      "grad_norm": 0.6947157355311014,
+      "learning_rate": 6.993828280594983e-07,
+      "loss": 1.6025,
+      "step": 286
+    },
+    {
+      "epoch": 0.01999790962617148,
+      "grad_norm": 0.7435927158927187,
+      "learning_rate": 6.993785060487901e-07,
+      "loss": 1.4261,
+      "step": 287
+    },
+    {
+      "epoch": 0.020067588753788803,
+      "grad_norm": 0.7142473424703729,
+      "learning_rate": 6.993741689724185e-07,
+      "loss": 1.5124,
+      "step": 288
+    },
+    {
+      "epoch": 0.020137267881406126,
+      "grad_norm": 0.7008072048840218,
+      "learning_rate": 6.993698168305914e-07,
+      "loss": 1.5411,
+      "step": 289
+    },
+    {
+      "epoch": 0.020206947009023448,
+      "grad_norm": 0.7435854653268873,
+      "learning_rate": 6.993654496235172e-07,
+      "loss": 1.6077,
+      "step": 290
+    },
+    {
+      "epoch": 0.02027662613664077,
+      "grad_norm": 0.725379113270835,
+      "learning_rate": 6.993610673514052e-07,
+      "loss": 1.5128,
+      "step": 291
+    },
+    {
+      "epoch": 0.020346305264258093,
+      "grad_norm": 0.7117406405072563,
+      "learning_rate": 6.993566700144656e-07,
+      "loss": 1.6271,
+      "step": 292
+    },
+    {
+      "epoch": 0.020415984391875412,
+      "grad_norm": 0.702802060197601,
+      "learning_rate": 6.99352257612909e-07,
+      "loss": 1.6355,
+      "step": 293
+    },
+    {
+      "epoch": 0.020485663519492735,
+      "grad_norm": 0.7416404216373786,
+      "learning_rate": 6.993478301469469e-07,
+      "loss": 1.5558,
+      "step": 294
+    },
+    {
+      "epoch": 0.020555342647110057,
+      "grad_norm": 6.474567925910475,
+      "learning_rate": 6.993433876167913e-07,
+      "loss": 1.6586,
+      "step": 295
+    },
+    {
+      "epoch": 0.02062502177472738,
+      "grad_norm": 0.7080837056054183,
+      "learning_rate": 6.993389300226553e-07,
+      "loss": 1.6812,
+      "step": 296
+    },
+    {
+      "epoch": 0.020694700902344702,
+      "grad_norm": 0.6653198180715081,
+      "learning_rate": 6.993344573647524e-07,
+      "loss": 1.4974,
+      "step": 297
+    },
+    {
+      "epoch": 0.020764380029962025,
+      "grad_norm": 0.7204476403446304,
+      "learning_rate": 6.99329969643297e-07,
+      "loss": 1.5675,
+      "step": 298
+    },
+    {
+      "epoch": 0.020834059157579347,
+      "grad_norm": 0.7445246032941663,
+      "learning_rate": 6.993254668585042e-07,
+      "loss": 1.5476,
+      "step": 299
+    },
+    {
+      "epoch": 0.02090373828519667,
+      "grad_norm": 0.7460425723778988,
+      "learning_rate": 6.993209490105897e-07,
+      "loss": 1.4898,
+      "step": 300
+    },
+    {
+      "epoch": 0.020973417412813992,
+      "grad_norm": 0.7157641517249568,
+      "learning_rate": 6.9931641609977e-07,
+      "loss": 1.5786,
+      "step": 301
+    },
+    {
+      "epoch": 0.021043096540431315,
+      "grad_norm": 0.7016023826923905,
+      "learning_rate": 6.993118681262625e-07,
+      "loss": 1.5981,
+      "step": 302
+    },
+    {
+      "epoch": 0.021112775668048638,
+      "grad_norm": 0.7382367733171915,
+      "learning_rate": 6.993073050902849e-07,
+      "loss": 1.5072,
+      "step": 303
+    },
+    {
+      "epoch": 0.021182454795665957,
+      "grad_norm": 0.7606065005113284,
+      "learning_rate": 6.99302726992056e-07,
+      "loss": 1.4621,
+      "step": 304
+    },
+    {
+      "epoch": 0.02125213392328328,
+      "grad_norm": 0.6813739822883382,
+      "learning_rate": 6.992981338317952e-07,
+      "loss": 1.4303,
+      "step": 305
+    },
+    {
+      "epoch": 0.0213218130509006,
+      "grad_norm": 9.560327004767403,
+      "learning_rate": 6.992935256097225e-07,
+      "loss": 1.6614,
+      "step": 306
+    },
+    {
+      "epoch": 0.021391492178517924,
+      "grad_norm": 0.7301189610598565,
+      "learning_rate": 6.992889023260589e-07,
+      "loss": 1.7052,
+      "step": 307
+    },
+    {
+      "epoch": 0.021461171306135247,
+      "grad_norm": 0.7407277030456509,
+      "learning_rate": 6.992842639810258e-07,
+      "loss": 1.6177,
+      "step": 308
+    },
+    {
+      "epoch": 0.02153085043375257,
+      "grad_norm": 0.6916162651462537,
+      "learning_rate": 6.992796105748455e-07,
+      "loss": 1.3961,
+      "step": 309
+    },
+    {
+      "epoch": 0.021600529561369892,
+      "grad_norm": 0.7318959820786545,
+      "learning_rate": 6.992749421077412e-07,
+      "loss": 1.6585,
+      "step": 310
+    },
+    {
+      "epoch": 0.021670208688987214,
+      "grad_norm": 0.7366437888923923,
+      "learning_rate": 6.992702585799365e-07,
+      "loss": 1.6567,
+      "step": 311
+    },
+    {
+      "epoch": 0.021739887816604537,
+      "grad_norm": 0.7091001881005623,
+      "learning_rate": 6.992655599916557e-07,
+      "loss": 1.4418,
+      "step": 312
+    },
+    {
+      "epoch": 0.02180956694422186,
+      "grad_norm": 0.7275485003563671,
+      "learning_rate": 6.992608463431242e-07,
+      "loss": 1.5372,
+      "step": 313
+    },
+    {
+      "epoch": 0.021879246071839182,
+      "grad_norm": 0.6704285285994563,
+      "learning_rate": 6.992561176345678e-07,
+      "loss": 1.4605,
+      "step": 314
+    },
+    {
+      "epoch": 0.021948925199456504,
+      "grad_norm": 0.7125544818193529,
+      "learning_rate": 6.99251373866213e-07,
+      "loss": 1.5262,
+      "step": 315
+    },
+    {
+      "epoch": 0.022018604327073824,
+      "grad_norm": 0.7509609338377493,
+      "learning_rate": 6.992466150382873e-07,
+      "loss": 1.5833,
+      "step": 316
+    },
+    {
+      "epoch": 0.022088283454691146,
+      "grad_norm": 0.747340093745921,
+      "learning_rate": 6.992418411510185e-07,
+      "loss": 1.6504,
+      "step": 317
+    },
+    {
+      "epoch": 0.02215796258230847,
+      "grad_norm": 0.7046299436255133,
+      "learning_rate": 6.992370522046357e-07,
+      "loss": 1.4034,
+      "step": 318
+    },
+    {
+      "epoch": 0.02222764170992579,
+      "grad_norm": 0.7257162097711396,
+      "learning_rate": 6.992322481993683e-07,
+      "loss": 1.5848,
+      "step": 319
+    },
+    {
+      "epoch": 0.022297320837543114,
+      "grad_norm": 0.6691152644528866,
+      "learning_rate": 6.992274291354463e-07,
+      "loss": 1.4598,
+      "step": 320
+    },
+    {
+      "epoch": 0.022366999965160436,
+      "grad_norm": 0.7457614871946824,
+      "learning_rate": 6.992225950131009e-07,
+      "loss": 1.5387,
+      "step": 321
+    },
+    {
+      "epoch": 0.02243667909277776,
+      "grad_norm": 0.689307701631239,
+      "learning_rate": 6.992177458325635e-07,
+      "loss": 1.5808,
+      "step": 322
+    },
+    {
+      "epoch": 0.02250635822039508,
+      "grad_norm": 0.7335991464373063,
+      "learning_rate": 6.992128815940668e-07,
+      "loss": 1.6553,
+      "step": 323
+    },
+    {
+      "epoch": 0.022576037348012404,
+      "grad_norm": 0.6899224219200941,
+      "learning_rate": 6.992080022978437e-07,
+      "loss": 1.5265,
+      "step": 324
+    },
+    {
+      "epoch": 0.022645716475629726,
+      "grad_norm": 0.6806245477363745,
+      "learning_rate": 6.992031079441282e-07,
+      "loss": 1.4794,
+      "step": 325
+    },
+    {
+      "epoch": 0.02271539560324705,
+      "grad_norm": 0.8141536466407525,
+      "learning_rate": 6.991981985331546e-07,
+      "loss": 1.5921,
+      "step": 326
+    },
+    {
+      "epoch": 0.022785074730864368,
+      "grad_norm": 0.6788755156136146,
+      "learning_rate": 6.991932740651582e-07,
+      "loss": 1.5735,
+      "step": 327
+    },
+    {
+      "epoch": 0.02285475385848169,
+      "grad_norm": 0.7272500675836373,
+      "learning_rate": 6.991883345403751e-07,
+      "loss": 1.55,
+      "step": 328
+    },
+    {
+      "epoch": 0.022924432986099013,
+      "grad_norm": 0.7009298523819532,
+      "learning_rate": 6.991833799590422e-07,
+      "loss": 1.4261,
+      "step": 329
+    },
+    {
+      "epoch": 0.022994112113716336,
+      "grad_norm": 0.6948863685694087,
+      "learning_rate": 6.991784103213965e-07,
+      "loss": 1.5659,
+      "step": 330
+    },
+    {
+      "epoch": 0.023063791241333658,
+      "grad_norm": 0.7063327387262183,
+      "learning_rate": 6.991734256276766e-07,
+      "loss": 1.5343,
+      "step": 331
+    },
+    {
+      "epoch": 0.02313347036895098,
+      "grad_norm": 0.7099461836572424,
+      "learning_rate": 6.99168425878121e-07,
+      "loss": 1.5125,
+      "step": 332
+    },
+    {
+      "epoch": 0.023203149496568303,
+      "grad_norm": 0.7796603493076472,
+      "learning_rate": 6.991634110729694e-07,
+      "loss": 1.6092,
+      "step": 333
+    },
+    {
+      "epoch": 0.023272828624185626,
+      "grad_norm": 0.7296219083016224,
+      "learning_rate": 6.991583812124623e-07,
+      "loss": 1.6215,
+      "step": 334
+    },
+    {
+      "epoch": 0.023342507751802948,
+      "grad_norm": 0.72574929831423,
+      "learning_rate": 6.991533362968406e-07,
+      "loss": 1.6325,
+      "step": 335
+    },
+    {
+      "epoch": 0.02341218687942027,
+      "grad_norm": 0.7019432899693439,
+      "learning_rate": 6.99148276326346e-07,
+      "loss": 1.5221,
+      "step": 336
+    },
+    {
+      "epoch": 0.023481866007037593,
+      "grad_norm": 0.7492264986012627,
+      "learning_rate": 6.991432013012211e-07,
+      "loss": 1.6158,
+      "step": 337
+    },
+    {
+      "epoch": 0.023551545134654916,
+      "grad_norm": 0.692588523766954,
+      "learning_rate": 6.991381112217092e-07,
+      "loss": 1.5078,
+      "step": 338
+    },
+    {
+      "epoch": 0.023621224262272235,
+      "grad_norm": 0.6981979538678789,
+      "learning_rate": 6.991330060880539e-07,
+      "loss": 1.6101,
+      "step": 339
+    },
+    {
+      "epoch": 0.023690903389889557,
+      "grad_norm": 0.7194235046945815,
+      "learning_rate": 6.991278859005003e-07,
+      "loss": 1.5601,
+      "step": 340
+    },
+    {
+      "epoch": 0.02376058251750688,
+      "grad_norm": 0.7420959635718415,
+      "learning_rate": 6.991227506592932e-07,
+      "loss": 1.6685,
+      "step": 341
+    },
+    {
+      "epoch": 0.023830261645124202,
+      "grad_norm": 0.7089125015154109,
+      "learning_rate": 6.991176003646792e-07,
+      "loss": 1.6257,
+      "step": 342
+    },
+    {
+      "epoch": 0.023899940772741525,
+      "grad_norm": 0.7210440382167659,
+      "learning_rate": 6.991124350169049e-07,
+      "loss": 1.5857,
+      "step": 343
+    },
+    {
+      "epoch": 0.023969619900358848,
+      "grad_norm": 0.7827481182453416,
+      "learning_rate": 6.991072546162178e-07,
+      "loss": 1.6987,
+      "step": 344
+    },
+    {
+      "epoch": 0.02403929902797617,
+      "grad_norm": 0.7395681382788681,
+      "learning_rate": 6.991020591628663e-07,
+      "loss": 1.6111,
+      "step": 345
+    },
+    {
+      "epoch": 0.024108978155593493,
+      "grad_norm": 0.800887301847492,
+      "learning_rate": 6.990968486570993e-07,
+      "loss": 1.5495,
+      "step": 346
+    },
+    {
+      "epoch": 0.024178657283210815,
+      "grad_norm": 0.7185838722342582,
+      "learning_rate": 6.990916230991663e-07,
+      "loss": 1.6471,
+      "step": 347
+    },
+    {
+      "epoch": 0.024248336410828138,
+      "grad_norm": 0.6988928940290781,
+      "learning_rate": 6.990863824893181e-07,
+      "loss": 1.5798,
+      "step": 348
+    },
+    {
+      "epoch": 0.02431801553844546,
+      "grad_norm": 0.7670273052980163,
+      "learning_rate": 6.990811268278056e-07,
+      "loss": 1.6012,
+      "step": 349
+    },
+    {
+      "epoch": 0.02438769466606278,
+      "grad_norm": 0.6956525844641243,
+      "learning_rate": 6.990758561148806e-07,
+      "loss": 1.6228,
+      "step": 350
+    },
+    {
+      "epoch": 0.024457373793680102,
+      "grad_norm": 0.7380720675968462,
+      "learning_rate": 6.990705703507959e-07,
+      "loss": 1.6031,
+      "step": 351
+    },
+    {
+      "epoch": 0.024527052921297424,
+      "grad_norm": 0.7351909737787244,
+      "learning_rate": 6.990652695358046e-07,
+      "loss": 1.5208,
+      "step": 352
+    },
+    {
+      "epoch": 0.024596732048914747,
+      "grad_norm": 0.740389363327584,
+      "learning_rate": 6.990599536701608e-07,
+      "loss": 1.5537,
+      "step": 353
+    },
+    {
+      "epoch": 0.02466641117653207,
+      "grad_norm": 0.8039919377703592,
+      "learning_rate": 6.990546227541194e-07,
+      "loss": 1.7098,
+      "step": 354
+    },
+    {
+      "epoch": 0.024736090304149392,
+      "grad_norm": 0.7304226595911018,
+      "learning_rate": 6.990492767879357e-07,
+      "loss": 1.605,
+      "step": 355
+    },
+    {
+      "epoch": 0.024805769431766714,
+      "grad_norm": 0.741024497237045,
+      "learning_rate": 6.990439157718658e-07,
+      "loss": 1.5043,
+      "step": 356
+    },
+    {
+      "epoch": 0.024875448559384037,
+      "grad_norm": 0.7024368053127528,
+      "learning_rate": 6.990385397061669e-07,
+      "loss": 1.5534,
+      "step": 357
+    },
+    {
+      "epoch": 0.02494512768700136,
+      "grad_norm": 0.7323425435804326,
+      "learning_rate": 6.990331485910965e-07,
+      "loss": 1.6298,
+      "step": 358
+    },
+    {
+      "epoch": 0.025014806814618682,
+      "grad_norm": 0.7290774888294044,
+      "learning_rate": 6.990277424269127e-07,
+      "loss": 1.4802,
+      "step": 359
+    },
+    {
+      "epoch": 0.025084485942236005,
+      "grad_norm": 0.6658582153329362,
+      "learning_rate": 6.990223212138749e-07,
+      "loss": 1.5377,
+      "step": 360
+    },
+    {
+      "epoch": 0.025154165069853327,
+      "grad_norm": 0.739800561775487,
+      "learning_rate": 6.990168849522429e-07,
+      "loss": 1.5655,
+      "step": 361
+    },
+    {
+      "epoch": 0.025223844197470646,
+      "grad_norm": 0.6911945679125883,
+      "learning_rate": 6.99011433642277e-07,
+      "loss": 1.6444,
+      "step": 362
+    },
+    {
+      "epoch": 0.02529352332508797,
+      "grad_norm": 0.7460442162123894,
+      "learning_rate": 6.990059672842386e-07,
+      "loss": 1.4954,
+      "step": 363
+    },
+    {
+      "epoch": 0.02536320245270529,
+      "grad_norm": 0.8406791490688356,
+      "learning_rate": 6.990004858783895e-07,
+      "loss": 1.683,
+      "step": 364
+    },
+    {
+      "epoch": 0.025432881580322614,
+      "grad_norm": 0.6825887722241889,
+      "learning_rate": 6.989949894249926e-07,
+      "loss": 1.4904,
+      "step": 365
+    },
+    {
+      "epoch": 0.025502560707939936,
+      "grad_norm": 0.734984485601963,
+      "learning_rate": 6.989894779243111e-07,
+      "loss": 1.57,
+      "step": 366
+    },
+    {
+      "epoch": 0.02557223983555726,
+      "grad_norm": 0.802220464323417,
+      "learning_rate": 6.989839513766093e-07,
+      "loss": 1.5195,
+      "step": 367
+    },
+    {
+      "epoch": 0.02564191896317458,
+      "grad_norm": 0.7037489151381637,
+      "learning_rate": 6.989784097821519e-07,
+      "loss": 1.4717,
+      "step": 368
+    },
+    {
+      "epoch": 0.025711598090791904,
+      "grad_norm": 0.7343259525355395,
+      "learning_rate": 6.989728531412046e-07,
+      "loss": 1.5571,
+      "step": 369
+    },
+    {
+      "epoch": 0.025781277218409226,
+      "grad_norm": 0.6747397697253835,
+      "learning_rate": 6.989672814540335e-07,
+      "loss": 1.5264,
+      "step": 370
+    },
+    {
+      "epoch": 0.02585095634602655,
+      "grad_norm": 0.7568976619670307,
+      "learning_rate": 6.989616947209057e-07,
+      "loss": 1.6249,
+      "step": 371
+    },
+    {
+      "epoch": 0.02592063547364387,
+      "grad_norm": 0.6759327486428545,
+      "learning_rate": 6.98956092942089e-07,
+      "loss": 1.5019,
+      "step": 372
+    },
+    {
+      "epoch": 0.02599031460126119,
+      "grad_norm": 0.709291267177882,
+      "learning_rate": 6.989504761178519e-07,
+      "loss": 1.5009,
+      "step": 373
+    },
+    {
+      "epoch": 0.026059993728878513,
+      "grad_norm": 0.7099006099654483,
+      "learning_rate": 6.989448442484632e-07,
+      "loss": 1.5475,
+      "step": 374
+    },
+    {
+      "epoch": 0.026129672856495836,
+      "grad_norm": 0.690112841906305,
+      "learning_rate": 6.989391973341932e-07,
+      "loss": 1.4402,
+      "step": 375
+    },
+    {
+      "epoch": 0.026199351984113158,
+      "grad_norm": 0.7427387751223772,
+      "learning_rate": 6.989335353753122e-07,
+      "loss": 1.5491,
+      "step": 376
+    },
+    {
+      "epoch": 0.02626903111173048,
+      "grad_norm": 0.6770409349734714,
+      "learning_rate": 6.98927858372092e-07,
+      "loss": 1.4618,
+      "step": 377
+    },
+    {
+      "epoch": 0.026338710239347803,
+      "grad_norm": 0.7358590438958797,
+      "learning_rate": 6.989221663248041e-07,
+      "loss": 1.5153,
+      "step": 378
+    },
+    {
+      "epoch": 0.026408389366965126,
+      "grad_norm": 0.6659329097898178,
+      "learning_rate": 6.989164592337216e-07,
+      "loss": 1.5076,
+      "step": 379
+    },
+    {
+      "epoch": 0.02647806849458245,
+      "grad_norm": 0.6891744971335575,
+      "learning_rate": 6.989107370991179e-07,
+      "loss": 1.6146,
+      "step": 380
+    },
+    {
+      "epoch": 0.02654774762219977,
+      "grad_norm": 0.7550006904638933,
+      "learning_rate": 6.989049999212671e-07,
+      "loss": 1.5802,
+      "step": 381
+    },
+    {
+      "epoch": 0.026617426749817093,
+      "grad_norm": 0.7002665501804611,
+      "learning_rate": 6.988992477004446e-07,
+      "loss": 1.5355,
+      "step": 382
+    },
+    {
+      "epoch": 0.026687105877434416,
+      "grad_norm": 0.7009921240154263,
+      "learning_rate": 6.988934804369254e-07,
+      "loss": 1.5517,
+      "step": 383
+    },
+    {
+      "epoch": 0.02675678500505174,
+      "grad_norm": 0.7407286371939378,
+      "learning_rate": 6.988876981309864e-07,
+      "loss": 1.616,
+      "step": 384
+    },
+    {
+      "epoch": 0.026826464132669058,
+      "grad_norm": 0.7137375904069622,
+      "learning_rate": 6.988819007829045e-07,
+      "loss": 1.569,
+      "step": 385
+    },
+    {
+      "epoch": 0.02689614326028638,
+      "grad_norm": 0.6968322471953381,
+      "learning_rate": 6.988760883929575e-07,
+      "loss": 1.4872,
+      "step": 386
+    },
+    {
+      "epoch": 0.026965822387903703,
+      "grad_norm": 0.7335852919010041,
+      "learning_rate": 6.988702609614239e-07,
+      "loss": 1.5747,
+      "step": 387
+    },
+    {
+      "epoch": 0.027035501515521025,
+      "grad_norm": 0.7563187176836982,
+      "learning_rate": 6.98864418488583e-07,
+      "loss": 1.6437,
+      "step": 388
+    },
+    {
+      "epoch": 0.027105180643138348,
+      "grad_norm": 0.7615857240277533,
+      "learning_rate": 6.988585609747149e-07,
+      "loss": 1.6221,
+      "step": 389
+    },
+    {
+      "epoch": 0.02717485977075567,
+      "grad_norm": 0.7603071022901849,
+      "learning_rate": 6.988526884201002e-07,
+      "loss": 1.6949,
+      "step": 390
+    },
+    {
+      "epoch": 0.027244538898372993,
+      "grad_norm": 0.6968767287992647,
+      "learning_rate": 6.988468008250203e-07,
+      "loss": 1.6981,
+      "step": 391
+    },
+    {
+      "epoch": 0.027314218025990315,
+      "grad_norm": 0.7112654333278738,
+      "learning_rate": 6.988408981897575e-07,
+      "loss": 1.5906,
+      "step": 392
+    },
+    {
+      "epoch": 0.027383897153607638,
+      "grad_norm": 0.7298025420013617,
+      "learning_rate": 6.988349805145946e-07,
+      "loss": 1.466,
+      "step": 393
+    },
+    {
+      "epoch": 0.02745357628122496,
+      "grad_norm": 0.710177304695451,
+      "learning_rate": 6.98829047799815e-07,
+      "loss": 1.6849,
+      "step": 394
+    },
+    {
+      "epoch": 0.027523255408842283,
+      "grad_norm": 0.757594638471493,
+      "learning_rate": 6.988231000457031e-07,
+      "loss": 1.5379,
+      "step": 395
+    },
+    {
+      "epoch": 0.027592934536459602,
+      "grad_norm": 0.666881395058542,
+      "learning_rate": 6.98817137252544e-07,
+      "loss": 1.3365,
+      "step": 396
+    },
+    {
+      "epoch": 0.027662613664076924,
+      "grad_norm": 0.7591578570938375,
+      "learning_rate": 6.988111594206236e-07,
+      "loss": 1.621,
+      "step": 397
+    },
+    {
+      "epoch": 0.027732292791694247,
+      "grad_norm": 0.7175961360945237,
+      "learning_rate": 6.988051665502281e-07,
+      "loss": 1.6332,
+      "step": 398
+    },
+    {
+      "epoch": 0.02780197191931157,
+      "grad_norm": 0.6804206351265298,
+      "learning_rate": 6.987991586416447e-07,
+      "loss": 1.5439,
+      "step": 399
+    },
+    {
+      "epoch": 0.027871651046928892,
+      "grad_norm": 0.7399865885593209,
+      "learning_rate": 6.987931356951616e-07,
+      "loss": 1.6571,
+      "step": 400
+    },
+    {
+      "epoch": 0.027941330174546215,
+      "grad_norm": 0.6659744696681877,
+      "learning_rate": 6.987870977110671e-07,
+      "loss": 1.4773,
+      "step": 401
+    },
+    {
+      "epoch": 0.028011009302163537,
+      "grad_norm": 0.7512620715969653,
+      "learning_rate": 6.987810446896507e-07,
+      "loss": 1.473,
+      "step": 402
+    },
+    {
+      "epoch": 0.02808068842978086,
+      "grad_norm": 0.6700099927086145,
+      "learning_rate": 6.987749766312023e-07,
+      "loss": 1.4796,
+      "step": 403
+    },
+    {
+      "epoch": 0.028150367557398182,
+      "grad_norm": 0.7135736477251078,
+      "learning_rate": 6.98768893536013e-07,
+      "loss": 1.5236,
+      "step": 404
+    },
+    {
+      "epoch": 0.028220046685015505,
+      "grad_norm": 0.6892839109346525,
+      "learning_rate": 6.987627954043743e-07,
+      "loss": 1.6111,
+      "step": 405
+    },
+    {
+      "epoch": 0.028289725812632827,
+      "grad_norm": 0.6767335993021136,
+      "learning_rate": 6.987566822365781e-07,
+      "loss": 1.5699,
+      "step": 406
+    },
+    {
+      "epoch": 0.028359404940250146,
+      "grad_norm": 0.7433392074067198,
+      "learning_rate": 6.987505540329176e-07,
+      "loss": 1.4527,
+      "step": 407
+    },
+    {
+      "epoch": 0.02842908406786747,
+      "grad_norm": 0.6556767958241513,
+      "learning_rate": 6.987444107936865e-07,
+      "loss": 1.6098,
+      "step": 408
+    },
+    {
+      "epoch": 0.02849876319548479,
+      "grad_norm": 0.7215818443504534,
+      "learning_rate": 6.987382525191791e-07,
+      "loss": 1.4621,
+      "step": 409
+    },
+    {
+      "epoch": 0.028568442323102114,
+      "grad_norm": 0.7477993104746903,
+      "learning_rate": 6.987320792096907e-07,
+      "loss": 1.6124,
+      "step": 410
+    },
+    {
+      "epoch": 0.028638121450719436,
+      "grad_norm": 0.7201733172736332,
+      "learning_rate": 6.987258908655168e-07,
+      "loss": 1.4661,
+      "step": 411
+    },
+    {
+      "epoch": 0.02870780057833676,
+      "grad_norm": 0.72337434724977,
+      "learning_rate": 6.987196874869541e-07,
+      "loss": 1.8324,
+      "step": 412
+    },
+    {
+      "epoch": 0.02877747970595408,
+      "grad_norm": 0.7097137381375397,
+      "learning_rate": 6.987134690743e-07,
+      "loss": 1.5459,
+      "step": 413
+    },
+    {
+      "epoch": 0.028847158833571404,
+      "grad_norm": 0.7173005147308623,
+      "learning_rate": 6.987072356278523e-07,
+      "loss": 1.4824,
+      "step": 414
+    },
+    {
+      "epoch": 0.028916837961188727,
+      "grad_norm": 0.7879220907659518,
+      "learning_rate": 6.987009871479101e-07,
+      "loss": 1.5851,
+      "step": 415
+    },
+    {
+      "epoch": 0.02898651708880605,
+      "grad_norm": 0.6911412058391415,
+      "learning_rate": 6.986947236347724e-07,
+      "loss": 1.4977,
+      "step": 416
+    },
+    {
+      "epoch": 0.02905619621642337,
+      "grad_norm": 0.7012733331107831,
+      "learning_rate": 6.986884450887396e-07,
+      "loss": 1.565,
+      "step": 417
+    },
+    {
+      "epoch": 0.029125875344040694,
+      "grad_norm": 0.7840924933610383,
+      "learning_rate": 6.986821515101124e-07,
+      "loss": 1.55,
+      "step": 418
+    },
+    {
+      "epoch": 0.029195554471658013,
+      "grad_norm": 0.6964388522029603,
+      "learning_rate": 6.986758428991927e-07,
+      "loss": 1.6359,
+      "step": 419
+    },
+    {
+      "epoch": 0.029265233599275336,
+      "grad_norm": 0.7336109790356905,
+      "learning_rate": 6.986695192562826e-07,
+      "loss": 1.5707,
+      "step": 420
+    },
+    {
+      "epoch": 0.02933491272689266,
+      "grad_norm": 0.7719512610826261,
+      "learning_rate": 6.986631805816851e-07,
+      "loss": 1.6261,
+      "step": 421
+    },
+    {
+      "epoch": 0.02940459185450998,
+      "grad_norm": 0.7301564052254476,
+      "learning_rate": 6.986568268757041e-07,
+      "loss": 1.5733,
+      "step": 422
+    },
+    {
+      "epoch": 0.029474270982127303,
+      "grad_norm": 0.7533907159721789,
+      "learning_rate": 6.98650458138644e-07,
+      "loss": 1.4902,
+      "step": 423
+    },
+    {
+      "epoch": 0.029543950109744626,
+      "grad_norm": 0.7073959569566014,
+      "learning_rate": 6.986440743708101e-07,
+      "loss": 1.5164,
+      "step": 424
+    },
+    {
+      "epoch": 0.02961362923736195,
+      "grad_norm": 0.7145941743912411,
+      "learning_rate": 6.986376755725082e-07,
+      "loss": 1.6115,
+      "step": 425
+    },
+    {
+      "epoch": 0.02968330836497927,
+      "grad_norm": 0.6923721591358505,
+      "learning_rate": 6.98631261744045e-07,
+      "loss": 1.5826,
+      "step": 426
+    },
+    {
+      "epoch": 0.029752987492596594,
+      "grad_norm": 0.6839320727296647,
+      "learning_rate": 6.986248328857279e-07,
+      "loss": 1.4239,
+      "step": 427
+    },
+    {
+      "epoch": 0.029822666620213916,
+      "grad_norm": 0.7210066398143022,
+      "learning_rate": 6.986183889978649e-07,
+      "loss": 1.6437,
+      "step": 428
+    },
+    {
+      "epoch": 0.02989234574783124,
+      "grad_norm": 0.7008115162017041,
+      "learning_rate": 6.98611930080765e-07,
+      "loss": 1.5053,
+      "step": 429
+    },
+    {
+      "epoch": 0.029962024875448558,
+      "grad_norm": 0.6942891542494432,
+      "learning_rate": 6.986054561347374e-07,
+      "loss": 1.5146,
+      "step": 430
+    },
+    {
+      "epoch": 0.03003170400306588,
+      "grad_norm": 0.6635542554591299,
+      "learning_rate": 6.985989671600925e-07,
+      "loss": 1.4483,
+      "step": 431
+    },
+    {
+      "epoch": 0.030101383130683203,
+      "grad_norm": 0.7115892340457141,
+      "learning_rate": 6.985924631571414e-07,
+      "loss": 1.5982,
+      "step": 432
+    },
+    {
+      "epoch": 0.030171062258300525,
+      "grad_norm": 0.6854278325150915,
+      "learning_rate": 6.985859441261956e-07,
+      "loss": 1.437,
+      "step": 433
+    },
+    {
+      "epoch": 0.030240741385917848,
+      "grad_norm": 0.6759620583516114,
+      "learning_rate": 6.985794100675676e-07,
+      "loss": 1.5435,
+      "step": 434
+    },
+    {
+      "epoch": 0.03031042051353517,
+      "grad_norm": 0.6933685069524756,
+      "learning_rate": 6.985728609815706e-07,
+      "loss": 1.6306,
+      "step": 435
+    },
+    {
+      "epoch": 0.030380099641152493,
+      "grad_norm": 0.7146245192057525,
+      "learning_rate": 6.985662968685184e-07,
+      "loss": 1.6303,
+      "step": 436
+    },
+    {
+      "epoch": 0.030449778768769815,
+      "grad_norm": 0.7232446855847888,
+      "learning_rate": 6.985597177287253e-07,
+      "loss": 1.5931,
+      "step": 437
+    },
+    {
+      "epoch": 0.030519457896387138,
+      "grad_norm": 0.7184645069885551,
+      "learning_rate": 6.985531235625069e-07,
+      "loss": 1.5508,
+      "step": 438
+    },
+    {
+      "epoch": 0.03058913702400446,
+      "grad_norm": 0.7214958357639737,
+      "learning_rate": 6.98546514370179e-07,
+      "loss": 1.4922,
+      "step": 439
+    },
+    {
+      "epoch": 0.030658816151621783,
+      "grad_norm": 0.6782468340100151,
+      "learning_rate": 6.985398901520586e-07,
+      "loss": 1.5012,
+      "step": 440
+    },
+    {
+      "epoch": 0.030728495279239106,
+      "grad_norm": 0.7311905689952574,
+      "learning_rate": 6.985332509084629e-07,
+      "loss": 1.5831,
+      "step": 441
+    },
+    {
+      "epoch": 0.030798174406856425,
+      "grad_norm": 0.698328027036214,
+      "learning_rate": 6.985265966397102e-07,
+      "loss": 1.4995,
+      "step": 442
+    },
+    {
+      "epoch": 0.030867853534473747,
+      "grad_norm": 0.7368612541969222,
+      "learning_rate": 6.985199273461193e-07,
+      "loss": 1.4867,
+      "step": 443
+    },
+    {
+      "epoch": 0.03093753266209107,
+      "grad_norm": 0.7327398721151145,
+      "learning_rate": 6.985132430280098e-07,
+      "loss": 1.5259,
+      "step": 444
+    },
+    {
+      "epoch": 0.031007211789708392,
+      "grad_norm": 0.7331724114655763,
+      "learning_rate": 6.985065436857021e-07,
+      "loss": 1.5916,
+      "step": 445
+    },
+    {
+      "epoch": 0.031076890917325715,
+      "grad_norm": 0.7182699009047467,
+      "learning_rate": 6.984998293195171e-07,
+      "loss": 1.654,
+      "step": 446
+    },
+    {
+      "epoch": 0.031146570044943037,
+      "grad_norm": 0.6931250303313554,
+      "learning_rate": 6.984930999297769e-07,
+      "loss": 1.4452,
+      "step": 447
+    },
+    {
+      "epoch": 0.03121624917256036,
+      "grad_norm": 0.7721643708509517,
+      "learning_rate": 6.984863555168036e-07,
+      "loss": 1.5864,
+      "step": 448
+    },
+    {
+      "epoch": 0.03128592830017768,
+      "grad_norm": 0.7545867751828237,
+      "learning_rate": 6.984795960809205e-07,
+      "loss": 1.5629,
+      "step": 449
+    },
+    {
+      "epoch": 0.031355607427795,
+      "grad_norm": 0.7269873867725736,
+      "learning_rate": 6.984728216224517e-07,
+      "loss": 1.6225,
+      "step": 450
+    },
+    {
+      "epoch": 0.031425286555412324,
+      "grad_norm": 0.722602283050896,
+      "learning_rate": 6.984660321417218e-07,
+      "loss": 1.4438,
+      "step": 451
+    },
+    {
+      "epoch": 0.031494965683029646,
+      "grad_norm": 0.7384727981711023,
+      "learning_rate": 6.98459227639056e-07,
+      "loss": 1.5485,
+      "step": 452
+    },
+    {
+      "epoch": 0.03156464481064697,
+      "grad_norm": 0.7018161474292629,
+      "learning_rate": 6.984524081147805e-07,
+      "loss": 1.5858,
+      "step": 453
+    },
+    {
+      "epoch": 0.03163432393826429,
+      "grad_norm": 0.7169485562565262,
+      "learning_rate": 6.984455735692222e-07,
+      "loss": 1.6247,
+      "step": 454
+    },
+    {
+      "epoch": 0.031704003065881614,
+      "grad_norm": 0.7308662225701769,
+      "learning_rate": 6.984387240027084e-07,
+      "loss": 1.5516,
+      "step": 455
+    },
+    {
+      "epoch": 0.03177368219349894,
+      "grad_norm": 0.7209362918239302,
+      "learning_rate": 6.984318594155676e-07,
+      "loss": 1.4468,
+      "step": 456
+    },
+    {
+      "epoch": 0.03184336132111626,
+      "grad_norm": 0.7890065268272298,
+      "learning_rate": 6.984249798081286e-07,
+      "loss": 1.5742,
+      "step": 457
+    },
+    {
+      "epoch": 0.03191304044873358,
+      "grad_norm": 0.7443926307266672,
+      "learning_rate": 6.98418085180721e-07,
+      "loss": 1.5981,
+      "step": 458
+    },
+    {
+      "epoch": 0.031982719576350904,
+      "grad_norm": 0.7686644197473939,
+      "learning_rate": 6.984111755336755e-07,
+      "loss": 1.5131,
+      "step": 459
+    },
+    {
+      "epoch": 0.03205239870396823,
+      "grad_norm": 0.7515630377431824,
+      "learning_rate": 6.984042508673228e-07,
+      "loss": 1.5627,
+      "step": 460
+    },
+    {
+      "epoch": 0.03212207783158555,
+      "grad_norm": 0.8071194750966978,
+      "learning_rate": 6.983973111819951e-07,
+      "loss": 1.5605,
+      "step": 461
+    },
+    {
+      "epoch": 0.03219175695920287,
+      "grad_norm": 0.7068935640984225,
+      "learning_rate": 6.98390356478025e-07,
+      "loss": 1.5796,
+      "step": 462
+    },
+    {
+      "epoch": 0.032261436086820194,
+      "grad_norm": 0.7079434375614534,
+      "learning_rate": 6.983833867557455e-07,
+      "loss": 1.544,
+      "step": 463
+    },
+    {
+      "epoch": 0.03233111521443752,
+      "grad_norm": 0.6875118632785433,
+      "learning_rate": 6.983764020154909e-07,
+      "loss": 1.5442,
+      "step": 464
+    },
+    {
+      "epoch": 0.03240079434205484,
+      "grad_norm": 0.7374325063772078,
+      "learning_rate": 6.983694022575955e-07,
+      "loss": 1.4734,
+      "step": 465
+    },
+    {
+      "epoch": 0.03247047346967216,
+      "grad_norm": 0.68601080519657,
+      "learning_rate": 6.983623874823952e-07,
+      "loss": 1.4625,
+      "step": 466
+    },
+    {
+      "epoch": 0.032540152597289485,
+      "grad_norm": 0.7015382118886788,
+      "learning_rate": 6.983553576902259e-07,
+      "loss": 1.4303,
+      "step": 467
+    },
+    {
+      "epoch": 0.03260983172490681,
+      "grad_norm": 0.7230971366494634,
+      "learning_rate": 6.983483128814246e-07,
+      "loss": 1.5003,
+      "step": 468
+    },
+    {
+      "epoch": 0.03267951085252413,
+      "grad_norm": 0.7052403587790288,
+      "learning_rate": 6.983412530563287e-07,
+      "loss": 1.5526,
+      "step": 469
+    },
+    {
+      "epoch": 0.032749189980141445,
+      "grad_norm": 0.7742076645158189,
+      "learning_rate": 6.98334178215277e-07,
+      "loss": 1.516,
+      "step": 470
+    },
+    {
+      "epoch": 0.03281886910775877,
+      "grad_norm": 0.7818787556440743,
+      "learning_rate": 6.983270883586081e-07,
+      "loss": 1.6288,
+      "step": 471
+    },
+    {
+      "epoch": 0.03288854823537609,
+      "grad_norm": 0.686648879518541,
+      "learning_rate": 6.983199834866616e-07,
+      "loss": 1.5191,
+      "step": 472
+    },
+    {
+      "epoch": 0.03295822736299341,
+      "grad_norm": 0.75638490928715,
+      "learning_rate": 6.983128635997785e-07,
+      "loss": 1.6619,
+      "step": 473
+    },
+    {
+      "epoch": 0.033027906490610735,
+      "grad_norm": 0.6944982743137997,
+      "learning_rate": 6.983057286982998e-07,
+      "loss": 1.504,
+      "step": 474
+    },
+    {
+      "epoch": 0.03309758561822806,
+      "grad_norm": 0.7297694980053806,
+      "learning_rate": 6.982985787825673e-07,
+      "loss": 1.7033,
+      "step": 475
+    },
+    {
+      "epoch": 0.03316726474584538,
+      "grad_norm": 0.7814280477752864,
+      "learning_rate": 6.982914138529237e-07,
+      "loss": 1.6998,
+      "step": 476
+    },
+    {
+      "epoch": 0.0332369438734627,
+      "grad_norm": 0.7592691279743546,
+      "learning_rate": 6.982842339097124e-07,
+      "loss": 1.6168,
+      "step": 477
+    },
+    {
+      "epoch": 0.033306623001080025,
+      "grad_norm": 0.6950822644790459,
+      "learning_rate": 6.982770389532773e-07,
+      "loss": 1.6014,
+      "step": 478
+    },
+    {
+      "epoch": 0.03337630212869735,
+      "grad_norm": 0.8258391482075685,
+      "learning_rate": 6.982698289839635e-07,
+      "loss": 1.6586,
+      "step": 479
+    },
+    {
+      "epoch": 0.03344598125631467,
+      "grad_norm": 0.6971077588869111,
+      "learning_rate": 6.982626040021164e-07,
+      "loss": 1.5408,
+      "step": 480
+    },
+    {
+      "epoch": 0.03351566038393199,
+      "grad_norm": 0.7228631898409523,
+      "learning_rate": 6.982553640080821e-07,
+      "loss": 1.5811,
+      "step": 481
+    },
+    {
+      "epoch": 0.033585339511549316,
+      "grad_norm": 0.7282650250649347,
+      "learning_rate": 6.982481090022077e-07,
+      "loss": 1.6308,
+      "step": 482
+    },
+    {
+      "epoch": 0.03365501863916664,
+      "grad_norm": 0.7363507876647463,
+      "learning_rate": 6.982408389848408e-07,
+      "loss": 1.6058,
+      "step": 483
+    },
+    {
+      "epoch": 0.03372469776678396,
+      "grad_norm": 0.7304548735709389,
+      "learning_rate": 6.982335539563299e-07,
+      "loss": 1.5704,
+      "step": 484
+    },
+    {
+      "epoch": 0.03379437689440128,
+      "grad_norm": 0.7182208438930545,
+      "learning_rate": 6.98226253917024e-07,
+      "loss": 1.559,
+      "step": 485
+    },
+    {
+      "epoch": 0.033864056022018606,
+      "grad_norm": 0.7095183394575739,
+      "learning_rate": 6.982189388672729e-07,
+      "loss": 1.5234,
+      "step": 486
+    },
+    {
+      "epoch": 0.03393373514963593,
+      "grad_norm": 0.7100337849855409,
+      "learning_rate": 6.982116088074274e-07,
+      "loss": 1.4691,
+      "step": 487
+    },
+    {
+      "epoch": 0.03400341427725325,
+      "grad_norm": 0.7187486302555179,
+      "learning_rate": 6.982042637378384e-07,
+      "loss": 1.4962,
+      "step": 488
+    },
+    {
+      "epoch": 0.03407309340487057,
+      "grad_norm": 0.738433028371621,
+      "learning_rate": 6.981969036588582e-07,
+      "loss": 1.5137,
+      "step": 489
+    },
+    {
+      "epoch": 0.034142772532487896,
+      "grad_norm": 0.7494164590053654,
+      "learning_rate": 6.981895285708394e-07,
+      "loss": 1.6279,
+      "step": 490
+    },
+    {
+      "epoch": 0.03421245166010522,
+      "grad_norm": 0.6714113417696089,
+      "learning_rate": 6.981821384741353e-07,
+      "loss": 1.5603,
+      "step": 491
+    },
+    {
+      "epoch": 0.03428213078772254,
+      "grad_norm": 0.7347104034043285,
+      "learning_rate": 6.981747333691003e-07,
+      "loss": 1.6593,
+      "step": 492
+    },
+    {
+      "epoch": 0.034351809915339857,
+      "grad_norm": 0.7315383543287005,
+      "learning_rate": 6.981673132560891e-07,
+      "loss": 1.6402,
+      "step": 493
+    },
+    {
+      "epoch": 0.03442148904295718,
+      "grad_norm": 0.7971154217795332,
+      "learning_rate": 6.981598781354574e-07,
+      "loss": 1.8423,
+      "step": 494
+    },
+    {
+      "epoch": 0.0344911681705745,
+      "grad_norm": 0.7651575324005175,
+      "learning_rate": 6.981524280075613e-07,
+      "loss": 1.4978,
+      "step": 495
+    },
+    {
+      "epoch": 0.034560847298191824,
+      "grad_norm": 0.7325811848983502,
+      "learning_rate": 6.981449628727581e-07,
+      "loss": 1.532,
+      "step": 496
+    },
+    {
+      "epoch": 0.03463052642580915,
+      "grad_norm": 0.7606852959640895,
+      "learning_rate": 6.981374827314053e-07,
+      "loss": 1.6605,
+      "step": 497
+    },
+    {
+      "epoch": 0.03470020555342647,
+      "grad_norm": 0.7965507881830842,
+      "learning_rate": 6.981299875838615e-07,
+      "loss": 1.7101,
+      "step": 498
+    },
+    {
+      "epoch": 0.03476988468104379,
+      "grad_norm": 0.7685993735694805,
+      "learning_rate": 6.981224774304859e-07,
+      "loss": 1.5262,
+      "step": 499
+    },
+    {
+      "epoch": 0.034839563808661114,
+      "grad_norm": 5.064186032538183,
+      "learning_rate": 6.981149522716382e-07,
+      "loss": 1.4328,
+      "step": 500
+    },
+    {
+      "epoch": 0.03490924293627844,
+      "grad_norm": 0.6818074661763434,
+      "learning_rate": 6.981074121076793e-07,
+      "loss": 1.5321,
+      "step": 501
+    },
+    {
+      "epoch": 0.03497892206389576,
+      "grad_norm": 0.7140914098818415,
+      "learning_rate": 6.980998569389705e-07,
+      "loss": 1.6084,
+      "step": 502
+    },
+    {
+      "epoch": 0.03504860119151308,
+      "grad_norm": 0.7070692585114468,
+      "learning_rate": 6.980922867658736e-07,
+      "loss": 1.5168,
+      "step": 503
+    },
+    {
+      "epoch": 0.035118280319130404,
+      "grad_norm": 0.7710204892635041,
+      "learning_rate": 6.980847015887516e-07,
+      "loss": 1.5205,
+      "step": 504
+    },
+    {
+      "epoch": 0.03518795944674773,
+      "grad_norm": 0.6981371493463051,
+      "learning_rate": 6.98077101407968e-07,
+      "loss": 1.525,
+      "step": 505
+    },
+    {
+      "epoch": 0.03525763857436505,
+      "grad_norm": 0.7402087236647792,
+      "learning_rate": 6.980694862238869e-07,
+      "loss": 1.5848,
+      "step": 506
+    },
+    {
+      "epoch": 0.03532731770198237,
+      "grad_norm": 0.7287932998192505,
+      "learning_rate": 6.980618560368733e-07,
+      "loss": 1.6185,
+      "step": 507
+    },
+    {
+      "epoch": 0.035396996829599695,
+      "grad_norm": 0.7641151729360003,
+      "learning_rate": 6.980542108472929e-07,
+      "loss": 1.5882,
+      "step": 508
+    },
+    {
+      "epoch": 0.03546667595721702,
+      "grad_norm": 0.7181253669244724,
+      "learning_rate": 6.98046550655512e-07,
+      "loss": 1.6705,
+      "step": 509
+    },
+    {
+      "epoch": 0.03553635508483434,
+      "grad_norm": 0.710146864055551,
+      "learning_rate": 6.980388754618978e-07,
+      "loss": 1.6129,
+      "step": 510
+    },
+    {
+      "epoch": 0.03560603421245166,
+      "grad_norm": 0.7452987549392524,
+      "learning_rate": 6.98031185266818e-07,
+      "loss": 1.5337,
+      "step": 511
+    },
+    {
+      "epoch": 0.035675713340068985,
+      "grad_norm": 0.7765641316168032,
+      "learning_rate": 6.980234800706411e-07,
+      "loss": 1.6418,
+      "step": 512
+    },
+    {
+      "epoch": 0.03574539246768631,
+      "grad_norm": 0.7175298447436895,
+      "learning_rate": 6.980157598737365e-07,
+      "loss": 1.5969,
+      "step": 513
+    },
+    {
+      "epoch": 0.03581507159530363,
+      "grad_norm": 0.6839168177575335,
+      "learning_rate": 6.98008024676474e-07,
+      "loss": 1.5608,
+      "step": 514
+    },
+    {
+      "epoch": 0.03588475072292095,
+      "grad_norm": 0.7179064031162559,
+      "learning_rate": 6.980002744792244e-07,
+      "loss": 1.5678,
+      "step": 515
+    },
+    {
+      "epoch": 0.03595442985053827,
+      "grad_norm": 0.7319035209158925,
+      "learning_rate": 6.97992509282359e-07,
+      "loss": 1.5515,
+      "step": 516
+    },
+    {
+      "epoch": 0.03602410897815559,
+      "grad_norm": 0.7417659390519656,
+      "learning_rate": 6.9798472908625e-07,
+      "loss": 1.5878,
+      "step": 517
+    },
+    {
+      "epoch": 0.03609378810577291,
+      "grad_norm": 0.7355893605239711,
+      "learning_rate": 6.979769338912703e-07,
+      "loss": 1.5392,
+      "step": 518
+    },
+    {
+      "epoch": 0.036163467233390235,
+      "grad_norm": 0.7046601739665954,
+      "learning_rate": 6.979691236977935e-07,
+      "loss": 1.559,
+      "step": 519
+    },
+    {
+      "epoch": 0.03623314636100756,
+      "grad_norm": 0.7097878162497628,
+      "learning_rate": 6.979612985061936e-07,
+      "loss": 1.6167,
+      "step": 520
+    },
+    {
+      "epoch": 0.03630282548862488,
+      "grad_norm": 0.7173533144523362,
+      "learning_rate": 6.979534583168458e-07,
+      "loss": 1.5876,
+      "step": 521
+    },
+    {
+      "epoch": 0.0363725046162422,
+      "grad_norm": 0.6869507707629723,
+      "learning_rate": 6.979456031301258e-07,
+      "loss": 1.5362,
+      "step": 522
+    },
+    {
+      "epoch": 0.036442183743859526,
+      "grad_norm": 0.7110753403068261,
+      "learning_rate": 6.9793773294641e-07,
+      "loss": 1.524,
+      "step": 523
+    },
+    {
+      "epoch": 0.03651186287147685,
+      "grad_norm": 0.6991914768564813,
+      "learning_rate": 6.979298477660757e-07,
+      "loss": 1.5218,
+      "step": 524
+    },
+    {
+      "epoch": 0.03658154199909417,
+      "grad_norm": 0.7184863855815552,
+      "learning_rate": 6.979219475895006e-07,
+      "loss": 1.4594,
+      "step": 525
+    },
+    {
+      "epoch": 0.03665122112671149,
+      "grad_norm": 0.6675660110790665,
+      "learning_rate": 6.979140324170635e-07,
+      "loss": 1.5326,
+      "step": 526
+    },
+    {
+      "epoch": 0.036720900254328816,
+      "grad_norm": 0.7247217590142752,
+      "learning_rate": 6.979061022491434e-07,
+      "loss": 1.5758,
+      "step": 527
+    },
+    {
+      "epoch": 0.03679057938194614,
+      "grad_norm": 0.7340681651786749,
+      "learning_rate": 6.978981570861205e-07,
+      "loss": 1.58,
+      "step": 528
+    },
+    {
+      "epoch": 0.03686025850956346,
+      "grad_norm": 0.6962545953214849,
+      "learning_rate": 6.978901969283756e-07,
+      "loss": 1.6094,
+      "step": 529
+    },
+    {
+      "epoch": 0.03692993763718078,
+      "grad_norm": 0.7616295983929137,
+      "learning_rate": 6.978822217762901e-07,
+      "loss": 1.5415,
+      "step": 530
+    },
+    {
+      "epoch": 0.036999616764798106,
+      "grad_norm": 0.6858684867940431,
+      "learning_rate": 6.978742316302462e-07,
+      "loss": 1.6065,
+      "step": 531
+    },
+    {
+      "epoch": 0.03706929589241543,
+      "grad_norm": 0.7566143279490591,
+      "learning_rate": 6.978662264906268e-07,
+      "loss": 1.6859,
+      "step": 532
+    },
+    {
+      "epoch": 0.03713897502003275,
+      "grad_norm": 0.7450351080732804,
+      "learning_rate": 6.978582063578154e-07,
+      "loss": 1.6706,
+      "step": 533
+    },
+    {
+      "epoch": 0.03720865414765007,
+      "grad_norm": 0.8200694743403997,
+      "learning_rate": 6.978501712321967e-07,
+      "loss": 1.5625,
+      "step": 534
+    },
+    {
+      "epoch": 0.037278333275267396,
+      "grad_norm": 0.7468795076646643,
+      "learning_rate": 6.978421211141554e-07,
+      "loss": 1.6703,
+      "step": 535
+    },
+    {
+      "epoch": 0.03734801240288472,
+      "grad_norm": 0.7180155939758754,
+      "learning_rate": 6.978340560040774e-07,
+      "loss": 1.5986,
+      "step": 536
+    },
+    {
+      "epoch": 0.03741769153050204,
+      "grad_norm": 0.7147435818894757,
+      "learning_rate": 6.978259759023493e-07,
+      "loss": 1.5683,
+      "step": 537
+    },
+    {
+      "epoch": 0.037487370658119364,
+      "grad_norm": 0.779891568987929,
+      "learning_rate": 6.978178808093581e-07,
+      "loss": 1.6255,
+      "step": 538
+    },
+    {
+      "epoch": 0.03755704978573668,
+      "grad_norm": 0.7618768803406499,
+      "learning_rate": 6.978097707254919e-07,
+      "loss": 1.7284,
+      "step": 539
+    },
+    {
+      "epoch": 0.037626728913354,
+      "grad_norm": 0.7298137167232285,
+      "learning_rate": 6.978016456511393e-07,
+      "loss": 1.6203,
+      "step": 540
+    },
+    {
+      "epoch": 0.037696408040971324,
+      "grad_norm": 0.7562638291509813,
+      "learning_rate": 6.977935055866896e-07,
+      "loss": 1.6245,
+      "step": 541
+    },
+    {
+      "epoch": 0.03776608716858865,
+      "grad_norm": 0.6874011048759046,
+      "learning_rate": 6.977853505325329e-07,
+      "loss": 1.4707,
+      "step": 542
+    },
+    {
+      "epoch": 0.03783576629620597,
+      "grad_norm": 0.7091168800804247,
+      "learning_rate": 6.977771804890601e-07,
+      "loss": 1.535,
+      "step": 543
+    },
+    {
+      "epoch": 0.03790544542382329,
+      "grad_norm": 0.7817069685386216,
+      "learning_rate": 6.97768995456663e-07,
+      "loss": 1.5988,
+      "step": 544
+    },
+    {
+      "epoch": 0.037975124551440614,
+      "grad_norm": 0.7041314844983713,
+      "learning_rate": 6.977607954357331e-07,
+      "loss": 1.6012,
+      "step": 545
+    },
+    {
+      "epoch": 0.03804480367905794,
+      "grad_norm": 0.7442509072892823,
+      "learning_rate": 6.977525804266641e-07,
+      "loss": 1.5457,
+      "step": 546
+    },
+    {
+      "epoch": 0.03811448280667526,
+      "grad_norm": 0.6605248898044932,
+      "learning_rate": 6.977443504298493e-07,
+      "loss": 1.5191,
+      "step": 547
+    },
+    {
+      "epoch": 0.03818416193429258,
+      "grad_norm": 0.7001973481764062,
+      "learning_rate": 6.977361054456831e-07,
+      "loss": 1.609,
+      "step": 548
+    },
+    {
+      "epoch": 0.038253841061909905,
+      "grad_norm": 0.7060005032563952,
+      "learning_rate": 6.977278454745608e-07,
+      "loss": 1.5307,
+      "step": 549
+    },
+    {
+      "epoch": 0.03832352018952723,
+      "grad_norm": 0.6902502138912355,
+      "learning_rate": 6.97719570516878e-07,
+      "loss": 1.5973,
+      "step": 550
+    },
+    {
+      "epoch": 0.03839319931714455,
+      "grad_norm": 0.7691013493985639,
+      "learning_rate": 6.977112805730315e-07,
+      "loss": 1.6391,
+      "step": 551
+    },
+    {
+      "epoch": 0.03846287844476187,
+      "grad_norm": 0.7352744642083536,
+      "learning_rate": 6.977029756434184e-07,
+      "loss": 1.6102,
+      "step": 552
+    },
+    {
+      "epoch": 0.038532557572379195,
+      "grad_norm": 0.7297849741047285,
+      "learning_rate": 6.976946557284367e-07,
+      "loss": 1.5189,
+      "step": 553
+    },
+    {
+      "epoch": 0.03860223669999652,
+      "grad_norm": 0.8050387561532055,
+      "learning_rate": 6.976863208284852e-07,
+      "loss": 1.6532,
+      "step": 554
+    },
+    {
+      "epoch": 0.03867191582761384,
+      "grad_norm": 0.7513047702526604,
+      "learning_rate": 6.976779709439633e-07,
+      "loss": 1.655,
+      "step": 555
+    },
+    {
+      "epoch": 0.03874159495523116,
+      "grad_norm": 0.7109447925987789,
+      "learning_rate": 6.97669606075271e-07,
+      "loss": 1.6917,
+      "step": 556
+    },
+    {
+      "epoch": 0.038811274082848485,
+      "grad_norm": 0.7066517914984015,
+      "learning_rate": 6.976612262228094e-07,
+      "loss": 1.548,
+      "step": 557
+    },
+    {
+      "epoch": 0.03888095321046581,
+      "grad_norm": 0.7926936156146592,
+      "learning_rate": 6.976528313869799e-07,
+      "loss": 1.6356,
+      "step": 558
+    },
+    {
+      "epoch": 0.03895063233808313,
+      "grad_norm": 0.7617651263509975,
+      "learning_rate": 6.976444215681848e-07,
+      "loss": 1.4892,
+      "step": 559
+    },
+    {
+      "epoch": 0.03902031146570045,
+      "grad_norm": 0.7338500441938097,
+      "learning_rate": 6.976359967668273e-07,
+      "loss": 1.5784,
+      "step": 560
+    },
+    {
+      "epoch": 0.039089990593317775,
+      "grad_norm": 0.7305585031017265,
+      "learning_rate": 6.97627556983311e-07,
+      "loss": 1.745,
+      "step": 561
+    },
+    {
+      "epoch": 0.03915966972093509,
+      "grad_norm": 0.712314154316217,
+      "learning_rate": 6.976191022180402e-07,
+      "loss": 1.5946,
+      "step": 562
+    },
+    {
+      "epoch": 0.03922934884855241,
+      "grad_norm": 0.6785678714420585,
+      "learning_rate": 6.976106324714204e-07,
+      "loss": 1.5513,
+      "step": 563
+    },
+    {
+      "epoch": 0.039299027976169736,
+      "grad_norm": 0.7179837146316714,
+      "learning_rate": 6.976021477438572e-07,
+      "loss": 1.5586,
+      "step": 564
+    },
+    {
+      "epoch": 0.03936870710378706,
+      "grad_norm": 0.6880995082747411,
+      "learning_rate": 6.975936480357574e-07,
+      "loss": 1.4726,
+      "step": 565
+    },
+    {
+      "epoch": 0.03943838623140438,
+      "grad_norm": 0.7490245409219959,
+      "learning_rate": 6.975851333475283e-07,
+      "loss": 1.5735,
+      "step": 566
+    },
+    {
+      "epoch": 0.0395080653590217,
+      "grad_norm": 0.7718344714659485,
+      "learning_rate": 6.975766036795778e-07,
+      "loss": 1.6744,
+      "step": 567
+    },
+    {
+      "epoch": 0.039577744486639026,
+      "grad_norm": 0.6999493242955485,
+      "learning_rate": 6.975680590323147e-07,
+      "loss": 1.5618,
+      "step": 568
+    },
+    {
+      "epoch": 0.03964742361425635,
+      "grad_norm": 0.7277339304673994,
+      "learning_rate": 6.975594994061485e-07,
+      "loss": 1.6268,
+      "step": 569
+    },
+    {
+      "epoch": 0.03971710274187367,
+      "grad_norm": 0.7224063955146929,
+      "learning_rate": 6.975509248014895e-07,
+      "loss": 1.6151,
+      "step": 570
+    },
+    {
+      "epoch": 0.03978678186949099,
+      "grad_norm": 0.7754705593906143,
+      "learning_rate": 6.975423352187485e-07,
+      "loss": 1.5736,
+      "step": 571
+    },
+    {
+      "epoch": 0.039856460997108316,
+      "grad_norm": 0.658621938146351,
+      "learning_rate": 6.975337306583371e-07,
+      "loss": 1.4863,
+      "step": 572
+    },
+    {
+      "epoch": 0.03992614012472564,
+      "grad_norm": 0.730264824789615,
+      "learning_rate": 6.975251111206678e-07,
+      "loss": 1.5592,
+      "step": 573
+    },
+    {
+      "epoch": 0.03999581925234296,
+      "grad_norm": 0.6657629477477796,
+      "learning_rate": 6.975164766061535e-07,
+      "loss": 1.3176,
+      "step": 574
+    },
+    {
+      "epoch": 0.04006549837996028,
+      "grad_norm": 0.7795968362382858,
+      "learning_rate": 6.975078271152082e-07,
+      "loss": 1.654,
+      "step": 575
+    },
+    {
+      "epoch": 0.040135177507577606,
+      "grad_norm": 0.7411465230498967,
+      "learning_rate": 6.974991626482462e-07,
+      "loss": 1.6269,
+      "step": 576
+    },
+    {
+      "epoch": 0.04020485663519493,
+      "grad_norm": 0.7291408994652216,
+      "learning_rate": 6.974904832056828e-07,
+      "loss": 1.4967,
+      "step": 577
+    },
+    {
+      "epoch": 0.04027453576281225,
+      "grad_norm": 0.6991182395286692,
+      "learning_rate": 6.974817887879338e-07,
+      "loss": 1.4528,
+      "step": 578
+    },
+    {
+      "epoch": 0.040344214890429574,
+      "grad_norm": 0.7457004375077715,
+      "learning_rate": 6.974730793954159e-07,
+      "loss": 1.586,
+      "step": 579
+    },
+    {
+      "epoch": 0.040413894018046896,
+      "grad_norm": 0.7067310028445386,
+      "learning_rate": 6.974643550285467e-07,
+      "loss": 1.638,
+      "step": 580
+    },
+    {
+      "epoch": 0.04048357314566422,
+      "grad_norm": 0.6841122043735114,
+      "learning_rate": 6.974556156877441e-07,
+      "loss": 1.5638,
+      "step": 581
+    },
+    {
+      "epoch": 0.04055325227328154,
+      "grad_norm": 0.7077425525133038,
+      "learning_rate": 6.974468613734269e-07,
+      "loss": 1.4724,
+      "step": 582
+    },
+    {
+      "epoch": 0.040622931400898864,
+      "grad_norm": 0.7009427894499272,
+      "learning_rate": 6.974380920860147e-07,
+      "loss": 1.6103,
+      "step": 583
+    },
+    {
+      "epoch": 0.040692610528516186,
+      "grad_norm": 0.7608036022452415,
+      "learning_rate": 6.974293078259277e-07,
+      "loss": 1.6457,
+      "step": 584
+    },
+    {
+      "epoch": 0.0407622896561335,
+      "grad_norm": 0.704166945656348,
+      "learning_rate": 6.974205085935869e-07,
+      "loss": 1.5455,
+      "step": 585
+    },
+    {
+      "epoch": 0.040831968783750824,
+      "grad_norm": 0.6946949545881428,
+      "learning_rate": 6.974116943894139e-07,
+      "loss": 1.666,
+      "step": 586
+    },
+    {
+      "epoch": 0.04090164791136815,
+      "grad_norm": 0.716100071019257,
+      "learning_rate": 6.974028652138311e-07,
+      "loss": 1.5197,
+      "step": 587
+    },
+    {
+      "epoch": 0.04097132703898547,
+      "grad_norm": 0.6838107620585757,
+      "learning_rate": 6.973940210672617e-07,
+      "loss": 1.4886,
+      "step": 588
+    },
+    {
+      "epoch": 0.04104100616660279,
+      "grad_norm": 0.6697472310234589,
+      "learning_rate": 6.973851619501295e-07,
+      "loss": 1.5473,
+      "step": 589
+    },
+    {
+      "epoch": 0.041110685294220115,
+      "grad_norm": 0.774401037898998,
+      "learning_rate": 6.973762878628589e-07,
+      "loss": 1.5531,
+      "step": 590
+    },
+    {
+      "epoch": 0.04118036442183744,
+      "grad_norm": 0.6962775521829606,
+      "learning_rate": 6.973673988058754e-07,
+      "loss": 1.621,
+      "step": 591
+    },
+    {
+      "epoch": 0.04125004354945476,
+      "grad_norm": 0.7465537441237275,
+      "learning_rate": 6.973584947796049e-07,
+      "loss": 1.3792,
+      "step": 592
+    },
+    {
+      "epoch": 0.04131972267707208,
+      "grad_norm": 0.7728818872843954,
+      "learning_rate": 6.973495757844739e-07,
+      "loss": 1.4883,
+      "step": 593
+    },
+    {
+      "epoch": 0.041389401804689405,
+      "grad_norm": 0.733610207501176,
+      "learning_rate": 6.973406418209102e-07,
+      "loss": 1.647,
+      "step": 594
+    },
+    {
+      "epoch": 0.04145908093230673,
+      "grad_norm": 0.8035948982021359,
+      "learning_rate": 6.973316928893416e-07,
+      "loss": 1.5647,
+      "step": 595
+    },
+    {
+      "epoch": 0.04152876005992405,
+      "grad_norm": 0.6768018321794376,
+      "learning_rate": 6.97322728990197e-07,
+      "loss": 1.4769,
+      "step": 596
+    },
+    {
+      "epoch": 0.04159843918754137,
+      "grad_norm": 0.7482483315404442,
+      "learning_rate": 6.973137501239061e-07,
+      "loss": 1.6266,
+      "step": 597
+    },
+    {
+      "epoch": 0.041668118315158695,
+      "grad_norm": 0.711191857418,
+      "learning_rate": 6.973047562908992e-07,
+      "loss": 1.4338,
+      "step": 598
+    },
+    {
+      "epoch": 0.04173779744277602,
+      "grad_norm": 0.7248513304507356,
+      "learning_rate": 6.972957474916072e-07,
+      "loss": 1.5761,
+      "step": 599
+    },
+    {
+      "epoch": 0.04180747657039334,
+      "grad_norm": 0.7389307964740751,
+      "learning_rate": 6.972867237264619e-07,
+      "loss": 1.5608,
+      "step": 600
+    },
+    {
+      "epoch": 0.04187715569801066,
+      "grad_norm": 0.7421731119566499,
+      "learning_rate": 6.972776849958957e-07,
+      "loss": 1.6273,
+      "step": 601
+    },
+    {
+      "epoch": 0.041946834825627985,
+      "grad_norm": 0.7836075175702004,
+      "learning_rate": 6.972686313003416e-07,
+      "loss": 1.594,
+      "step": 602
+    },
+    {
+      "epoch": 0.04201651395324531,
+      "grad_norm": 0.7090879903195844,
+      "learning_rate": 6.972595626402337e-07,
+      "loss": 1.4985,
+      "step": 603
+    },
+    {
+      "epoch": 0.04208619308086263,
+      "grad_norm": 0.6658000549796161,
+      "learning_rate": 6.972504790160064e-07,
+      "loss": 1.518,
+      "step": 604
+    },
+    {
+      "epoch": 0.04215587220847995,
+      "grad_norm": 0.7349695726566142,
+      "learning_rate": 6.972413804280953e-07,
+      "loss": 1.5596,
+      "step": 605
+    },
+    {
+      "epoch": 0.042225551336097275,
+      "grad_norm": 0.7483074717752025,
+      "learning_rate": 6.972322668769361e-07,
+      "loss": 1.5794,
+      "step": 606
+    },
+    {
+      "epoch": 0.0422952304637146,
+      "grad_norm": 0.7038158186966939,
+      "learning_rate": 6.972231383629657e-07,
+      "loss": 1.5117,
+      "step": 607
+    },
+    {
+      "epoch": 0.04236490959133191,
+      "grad_norm": 0.6751541816081398,
+      "learning_rate": 6.972139948866215e-07,
+      "loss": 1.5732,
+      "step": 608
+    },
+    {
+      "epoch": 0.042434588718949236,
+      "grad_norm": 0.747301063418828,
+      "learning_rate": 6.972048364483418e-07,
+      "loss": 1.5431,
+      "step": 609
+    },
+    {
+      "epoch": 0.04250426784656656,
+      "grad_norm": 0.7420324379102797,
+      "learning_rate": 6.971956630485652e-07,
+      "loss": 1.6458,
+      "step": 610
+    },
+    {
+      "epoch": 0.04257394697418388,
+      "grad_norm": 0.6698738775177508,
+      "learning_rate": 6.971864746877316e-07,
+      "loss": 1.5838,
+      "step": 611
+    },
+    {
+      "epoch": 0.0426436261018012,
+      "grad_norm": 0.7080826823481463,
+      "learning_rate": 6.971772713662812e-07,
+      "loss": 1.5728,
+      "step": 612
+    },
+    {
+      "epoch": 0.042713305229418526,
+      "grad_norm": 0.7130688284022854,
+      "learning_rate": 6.971680530846551e-07,
+      "loss": 1.6254,
+      "step": 613
+    },
+    {
+      "epoch": 0.04278298435703585,
+      "grad_norm": 0.6664129156822504,
+      "learning_rate": 6.971588198432952e-07,
+      "loss": 1.4956,
+      "step": 614
+    },
+    {
+      "epoch": 0.04285266348465317,
+      "grad_norm": 0.7132385041089505,
+      "learning_rate": 6.971495716426435e-07,
+      "loss": 1.5515,
+      "step": 615
+    },
+    {
+      "epoch": 0.04292234261227049,
+      "grad_norm": 0.73710904278707,
+      "learning_rate": 6.971403084831436e-07,
+      "loss": 1.6063,
+      "step": 616
+    },
+    {
+      "epoch": 0.042992021739887816,
+      "grad_norm": 0.8132005846012044,
+      "learning_rate": 6.971310303652395e-07,
+      "loss": 1.6231,
+      "step": 617
+    },
+    {
+      "epoch": 0.04306170086750514,
+      "grad_norm": 0.6973241807952938,
+      "learning_rate": 6.971217372893753e-07,
+      "loss": 1.6154,
+      "step": 618
+    },
+    {
+      "epoch": 0.04313137999512246,
+      "grad_norm": 2.8190614475050513,
+      "learning_rate": 6.971124292559969e-07,
+      "loss": 1.6138,
+      "step": 619
+    },
+    {
+      "epoch": 0.043201059122739784,
+      "grad_norm": 0.7236203087033882,
+      "learning_rate": 6.971031062655502e-07,
+      "loss": 1.46,
+      "step": 620
+    },
+    {
+      "epoch": 0.043270738250357106,
+      "grad_norm": 0.7721528246703844,
+      "learning_rate": 6.970937683184816e-07,
+      "loss": 1.651,
+      "step": 621
+    },
+    {
+      "epoch": 0.04334041737797443,
+      "grad_norm": 0.7207721016116038,
+      "learning_rate": 6.970844154152392e-07,
+      "loss": 1.6102,
+      "step": 622
+    },
+    {
+      "epoch": 0.04341009650559175,
+      "grad_norm": 0.7470132174851035,
+      "learning_rate": 6.970750475562709e-07,
+      "loss": 1.5735,
+      "step": 623
+    },
+    {
+      "epoch": 0.043479775633209074,
+      "grad_norm": 0.682052163775226,
+      "learning_rate": 6.970656647420255e-07,
+      "loss": 1.5302,
+      "step": 624
+    },
+    {
+      "epoch": 0.043549454760826396,
+      "grad_norm": 0.7618764660275008,
+      "learning_rate": 6.970562669729528e-07,
+      "loss": 1.6138,
+      "step": 625
+    },
+    {
+      "epoch": 0.04361913388844372,
+      "grad_norm": 0.6772064059244419,
+      "learning_rate": 6.970468542495033e-07,
+      "loss": 1.4087,
+      "step": 626
+    },
+    {
+      "epoch": 0.04368881301606104,
+      "grad_norm": 0.7376677084023966,
+      "learning_rate": 6.970374265721277e-07,
+      "loss": 1.5959,
+      "step": 627
+    },
+    {
+      "epoch": 0.043758492143678364,
+      "grad_norm": 0.8303204432724274,
+      "learning_rate": 6.970279839412782e-07,
+      "loss": 1.6502,
+      "step": 628
+    },
+    {
+      "epoch": 0.043828171271295686,
+      "grad_norm": 0.7786738881941985,
+      "learning_rate": 6.970185263574071e-07,
+      "loss": 1.6448,
+      "step": 629
+    },
+    {
+      "epoch": 0.04389785039891301,
+      "grad_norm": 0.7128585750577182,
+      "learning_rate": 6.970090538209676e-07,
+      "loss": 1.44,
+      "step": 630
+    },
+    {
+      "epoch": 0.043967529526530325,
+      "grad_norm": 0.6627090163476095,
+      "learning_rate": 6.969995663324138e-07,
+      "loss": 1.3244,
+      "step": 631
+    },
+    {
+      "epoch": 0.04403720865414765,
+      "grad_norm": 0.7745287325391855,
+      "learning_rate": 6.969900638922e-07,
+      "loss": 1.7033,
+      "step": 632
+    },
+    {
+      "epoch": 0.04410688778176497,
+      "grad_norm": 0.733220507077979,
+      "learning_rate": 6.969805465007822e-07,
+      "loss": 1.6772,
+      "step": 633
+    },
+    {
+      "epoch": 0.04417656690938229,
+      "grad_norm": 0.6960784033598787,
+      "learning_rate": 6.969710141586159e-07,
+      "loss": 1.5492,
+      "step": 634
+    },
+    {
+      "epoch": 0.044246246036999615,
+      "grad_norm": 0.6881665503401039,
+      "learning_rate": 6.96961466866158e-07,
+      "loss": 1.5384,
+      "step": 635
+    },
+    {
+      "epoch": 0.04431592516461694,
+      "grad_norm": 0.7093423611533153,
+      "learning_rate": 6.969519046238665e-07,
+      "loss": 1.5778,
+      "step": 636
+    },
+    {
+      "epoch": 0.04438560429223426,
+      "grad_norm": 0.6950788938005011,
+      "learning_rate": 6.969423274321992e-07,
+      "loss": 1.4672,
+      "step": 637
+    },
+    {
+      "epoch": 0.04445528341985158,
+      "grad_norm": 0.6971933321048744,
+      "learning_rate": 6.969327352916151e-07,
+      "loss": 1.5282,
+      "step": 638
+    },
+    {
+      "epoch": 0.044524962547468905,
+      "grad_norm": 0.6700704096045089,
+      "learning_rate": 6.96923128202574e-07,
+      "loss": 1.4895,
+      "step": 639
+    },
+    {
+      "epoch": 0.04459464167508623,
+      "grad_norm": 0.6942860808996856,
+      "learning_rate": 6.969135061655361e-07,
+      "loss": 1.4508,
+      "step": 640
+    },
+    {
+      "epoch": 0.04466432080270355,
+      "grad_norm": 0.7804476942861223,
+      "learning_rate": 6.969038691809628e-07,
+      "loss": 1.509,
+      "step": 641
+    },
+    {
+      "epoch": 0.04473399993032087,
+      "grad_norm": 0.6853905689405291,
+      "learning_rate": 6.968942172493156e-07,
+      "loss": 1.4684,
+      "step": 642
+    },
+    {
+      "epoch": 0.044803679057938195,
+      "grad_norm": 0.6764348433542908,
+      "learning_rate": 6.968845503710572e-07,
+      "loss": 1.5356,
+      "step": 643
+    },
+    {
+      "epoch": 0.04487335818555552,
+      "grad_norm": 0.6773570660436679,
+      "learning_rate": 6.96874868546651e-07,
+      "loss": 1.5618,
+      "step": 644
+    },
+    {
+      "epoch": 0.04494303731317284,
+      "grad_norm": 0.7622466312475294,
+      "learning_rate": 6.968651717765608e-07,
+      "loss": 1.5596,
+      "step": 645
+    },
+    {
+      "epoch": 0.04501271644079016,
+      "grad_norm": 0.7254213570087428,
+      "learning_rate": 6.968554600612512e-07,
+      "loss": 1.5213,
+      "step": 646
+    },
+    {
+      "epoch": 0.045082395568407485,
+      "grad_norm": 0.6885774067075373,
+      "learning_rate": 6.968457334011879e-07,
+      "loss": 1.5261,
+      "step": 647
+    },
+    {
+      "epoch": 0.04515207469602481,
+      "grad_norm": 0.7101552502470694,
+      "learning_rate": 6.968359917968368e-07,
+      "loss": 1.6694,
+      "step": 648
+    },
+    {
+      "epoch": 0.04522175382364213,
+      "grad_norm": 0.7369871559845563,
+      "learning_rate": 6.968262352486649e-07,
+      "loss": 1.5304,
+      "step": 649
+    },
+    {
+      "epoch": 0.04529143295125945,
+      "grad_norm": 0.7875133371881571,
+      "learning_rate": 6.968164637571393e-07,
+      "loss": 1.4934,
+      "step": 650
+    },
+    {
+      "epoch": 0.045361112078876775,
+      "grad_norm": 0.7019279683983147,
+      "learning_rate": 6.968066773227289e-07,
+      "loss": 1.5468,
+      "step": 651
+    },
+    {
+      "epoch": 0.0454307912064941,
+      "grad_norm": 0.7341904116170085,
+      "learning_rate": 6.967968759459023e-07,
+      "loss": 1.5338,
+      "step": 652
+    },
+    {
+      "epoch": 0.04550047033411142,
+      "grad_norm": 0.6810202405701805,
+      "learning_rate": 6.967870596271292e-07,
+      "loss": 1.6357,
+      "step": 653
+    },
+    {
+      "epoch": 0.045570149461728736,
+      "grad_norm": 0.699448200025997,
+      "learning_rate": 6.967772283668803e-07,
+      "loss": 1.5453,
+      "step": 654
+    },
+    {
+      "epoch": 0.04563982858934606,
+      "grad_norm": 0.7158382501901476,
+      "learning_rate": 6.967673821656265e-07,
+      "loss": 1.5505,
+      "step": 655
+    },
+    {
+      "epoch": 0.04570950771696338,
+      "grad_norm": 0.6535608334622804,
+      "learning_rate": 6.967575210238395e-07,
+      "loss": 1.5453,
+      "step": 656
+    },
+    {
+      "epoch": 0.045779186844580703,
+      "grad_norm": 0.7080634763231599,
+      "learning_rate": 6.967476449419924e-07,
+      "loss": 1.5479,
+      "step": 657
+    },
+    {
+      "epoch": 0.045848865972198026,
+      "grad_norm": 0.7659320955527936,
+      "learning_rate": 6.967377539205579e-07,
+      "loss": 1.6749,
+      "step": 658
+    },
+    {
+      "epoch": 0.04591854509981535,
+      "grad_norm": 0.6686160543798587,
+      "learning_rate": 6.967278479600104e-07,
+      "loss": 1.4683,
+      "step": 659
+    },
+    {
+      "epoch": 0.04598822422743267,
+      "grad_norm": 0.7537119319439501,
+      "learning_rate": 6.967179270608243e-07,
+      "loss": 1.45,
+      "step": 660
+    },
+    {
+      "epoch": 0.046057903355049994,
+      "grad_norm": 0.7349118447650984,
+      "learning_rate": 6.967079912234754e-07,
+      "loss": 1.595,
+      "step": 661
+    },
+    {
+      "epoch": 0.046127582482667316,
+      "grad_norm": 0.7293110394952392,
+      "learning_rate": 6.966980404484395e-07,
+      "loss": 1.4752,
+      "step": 662
+    },
+    {
+      "epoch": 0.04619726161028464,
+      "grad_norm": 0.7387902011995563,
+      "learning_rate": 6.966880747361936e-07,
+      "loss": 1.4697,
+      "step": 663
+    },
+    {
+      "epoch": 0.04626694073790196,
+      "grad_norm": 0.6765249539707564,
+      "learning_rate": 6.966780940872153e-07,
+      "loss": 1.4381,
+      "step": 664
+    },
+    {
+      "epoch": 0.046336619865519284,
+      "grad_norm": 0.689217791498506,
+      "learning_rate": 6.966680985019828e-07,
+      "loss": 1.4685,
+      "step": 665
+    },
+    {
+      "epoch": 0.046406298993136606,
+      "grad_norm": 0.6841331183811533,
+      "learning_rate": 6.966580879809752e-07,
+      "loss": 1.4852,
+      "step": 666
+    },
+    {
+      "epoch": 0.04647597812075393,
+      "grad_norm": 0.743925290927153,
+      "learning_rate": 6.966480625246722e-07,
+      "loss": 1.5746,
+      "step": 667
+    },
+    {
+      "epoch": 0.04654565724837125,
+      "grad_norm": 0.6890299457883376,
+      "learning_rate": 6.966380221335544e-07,
+      "loss": 1.6132,
+      "step": 668
+    },
+    {
+      "epoch": 0.046615336375988574,
+      "grad_norm": 0.7300586622886946,
+      "learning_rate": 6.966279668081026e-07,
+      "loss": 1.5986,
+      "step": 669
+    },
+    {
+      "epoch": 0.046685015503605896,
+      "grad_norm": 0.7367747555470463,
+      "learning_rate": 6.966178965487989e-07,
+      "loss": 1.6231,
+      "step": 670
+    },
+    {
+      "epoch": 0.04675469463122322,
+      "grad_norm": 0.6836238330355955,
+      "learning_rate": 6.96607811356126e-07,
+      "loss": 1.576,
+      "step": 671
+    },
+    {
+      "epoch": 0.04682437375884054,
+      "grad_norm": 0.7033258443880477,
+      "learning_rate": 6.96597711230567e-07,
+      "loss": 1.6208,
+      "step": 672
+    },
+    {
+      "epoch": 0.046894052886457864,
+      "grad_norm": 0.8339698614381776,
+      "learning_rate": 6.96587596172606e-07,
+      "loss": 1.5455,
+      "step": 673
+    },
+    {
+      "epoch": 0.04696373201407519,
+      "grad_norm": 0.7085764029952359,
+      "learning_rate": 6.965774661827277e-07,
+      "loss": 1.4264,
+      "step": 674
+    },
+    {
+      "epoch": 0.04703341114169251,
+      "grad_norm": 0.7646037417894928,
+      "learning_rate": 6.965673212614174e-07,
+      "loss": 1.4358,
+      "step": 675
+    },
+    {
+      "epoch": 0.04710309026930983,
+      "grad_norm": 0.7532185100268795,
+      "learning_rate": 6.965571614091616e-07,
+      "loss": 1.6341,
+      "step": 676
+    },
+    {
+      "epoch": 0.04717276939692715,
+      "grad_norm": 0.6685706703379944,
+      "learning_rate": 6.96546986626447e-07,
+      "loss": 1.546,
+      "step": 677
+    },
+    {
+      "epoch": 0.04724244852454447,
+      "grad_norm": 0.680236955536486,
+      "learning_rate": 6.965367969137611e-07,
+      "loss": 1.4986,
+      "step": 678
+    },
+    {
+      "epoch": 0.04731212765216179,
+      "grad_norm": 0.6986156128855948,
+      "learning_rate": 6.965265922715925e-07,
+      "loss": 1.5194,
+      "step": 679
+    },
+    {
+      "epoch": 0.047381806779779115,
+      "grad_norm": 0.7329077948281687,
+      "learning_rate": 6.9651637270043e-07,
+      "loss": 1.5835,
+      "step": 680
+    },
+    {
+      "epoch": 0.04745148590739644,
+      "grad_norm": 0.67737004024086,
+      "learning_rate": 6.965061382007632e-07,
+      "loss": 1.5043,
+      "step": 681
+    },
+    {
+      "epoch": 0.04752116503501376,
+      "grad_norm": 0.7124756190212052,
+      "learning_rate": 6.964958887730829e-07,
+      "loss": 1.5287,
+      "step": 682
+    },
+    {
+      "epoch": 0.04759084416263108,
+      "grad_norm": 0.7089711249978385,
+      "learning_rate": 6.964856244178801e-07,
+      "loss": 1.5851,
+      "step": 683
+    },
+    {
+      "epoch": 0.047660523290248405,
+      "grad_norm": 0.6927438151246575,
+      "learning_rate": 6.964753451356467e-07,
+      "loss": 1.5052,
+      "step": 684
+    },
+    {
+      "epoch": 0.04773020241786573,
+      "grad_norm": 0.7749472284337433,
+      "learning_rate": 6.964650509268753e-07,
+      "loss": 1.6473,
+      "step": 685
+    },
+    {
+      "epoch": 0.04779988154548305,
+      "grad_norm": 0.6963609256556942,
+      "learning_rate": 6.964547417920593e-07,
+      "loss": 1.6013,
+      "step": 686
+    },
+    {
+      "epoch": 0.04786956067310037,
+      "grad_norm": 0.687383111340697,
+      "learning_rate": 6.964444177316926e-07,
+      "loss": 1.5295,
+      "step": 687
+    },
+    {
+      "epoch": 0.047939239800717695,
+      "grad_norm": 0.7180160100897451,
+      "learning_rate": 6.964340787462701e-07,
+      "loss": 1.4711,
+      "step": 688
+    },
+    {
+      "epoch": 0.04800891892833502,
+      "grad_norm": 0.7590607641255273,
+      "learning_rate": 6.964237248362871e-07,
+      "loss": 1.5769,
+      "step": 689
+    },
+    {
+      "epoch": 0.04807859805595234,
+      "grad_norm": 0.8004088368747236,
+      "learning_rate": 6.9641335600224e-07,
+      "loss": 1.5545,
+      "step": 690
+    },
+    {
+      "epoch": 0.04814827718356966,
+      "grad_norm": 0.7273413263118548,
+      "learning_rate": 6.964029722446253e-07,
+      "loss": 1.6128,
+      "step": 691
+    },
+    {
+      "epoch": 0.048217956311186985,
+      "grad_norm": 0.7066776812890259,
+      "learning_rate": 6.963925735639411e-07,
+      "loss": 1.5317,
+      "step": 692
+    },
+    {
+      "epoch": 0.04828763543880431,
+      "grad_norm": 0.7015857397206559,
+      "learning_rate": 6.963821599606854e-07,
+      "loss": 1.5972,
+      "step": 693
+    },
+    {
+      "epoch": 0.04835731456642163,
+      "grad_norm": 0.7232812425018523,
+      "learning_rate": 6.963717314353574e-07,
+      "loss": 1.5294,
+      "step": 694
+    },
+    {
+      "epoch": 0.04842699369403895,
+      "grad_norm": 0.7243849048748213,
+      "learning_rate": 6.963612879884567e-07,
+      "loss": 1.556,
+      "step": 695
+    },
+    {
+      "epoch": 0.048496672821656275,
+      "grad_norm": 0.7359746952905802,
+      "learning_rate": 6.96350829620484e-07,
+      "loss": 1.546,
+      "step": 696
+    },
+    {
+      "epoch": 0.0485663519492736,
+      "grad_norm": 0.7122154316120477,
+      "learning_rate": 6.963403563319402e-07,
+      "loss": 1.733,
+      "step": 697
+    },
+    {
+      "epoch": 0.04863603107689092,
+      "grad_norm": 0.7055779006893984,
+      "learning_rate": 6.963298681233274e-07,
+      "loss": 1.6171,
+      "step": 698
+    },
+    {
+      "epoch": 0.04870571020450824,
+      "grad_norm": 0.716296197133196,
+      "learning_rate": 6.963193649951483e-07,
+      "loss": 1.752,
+      "step": 699
+    },
+    {
+      "epoch": 0.04877538933212556,
+      "grad_norm": 0.6713758234296247,
+      "learning_rate": 6.96308846947906e-07,
+      "loss": 1.4199,
+      "step": 700
+    },
+    {
+      "epoch": 0.04884506845974288,
+      "grad_norm": 0.7743784620060852,
+      "learning_rate": 6.962983139821047e-07,
+      "loss": 1.4771,
+      "step": 701
+    },
+    {
+      "epoch": 0.048914747587360204,
+      "grad_norm": 0.6641119216158364,
+      "learning_rate": 6.96287766098249e-07,
+      "loss": 1.5956,
+      "step": 702
+    },
+    {
+      "epoch": 0.048984426714977526,
+      "grad_norm": 0.6558222609838982,
+      "learning_rate": 6.962772032968446e-07,
+      "loss": 1.4149,
+      "step": 703
+    },
+    {
+      "epoch": 0.04905410584259485,
+      "grad_norm": 0.7450866187835455,
+      "learning_rate": 6.962666255783975e-07,
+      "loss": 1.5214,
+      "step": 704
+    },
+    {
+      "epoch": 0.04912378497021217,
+      "grad_norm": 0.7446602054101235,
+      "learning_rate": 6.962560329434148e-07,
+      "loss": 1.5909,
+      "step": 705
+    },
+    {
+      "epoch": 0.049193464097829494,
+      "grad_norm": 0.6912223067403324,
+      "learning_rate": 6.962454253924038e-07,
+      "loss": 1.5803,
+      "step": 706
+    },
+    {
+      "epoch": 0.049263143225446816,
+      "grad_norm": 0.7497854047755567,
+      "learning_rate": 6.962348029258732e-07,
+      "loss": 1.6816,
+      "step": 707
+    },
+    {
+      "epoch": 0.04933282235306414,
+      "grad_norm": 0.6944629046945672,
+      "learning_rate": 6.96224165544332e-07,
+      "loss": 1.4664,
+      "step": 708
+    },
+    {
+      "epoch": 0.04940250148068146,
+      "grad_norm": 0.6968358758121764,
+      "learning_rate": 6.962135132482896e-07,
+      "loss": 1.5491,
+      "step": 709
+    },
+    {
+      "epoch": 0.049472180608298784,
+      "grad_norm": 0.6771949433189489,
+      "learning_rate": 6.962028460382568e-07,
+      "loss": 1.4434,
+      "step": 710
+    },
+    {
+      "epoch": 0.049541859735916106,
+      "grad_norm": 0.7158393842253569,
+      "learning_rate": 6.961921639147448e-07,
+      "loss": 1.6615,
+      "step": 711
+    },
+    {
+      "epoch": 0.04961153886353343,
+      "grad_norm": 0.6926218608736993,
+      "learning_rate": 6.961814668782655e-07,
+      "loss": 1.551,
+      "step": 712
+    },
+    {
+      "epoch": 0.04968121799115075,
+      "grad_norm": 0.7354027473157468,
+      "learning_rate": 6.961707549293313e-07,
+      "loss": 1.5184,
+      "step": 713
+    },
+    {
+      "epoch": 0.049750897118768074,
+      "grad_norm": 0.6540498769668062,
+      "learning_rate": 6.961600280684558e-07,
+      "loss": 1.4642,
+      "step": 714
+    },
+    {
+      "epoch": 0.0498205762463854,
+      "grad_norm": 0.7328051651446971,
+      "learning_rate": 6.961492862961528e-07,
+      "loss": 1.5514,
+      "step": 715
+    },
+    {
+      "epoch": 0.04989025537400272,
+      "grad_norm": 0.7123550388724823,
+      "learning_rate": 6.961385296129375e-07,
+      "loss": 1.6047,
+      "step": 716
+    },
+    {
+      "epoch": 0.04995993450162004,
+      "grad_norm": 0.7147651353416852,
+      "learning_rate": 6.961277580193249e-07,
+      "loss": 1.6656,
+      "step": 717
+    },
+    {
+      "epoch": 0.050029613629237364,
+      "grad_norm": 0.6906496045277001,
+      "learning_rate": 6.961169715158317e-07,
+      "loss": 1.5934,
+      "step": 718
+    },
+    {
+      "epoch": 0.05009929275685469,
+      "grad_norm": 0.7541044654754127,
+      "learning_rate": 6.961061701029741e-07,
+      "loss": 1.4913,
+      "step": 719
+    },
+    {
+      "epoch": 0.05016897188447201,
+      "grad_norm": 0.7540942860771993,
+      "learning_rate": 6.960953537812703e-07,
+      "loss": 1.5373,
+      "step": 720
+    },
+    {
+      "epoch": 0.05023865101208933,
+      "grad_norm": 0.6781331883086441,
+      "learning_rate": 6.960845225512386e-07,
+      "loss": 1.5436,
+      "step": 721
+    },
+    {
+      "epoch": 0.050308330139706654,
+      "grad_norm": 0.8470452786448093,
+      "learning_rate": 6.960736764133978e-07,
+      "loss": 1.6208,
+      "step": 722
+    },
+    {
+      "epoch": 0.05037800926732397,
+      "grad_norm": 0.6760410770524181,
+      "learning_rate": 6.960628153682679e-07,
+      "loss": 1.4621,
+      "step": 723
+    },
+    {
+      "epoch": 0.05044768839494129,
+      "grad_norm": 0.7490727821128771,
+      "learning_rate": 6.960519394163692e-07,
+      "loss": 1.4257,
+      "step": 724
+    },
+    {
+      "epoch": 0.050517367522558615,
+      "grad_norm": 0.6857827107854004,
+      "learning_rate": 6.96041048558223e-07,
+      "loss": 1.6527,
+      "step": 725
+    },
+    {
+      "epoch": 0.05058704665017594,
+      "grad_norm": 0.7032073154704966,
+      "learning_rate": 6.960301427943513e-07,
+      "loss": 1.6024,
+      "step": 726
+    },
+    {
+      "epoch": 0.05065672577779326,
+      "grad_norm": 0.6849058234749557,
+      "learning_rate": 6.960192221252765e-07,
+      "loss": 1.4722,
+      "step": 727
+    },
+    {
+      "epoch": 0.05072640490541058,
+      "grad_norm": 0.7322091820332296,
+      "learning_rate": 6.960082865515221e-07,
+      "loss": 1.6492,
+      "step": 728
+    },
+    {
+      "epoch": 0.050796084033027905,
+      "grad_norm": 0.7224516190286429,
+      "learning_rate": 6.959973360736122e-07,
+      "loss": 1.6839,
+      "step": 729
+    },
+    {
+      "epoch": 0.05086576316064523,
+      "grad_norm": 0.686954801078055,
+      "learning_rate": 6.959863706920713e-07,
+      "loss": 1.5376,
+      "step": 730
+    },
+    {
+      "epoch": 0.05093544228826255,
+      "grad_norm": 0.6862780679407999,
+      "learning_rate": 6.959753904074253e-07,
+      "loss": 1.5023,
+      "step": 731
+    },
+    {
+      "epoch": 0.05100512141587987,
+      "grad_norm": 0.721010317984146,
+      "learning_rate": 6.959643952202001e-07,
+      "loss": 1.5141,
+      "step": 732
+    },
+    {
+      "epoch": 0.051074800543497195,
+      "grad_norm": 0.6967847500688505,
+      "learning_rate": 6.959533851309226e-07,
+      "loss": 1.4556,
+      "step": 733
+    },
+    {
+      "epoch": 0.05114447967111452,
+      "grad_norm": 0.7489626901360203,
+      "learning_rate": 6.959423601401205e-07,
+      "loss": 1.533,
+      "step": 734
+    },
+    {
+      "epoch": 0.05121415879873184,
+      "grad_norm": 0.7850799405099177,
+      "learning_rate": 6.959313202483222e-07,
+      "loss": 1.5717,
+      "step": 735
+    },
+    {
+      "epoch": 0.05128383792634916,
+      "grad_norm": 0.6962064304868032,
+      "learning_rate": 6.959202654560567e-07,
+      "loss": 1.4927,
+      "step": 736
+    },
+    {
+      "epoch": 0.051353517053966485,
+      "grad_norm": 0.6814157744096807,
+      "learning_rate": 6.959091957638539e-07,
+      "loss": 1.5426,
+      "step": 737
+    },
+    {
+      "epoch": 0.05142319618158381,
+      "grad_norm": 0.7046197816079532,
+      "learning_rate": 6.958981111722439e-07,
+      "loss": 1.5478,
+      "step": 738
+    },
+    {
+      "epoch": 0.05149287530920113,
+      "grad_norm": 0.730215604464463,
+      "learning_rate": 6.958870116817583e-07,
+      "loss": 1.5133,
+      "step": 739
+    },
+    {
+      "epoch": 0.05156255443681845,
+      "grad_norm": 0.7009966531083134,
+      "learning_rate": 6.958758972929289e-07,
+      "loss": 1.5814,
+      "step": 740
+    },
+    {
+      "epoch": 0.051632233564435775,
+      "grad_norm": 0.7164763075684121,
+      "learning_rate": 6.958647680062882e-07,
+      "loss": 1.54,
+      "step": 741
+    },
+    {
+      "epoch": 0.0517019126920531,
+      "grad_norm": 0.6961036208661276,
+      "learning_rate": 6.958536238223697e-07,
+      "loss": 1.5408,
+      "step": 742
+    },
+    {
+      "epoch": 0.05177159181967042,
+      "grad_norm": 0.7736476120671665,
+      "learning_rate": 6.958424647417074e-07,
+      "loss": 1.633,
+      "step": 743
+    },
+    {
+      "epoch": 0.05184127094728774,
+      "grad_norm": 0.7260767598944713,
+      "learning_rate": 6.958312907648358e-07,
+      "loss": 1.4184,
+      "step": 744
+    },
+    {
+      "epoch": 0.051910950074905066,
+      "grad_norm": 0.7292566831371331,
+      "learning_rate": 6.958201018922908e-07,
+      "loss": 1.4852,
+      "step": 745
+    },
+    {
+      "epoch": 0.05198062920252238,
+      "grad_norm": 0.6826555733917109,
+      "learning_rate": 6.958088981246085e-07,
+      "loss": 1.4976,
+      "step": 746
+    },
+    {
+      "epoch": 0.052050308330139704,
+      "grad_norm": 0.8057138550101121,
+      "learning_rate": 6.957976794623257e-07,
+      "loss": 1.6662,
+      "step": 747
+    },
+    {
+      "epoch": 0.052119987457757026,
+      "grad_norm": 0.7383595107784787,
+      "learning_rate": 6.9578644590598e-07,
+      "loss": 1.6895,
+      "step": 748
+    },
+    {
+      "epoch": 0.05218966658537435,
+      "grad_norm": 0.7623169165234596,
+      "learning_rate": 6.957751974561098e-07,
+      "loss": 1.5831,
+      "step": 749
+    },
+    {
+      "epoch": 0.05225934571299167,
+      "grad_norm": 0.7010872976902116,
+      "learning_rate": 6.95763934113254e-07,
+      "loss": 1.6176,
+      "step": 750
+    },
+    {
+      "epoch": 0.052329024840608994,
+      "grad_norm": 0.7652887108312698,
+      "learning_rate": 6.957526558779526e-07,
+      "loss": 1.5928,
+      "step": 751
+    },
+    {
+      "epoch": 0.052398703968226316,
+      "grad_norm": 0.8296288937742872,
+      "learning_rate": 6.95741362750746e-07,
+      "loss": 1.5899,
+      "step": 752
+    },
+    {
+      "epoch": 0.05246838309584364,
+      "grad_norm": 0.7128647360486586,
+      "learning_rate": 6.957300547321753e-07,
+      "loss": 1.55,
+      "step": 753
+    },
+    {
+      "epoch": 0.05253806222346096,
+      "grad_norm": 0.7111814164131637,
+      "learning_rate": 6.957187318227823e-07,
+      "loss": 1.7715,
+      "step": 754
+    },
+    {
+      "epoch": 0.052607741351078284,
+      "grad_norm": 0.7581371563133963,
+      "learning_rate": 6.9570739402311e-07,
+      "loss": 1.621,
+      "step": 755
+    },
+    {
+      "epoch": 0.05267742047869561,
+      "grad_norm": 0.7320911081083655,
+      "learning_rate": 6.956960413337015e-07,
+      "loss": 1.6028,
+      "step": 756
+    },
+    {
+      "epoch": 0.05274709960631293,
+      "grad_norm": 0.6811468195966736,
+      "learning_rate": 6.956846737551008e-07,
+      "loss": 1.4258,
+      "step": 757
+    },
+    {
+      "epoch": 0.05281677873393025,
+      "grad_norm": 0.7558283632455667,
+      "learning_rate": 6.956732912878528e-07,
+      "loss": 1.6908,
+      "step": 758
+    },
+    {
+      "epoch": 0.052886457861547574,
+      "grad_norm": 0.7401585909526207,
+      "learning_rate": 6.956618939325027e-07,
+      "loss": 1.5859,
+      "step": 759
+    },
+    {
+      "epoch": 0.0529561369891649,
+      "grad_norm": 0.6869688342909643,
+      "learning_rate": 6.95650481689597e-07,
+      "loss": 1.5995,
+      "step": 760
+    },
+    {
+      "epoch": 0.05302581611678222,
+      "grad_norm": 0.732390242062926,
+      "learning_rate": 6.956390545596824e-07,
+      "loss": 1.5717,
+      "step": 761
+    },
+    {
+      "epoch": 0.05309549524439954,
+      "grad_norm": 0.7217401642738522,
+      "learning_rate": 6.956276125433066e-07,
+      "loss": 1.6592,
+      "step": 762
+    },
+    {
+      "epoch": 0.053165174372016864,
+      "grad_norm": 0.7026097108747728,
+      "learning_rate": 6.956161556410179e-07,
+      "loss": 1.4727,
+      "step": 763
+    },
+    {
+      "epoch": 0.05323485349963419,
+      "grad_norm": 0.7104570430746965,
+      "learning_rate": 6.956046838533654e-07,
+      "loss": 1.548,
+      "step": 764
+    },
+    {
+      "epoch": 0.05330453262725151,
+      "grad_norm": 0.6794024019476335,
+      "learning_rate": 6.955931971808987e-07,
+      "loss": 1.3615,
+      "step": 765
+    },
+    {
+      "epoch": 0.05337421175486883,
+      "grad_norm": 0.7494295020279464,
+      "learning_rate": 6.955816956241684e-07,
+      "loss": 1.5749,
+      "step": 766
+    },
+    {
+      "epoch": 0.053443890882486154,
+      "grad_norm": 0.6688709082207913,
+      "learning_rate": 6.955701791837256e-07,
+      "loss": 1.5078,
+      "step": 767
+    },
+    {
+      "epoch": 0.05351357001010348,
+      "grad_norm": 0.710494094903225,
+      "learning_rate": 6.955586478601222e-07,
+      "loss": 1.4936,
+      "step": 768
+    },
+    {
+      "epoch": 0.05358324913772079,
+      "grad_norm": 0.711894193384377,
+      "learning_rate": 6.955471016539109e-07,
+      "loss": 1.563,
+      "step": 769
+    },
+    {
+      "epoch": 0.053652928265338115,
+      "grad_norm": 0.7323768349826295,
+      "learning_rate": 6.95535540565645e-07,
+      "loss": 1.5585,
+      "step": 770
+    },
+    {
+      "epoch": 0.05372260739295544,
+      "grad_norm": 0.7329434747907027,
+      "learning_rate": 6.955239645958784e-07,
+      "loss": 1.617,
+      "step": 771
+    },
+    {
+      "epoch": 0.05379228652057276,
+      "grad_norm": 0.748227642697152,
+      "learning_rate": 6.95512373745166e-07,
+      "loss": 1.5454,
+      "step": 772
+    },
+    {
+      "epoch": 0.05386196564819008,
+      "grad_norm": 0.7328425554339898,
+      "learning_rate": 6.95500768014063e-07,
+      "loss": 1.5224,
+      "step": 773
+    },
+    {
+      "epoch": 0.053931644775807405,
+      "grad_norm": 0.7466917176783684,
+      "learning_rate": 6.954891474031259e-07,
+      "loss": 1.5347,
+      "step": 774
+    },
+    {
+      "epoch": 0.05400132390342473,
+      "grad_norm": 0.7585007520137389,
+      "learning_rate": 6.954775119129114e-07,
+      "loss": 1.7438,
+      "step": 775
+    },
+    {
+      "epoch": 0.05407100303104205,
+      "grad_norm": 0.7439885599579423,
+      "learning_rate": 6.95465861543977e-07,
+      "loss": 1.6233,
+      "step": 776
+    },
+    {
+      "epoch": 0.05414068215865937,
+      "grad_norm": 0.6937689954958782,
+      "learning_rate": 6.954541962968813e-07,
+      "loss": 1.5569,
+      "step": 777
+    },
+    {
+      "epoch": 0.054210361286276695,
+      "grad_norm": 0.6808608820879647,
+      "learning_rate": 6.954425161721832e-07,
+      "loss": 1.5482,
+      "step": 778
+    },
+    {
+      "epoch": 0.05428004041389402,
+      "grad_norm": 0.7549169966464018,
+      "learning_rate": 6.954308211704422e-07,
+      "loss": 1.5517,
+      "step": 779
+    },
+    {
+      "epoch": 0.05434971954151134,
+      "grad_norm": 0.6734510329316952,
+      "learning_rate": 6.954191112922192e-07,
+      "loss": 1.5492,
+      "step": 780
+    },
+    {
+      "epoch": 0.05441939866912866,
+      "grad_norm": 0.6877325247570651,
+      "learning_rate": 6.954073865380749e-07,
+      "loss": 1.4862,
+      "step": 781
+    },
+    {
+      "epoch": 0.054489077796745986,
+      "grad_norm": 0.6788147898970631,
+      "learning_rate": 6.953956469085715e-07,
+      "loss": 1.5726,
+      "step": 782
+    },
+    {
+      "epoch": 0.05455875692436331,
+      "grad_norm": 0.7305927316424901,
+      "learning_rate": 6.953838924042714e-07,
+      "loss": 1.4975,
+      "step": 783
+    },
+    {
+      "epoch": 0.05462843605198063,
+      "grad_norm": 0.7132269277569465,
+      "learning_rate": 6.95372123025738e-07,
+      "loss": 1.4406,
+      "step": 784
+    },
+    {
+      "epoch": 0.05469811517959795,
+      "grad_norm": 0.7020412658784635,
+      "learning_rate": 6.953603387735353e-07,
+      "loss": 1.587,
+      "step": 785
+    },
+    {
+      "epoch": 0.054767794307215276,
+      "grad_norm": 0.7528096373864551,
+      "learning_rate": 6.953485396482281e-07,
+      "loss": 1.601,
+      "step": 786
+    },
+    {
+      "epoch": 0.0548374734348326,
+      "grad_norm": 0.6512215101888548,
+      "learning_rate": 6.953367256503816e-07,
+      "loss": 1.5033,
+      "step": 787
+    },
+    {
+      "epoch": 0.05490715256244992,
+      "grad_norm": 0.6788130061609902,
+      "learning_rate": 6.953248967805621e-07,
+      "loss": 1.4305,
+      "step": 788
+    },
+    {
+      "epoch": 0.05497683169006724,
+      "grad_norm": 0.7446892987199926,
+      "learning_rate": 6.953130530393365e-07,
+      "loss": 1.7707,
+      "step": 789
+    },
+    {
+      "epoch": 0.055046510817684566,
+      "grad_norm": 0.8206689971928751,
+      "learning_rate": 6.953011944272724e-07,
+      "loss": 1.697,
+      "step": 790
+    },
+    {
+      "epoch": 0.05511618994530188,
+      "grad_norm": 0.7121618215215683,
+      "learning_rate": 6.952893209449378e-07,
+      "loss": 1.5926,
+      "step": 791
+    },
+    {
+      "epoch": 0.055185869072919204,
+      "grad_norm": 0.6684685055821509,
+      "learning_rate": 6.952774325929022e-07,
+      "loss": 1.5346,
+      "step": 792
+    },
+    {
+      "epoch": 0.055255548200536526,
+      "grad_norm": 0.7876441568581036,
+      "learning_rate": 6.95265529371735e-07,
+      "loss": 1.6429,
+      "step": 793
+    },
+    {
+      "epoch": 0.05532522732815385,
+      "grad_norm": 0.6966735167203888,
+      "learning_rate": 6.952536112820066e-07,
+      "loss": 1.5847,
+      "step": 794
+    },
+    {
+      "epoch": 0.05539490645577117,
+      "grad_norm": 0.6584961268883045,
+      "learning_rate": 6.952416783242882e-07,
+      "loss": 1.4795,
+      "step": 795
+    },
+    {
+      "epoch": 0.055464585583388494,
+      "grad_norm": 0.8031283060641858,
+      "learning_rate": 6.952297304991516e-07,
+      "loss": 1.7066,
+      "step": 796
+    },
+    {
+      "epoch": 0.05553426471100582,
+      "grad_norm": 0.7839285381689913,
+      "learning_rate": 6.952177678071696e-07,
+      "loss": 1.7255,
+      "step": 797
+    },
+    {
+      "epoch": 0.05560394383862314,
+      "grad_norm": 0.7693006460738586,
+      "learning_rate": 6.952057902489152e-07,
+      "loss": 1.6596,
+      "step": 798
+    },
+    {
+      "epoch": 0.05567362296624046,
+      "grad_norm": 0.722826042731386,
+      "learning_rate": 6.951937978249624e-07,
+      "loss": 1.5677,
+      "step": 799
+    },
+    {
+      "epoch": 0.055743302093857784,
+      "grad_norm": 0.7117854955696307,
+      "learning_rate": 6.951817905358861e-07,
+      "loss": 1.5712,
+      "step": 800
+    },
+    {
+      "epoch": 0.05581298122147511,
+      "grad_norm": 0.7873681567228429,
+      "learning_rate": 6.951697683822617e-07,
+      "loss": 1.5217,
+      "step": 801
+    },
+    {
+      "epoch": 0.05588266034909243,
+      "grad_norm": 0.6990444943511518,
+      "learning_rate": 6.951577313646651e-07,
+      "loss": 1.4714,
+      "step": 802
+    },
+    {
+      "epoch": 0.05595233947670975,
+      "grad_norm": 0.7087438008306041,
+      "learning_rate": 6.951456794836733e-07,
+      "loss": 1.5843,
+      "step": 803
+    },
+    {
+      "epoch": 0.056022018604327074,
+      "grad_norm": 0.7466712931823629,
+      "learning_rate": 6.951336127398638e-07,
+      "loss": 1.5316,
+      "step": 804
+    },
+    {
+      "epoch": 0.0560916977319444,
+      "grad_norm": 0.7653458202685157,
+      "learning_rate": 6.951215311338148e-07,
+      "loss": 1.6453,
+      "step": 805
+    },
+    {
+      "epoch": 0.05616137685956172,
+      "grad_norm": 0.7236559613144168,
+      "learning_rate": 6.951094346661055e-07,
+      "loss": 1.6199,
+      "step": 806
+    },
+    {
+      "epoch": 0.05623105598717904,
+      "grad_norm": 0.6944459703744715,
+      "learning_rate": 6.950973233373155e-07,
+      "loss": 1.5263,
+      "step": 807
+    },
+    {
+      "epoch": 0.056300735114796364,
+      "grad_norm": 0.7816944407503731,
+      "learning_rate": 6.95085197148025e-07,
+      "loss": 1.6126,
+      "step": 808
+    },
+    {
+      "epoch": 0.05637041424241369,
+      "grad_norm": 0.7124399093435471,
+      "learning_rate": 6.950730560988153e-07,
+      "loss": 1.5162,
+      "step": 809
+    },
+    {
+      "epoch": 0.05644009337003101,
+      "grad_norm": 0.6988292198617666,
+      "learning_rate": 6.950609001902682e-07,
+      "loss": 1.4888,
+      "step": 810
+    },
+    {
+      "epoch": 0.05650977249764833,
+      "grad_norm": 0.7329883842362056,
+      "learning_rate": 6.950487294229662e-07,
+      "loss": 1.6594,
+      "step": 811
+    },
+    {
+      "epoch": 0.056579451625265655,
+      "grad_norm": 0.6939338590714141,
+      "learning_rate": 6.950365437974927e-07,
+      "loss": 1.3898,
+      "step": 812
+    },
+    {
+      "epoch": 0.05664913075288298,
+      "grad_norm": 0.7188599302209264,
+      "learning_rate": 6.950243433144314e-07,
+      "loss": 1.4821,
+      "step": 813
+    },
+    {
+      "epoch": 0.05671880988050029,
+      "grad_norm": 0.6823465353383991,
+      "learning_rate": 6.950121279743672e-07,
+      "loss": 1.5173,
+      "step": 814
+    },
+    {
+      "epoch": 0.056788489008117615,
+      "grad_norm": 0.8750761366460855,
+      "learning_rate": 6.949998977778852e-07,
+      "loss": 1.7818,
+      "step": 815
+    },
+    {
+      "epoch": 0.05685816813573494,
+      "grad_norm": 0.7137120708518332,
+      "learning_rate": 6.949876527255718e-07,
+      "loss": 1.6433,
+      "step": 816
+    },
+    {
+      "epoch": 0.05692784726335226,
+      "grad_norm": 0.6619871776167325,
+      "learning_rate": 6.949753928180137e-07,
+      "loss": 1.5437,
+      "step": 817
+    },
+    {
+      "epoch": 0.05699752639096958,
+      "grad_norm": 0.698883982747292,
+      "learning_rate": 6.949631180557984e-07,
+      "loss": 1.5743,
+      "step": 818
+    },
+    {
+      "epoch": 0.057067205518586905,
+      "grad_norm": 0.7606502997566646,
+      "learning_rate": 6.949508284395141e-07,
+      "loss": 1.6545,
+      "step": 819
+    },
+    {
+      "epoch": 0.05713688464620423,
+      "grad_norm": 0.714118076308743,
+      "learning_rate": 6.949385239697498e-07,
+      "loss": 1.4312,
+      "step": 820
+    },
+    {
+      "epoch": 0.05720656377382155,
+      "grad_norm": 0.7478782433732383,
+      "learning_rate": 6.949262046470951e-07,
+      "loss": 1.5869,
+      "step": 821
+    },
+    {
+      "epoch": 0.05727624290143887,
+      "grad_norm": 0.7041434815718658,
+      "learning_rate": 6.949138704721405e-07,
+      "loss": 1.5026,
+      "step": 822
+    },
+    {
+      "epoch": 0.057345922029056196,
+      "grad_norm": 0.7290786553345412,
+      "learning_rate": 6.94901521445477e-07,
+      "loss": 1.8279,
+      "step": 823
+    },
+    {
+      "epoch": 0.05741560115667352,
+      "grad_norm": 0.6990004793808002,
+      "learning_rate": 6.948891575676963e-07,
+      "loss": 1.4113,
+      "step": 824
+    },
+    {
+      "epoch": 0.05748528028429084,
+      "grad_norm": 0.7811171957869557,
+      "learning_rate": 6.94876778839391e-07,
+      "loss": 1.558,
+      "step": 825
+    },
+    {
+      "epoch": 0.05755495941190816,
+      "grad_norm": 0.739030141597572,
+      "learning_rate": 6.948643852611543e-07,
+      "loss": 1.5298,
+      "step": 826
+    },
+    {
+      "epoch": 0.057624638539525486,
+      "grad_norm": 0.7219462250643431,
+      "learning_rate": 6.948519768335801e-07,
+      "loss": 1.5291,
+      "step": 827
+    },
+    {
+      "epoch": 0.05769431766714281,
+      "grad_norm": 0.7120923451377806,
+      "learning_rate": 6.948395535572631e-07,
+      "loss": 1.4506,
+      "step": 828
+    },
+    {
+      "epoch": 0.05776399679476013,
+      "grad_norm": 0.6975168518442334,
+      "learning_rate": 6.948271154327985e-07,
+      "loss": 1.4987,
+      "step": 829
+    },
+    {
+      "epoch": 0.05783367592237745,
+      "grad_norm": 0.7138373985437595,
+      "learning_rate": 6.948146624607826e-07,
+      "loss": 1.6882,
+      "step": 830
+    },
+    {
+      "epoch": 0.057903355049994776,
+      "grad_norm": 0.6708575706578195,
+      "learning_rate": 6.948021946418118e-07,
+      "loss": 1.4863,
+      "step": 831
+    },
+    {
+      "epoch": 0.0579730341776121,
+      "grad_norm": 0.6808372768351861,
+      "learning_rate": 6.947897119764841e-07,
+      "loss": 1.5298,
+      "step": 832
+    },
+    {
+      "epoch": 0.05804271330522942,
+      "grad_norm": 0.6852255133243262,
+      "learning_rate": 6.947772144653973e-07,
+      "loss": 1.5228,
+      "step": 833
+    },
+    {
+      "epoch": 0.05811239243284674,
+      "grad_norm": 0.7146018666757779,
+      "learning_rate": 6.947647021091504e-07,
+      "loss": 1.5371,
+      "step": 834
+    },
+    {
+      "epoch": 0.058182071560464066,
+      "grad_norm": 0.7270596691308416,
+      "learning_rate": 6.947521749083431e-07,
+      "loss": 1.6466,
+      "step": 835
+    },
+    {
+      "epoch": 0.05825175068808139,
+      "grad_norm": 0.7603407007056545,
+      "learning_rate": 6.947396328635757e-07,
+      "loss": 1.6199,
+      "step": 836
+    },
+    {
+      "epoch": 0.058321429815698704,
+      "grad_norm": 0.6738987574655745,
+      "learning_rate": 6.947270759754491e-07,
+      "loss": 1.5234,
+      "step": 837
+    },
+    {
+      "epoch": 0.05839110894331603,
+      "grad_norm": 0.6897399123926644,
+      "learning_rate": 6.947145042445652e-07,
+      "loss": 1.4501,
+      "step": 838
+    },
+    {
+      "epoch": 0.05846078807093335,
+      "grad_norm": 0.7264844681997801,
+      "learning_rate": 6.947019176715265e-07,
+      "loss": 1.5785,
+      "step": 839
+    },
+    {
+      "epoch": 0.05853046719855067,
+      "grad_norm": 0.71216329594823,
+      "learning_rate": 6.94689316256936e-07,
+      "loss": 1.5699,
+      "step": 840
+    },
+    {
+      "epoch": 0.058600146326167994,
+      "grad_norm": 0.7269442447044326,
+      "learning_rate": 6.946767000013978e-07,
+      "loss": 1.5217,
+      "step": 841
+    },
+    {
+      "epoch": 0.05866982545378532,
+      "grad_norm": 0.7213513641430525,
+      "learning_rate": 6.946640689055163e-07,
+      "loss": 1.4736,
+      "step": 842
+    },
+    {
+      "epoch": 0.05873950458140264,
+      "grad_norm": 0.6911765142958883,
+      "learning_rate": 6.946514229698968e-07,
+      "loss": 1.5468,
+      "step": 843
+    },
+    {
+      "epoch": 0.05880918370901996,
+      "grad_norm": 0.7312402973375173,
+      "learning_rate": 6.946387621951456e-07,
+      "loss": 1.5845,
+      "step": 844
+    },
+    {
+      "epoch": 0.058878862836637284,
+      "grad_norm": 0.8766300017445524,
+      "learning_rate": 6.946260865818691e-07,
+      "loss": 1.6933,
+      "step": 845
+    },
+    {
+      "epoch": 0.05894854196425461,
+      "grad_norm": 0.7437506934142784,
+      "learning_rate": 6.946133961306748e-07,
+      "loss": 1.5705,
+      "step": 846
+    },
+    {
+      "epoch": 0.05901822109187193,
+      "grad_norm": 0.7250232099606774,
+      "learning_rate": 6.946006908421711e-07,
+      "loss": 1.6298,
+      "step": 847
+    },
+    {
+      "epoch": 0.05908790021948925,
+      "grad_norm": 0.7468248406429453,
+      "learning_rate": 6.945879707169668e-07,
+      "loss": 1.5957,
+      "step": 848
+    },
+    {
+      "epoch": 0.059157579347106574,
+      "grad_norm": 0.7307135773660111,
+      "learning_rate": 6.945752357556712e-07,
+      "loss": 1.6783,
+      "step": 849
+    },
+    {
+      "epoch": 0.0592272584747239,
+      "grad_norm": 0.7462954093323562,
+      "learning_rate": 6.945624859588947e-07,
+      "loss": 1.6313,
+      "step": 850
+    },
+    {
+      "epoch": 0.05929693760234122,
+      "grad_norm": 0.6836820982793157,
+      "learning_rate": 6.945497213272485e-07,
+      "loss": 1.5552,
+      "step": 851
+    },
+    {
+      "epoch": 0.05936661672995854,
+      "grad_norm": 0.7349495347711605,
+      "learning_rate": 6.94536941861344e-07,
+      "loss": 1.5534,
+      "step": 852
+    },
+    {
+      "epoch": 0.059436295857575865,
+      "grad_norm": 0.7393307515575156,
+      "learning_rate": 6.945241475617939e-07,
+      "loss": 1.5919,
+      "step": 853
+    },
+    {
+      "epoch": 0.05950597498519319,
+      "grad_norm": 0.7581161323825397,
+      "learning_rate": 6.945113384292112e-07,
+      "loss": 1.5157,
+      "step": 854
+    },
+    {
+      "epoch": 0.05957565411281051,
+      "grad_norm": 0.7811029541452695,
+      "learning_rate": 6.944985144642097e-07,
+      "loss": 1.5304,
+      "step": 855
+    },
+    {
+      "epoch": 0.05964533324042783,
+      "grad_norm": 0.6479169819808708,
+      "learning_rate": 6.94485675667404e-07,
+      "loss": 1.5384,
+      "step": 856
+    },
+    {
+      "epoch": 0.059715012368045155,
+      "grad_norm": 0.6754137449735433,
+      "learning_rate": 6.944728220394094e-07,
+      "loss": 1.5514,
+      "step": 857
+    },
+    {
+      "epoch": 0.05978469149566248,
+      "grad_norm": 0.7010515414317577,
+      "learning_rate": 6.944599535808418e-07,
+      "loss": 1.6216,
+      "step": 858
+    },
+    {
+      "epoch": 0.0598543706232798,
+      "grad_norm": 0.7420195687781036,
+      "learning_rate": 6.944470702923181e-07,
+      "loss": 1.4931,
+      "step": 859
+    },
+    {
+      "epoch": 0.059924049750897115,
+      "grad_norm": 0.6985723055107013,
+      "learning_rate": 6.944341721744553e-07,
+      "loss": 1.5086,
+      "step": 860
+    },
+    {
+      "epoch": 0.05999372887851444,
+      "grad_norm": 0.7496262938898509,
+      "learning_rate": 6.944212592278718e-07,
+      "loss": 1.5999,
+      "step": 861
+    },
+    {
+      "epoch": 0.06006340800613176,
+      "grad_norm": 0.6981277967509739,
+      "learning_rate": 6.944083314531863e-07,
+      "loss": 1.6053,
+      "step": 862
+    },
+    {
+      "epoch": 0.06013308713374908,
+      "grad_norm": 0.6804099181252625,
+      "learning_rate": 6.943953888510182e-07,
+      "loss": 1.5734,
+      "step": 863
+    },
+    {
+      "epoch": 0.060202766261366406,
+      "grad_norm": 0.6731032371257547,
+      "learning_rate": 6.943824314219881e-07,
+      "loss": 1.5698,
+      "step": 864
+    },
+    {
+      "epoch": 0.06027244538898373,
+      "grad_norm": 0.7425209366343755,
+      "learning_rate": 6.943694591667166e-07,
+      "loss": 1.5785,
+      "step": 865
+    },
+    {
+      "epoch": 0.06034212451660105,
+      "grad_norm": 0.7528653762418056,
+      "learning_rate": 6.943564720858257e-07,
+      "loss": 1.6388,
+      "step": 866
+    },
+    {
+      "epoch": 0.06041180364421837,
+      "grad_norm": 0.8122728565845222,
+      "learning_rate": 6.943434701799373e-07,
+      "loss": 1.625,
+      "step": 867
+    },
+    {
+      "epoch": 0.060481482771835696,
+      "grad_norm": 0.7019858720037662,
+      "learning_rate": 6.943304534496749e-07,
+      "loss": 1.564,
+      "step": 868
+    },
+    {
+      "epoch": 0.06055116189945302,
+      "grad_norm": 0.7177275391866323,
+      "learning_rate": 6.943174218956621e-07,
+      "loss": 1.6123,
+      "step": 869
+    },
+    {
+      "epoch": 0.06062084102707034,
+      "grad_norm": 0.7030350752558858,
+      "learning_rate": 6.943043755185235e-07,
+      "loss": 1.5799,
+      "step": 870
+    },
+    {
+      "epoch": 0.06069052015468766,
+      "grad_norm": 0.7320624667675142,
+      "learning_rate": 6.942913143188841e-07,
+      "loss": 1.626,
+      "step": 871
+    },
+    {
+      "epoch": 0.060760199282304986,
+      "grad_norm": 0.7741178657105833,
+      "learning_rate": 6.9427823829737e-07,
+      "loss": 1.6211,
+      "step": 872
+    },
+    {
+      "epoch": 0.06082987840992231,
+      "grad_norm": 0.7034164705948763,
+      "learning_rate": 6.942651474546077e-07,
+      "loss": 1.5162,
+      "step": 873
+    },
+    {
+      "epoch": 0.06089955753753963,
+      "grad_norm": 0.6909381638542067,
+      "learning_rate": 6.942520417912248e-07,
+      "loss": 1.4292,
+      "step": 874
+    },
+    {
+      "epoch": 0.06096923666515695,
+      "grad_norm": 0.7115351412753147,
+      "learning_rate": 6.94238921307849e-07,
+      "loss": 1.6299,
+      "step": 875
+    },
+    {
+      "epoch": 0.061038915792774276,
+      "grad_norm": 0.6704973088007256,
+      "learning_rate": 6.942257860051093e-07,
+      "loss": 1.6049,
+      "step": 876
+    },
+    {
+      "epoch": 0.0611085949203916,
+      "grad_norm": 0.7460113523362771,
+      "learning_rate": 6.942126358836352e-07,
+      "loss": 1.6023,
+      "step": 877
+    },
+    {
+      "epoch": 0.06117827404800892,
+      "grad_norm": 0.740997449770298,
+      "learning_rate": 6.941994709440567e-07,
+      "loss": 1.461,
+      "step": 878
+    },
+    {
+      "epoch": 0.061247953175626244,
+      "grad_norm": 0.7344001816527201,
+      "learning_rate": 6.941862911870047e-07,
+      "loss": 1.6357,
+      "step": 879
+    },
+    {
+      "epoch": 0.061317632303243566,
+      "grad_norm": 0.842515805713784,
+      "learning_rate": 6.941730966131111e-07,
+      "loss": 1.6079,
+      "step": 880
+    },
+    {
+      "epoch": 0.06138731143086089,
+      "grad_norm": 0.7541684280459339,
+      "learning_rate": 6.941598872230078e-07,
+      "loss": 1.4737,
+      "step": 881
+    },
+    {
+      "epoch": 0.06145699055847821,
+      "grad_norm": 0.7780595951607582,
+      "learning_rate": 6.941466630173281e-07,
+      "loss": 1.5857,
+      "step": 882
+    },
+    {
+      "epoch": 0.06152666968609553,
+      "grad_norm": 0.7922696893069073,
+      "learning_rate": 6.941334239967056e-07,
+      "loss": 1.6614,
+      "step": 883
+    },
+    {
+      "epoch": 0.06159634881371285,
+      "grad_norm": 0.7049532795712117,
+      "learning_rate": 6.941201701617749e-07,
+      "loss": 1.5411,
+      "step": 884
+    },
+    {
+      "epoch": 0.06166602794133017,
+      "grad_norm": 0.7030659598960232,
+      "learning_rate": 6.941069015131709e-07,
+      "loss": 1.6289,
+      "step": 885
+    },
+    {
+      "epoch": 0.061735707068947494,
+      "grad_norm": 0.7093629644917849,
+      "learning_rate": 6.940936180515296e-07,
+      "loss": 1.4986,
+      "step": 886
+    },
+    {
+      "epoch": 0.06180538619656482,
+      "grad_norm": 0.7310013628107739,
+      "learning_rate": 6.940803197774875e-07,
+      "loss": 1.7331,
+      "step": 887
+    },
+    {
+      "epoch": 0.06187506532418214,
+      "grad_norm": 0.7085823814611059,
+      "learning_rate": 6.94067006691682e-07,
+      "loss": 1.4859,
+      "step": 888
+    },
+    {
+      "epoch": 0.06194474445179946,
+      "grad_norm": 0.6963722191996505,
+      "learning_rate": 6.940536787947512e-07,
+      "loss": 1.6608,
+      "step": 889
+    },
+    {
+      "epoch": 0.062014423579416784,
+      "grad_norm": 0.7409922933496068,
+      "learning_rate": 6.940403360873335e-07,
+      "loss": 1.5919,
+      "step": 890
+    },
+    {
+      "epoch": 0.06208410270703411,
+      "grad_norm": 0.6655268228790038,
+      "learning_rate": 6.940269785700685e-07,
+      "loss": 1.507,
+      "step": 891
+    },
+    {
+      "epoch": 0.06215378183465143,
+      "grad_norm": 0.6821234570236667,
+      "learning_rate": 6.940136062435963e-07,
+      "loss": 1.6311,
+      "step": 892
+    },
+    {
+      "epoch": 0.06222346096226875,
+      "grad_norm": 0.6902129271155977,
+      "learning_rate": 6.940002191085575e-07,
+      "loss": 1.533,
+      "step": 893
+    },
+    {
+      "epoch": 0.062293140089886075,
+      "grad_norm": 0.7606556884733682,
+      "learning_rate": 6.93986817165594e-07,
+      "loss": 1.6512,
+      "step": 894
+    },
+    {
+      "epoch": 0.0623628192175034,
+      "grad_norm": 0.7289684165848465,
+      "learning_rate": 6.939734004153479e-07,
+      "loss": 1.5418,
+      "step": 895
+    },
+    {
+      "epoch": 0.06243249834512072,
+      "grad_norm": 0.6811870960828804,
+      "learning_rate": 6.939599688584621e-07,
+      "loss": 1.5511,
+      "step": 896
+    },
+    {
+      "epoch": 0.06250217747273804,
+      "grad_norm": 0.7242835339866917,
+      "learning_rate": 6.939465224955802e-07,
+      "loss": 1.4901,
+      "step": 897
+    },
+    {
+      "epoch": 0.06257185660035536,
+      "grad_norm": 0.7231645951362123,
+      "learning_rate": 6.939330613273468e-07,
+      "loss": 1.5142,
+      "step": 898
+    },
+    {
+      "epoch": 0.06264153572797268,
+      "grad_norm": 0.7434077826308495,
+      "learning_rate": 6.939195853544069e-07,
+      "loss": 1.5778,
+      "step": 899
+    },
+    {
+      "epoch": 0.06271121485559,
+      "grad_norm": 0.7528424466845036,
+      "learning_rate": 6.939060945774062e-07,
+      "loss": 1.5573,
+      "step": 900
+    },
+    {
+      "epoch": 0.06278089398320733,
+      "grad_norm": 0.6960916907060996,
+      "learning_rate": 6.938925889969913e-07,
+      "loss": 1.5107,
+      "step": 901
+    },
+    {
+      "epoch": 0.06285057311082465,
+      "grad_norm": 0.741434891143874,
+      "learning_rate": 6.938790686138093e-07,
+      "loss": 1.4976,
+      "step": 902
+    },
+    {
+      "epoch": 0.06292025223844197,
+      "grad_norm": 0.7758276506408398,
+      "learning_rate": 6.938655334285084e-07,
+      "loss": 1.7011,
+      "step": 903
+    },
+    {
+      "epoch": 0.06298993136605929,
+      "grad_norm": 0.7923740403380025,
+      "learning_rate": 6.938519834417369e-07,
+      "loss": 1.6621,
+      "step": 904
+    },
+    {
+      "epoch": 0.06305961049367662,
+      "grad_norm": 0.6856901148311233,
+      "learning_rate": 6.938384186541444e-07,
+      "loss": 1.6103,
+      "step": 905
+    },
+    {
+      "epoch": 0.06312928962129394,
+      "grad_norm": 0.702310601086839,
+      "learning_rate": 6.938248390663807e-07,
+      "loss": 1.5056,
+      "step": 906
+    },
+    {
+      "epoch": 0.06319896874891126,
+      "grad_norm": 0.7810829024837994,
+      "learning_rate": 6.938112446790969e-07,
+      "loss": 1.5446,
+      "step": 907
+    },
+    {
+      "epoch": 0.06326864787652858,
+      "grad_norm": 0.7307945024026805,
+      "learning_rate": 6.937976354929442e-07,
+      "loss": 1.6072,
+      "step": 908
+    },
+    {
+      "epoch": 0.0633383270041459,
+      "grad_norm": 0.7100956099681816,
+      "learning_rate": 6.937840115085747e-07,
+      "loss": 1.5953,
+      "step": 909
+    },
+    {
+      "epoch": 0.06340800613176323,
+      "grad_norm": 0.7112718468289873,
+      "learning_rate": 6.937703727266416e-07,
+      "loss": 1.5293,
+      "step": 910
+    },
+    {
+      "epoch": 0.06347768525938055,
+      "grad_norm": 0.6940826349677348,
+      "learning_rate": 6.937567191477984e-07,
+      "loss": 1.52,
+      "step": 911
+    },
+    {
+      "epoch": 0.06354736438699787,
+      "grad_norm": 0.7615667933385247,
+      "learning_rate": 6.937430507726993e-07,
+      "loss": 1.5318,
+      "step": 912
+    },
+    {
+      "epoch": 0.0636170435146152,
+      "grad_norm": 0.7576711542574699,
+      "learning_rate": 6.937293676019993e-07,
+      "loss": 1.6737,
+      "step": 913
+    },
+    {
+      "epoch": 0.06368672264223252,
+      "grad_norm": 0.7429248277684565,
+      "learning_rate": 6.937156696363543e-07,
+      "loss": 1.6404,
+      "step": 914
+    },
+    {
+      "epoch": 0.06375640176984984,
+      "grad_norm": 0.696155637296429,
+      "learning_rate": 6.937019568764206e-07,
+      "loss": 1.5436,
+      "step": 915
+    },
+    {
+      "epoch": 0.06382608089746716,
+      "grad_norm": 0.6842590034765481,
+      "learning_rate": 6.936882293228554e-07,
+      "loss": 1.5392,
+      "step": 916
+    },
+    {
+      "epoch": 0.06389576002508449,
+      "grad_norm": 0.696787602226188,
+      "learning_rate": 6.936744869763163e-07,
+      "loss": 1.5912,
+      "step": 917
+    },
+    {
+      "epoch": 0.06396543915270181,
+      "grad_norm": 0.7528902720982941,
+      "learning_rate": 6.936607298374624e-07,
+      "loss": 1.6004,
+      "step": 918
+    },
+    {
+      "epoch": 0.06403511828031913,
+      "grad_norm": 0.6598951020939933,
+      "learning_rate": 6.936469579069525e-07,
+      "loss": 1.5006,
+      "step": 919
+    },
+    {
+      "epoch": 0.06410479740793645,
+      "grad_norm": 0.6731494738244983,
+      "learning_rate": 6.936331711854467e-07,
+      "loss": 1.4561,
+      "step": 920
+    },
+    {
+      "epoch": 0.06417447653555378,
+      "grad_norm": 0.6743817096436955,
+      "learning_rate": 6.936193696736058e-07,
+      "loss": 1.5704,
+      "step": 921
+    },
+    {
+      "epoch": 0.0642441556631711,
+      "grad_norm": 0.7051720087157666,
+      "learning_rate": 6.936055533720911e-07,
+      "loss": 1.4952,
+      "step": 922
+    },
+    {
+      "epoch": 0.06431383479078842,
+      "grad_norm": 0.7103238647333436,
+      "learning_rate": 6.935917222815648e-07,
+      "loss": 1.5239,
+      "step": 923
+    },
+    {
+      "epoch": 0.06438351391840574,
+      "grad_norm": 0.7235088782382487,
+      "learning_rate": 6.935778764026895e-07,
+      "loss": 1.6262,
+      "step": 924
+    },
+    {
+      "epoch": 0.06445319304602307,
+      "grad_norm": 0.6819119127698352,
+      "learning_rate": 6.935640157361289e-07,
+      "loss": 1.5787,
+      "step": 925
+    },
+    {
+      "epoch": 0.06452287217364039,
+      "grad_norm": 0.7374359593151948,
+      "learning_rate": 6.935501402825473e-07,
+      "loss": 1.4686,
+      "step": 926
+    },
+    {
+      "epoch": 0.06459255130125771,
+      "grad_norm": 0.7116900330828307,
+      "learning_rate": 6.935362500426095e-07,
+      "loss": 1.5711,
+      "step": 927
+    },
+    {
+      "epoch": 0.06466223042887503,
+      "grad_norm": 0.6973009589376816,
+      "learning_rate": 6.935223450169812e-07,
+      "loss": 1.6242,
+      "step": 928
+    },
+    {
+      "epoch": 0.06473190955649236,
+      "grad_norm": 0.721322375011435,
+      "learning_rate": 6.935084252063286e-07,
+      "loss": 1.5486,
+      "step": 929
+    },
+    {
+      "epoch": 0.06480158868410968,
+      "grad_norm": 0.6756510474295269,
+      "learning_rate": 6.934944906113191e-07,
+      "loss": 1.4674,
+      "step": 930
+    },
+    {
+      "epoch": 0.064871267811727,
+      "grad_norm": 0.7197319416399377,
+      "learning_rate": 6.934805412326201e-07,
+      "loss": 1.5464,
+      "step": 931
+    },
+    {
+      "epoch": 0.06494094693934432,
+      "grad_norm": 0.7561165970179851,
+      "learning_rate": 6.934665770709004e-07,
+      "loss": 1.4934,
+      "step": 932
+    },
+    {
+      "epoch": 0.06501062606696165,
+      "grad_norm": 0.7184946003658444,
+      "learning_rate": 6.93452598126829e-07,
+      "loss": 1.39,
+      "step": 933
+    },
+    {
+      "epoch": 0.06508030519457897,
+      "grad_norm": 0.7180786135022353,
+      "learning_rate": 6.934386044010759e-07,
+      "loss": 1.6113,
+      "step": 934
+    },
+    {
+      "epoch": 0.06514998432219629,
+      "grad_norm": 0.7796809708168528,
+      "learning_rate": 6.934245958943115e-07,
+      "loss": 1.6264,
+      "step": 935
+    },
+    {
+      "epoch": 0.06521966344981361,
+      "grad_norm": 0.725538535253128,
+      "learning_rate": 6.934105726072076e-07,
+      "loss": 1.5312,
+      "step": 936
+    },
+    {
+      "epoch": 0.06528934257743094,
+      "grad_norm": 0.7304382528288118,
+      "learning_rate": 6.933965345404356e-07,
+      "loss": 1.5362,
+      "step": 937
+    },
+    {
+      "epoch": 0.06535902170504826,
+      "grad_norm": 0.700294818967333,
+      "learning_rate": 6.933824816946687e-07,
+      "loss": 1.4986,
+      "step": 938
+    },
+    {
+      "epoch": 0.06542870083266558,
+      "grad_norm": 0.6997315562329343,
+      "learning_rate": 6.933684140705801e-07,
+      "loss": 1.4747,
+      "step": 939
+    },
+    {
+      "epoch": 0.06549837996028289,
+      "grad_norm": 0.7567625481144566,
+      "learning_rate": 6.933543316688441e-07,
+      "loss": 1.5696,
+      "step": 940
+    },
+    {
+      "epoch": 0.06556805908790021,
+      "grad_norm": 41.44909914705615,
+      "learning_rate": 6.933402344901354e-07,
+      "loss": 1.6127,
+      "step": 941
+    },
+    {
+      "epoch": 0.06563773821551754,
+      "grad_norm": 0.6679798181768342,
+      "learning_rate": 6.933261225351298e-07,
+      "loss": 1.585,
+      "step": 942
+    },
+    {
+      "epoch": 0.06570741734313486,
+      "grad_norm": 0.7000648730835969,
+      "learning_rate": 6.933119958045033e-07,
+      "loss": 1.6216,
+      "step": 943
+    },
+    {
+      "epoch": 0.06577709647075218,
+      "grad_norm": 0.7596264387177684,
+      "learning_rate": 6.93297854298933e-07,
+      "loss": 1.6523,
+      "step": 944
+    },
+    {
+      "epoch": 0.0658467755983695,
+      "grad_norm": 0.7193619041313796,
+      "learning_rate": 6.932836980190967e-07,
+      "loss": 1.5485,
+      "step": 945
+    },
+    {
+      "epoch": 0.06591645472598683,
+      "grad_norm": 0.6980772094913987,
+      "learning_rate": 6.932695269656726e-07,
+      "loss": 1.5319,
+      "step": 946
+    },
+    {
+      "epoch": 0.06598613385360415,
+      "grad_norm": 0.7442746561075959,
+      "learning_rate": 6.9325534113934e-07,
+      "loss": 1.5161,
+      "step": 947
+    },
+    {
+      "epoch": 0.06605581298122147,
+      "grad_norm": 0.6777835337823753,
+      "learning_rate": 6.932411405407785e-07,
+      "loss": 1.4636,
+      "step": 948
+    },
+    {
+      "epoch": 0.0661254921088388,
+      "grad_norm": 0.6727156937868579,
+      "learning_rate": 6.932269251706688e-07,
+      "loss": 1.5201,
+      "step": 949
+    },
+    {
+      "epoch": 0.06619517123645612,
+      "grad_norm": 0.7245330036989707,
+      "learning_rate": 6.932126950296921e-07,
+      "loss": 1.5595,
+      "step": 950
+    },
+    {
+      "epoch": 0.06626485036407344,
+      "grad_norm": 0.7125701635864038,
+      "learning_rate": 6.931984501185303e-07,
+      "loss": 1.6323,
+      "step": 951
+    },
+    {
+      "epoch": 0.06633452949169076,
+      "grad_norm": 0.6963101718771975,
+      "learning_rate": 6.93184190437866e-07,
+      "loss": 1.513,
+      "step": 952
+    },
+    {
+      "epoch": 0.06640420861930808,
+      "grad_norm": 0.7046379018841911,
+      "learning_rate": 6.931699159883825e-07,
+      "loss": 1.5284,
+      "step": 953
+    },
+    {
+      "epoch": 0.0664738877469254,
+      "grad_norm": 0.678439137321281,
+      "learning_rate": 6.931556267707642e-07,
+      "loss": 1.553,
+      "step": 954
+    },
+    {
+      "epoch": 0.06654356687454273,
+      "grad_norm": 0.7410253132305197,
+      "learning_rate": 6.931413227856954e-07,
+      "loss": 1.5533,
+      "step": 955
+    },
+    {
+      "epoch": 0.06661324600216005,
+      "grad_norm": 0.6908842566143348,
+      "learning_rate": 6.93127004033862e-07,
+      "loss": 1.6341,
+      "step": 956
+    },
+    {
+      "epoch": 0.06668292512977737,
+      "grad_norm": 0.6651535227538429,
+      "learning_rate": 6.931126705159499e-07,
+      "loss": 1.5806,
+      "step": 957
+    },
+    {
+      "epoch": 0.0667526042573947,
+      "grad_norm": 0.6920497807137774,
+      "learning_rate": 6.930983222326462e-07,
+      "loss": 1.567,
+      "step": 958
+    },
+    {
+      "epoch": 0.06682228338501202,
+      "grad_norm": 0.7078371963892504,
+      "learning_rate": 6.930839591846383e-07,
+      "loss": 1.6668,
+      "step": 959
+    },
+    {
+      "epoch": 0.06689196251262934,
+      "grad_norm": 0.7606381115394008,
+      "learning_rate": 6.930695813726146e-07,
+      "loss": 1.4752,
+      "step": 960
+    },
+    {
+      "epoch": 0.06696164164024666,
+      "grad_norm": 0.8536717759212689,
+      "learning_rate": 6.93055188797264e-07,
+      "loss": 1.5652,
+      "step": 961
+    },
+    {
+      "epoch": 0.06703132076786399,
+      "grad_norm": 0.6653573860861989,
+      "learning_rate": 6.930407814592765e-07,
+      "loss": 1.5914,
+      "step": 962
+    },
+    {
+      "epoch": 0.06710099989548131,
+      "grad_norm": 0.784194439397854,
+      "learning_rate": 6.930263593593424e-07,
+      "loss": 1.5406,
+      "step": 963
+    },
+    {
+      "epoch": 0.06717067902309863,
+      "grad_norm": 0.7610703370050599,
+      "learning_rate": 6.930119224981526e-07,
+      "loss": 1.6075,
+      "step": 964
+    },
+    {
+      "epoch": 0.06724035815071595,
+      "grad_norm": 0.6764485214707169,
+      "learning_rate": 6.929974708763992e-07,
+      "loss": 1.504,
+      "step": 965
+    },
+    {
+      "epoch": 0.06731003727833328,
+      "grad_norm": 0.707512885228743,
+      "learning_rate": 6.929830044947746e-07,
+      "loss": 1.4592,
+      "step": 966
+    },
+    {
+      "epoch": 0.0673797164059506,
+      "grad_norm": 0.7081374537556824,
+      "learning_rate": 6.929685233539723e-07,
+      "loss": 1.6566,
+      "step": 967
+    },
+    {
+      "epoch": 0.06744939553356792,
+      "grad_norm": 0.7122482694793018,
+      "learning_rate": 6.929540274546861e-07,
+      "loss": 1.6348,
+      "step": 968
+    },
+    {
+      "epoch": 0.06751907466118524,
+      "grad_norm": 0.6901522900821034,
+      "learning_rate": 6.929395167976105e-07,
+      "loss": 1.5057,
+      "step": 969
+    },
+    {
+      "epoch": 0.06758875378880257,
+      "grad_norm": 0.7160908180373641,
+      "learning_rate": 6.929249913834413e-07,
+      "loss": 1.5376,
+      "step": 970
+    },
+    {
+      "epoch": 0.06765843291641989,
+      "grad_norm": 0.7388489184952017,
+      "learning_rate": 6.929104512128743e-07,
+      "loss": 1.4687,
+      "step": 971
+    },
+    {
+      "epoch": 0.06772811204403721,
+      "grad_norm": 0.7003909631678097,
+      "learning_rate": 6.928958962866063e-07,
+      "loss": 1.6329,
+      "step": 972
+    },
+    {
+      "epoch": 0.06779779117165453,
+      "grad_norm": 0.6912891435110838,
+      "learning_rate": 6.928813266053349e-07,
+      "loss": 1.4846,
+      "step": 973
+    },
+    {
+      "epoch": 0.06786747029927186,
+      "grad_norm": 0.6603982078975053,
+      "learning_rate": 6.928667421697582e-07,
+      "loss": 1.3881,
+      "step": 974
+    },
+    {
+      "epoch": 0.06793714942688918,
+      "grad_norm": 0.7235369345078301,
+      "learning_rate": 6.928521429805752e-07,
+      "loss": 1.6745,
+      "step": 975
+    },
+    {
+      "epoch": 0.0680068285545065,
+      "grad_norm": 0.7482907027017797,
+      "learning_rate": 6.928375290384856e-07,
+      "loss": 1.5114,
+      "step": 976
+    },
+    {
+      "epoch": 0.06807650768212382,
+      "grad_norm": 0.786232207346211,
+      "learning_rate": 6.928229003441894e-07,
+      "loss": 1.5673,
+      "step": 977
+    },
+    {
+      "epoch": 0.06814618680974115,
+      "grad_norm": 0.6986408216965335,
+      "learning_rate": 6.928082568983882e-07,
+      "loss": 1.6475,
+      "step": 978
+    },
+    {
+      "epoch": 0.06821586593735847,
+      "grad_norm": 0.7219767354753639,
+      "learning_rate": 6.927935987017831e-07,
+      "loss": 1.4191,
+      "step": 979
+    },
+    {
+      "epoch": 0.06828554506497579,
+      "grad_norm": 0.7733751013809093,
+      "learning_rate": 6.927789257550769e-07,
+      "loss": 1.4901,
+      "step": 980
+    },
+    {
+      "epoch": 0.06835522419259311,
+      "grad_norm": 0.7088273005920183,
+      "learning_rate": 6.927642380589728e-07,
+      "loss": 1.5104,
+      "step": 981
+    },
+    {
+      "epoch": 0.06842490332021044,
+      "grad_norm": 0.7231089020272345,
+      "learning_rate": 6.927495356141747e-07,
+      "loss": 1.565,
+      "step": 982
+    },
+    {
+      "epoch": 0.06849458244782776,
+      "grad_norm": 0.7465251630697766,
+      "learning_rate": 6.927348184213869e-07,
+      "loss": 1.6297,
+      "step": 983
+    },
+    {
+      "epoch": 0.06856426157544508,
+      "grad_norm": 0.7028021616887362,
+      "learning_rate": 6.927200864813149e-07,
+      "loss": 1.5676,
+      "step": 984
+    },
+    {
+      "epoch": 0.0686339407030624,
+      "grad_norm": 0.7241553816634947,
+      "learning_rate": 6.927053397946644e-07,
+      "loss": 1.5426,
+      "step": 985
+    },
+    {
+      "epoch": 0.06870361983067971,
+      "grad_norm": 0.6631488358648268,
+      "learning_rate": 6.926905783621427e-07,
+      "loss": 1.426,
+      "step": 986
+    },
+    {
+      "epoch": 0.06877329895829704,
+      "grad_norm": 0.7238485756150155,
+      "learning_rate": 6.926758021844565e-07,
+      "loss": 1.6008,
+      "step": 987
+    },
+    {
+      "epoch": 0.06884297808591436,
+      "grad_norm": 0.6720576579169943,
+      "learning_rate": 6.926610112623144e-07,
+      "loss": 1.4347,
+      "step": 988
+    },
+    {
+      "epoch": 0.06891265721353168,
+      "grad_norm": 0.73334548757339,
+      "learning_rate": 6.926462055964249e-07,
+      "loss": 1.5721,
+      "step": 989
+    },
+    {
+      "epoch": 0.068982336341149,
+      "grad_norm": 0.7137428718304452,
+      "learning_rate": 6.926313851874977e-07,
+      "loss": 1.6358,
+      "step": 990
+    },
+    {
+      "epoch": 0.06905201546876633,
+      "grad_norm": 0.7729856075932263,
+      "learning_rate": 6.92616550036243e-07,
+      "loss": 1.5841,
+      "step": 991
+    },
+    {
+      "epoch": 0.06912169459638365,
+      "grad_norm": 0.720906050858157,
+      "learning_rate": 6.926017001433716e-07,
+      "loss": 1.5314,
+      "step": 992
+    },
+    {
+      "epoch": 0.06919137372400097,
+      "grad_norm": 0.7376452297137691,
+      "learning_rate": 6.925868355095953e-07,
+      "loss": 1.5631,
+      "step": 993
+    },
+    {
+      "epoch": 0.0692610528516183,
+      "grad_norm": 0.7351626865087098,
+      "learning_rate": 6.925719561356263e-07,
+      "loss": 1.4988,
+      "step": 994
+    },
+    {
+      "epoch": 0.06933073197923562,
+      "grad_norm": 0.6997542125412789,
+      "learning_rate": 6.925570620221779e-07,
+      "loss": 1.5988,
+      "step": 995
+    },
+    {
+      "epoch": 0.06940041110685294,
+      "grad_norm": 0.706269039215262,
+      "learning_rate": 6.925421531699636e-07,
+      "loss": 1.5851,
+      "step": 996
+    },
+    {
+      "epoch": 0.06947009023447026,
+      "grad_norm": 0.6832486756386302,
+      "learning_rate": 6.925272295796979e-07,
+      "loss": 1.4754,
+      "step": 997
+    },
+    {
+      "epoch": 0.06953976936208758,
+      "grad_norm": 0.7294064047869246,
+      "learning_rate": 6.92512291252096e-07,
+      "loss": 1.5658,
+      "step": 998
+    },
+    {
+      "epoch": 0.0696094484897049,
+      "grad_norm": 0.7209483681033394,
+      "learning_rate": 6.924973381878738e-07,
+      "loss": 1.4977,
+      "step": 999
+    },
+    {
+      "epoch": 0.06967912761732223,
+      "grad_norm": 0.6482722642545093,
+      "learning_rate": 6.92482370387748e-07,
+      "loss": 1.5456,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06974880674493955,
+      "grad_norm": 0.6911923296637908,
+      "learning_rate": 6.924673878524356e-07,
+      "loss": 1.5774,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06981848587255687,
+      "grad_norm": 0.6806494876420273,
+      "learning_rate": 6.924523905826549e-07,
+      "loss": 1.6226,
+      "step": 1002
+    },
+    {
+      "epoch": 0.0698881650001742,
+      "grad_norm": 0.7206441347495618,
+      "learning_rate": 6.924373785791244e-07,
+      "loss": 1.5996,
+      "step": 1003
+    },
+    {
+      "epoch": 0.06995784412779152,
+      "grad_norm": 0.6568982429262141,
+      "learning_rate": 6.924223518425635e-07,
+      "loss": 1.5396,
+      "step": 1004
+    },
+    {
+      "epoch": 0.07002752325540884,
+      "grad_norm": 0.718836906784285,
+      "learning_rate": 6.924073103736925e-07,
+      "loss": 1.4372,
+      "step": 1005
+    },
+    {
+      "epoch": 0.07009720238302616,
+      "grad_norm": 0.7071783051387326,
+      "learning_rate": 6.92392254173232e-07,
+      "loss": 1.5001,
+      "step": 1006
+    },
+    {
+      "epoch": 0.07016688151064349,
+      "grad_norm": 0.722945628224867,
+      "learning_rate": 6.923771832419036e-07,
+      "loss": 1.691,
+      "step": 1007
+    },
+    {
+      "epoch": 0.07023656063826081,
+      "grad_norm": 0.783086373789463,
+      "learning_rate": 6.923620975804296e-07,
+      "loss": 1.568,
+      "step": 1008
+    },
+    {
+      "epoch": 0.07030623976587813,
+      "grad_norm": 0.6976889026374747,
+      "learning_rate": 6.923469971895328e-07,
+      "loss": 1.6341,
+      "step": 1009
+    },
+    {
+      "epoch": 0.07037591889349545,
+      "grad_norm": 0.7523424135440696,
+      "learning_rate": 6.923318820699369e-07,
+      "loss": 1.5886,
+      "step": 1010
+    },
+    {
+      "epoch": 0.07044559802111278,
+      "grad_norm": 0.7173214160819936,
+      "learning_rate": 6.923167522223664e-07,
+      "loss": 1.6562,
+      "step": 1011
+    },
+    {
+      "epoch": 0.0705152771487301,
+      "grad_norm": 0.7123339152044518,
+      "learning_rate": 6.923016076475462e-07,
+      "loss": 1.5676,
+      "step": 1012
+    },
+    {
+      "epoch": 0.07058495627634742,
+      "grad_norm": 0.7333638244874635,
+      "learning_rate": 6.92286448346202e-07,
+      "loss": 1.6151,
+      "step": 1013
+    },
+    {
+      "epoch": 0.07065463540396474,
+      "grad_norm": 0.7004839329730347,
+      "learning_rate": 6.922712743190605e-07,
+      "loss": 1.5125,
+      "step": 1014
+    },
+    {
+      "epoch": 0.07072431453158207,
+      "grad_norm": 0.744033569156811,
+      "learning_rate": 6.922560855668486e-07,
+      "loss": 1.579,
+      "step": 1015
+    },
+    {
+      "epoch": 0.07079399365919939,
+      "grad_norm": 0.6842826563152172,
+      "learning_rate": 6.922408820902942e-07,
+      "loss": 1.4839,
+      "step": 1016
+    },
+    {
+      "epoch": 0.07086367278681671,
+      "grad_norm": 0.6780606922456736,
+      "learning_rate": 6.922256638901262e-07,
+      "loss": 1.478,
+      "step": 1017
+    },
+    {
+      "epoch": 0.07093335191443403,
+      "grad_norm": 0.704981494096453,
+      "learning_rate": 6.922104309670736e-07,
+      "loss": 1.6119,
+      "step": 1018
+    },
+    {
+      "epoch": 0.07100303104205136,
+      "grad_norm": 0.7637589997063822,
+      "learning_rate": 6.921951833218664e-07,
+      "loss": 1.6386,
+      "step": 1019
+    },
+    {
+      "epoch": 0.07107271016966868,
+      "grad_norm": 0.6944249613797556,
+      "learning_rate": 6.921799209552354e-07,
+      "loss": 1.428,
+      "step": 1020
+    },
+    {
+      "epoch": 0.071142389297286,
+      "grad_norm": 0.6580765696200459,
+      "learning_rate": 6.921646438679119e-07,
+      "loss": 1.6064,
+      "step": 1021
+    },
+    {
+      "epoch": 0.07121206842490332,
+      "grad_norm": 0.6919362421338939,
+      "learning_rate": 6.921493520606281e-07,
+      "loss": 1.5615,
+      "step": 1022
+    },
+    {
+      "epoch": 0.07128174755252065,
+      "grad_norm": 0.6524789983946453,
+      "learning_rate": 6.921340455341168e-07,
+      "loss": 1.4955,
+      "step": 1023
+    },
+    {
+      "epoch": 0.07135142668013797,
+      "grad_norm": 0.7358174972537268,
+      "learning_rate": 6.921187242891115e-07,
+      "loss": 1.5882,
+      "step": 1024
+    },
+    {
+      "epoch": 0.07142110580775529,
+      "grad_norm": 0.7351799396329001,
+      "learning_rate": 6.921033883263464e-07,
+      "loss": 1.5962,
+      "step": 1025
+    },
+    {
+      "epoch": 0.07149078493537261,
+      "grad_norm": 0.7773506925419971,
+      "learning_rate": 6.920880376465565e-07,
+      "loss": 1.5903,
+      "step": 1026
+    },
+    {
+      "epoch": 0.07156046406298994,
+      "grad_norm": 0.720052376496024,
+      "learning_rate": 6.920726722504773e-07,
+      "loss": 1.5835,
+      "step": 1027
+    },
+    {
+      "epoch": 0.07163014319060726,
+      "grad_norm": 0.7122596855425408,
+      "learning_rate": 6.920572921388453e-07,
+      "loss": 1.4811,
+      "step": 1028
+    },
+    {
+      "epoch": 0.07169982231822458,
+      "grad_norm": 0.7530822747200974,
+      "learning_rate": 6.920418973123976e-07,
+      "loss": 1.6935,
+      "step": 1029
+    },
+    {
+      "epoch": 0.0717695014458419,
+      "grad_norm": 0.6738093485439884,
+      "learning_rate": 6.920264877718716e-07,
+      "loss": 1.5866,
+      "step": 1030
+    },
+    {
+      "epoch": 0.07183918057345923,
+      "grad_norm": 0.7484468258192855,
+      "learning_rate": 6.920110635180063e-07,
+      "loss": 1.6671,
+      "step": 1031
+    },
+    {
+      "epoch": 0.07190885970107654,
+      "grad_norm": 0.7419311209570713,
+      "learning_rate": 6.919956245515402e-07,
+      "loss": 1.6695,
+      "step": 1032
+    },
+    {
+      "epoch": 0.07197853882869386,
+      "grad_norm": 0.6826743849739504,
+      "learning_rate": 6.919801708732137e-07,
+      "loss": 1.5021,
+      "step": 1033
+    },
+    {
+      "epoch": 0.07204821795631118,
+      "grad_norm": 0.7000484130543231,
+      "learning_rate": 6.919647024837673e-07,
+      "loss": 1.5443,
+      "step": 1034
+    },
+    {
+      "epoch": 0.0721178970839285,
+      "grad_norm": 0.6954818658655855,
+      "learning_rate": 6.91949219383942e-07,
+      "loss": 1.4374,
+      "step": 1035
+    },
+    {
+      "epoch": 0.07218757621154583,
+      "grad_norm": 0.715842381283248,
+      "learning_rate": 6.9193372157448e-07,
+      "loss": 1.5712,
+      "step": 1036
+    },
+    {
+      "epoch": 0.07225725533916315,
+      "grad_norm": 0.6767161402869131,
+      "learning_rate": 6.919182090561241e-07,
+      "loss": 1.3916,
+      "step": 1037
+    },
+    {
+      "epoch": 0.07232693446678047,
+      "grad_norm": 0.6910566641237087,
+      "learning_rate": 6.919026818296173e-07,
+      "loss": 1.5389,
+      "step": 1038
+    },
+    {
+      "epoch": 0.0723966135943978,
+      "grad_norm": 0.7341954371107908,
+      "learning_rate": 6.91887139895704e-07,
+      "loss": 1.5189,
+      "step": 1039
+    },
+    {
+      "epoch": 0.07246629272201512,
+      "grad_norm": 0.7214373950468113,
+      "learning_rate": 6.91871583255129e-07,
+      "loss": 1.5847,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07253597184963244,
+      "grad_norm": 0.7014361915993079,
+      "learning_rate": 6.918560119086376e-07,
+      "loss": 1.4826,
+      "step": 1041
+    },
+    {
+      "epoch": 0.07260565097724976,
+      "grad_norm": 0.7704601014028651,
+      "learning_rate": 6.918404258569763e-07,
+      "loss": 1.468,
+      "step": 1042
+    },
+    {
+      "epoch": 0.07267533010486708,
+      "grad_norm": 0.808693366876018,
+      "learning_rate": 6.918248251008917e-07,
+      "loss": 1.6654,
+      "step": 1043
+    },
+    {
+      "epoch": 0.0727450092324844,
+      "grad_norm": 0.7110048932554214,
+      "learning_rate": 6.918092096411318e-07,
+      "loss": 1.5662,
+      "step": 1044
+    },
+    {
+      "epoch": 0.07281468836010173,
+      "grad_norm": 0.6982849324228914,
+      "learning_rate": 6.917935794784445e-07,
+      "loss": 1.5331,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07288436748771905,
+      "grad_norm": 0.7381550381399294,
+      "learning_rate": 6.917779346135791e-07,
+      "loss": 1.5555,
+      "step": 1046
+    },
+    {
+      "epoch": 0.07295404661533637,
+      "grad_norm": 0.704548114293804,
+      "learning_rate": 6.917622750472852e-07,
+      "loss": 1.6146,
+      "step": 1047
+    },
+    {
+      "epoch": 0.0730237257429537,
+      "grad_norm": 0.7042543974542429,
+      "learning_rate": 6.917466007803135e-07,
+      "loss": 1.3595,
+      "step": 1048
+    },
+    {
+      "epoch": 0.07309340487057102,
+      "grad_norm": 0.7267693860346068,
+      "learning_rate": 6.917309118134148e-07,
+      "loss": 1.6352,
+      "step": 1049
+    },
+    {
+      "epoch": 0.07316308399818834,
+      "grad_norm": 0.7426775788118576,
+      "learning_rate": 6.917152081473412e-07,
+      "loss": 1.6405,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07323276312580566,
+      "grad_norm": 0.7070458152085065,
+      "learning_rate": 6.91699489782845e-07,
+      "loss": 1.4703,
+      "step": 1051
+    },
+    {
+      "epoch": 0.07330244225342299,
+      "grad_norm": 0.7417121840086736,
+      "learning_rate": 6.916837567206797e-07,
+      "loss": 1.5257,
+      "step": 1052
+    },
+    {
+      "epoch": 0.07337212138104031,
+      "grad_norm": 0.7486626599689458,
+      "learning_rate": 6.916680089615992e-07,
+      "loss": 1.469,
+      "step": 1053
+    },
+    {
+      "epoch": 0.07344180050865763,
+      "grad_norm": 0.6854908185511666,
+      "learning_rate": 6.916522465063581e-07,
+      "loss": 1.487,
+      "step": 1054
+    },
+    {
+      "epoch": 0.07351147963627495,
+      "grad_norm": 0.7121319583710837,
+      "learning_rate": 6.916364693557117e-07,
+      "loss": 1.5364,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07358115876389228,
+      "grad_norm": 0.7287291289514267,
+      "learning_rate": 6.916206775104164e-07,
+      "loss": 1.5476,
+      "step": 1056
+    },
+    {
+      "epoch": 0.0736508378915096,
+      "grad_norm": 0.744317342873817,
+      "learning_rate": 6.916048709712286e-07,
+      "loss": 1.5123,
+      "step": 1057
+    },
+    {
+      "epoch": 0.07372051701912692,
+      "grad_norm": 0.7352301107365474,
+      "learning_rate": 6.915890497389059e-07,
+      "loss": 1.5221,
+      "step": 1058
+    },
+    {
+      "epoch": 0.07379019614674424,
+      "grad_norm": 0.667449801712433,
+      "learning_rate": 6.915732138142066e-07,
+      "loss": 1.5428,
+      "step": 1059
+    },
+    {
+      "epoch": 0.07385987527436157,
+      "grad_norm": 0.8240861662656697,
+      "learning_rate": 6.915573631978896e-07,
+      "loss": 1.5278,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07392955440197889,
+      "grad_norm": 0.8115926103137241,
+      "learning_rate": 6.915414978907143e-07,
+      "loss": 1.5511,
+      "step": 1061
+    },
+    {
+      "epoch": 0.07399923352959621,
+      "grad_norm": 0.7115918276627143,
+      "learning_rate": 6.915256178934411e-07,
+      "loss": 1.5539,
+      "step": 1062
+    },
+    {
+      "epoch": 0.07406891265721353,
+      "grad_norm": 0.7231534378390903,
+      "learning_rate": 6.915097232068309e-07,
+      "loss": 1.4308,
+      "step": 1063
+    },
+    {
+      "epoch": 0.07413859178483086,
+      "grad_norm": 0.6849196644707507,
+      "learning_rate": 6.914938138316456e-07,
+      "loss": 1.5336,
+      "step": 1064
+    },
+    {
+      "epoch": 0.07420827091244818,
+      "grad_norm": 0.712647033269892,
+      "learning_rate": 6.914778897686477e-07,
+      "loss": 1.4848,
+      "step": 1065
+    },
+    {
+      "epoch": 0.0742779500400655,
+      "grad_norm": 0.6841995015789953,
+      "learning_rate": 6.914619510185999e-07,
+      "loss": 1.5199,
+      "step": 1066
+    },
+    {
+      "epoch": 0.07434762916768282,
+      "grad_norm": 0.7032621096597862,
+      "learning_rate": 6.914459975822664e-07,
+      "loss": 1.5288,
+      "step": 1067
+    },
+    {
+      "epoch": 0.07441730829530015,
+      "grad_norm": 0.8064800483185524,
+      "learning_rate": 6.914300294604115e-07,
+      "loss": 1.5849,
+      "step": 1068
+    },
+    {
+      "epoch": 0.07448698742291747,
+      "grad_norm": 0.6616858768991769,
+      "learning_rate": 6.914140466538005e-07,
+      "loss": 1.5662,
+      "step": 1069
+    },
+    {
+      "epoch": 0.07455666655053479,
+      "grad_norm": 0.6981029713873405,
+      "learning_rate": 6.913980491631993e-07,
+      "loss": 1.5105,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07462634567815211,
+      "grad_norm": 0.7960162371795243,
+      "learning_rate": 6.913820369893746e-07,
+      "loss": 1.511,
+      "step": 1071
+    },
+    {
+      "epoch": 0.07469602480576944,
+      "grad_norm": 0.696388376902543,
+      "learning_rate": 6.913660101330937e-07,
+      "loss": 1.4846,
+      "step": 1072
+    },
+    {
+      "epoch": 0.07476570393338676,
+      "grad_norm": 0.7920768720605448,
+      "learning_rate": 6.913499685951247e-07,
+      "loss": 1.5168,
+      "step": 1073
+    },
+    {
+      "epoch": 0.07483538306100408,
+      "grad_norm": 0.694588502839473,
+      "learning_rate": 6.913339123762361e-07,
+      "loss": 1.5247,
+      "step": 1074
+    },
+    {
+      "epoch": 0.0749050621886214,
+      "grad_norm": 0.7466383220743256,
+      "learning_rate": 6.913178414771977e-07,
+      "loss": 1.6722,
+      "step": 1075
+    },
+    {
+      "epoch": 0.07497474131623873,
+      "grad_norm": 0.7403064737880765,
+      "learning_rate": 6.913017558987794e-07,
+      "loss": 1.5448,
+      "step": 1076
+    },
+    {
+      "epoch": 0.07504442044385605,
+      "grad_norm": 0.7088682271914373,
+      "learning_rate": 6.912856556417521e-07,
+      "loss": 1.518,
+      "step": 1077
+    },
+    {
+      "epoch": 0.07511409957147336,
+      "grad_norm": 0.7251852788383186,
+      "learning_rate": 6.912695407068875e-07,
+      "loss": 1.4666,
+      "step": 1078
+    },
+    {
+      "epoch": 0.07518377869909068,
+      "grad_norm": 0.7118843231495031,
+      "learning_rate": 6.912534110949577e-07,
+      "loss": 1.5608,
+      "step": 1079
+    },
+    {
+      "epoch": 0.075253457826708,
+      "grad_norm": 0.7373827534866939,
+      "learning_rate": 6.912372668067356e-07,
+      "loss": 1.56,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07532313695432533,
+      "grad_norm": 0.6652634128463273,
+      "learning_rate": 6.912211078429952e-07,
+      "loss": 1.4923,
+      "step": 1081
+    },
+    {
+      "epoch": 0.07539281608194265,
+      "grad_norm": 0.7210995227862197,
+      "learning_rate": 6.912049342045104e-07,
+      "loss": 1.6125,
+      "step": 1082
+    },
+    {
+      "epoch": 0.07546249520955997,
+      "grad_norm": 0.7873573632536712,
+      "learning_rate": 6.911887458920568e-07,
+      "loss": 1.6104,
+      "step": 1083
+    },
+    {
+      "epoch": 0.0755321743371773,
+      "grad_norm": 0.670639410874111,
+      "learning_rate": 6.911725429064096e-07,
+      "loss": 1.5491,
+      "step": 1084
+    },
+    {
+      "epoch": 0.07560185346479462,
+      "grad_norm": 0.7412530911154177,
+      "learning_rate": 6.911563252483458e-07,
+      "loss": 1.5383,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07567153259241194,
+      "grad_norm": 0.8068313122574107,
+      "learning_rate": 6.911400929186425e-07,
+      "loss": 1.5513,
+      "step": 1086
+    },
+    {
+      "epoch": 0.07574121172002926,
+      "grad_norm": 0.721225693696014,
+      "learning_rate": 6.911238459180772e-07,
+      "loss": 1.5331,
+      "step": 1087
+    },
+    {
+      "epoch": 0.07581089084764658,
+      "grad_norm": 0.7167014213265765,
+      "learning_rate": 6.911075842474287e-07,
+      "loss": 1.6393,
+      "step": 1088
+    },
+    {
+      "epoch": 0.0758805699752639,
+      "grad_norm": 0.7171069278949371,
+      "learning_rate": 6.910913079074765e-07,
+      "loss": 1.549,
+      "step": 1089
+    },
+    {
+      "epoch": 0.07595024910288123,
+      "grad_norm": 0.6934862794176397,
+      "learning_rate": 6.910750168990005e-07,
+      "loss": 1.4912,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07601992823049855,
+      "grad_norm": 0.7197521911287685,
+      "learning_rate": 6.910587112227811e-07,
+      "loss": 1.5344,
+      "step": 1091
+    },
+    {
+      "epoch": 0.07608960735811587,
+      "grad_norm": 0.7464831626664131,
+      "learning_rate": 6.910423908796001e-07,
+      "loss": 1.4993,
+      "step": 1092
+    },
+    {
+      "epoch": 0.0761592864857332,
+      "grad_norm": 0.7137983125771801,
+      "learning_rate": 6.910260558702393e-07,
+      "loss": 1.6077,
+      "step": 1093
+    },
+    {
+      "epoch": 0.07622896561335052,
+      "grad_norm": 0.7084434020338641,
+      "learning_rate": 6.910097061954817e-07,
+      "loss": 1.5826,
+      "step": 1094
+    },
+    {
+      "epoch": 0.07629864474096784,
+      "grad_norm": 0.6707576740728572,
+      "learning_rate": 6.909933418561109e-07,
+      "loss": 1.5207,
+      "step": 1095
+    },
+    {
+      "epoch": 0.07636832386858516,
+      "grad_norm": 0.701643205980676,
+      "learning_rate": 6.909769628529107e-07,
+      "loss": 1.5677,
+      "step": 1096
+    },
+    {
+      "epoch": 0.07643800299620249,
+      "grad_norm": 0.6960908138813413,
+      "learning_rate": 6.909605691866665e-07,
+      "loss": 1.5405,
+      "step": 1097
+    },
+    {
+      "epoch": 0.07650768212381981,
+      "grad_norm": 0.7409390432569176,
+      "learning_rate": 6.909441608581636e-07,
+      "loss": 1.6348,
+      "step": 1098
+    },
+    {
+      "epoch": 0.07657736125143713,
+      "grad_norm": 0.7330616230821628,
+      "learning_rate": 6.909277378681885e-07,
+      "loss": 1.6215,
+      "step": 1099
+    },
+    {
+      "epoch": 0.07664704037905445,
+      "grad_norm": 0.706598439534836,
+      "learning_rate": 6.909113002175281e-07,
+      "loss": 1.5915,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07671671950667178,
+      "grad_norm": 0.7230263429802867,
+      "learning_rate": 6.908948479069701e-07,
+      "loss": 1.5836,
+      "step": 1101
+    },
+    {
+      "epoch": 0.0767863986342891,
+      "grad_norm": 0.6871898211556535,
+      "learning_rate": 6.908783809373031e-07,
+      "loss": 1.5958,
+      "step": 1102
+    },
+    {
+      "epoch": 0.07685607776190642,
+      "grad_norm": 0.7139068952007671,
+      "learning_rate": 6.908618993093161e-07,
+      "loss": 1.6262,
+      "step": 1103
+    },
+    {
+      "epoch": 0.07692575688952374,
+      "grad_norm": 0.6569664827836058,
+      "learning_rate": 6.90845403023799e-07,
+      "loss": 1.5115,
+      "step": 1104
+    },
+    {
+      "epoch": 0.07699543601714107,
+      "grad_norm": 0.7253554862037589,
+      "learning_rate": 6.908288920815422e-07,
+      "loss": 1.529,
+      "step": 1105
+    },
+    {
+      "epoch": 0.07706511514475839,
+      "grad_norm": 0.6970591877647012,
+      "learning_rate": 6.90812366483337e-07,
+      "loss": 1.6525,
+      "step": 1106
+    },
+    {
+      "epoch": 0.07713479427237571,
+      "grad_norm": 0.6659396879644911,
+      "learning_rate": 6.907958262299755e-07,
+      "loss": 1.5162,
+      "step": 1107
+    },
+    {
+      "epoch": 0.07720447339999303,
+      "grad_norm": 0.6871713539255694,
+      "learning_rate": 6.907792713222501e-07,
+      "loss": 1.6047,
+      "step": 1108
+    },
+    {
+      "epoch": 0.07727415252761036,
+      "grad_norm": 0.7116167931927397,
+      "learning_rate": 6.907627017609543e-07,
+      "loss": 1.485,
+      "step": 1109
+    },
+    {
+      "epoch": 0.07734383165522768,
+      "grad_norm": 0.7053043605198956,
+      "learning_rate": 6.907461175468822e-07,
+      "loss": 1.4887,
+      "step": 1110
+    },
+    {
+      "epoch": 0.077413510782845,
+      "grad_norm": 0.6875986075874049,
+      "learning_rate": 6.907295186808285e-07,
+      "loss": 1.5854,
+      "step": 1111
+    },
+    {
+      "epoch": 0.07748318991046232,
+      "grad_norm": 0.7028170272289873,
+      "learning_rate": 6.907129051635885e-07,
+      "loss": 1.6719,
+      "step": 1112
+    },
+    {
+      "epoch": 0.07755286903807965,
+      "grad_norm": 0.7086162175094449,
+      "learning_rate": 6.906962769959585e-07,
+      "loss": 1.5725,
+      "step": 1113
+    },
+    {
+      "epoch": 0.07762254816569697,
+      "grad_norm": 0.7438672634911797,
+      "learning_rate": 6.906796341787353e-07,
+      "loss": 1.5896,
+      "step": 1114
+    },
+    {
+      "epoch": 0.07769222729331429,
+      "grad_norm": 0.6975872227510016,
+      "learning_rate": 6.906629767127165e-07,
+      "loss": 1.5404,
+      "step": 1115
+    },
+    {
+      "epoch": 0.07776190642093161,
+      "grad_norm": 0.7077339896057622,
+      "learning_rate": 6.906463045987004e-07,
+      "loss": 1.6565,
+      "step": 1116
+    },
+    {
+      "epoch": 0.07783158554854894,
+      "grad_norm": 0.7147410083240348,
+      "learning_rate": 6.906296178374858e-07,
+      "loss": 1.4631,
+      "step": 1117
+    },
+    {
+      "epoch": 0.07790126467616626,
+      "grad_norm": 0.7178738712503007,
+      "learning_rate": 6.906129164298726e-07,
+      "loss": 1.6519,
+      "step": 1118
+    },
+    {
+      "epoch": 0.07797094380378358,
+      "grad_norm": 0.7567807790555247,
+      "learning_rate": 6.905962003766609e-07,
+      "loss": 1.606,
+      "step": 1119
+    },
+    {
+      "epoch": 0.0780406229314009,
+      "grad_norm": 0.7677258840305423,
+      "learning_rate": 6.90579469678652e-07,
+      "loss": 1.514,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07811030205901823,
+      "grad_norm": 0.7107671421573318,
+      "learning_rate": 6.905627243366474e-07,
+      "loss": 1.7219,
+      "step": 1121
+    },
+    {
+      "epoch": 0.07817998118663555,
+      "grad_norm": 0.7435836249390709,
+      "learning_rate": 6.905459643514499e-07,
+      "loss": 1.5213,
+      "step": 1122
+    },
+    {
+      "epoch": 0.07824966031425287,
+      "grad_norm": 0.7182999889998696,
+      "learning_rate": 6.905291897238625e-07,
+      "loss": 1.5407,
+      "step": 1123
+    },
+    {
+      "epoch": 0.07831933944187018,
+      "grad_norm": 0.7416359447410312,
+      "learning_rate": 6.905124004546891e-07,
+      "loss": 1.5425,
+      "step": 1124
+    },
+    {
+      "epoch": 0.0783890185694875,
+      "grad_norm": 0.757556656682122,
+      "learning_rate": 6.904955965447342e-07,
+      "loss": 1.5452,
+      "step": 1125
+    },
+    {
+      "epoch": 0.07845869769710483,
+      "grad_norm": 0.6606122586088298,
+      "learning_rate": 6.904787779948031e-07,
+      "loss": 1.4316,
+      "step": 1126
+    },
+    {
+      "epoch": 0.07852837682472215,
+      "grad_norm": 0.7167556172667429,
+      "learning_rate": 6.90461944805702e-07,
+      "loss": 1.5075,
+      "step": 1127
+    },
+    {
+      "epoch": 0.07859805595233947,
+      "grad_norm": 0.641614834382918,
+      "learning_rate": 6.904450969782374e-07,
+      "loss": 1.4476,
+      "step": 1128
+    },
+    {
+      "epoch": 0.0786677350799568,
+      "grad_norm": 0.7465073966135141,
+      "learning_rate": 6.904282345132164e-07,
+      "loss": 1.6159,
+      "step": 1129
+    },
+    {
+      "epoch": 0.07873741420757412,
+      "grad_norm": 0.7399215393549937,
+      "learning_rate": 6.904113574114476e-07,
+      "loss": 1.64,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07880709333519144,
+      "grad_norm": 0.6855077939067492,
+      "learning_rate": 6.903944656737396e-07,
+      "loss": 1.4624,
+      "step": 1131
+    },
+    {
+      "epoch": 0.07887677246280876,
+      "grad_norm": 0.7251989381893763,
+      "learning_rate": 6.903775593009017e-07,
+      "loss": 1.5619,
+      "step": 1132
+    },
+    {
+      "epoch": 0.07894645159042608,
+      "grad_norm": 0.6915517066303697,
+      "learning_rate": 6.903606382937443e-07,
+      "loss": 1.4697,
+      "step": 1133
+    },
+    {
+      "epoch": 0.0790161307180434,
+      "grad_norm": 0.7016250746878615,
+      "learning_rate": 6.903437026530782e-07,
+      "loss": 1.4792,
+      "step": 1134
+    },
+    {
+      "epoch": 0.07908580984566073,
+      "grad_norm": 0.7219157858272937,
+      "learning_rate": 6.90326752379715e-07,
+      "loss": 1.6381,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07915548897327805,
+      "grad_norm": 0.7397081395480988,
+      "learning_rate": 6.90309787474467e-07,
+      "loss": 1.5577,
+      "step": 1136
+    },
+    {
+      "epoch": 0.07922516810089537,
+      "grad_norm": 0.7266051003891588,
+      "learning_rate": 6.902928079381473e-07,
+      "loss": 1.6519,
+      "step": 1137
+    },
+    {
+      "epoch": 0.0792948472285127,
+      "grad_norm": 0.7290609479762402,
+      "learning_rate": 6.902758137715693e-07,
+      "loss": 1.5796,
+      "step": 1138
+    },
+    {
+      "epoch": 0.07936452635613002,
+      "grad_norm": 0.6567695028044657,
+      "learning_rate": 6.902588049755478e-07,
+      "loss": 1.501,
+      "step": 1139
+    },
+    {
+      "epoch": 0.07943420548374734,
+      "grad_norm": 0.7077836191905298,
+      "learning_rate": 6.902417815508975e-07,
+      "loss": 1.5554,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07950388461136466,
+      "grad_norm": 0.640089811517191,
+      "learning_rate": 6.902247434984345e-07,
+      "loss": 1.4511,
+      "step": 1141
+    },
+    {
+      "epoch": 0.07957356373898199,
+      "grad_norm": 0.7397155755165485,
+      "learning_rate": 6.902076908189751e-07,
+      "loss": 1.6186,
+      "step": 1142
+    },
+    {
+      "epoch": 0.07964324286659931,
+      "grad_norm": 0.7370105780872295,
+      "learning_rate": 6.901906235133365e-07,
+      "loss": 1.5568,
+      "step": 1143
+    },
+    {
+      "epoch": 0.07971292199421663,
+      "grad_norm": 0.678857242792769,
+      "learning_rate": 6.90173541582337e-07,
+      "loss": 1.5696,
+      "step": 1144
+    },
+    {
+      "epoch": 0.07978260112183395,
+      "grad_norm": 0.7033520068281577,
+      "learning_rate": 6.901564450267946e-07,
+      "loss": 1.6165,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07985228024945128,
+      "grad_norm": 0.7137263676325256,
+      "learning_rate": 6.901393338475288e-07,
+      "loss": 1.6099,
+      "step": 1146
+    },
+    {
+      "epoch": 0.0799219593770686,
+      "grad_norm": 0.7144404342321975,
+      "learning_rate": 6.901222080453598e-07,
+      "loss": 1.5078,
+      "step": 1147
+    },
+    {
+      "epoch": 0.07999163850468592,
+      "grad_norm": 0.718330695095129,
+      "learning_rate": 6.901050676211082e-07,
+      "loss": 1.5648,
+      "step": 1148
+    },
+    {
+      "epoch": 0.08006131763230324,
+      "grad_norm": 0.7194005716639188,
+      "learning_rate": 6.900879125755955e-07,
+      "loss": 1.511,
+      "step": 1149
+    },
+    {
+      "epoch": 0.08013099675992057,
+      "grad_norm": 0.7678458971817134,
+      "learning_rate": 6.900707429096435e-07,
+      "loss": 1.5421,
+      "step": 1150
+    },
+    {
+      "epoch": 0.08020067588753789,
+      "grad_norm": 0.6985402111375227,
+      "learning_rate": 6.900535586240753e-07,
+      "loss": 1.539,
+      "step": 1151
+    },
+    {
+      "epoch": 0.08027035501515521,
+      "grad_norm": 0.6740543021552393,
+      "learning_rate": 6.900363597197144e-07,
+      "loss": 1.544,
+      "step": 1152
+    },
+    {
+      "epoch": 0.08034003414277253,
+      "grad_norm": 0.7044355580971693,
+      "learning_rate": 6.900191461973848e-07,
+      "loss": 1.5279,
+      "step": 1153
+    },
+    {
+      "epoch": 0.08040971327038986,
+      "grad_norm": 0.7132437628563113,
+      "learning_rate": 6.900019180579115e-07,
+      "loss": 1.5389,
+      "step": 1154
+    },
+    {
+      "epoch": 0.08047939239800718,
+      "grad_norm": 0.6845257281834255,
+      "learning_rate": 6.899846753021202e-07,
+      "loss": 1.576,
+      "step": 1155
+    },
+    {
+      "epoch": 0.0805490715256245,
+      "grad_norm": 0.7459578464387648,
+      "learning_rate": 6.899674179308371e-07,
+      "loss": 1.4598,
+      "step": 1156
+    },
+    {
+      "epoch": 0.08061875065324182,
+      "grad_norm": 0.7157920946244044,
+      "learning_rate": 6.899501459448892e-07,
+      "loss": 1.534,
+      "step": 1157
+    },
+    {
+      "epoch": 0.08068842978085915,
+      "grad_norm": 0.8475637704029476,
+      "learning_rate": 6.899328593451042e-07,
+      "loss": 1.5881,
+      "step": 1158
+    },
+    {
+      "epoch": 0.08075810890847647,
+      "grad_norm": 0.7428948647025614,
+      "learning_rate": 6.899155581323108e-07,
+      "loss": 1.5358,
+      "step": 1159
+    },
+    {
+      "epoch": 0.08082778803609379,
+      "grad_norm": 0.6865769571057071,
+      "learning_rate": 6.898982423073378e-07,
+      "loss": 1.5509,
+      "step": 1160
+    },
+    {
+      "epoch": 0.08089746716371111,
+      "grad_norm": 0.7284383625362557,
+      "learning_rate": 6.898809118710149e-07,
+      "loss": 1.6219,
+      "step": 1161
+    },
+    {
+      "epoch": 0.08096714629132844,
+      "grad_norm": 0.7222780249612296,
+      "learning_rate": 6.89863566824173e-07,
+      "loss": 1.6027,
+      "step": 1162
+    },
+    {
+      "epoch": 0.08103682541894576,
+      "grad_norm": 0.7075963278835353,
+      "learning_rate": 6.89846207167643e-07,
+      "loss": 1.5669,
+      "step": 1163
+    },
+    {
+      "epoch": 0.08110650454656308,
+      "grad_norm": 0.7514979660782378,
+      "learning_rate": 6.89828832902257e-07,
+      "loss": 1.6114,
+      "step": 1164
+    },
+    {
+      "epoch": 0.0811761836741804,
+      "grad_norm": 0.6928840372383243,
+      "learning_rate": 6.898114440288475e-07,
+      "loss": 1.5779,
+      "step": 1165
+    },
+    {
+      "epoch": 0.08124586280179773,
+      "grad_norm": 0.7625255343114417,
+      "learning_rate": 6.897940405482478e-07,
+      "loss": 1.5837,
+      "step": 1166
+    },
+    {
+      "epoch": 0.08131554192941505,
+      "grad_norm": 0.6853653653010325,
+      "learning_rate": 6.897766224612919e-07,
+      "loss": 1.5777,
+      "step": 1167
+    },
+    {
+      "epoch": 0.08138522105703237,
+      "grad_norm": 0.7381018328852454,
+      "learning_rate": 6.897591897688147e-07,
+      "loss": 1.5864,
+      "step": 1168
+    },
+    {
+      "epoch": 0.0814549001846497,
+      "grad_norm": 0.6959768347113379,
+      "learning_rate": 6.897417424716513e-07,
+      "loss": 1.6115,
+      "step": 1169
+    },
+    {
+      "epoch": 0.081524579312267,
+      "grad_norm": 0.6895660442037262,
+      "learning_rate": 6.897242805706381e-07,
+      "loss": 1.5186,
+      "step": 1170
+    },
+    {
+      "epoch": 0.08159425843988433,
+      "grad_norm": 0.7519857537232603,
+      "learning_rate": 6.897068040666117e-07,
+      "loss": 1.5396,
+      "step": 1171
+    },
+    {
+      "epoch": 0.08166393756750165,
+      "grad_norm": 0.718532609238401,
+      "learning_rate": 6.896893129604098e-07,
+      "loss": 1.5395,
+      "step": 1172
+    },
+    {
+      "epoch": 0.08173361669511897,
+      "grad_norm": 0.7399091934568263,
+      "learning_rate": 6.896718072528704e-07,
+      "loss": 1.621,
+      "step": 1173
+    },
+    {
+      "epoch": 0.0818032958227363,
+      "grad_norm": 0.6925568720224577,
+      "learning_rate": 6.896542869448327e-07,
+      "loss": 1.5752,
+      "step": 1174
+    },
+    {
+      "epoch": 0.08187297495035362,
+      "grad_norm": 0.7067599488817029,
+      "learning_rate": 6.896367520371359e-07,
+      "loss": 1.5278,
+      "step": 1175
+    },
+    {
+      "epoch": 0.08194265407797094,
+      "grad_norm": 0.7103145143633663,
+      "learning_rate": 6.896192025306206e-07,
+      "loss": 1.6391,
+      "step": 1176
+    },
+    {
+      "epoch": 0.08201233320558826,
+      "grad_norm": 0.7734153160510779,
+      "learning_rate": 6.896016384261277e-07,
+      "loss": 1.5156,
+      "step": 1177
+    },
+    {
+      "epoch": 0.08208201233320558,
+      "grad_norm": 0.6859669641724797,
+      "learning_rate": 6.895840597244991e-07,
+      "loss": 1.5193,
+      "step": 1178
+    },
+    {
+      "epoch": 0.0821516914608229,
+      "grad_norm": 0.673186710234392,
+      "learning_rate": 6.89566466426577e-07,
+      "loss": 1.3677,
+      "step": 1179
+    },
+    {
+      "epoch": 0.08222137058844023,
+      "grad_norm": 0.7640001453836751,
+      "learning_rate": 6.895488585332045e-07,
+      "loss": 1.6201,
+      "step": 1180
+    },
+    {
+      "epoch": 0.08229104971605755,
+      "grad_norm": 0.7061881267193482,
+      "learning_rate": 6.895312360452255e-07,
+      "loss": 1.5405,
+      "step": 1181
+    },
+    {
+      "epoch": 0.08236072884367487,
+      "grad_norm": 0.7576495015869381,
+      "learning_rate": 6.895135989634845e-07,
+      "loss": 1.6068,
+      "step": 1182
+    },
+    {
+      "epoch": 0.0824304079712922,
+      "grad_norm": 0.8049680580056225,
+      "learning_rate": 6.894959472888267e-07,
+      "loss": 1.6249,
+      "step": 1183
+    },
+    {
+      "epoch": 0.08250008709890952,
+      "grad_norm": 0.7441843872012226,
+      "learning_rate": 6.894782810220979e-07,
+      "loss": 1.6032,
+      "step": 1184
+    },
+    {
+      "epoch": 0.08256976622652684,
+      "grad_norm": 0.6633384020829896,
+      "learning_rate": 6.89460600164145e-07,
+      "loss": 1.5106,
+      "step": 1185
+    },
+    {
+      "epoch": 0.08263944535414416,
+      "grad_norm": 0.8483096471103987,
+      "learning_rate": 6.89442904715815e-07,
+      "loss": 1.5806,
+      "step": 1186
+    },
+    {
+      "epoch": 0.08270912448176149,
+      "grad_norm": 0.6666047476495981,
+      "learning_rate": 6.894251946779559e-07,
+      "loss": 1.442,
+      "step": 1187
+    },
+    {
+      "epoch": 0.08277880360937881,
+      "grad_norm": 0.7289293352898126,
+      "learning_rate": 6.894074700514165e-07,
+      "loss": 1.5653,
+      "step": 1188
+    },
+    {
+      "epoch": 0.08284848273699613,
+      "grad_norm": 0.7406788086107634,
+      "learning_rate": 6.893897308370463e-07,
+      "loss": 1.3979,
+      "step": 1189
+    },
+    {
+      "epoch": 0.08291816186461345,
+      "grad_norm": 0.6537157668650204,
+      "learning_rate": 6.893719770356951e-07,
+      "loss": 1.431,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08298784099223078,
+      "grad_norm": 0.7073784965936744,
+      "learning_rate": 6.893542086482141e-07,
+      "loss": 1.3773,
+      "step": 1191
+    },
+    {
+      "epoch": 0.0830575201198481,
+      "grad_norm": 26.82777241523961,
+      "learning_rate": 6.893364256754546e-07,
+      "loss": 1.5639,
+      "step": 1192
+    },
+    {
+      "epoch": 0.08312719924746542,
+      "grad_norm": 0.6623471303915018,
+      "learning_rate": 6.893186281182687e-07,
+      "loss": 1.5554,
+      "step": 1193
+    },
+    {
+      "epoch": 0.08319687837508274,
+      "grad_norm": 0.6809260214860811,
+      "learning_rate": 6.893008159775095e-07,
+      "loss": 1.4711,
+      "step": 1194
+    },
+    {
+      "epoch": 0.08326655750270007,
+      "grad_norm": 0.7157109529885805,
+      "learning_rate": 6.892829892540304e-07,
+      "loss": 1.4567,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08333623663031739,
+      "grad_norm": 0.7103122232716318,
+      "learning_rate": 6.892651479486856e-07,
+      "loss": 1.5585,
+      "step": 1196
+    },
+    {
+      "epoch": 0.08340591575793471,
+      "grad_norm": 0.7422514010124394,
+      "learning_rate": 6.892472920623305e-07,
+      "loss": 1.5075,
+      "step": 1197
+    },
+    {
+      "epoch": 0.08347559488555203,
+      "grad_norm": 0.6853896896273873,
+      "learning_rate": 6.892294215958206e-07,
+      "loss": 1.4262,
+      "step": 1198
+    },
+    {
+      "epoch": 0.08354527401316936,
+      "grad_norm": 0.706936413282032,
+      "learning_rate": 6.892115365500121e-07,
+      "loss": 1.5037,
+      "step": 1199
+    },
+    {
+      "epoch": 0.08361495314078668,
+      "grad_norm": 0.6749722628125229,
+      "learning_rate": 6.891936369257622e-07,
+      "loss": 1.668,
+      "step": 1200
+    },
+    {
+      "epoch": 0.083684632268404,
+      "grad_norm": 0.7119207196263206,
+      "learning_rate": 6.891757227239289e-07,
+      "loss": 1.6217,
+      "step": 1201
+    },
+    {
+      "epoch": 0.08375431139602132,
+      "grad_norm": 0.7160388119086022,
+      "learning_rate": 6.891577939453703e-07,
+      "loss": 1.4864,
+      "step": 1202
+    },
+    {
+      "epoch": 0.08382399052363865,
+      "grad_norm": 0.719587381416626,
+      "learning_rate": 6.891398505909459e-07,
+      "loss": 1.6268,
+      "step": 1203
+    },
+    {
+      "epoch": 0.08389366965125597,
+      "grad_norm": 0.7042976963968398,
+      "learning_rate": 6.891218926615155e-07,
+      "loss": 1.5495,
+      "step": 1204
+    },
+    {
+      "epoch": 0.08396334877887329,
+      "grad_norm": 0.704887862807588,
+      "learning_rate": 6.891039201579396e-07,
+      "loss": 1.5302,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08403302790649061,
+      "grad_norm": 0.6661954067466997,
+      "learning_rate": 6.890859330810796e-07,
+      "loss": 1.6247,
+      "step": 1206
+    },
+    {
+      "epoch": 0.08410270703410794,
+      "grad_norm": 0.6652903746555803,
+      "learning_rate": 6.890679314317973e-07,
+      "loss": 1.3775,
+      "step": 1207
+    },
+    {
+      "epoch": 0.08417238616172526,
+      "grad_norm": 0.6829149653109714,
+      "learning_rate": 6.890499152109555e-07,
+      "loss": 1.4859,
+      "step": 1208
+    },
+    {
+      "epoch": 0.08424206528934258,
+      "grad_norm": 0.7343751200022568,
+      "learning_rate": 6.890318844194177e-07,
+      "loss": 1.482,
+      "step": 1209
+    },
+    {
+      "epoch": 0.0843117444169599,
+      "grad_norm": 0.68418704479864,
+      "learning_rate": 6.890138390580477e-07,
+      "loss": 1.4719,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08438142354457723,
+      "grad_norm": 0.7197352748202582,
+      "learning_rate": 6.889957791277104e-07,
+      "loss": 1.6337,
+      "step": 1211
+    },
+    {
+      "epoch": 0.08445110267219455,
+      "grad_norm": 0.6649129082292412,
+      "learning_rate": 6.889777046292714e-07,
+      "loss": 1.6359,
+      "step": 1212
+    },
+    {
+      "epoch": 0.08452078179981187,
+      "grad_norm": 0.7587559411692023,
+      "learning_rate": 6.889596155635966e-07,
+      "loss": 1.5549,
+      "step": 1213
+    },
+    {
+      "epoch": 0.0845904609274292,
+      "grad_norm": 0.6806377337801913,
+      "learning_rate": 6.889415119315531e-07,
+      "loss": 1.496,
+      "step": 1214
+    },
+    {
+      "epoch": 0.08466014005504652,
+      "grad_norm": 0.776941258193978,
+      "learning_rate": 6.889233937340084e-07,
+      "loss": 1.5431,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08472981918266383,
+      "grad_norm": 0.6941217236555354,
+      "learning_rate": 6.889052609718306e-07,
+      "loss": 1.5759,
+      "step": 1216
+    },
+    {
+      "epoch": 0.08479949831028115,
+      "grad_norm": 0.7495506413349972,
+      "learning_rate": 6.888871136458888e-07,
+      "loss": 1.476,
+      "step": 1217
+    },
+    {
+      "epoch": 0.08486917743789847,
+      "grad_norm": 0.7426227429066157,
+      "learning_rate": 6.888689517570526e-07,
+      "loss": 1.4853,
+      "step": 1218
+    },
+    {
+      "epoch": 0.0849388565655158,
+      "grad_norm": 0.7340607548502726,
+      "learning_rate": 6.888507753061925e-07,
+      "loss": 1.5197,
+      "step": 1219
+    },
+    {
+      "epoch": 0.08500853569313312,
+      "grad_norm": 0.7516432731396169,
+      "learning_rate": 6.888325842941795e-07,
+      "loss": 1.5661,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08507821482075044,
+      "grad_norm": 0.690211007495237,
+      "learning_rate": 6.888143787218852e-07,
+      "loss": 1.4689,
+      "step": 1221
+    },
+    {
+      "epoch": 0.08514789394836776,
+      "grad_norm": 0.7315103283612352,
+      "learning_rate": 6.887961585901822e-07,
+      "loss": 1.5662,
+      "step": 1222
+    },
+    {
+      "epoch": 0.08521757307598508,
+      "grad_norm": 0.7556938120563503,
+      "learning_rate": 6.887779238999436e-07,
+      "loss": 1.4909,
+      "step": 1223
+    },
+    {
+      "epoch": 0.0852872522036024,
+      "grad_norm": 0.7979406139485317,
+      "learning_rate": 6.887596746520433e-07,
+      "loss": 1.6,
+      "step": 1224
+    },
+    {
+      "epoch": 0.08535693133121973,
+      "grad_norm": 0.7772957970517779,
+      "learning_rate": 6.887414108473558e-07,
+      "loss": 1.6837,
+      "step": 1225
+    },
+    {
+      "epoch": 0.08542661045883705,
+      "grad_norm": 0.7963373152044353,
+      "learning_rate": 6.887231324867562e-07,
+      "loss": 1.6797,
+      "step": 1226
+    },
+    {
+      "epoch": 0.08549628958645437,
+      "grad_norm": 0.6920973186236006,
+      "learning_rate": 6.887048395711207e-07,
+      "loss": 1.4087,
+      "step": 1227
+    },
+    {
+      "epoch": 0.0855659687140717,
+      "grad_norm": 0.6683888822025812,
+      "learning_rate": 6.886865321013255e-07,
+      "loss": 1.4964,
+      "step": 1228
+    },
+    {
+      "epoch": 0.08563564784168902,
+      "grad_norm": 0.7450329584753508,
+      "learning_rate": 6.886682100782485e-07,
+      "loss": 1.5268,
+      "step": 1229
+    },
+    {
+      "epoch": 0.08570532696930634,
+      "grad_norm": 0.6858244481120948,
+      "learning_rate": 6.886498735027673e-07,
+      "loss": 1.5805,
+      "step": 1230
+    },
+    {
+      "epoch": 0.08577500609692366,
+      "grad_norm": 0.7184719502519555,
+      "learning_rate": 6.88631522375761e-07,
+      "loss": 1.528,
+      "step": 1231
+    },
+    {
+      "epoch": 0.08584468522454099,
+      "grad_norm": 0.7760134696891086,
+      "learning_rate": 6.886131566981086e-07,
+      "loss": 1.5172,
+      "step": 1232
+    },
+    {
+      "epoch": 0.08591436435215831,
+      "grad_norm": 0.7105109119867613,
+      "learning_rate": 6.885947764706906e-07,
+      "loss": 1.532,
+      "step": 1233
+    },
+    {
+      "epoch": 0.08598404347977563,
+      "grad_norm": 0.7015865098791763,
+      "learning_rate": 6.885763816943875e-07,
+      "loss": 1.4887,
+      "step": 1234
+    },
+    {
+      "epoch": 0.08605372260739295,
+      "grad_norm": 0.6983077687068207,
+      "learning_rate": 6.885579723700809e-07,
+      "loss": 1.5157,
+      "step": 1235
+    },
+    {
+      "epoch": 0.08612340173501028,
+      "grad_norm": 0.7334632983600148,
+      "learning_rate": 6.885395484986532e-07,
+      "loss": 1.5426,
+      "step": 1236
+    },
+    {
+      "epoch": 0.0861930808626276,
+      "grad_norm": 0.693820895798714,
+      "learning_rate": 6.885211100809872e-07,
+      "loss": 1.4726,
+      "step": 1237
+    },
+    {
+      "epoch": 0.08626275999024492,
+      "grad_norm": 0.6706017501517098,
+      "learning_rate": 6.885026571179664e-07,
+      "loss": 1.553,
+      "step": 1238
+    },
+    {
+      "epoch": 0.08633243911786224,
+      "grad_norm": 0.6712469046406953,
+      "learning_rate": 6.884841896104753e-07,
+      "loss": 1.5283,
+      "step": 1239
+    },
+    {
+      "epoch": 0.08640211824547957,
+      "grad_norm": 0.7396947117151503,
+      "learning_rate": 6.884657075593987e-07,
+      "loss": 1.4582,
+      "step": 1240
+    },
+    {
+      "epoch": 0.08647179737309689,
+      "grad_norm": 0.7470920745785362,
+      "learning_rate": 6.884472109656224e-07,
+      "loss": 1.5739,
+      "step": 1241
+    },
+    {
+      "epoch": 0.08654147650071421,
+      "grad_norm": 5.5737115931864585,
+      "learning_rate": 6.884286998300328e-07,
+      "loss": 1.5824,
+      "step": 1242
+    },
+    {
+      "epoch": 0.08661115562833153,
+      "grad_norm": 0.6947509528605078,
+      "learning_rate": 6.88410174153517e-07,
+      "loss": 1.5973,
+      "step": 1243
+    },
+    {
+      "epoch": 0.08668083475594886,
+      "grad_norm": 0.7216219247735955,
+      "learning_rate": 6.883916339369627e-07,
+      "loss": 1.5319,
+      "step": 1244
+    },
+    {
+      "epoch": 0.08675051388356618,
+      "grad_norm": 0.7350172934004601,
+      "learning_rate": 6.883730791812585e-07,
+      "loss": 1.5269,
+      "step": 1245
+    },
+    {
+      "epoch": 0.0868201930111835,
+      "grad_norm": 0.6955094801968497,
+      "learning_rate": 6.883545098872935e-07,
+      "loss": 1.6165,
+      "step": 1246
+    },
+    {
+      "epoch": 0.08688987213880082,
+      "grad_norm": 0.7538227602808574,
+      "learning_rate": 6.883359260559576e-07,
+      "loss": 1.4654,
+      "step": 1247
+    },
+    {
+      "epoch": 0.08695955126641815,
+      "grad_norm": 0.789486381100046,
+      "learning_rate": 6.883173276881414e-07,
+      "loss": 1.5613,
+      "step": 1248
+    },
+    {
+      "epoch": 0.08702923039403547,
+      "grad_norm": 0.6975860457622158,
+      "learning_rate": 6.882987147847363e-07,
+      "loss": 1.5203,
+      "step": 1249
+    },
+    {
+      "epoch": 0.08709890952165279,
+      "grad_norm": 0.7033421302463916,
+      "learning_rate": 6.88280087346634e-07,
+      "loss": 1.5222,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08716858864927012,
+      "grad_norm": 0.7153412252377012,
+      "learning_rate": 6.882614453747272e-07,
+      "loss": 1.4739,
+      "step": 1251
+    },
+    {
+      "epoch": 0.08723826777688744,
+      "grad_norm": 0.694186348996605,
+      "learning_rate": 6.882427888699094e-07,
+      "loss": 1.5958,
+      "step": 1252
+    },
+    {
+      "epoch": 0.08730794690450476,
+      "grad_norm": 0.6852337145570081,
+      "learning_rate": 6.882241178330748e-07,
+      "loss": 1.545,
+      "step": 1253
+    },
+    {
+      "epoch": 0.08737762603212208,
+      "grad_norm": 0.7840072464679235,
+      "learning_rate": 6.882054322651178e-07,
+      "loss": 1.5576,
+      "step": 1254
+    },
+    {
+      "epoch": 0.0874473051597394,
+      "grad_norm": 0.7009252831810354,
+      "learning_rate": 6.88186732166934e-07,
+      "loss": 1.5604,
+      "step": 1255
+    },
+    {
+      "epoch": 0.08751698428735673,
+      "grad_norm": 0.7162091483897209,
+      "learning_rate": 6.881680175394195e-07,
+      "loss": 1.4449,
+      "step": 1256
+    },
+    {
+      "epoch": 0.08758666341497405,
+      "grad_norm": 0.6789452720548425,
+      "learning_rate": 6.881492883834714e-07,
+      "loss": 1.5185,
+      "step": 1257
+    },
+    {
+      "epoch": 0.08765634254259137,
+      "grad_norm": 0.75702264090507,
+      "learning_rate": 6.881305446999869e-07,
+      "loss": 1.5161,
+      "step": 1258
+    },
+    {
+      "epoch": 0.0877260216702087,
+      "grad_norm": 0.6837204332937512,
+      "learning_rate": 6.881117864898646e-07,
+      "loss": 1.5034,
+      "step": 1259
+    },
+    {
+      "epoch": 0.08779570079782602,
+      "grad_norm": 0.70633577212723,
+      "learning_rate": 6.88093013754003e-07,
+      "loss": 1.4734,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08786537992544333,
+      "grad_norm": 0.7236570341818808,
+      "learning_rate": 6.880742264933021e-07,
+      "loss": 1.5184,
+      "step": 1261
+    },
+    {
+      "epoch": 0.08793505905306065,
+      "grad_norm": 0.7415268607505546,
+      "learning_rate": 6.880554247086619e-07,
+      "loss": 1.6711,
+      "step": 1262
+    },
+    {
+      "epoch": 0.08800473818067797,
+      "grad_norm": 0.6770620482470922,
+      "learning_rate": 6.880366084009836e-07,
+      "loss": 1.5219,
+      "step": 1263
+    },
+    {
+      "epoch": 0.0880744173082953,
+      "grad_norm": 0.7006824285806458,
+      "learning_rate": 6.880177775711691e-07,
+      "loss": 1.4605,
+      "step": 1264
+    },
+    {
+      "epoch": 0.08814409643591262,
+      "grad_norm": 0.6941678012435419,
+      "learning_rate": 6.879989322201204e-07,
+      "loss": 1.5493,
+      "step": 1265
+    },
+    {
+      "epoch": 0.08821377556352994,
+      "grad_norm": 0.6932515587732107,
+      "learning_rate": 6.879800723487409e-07,
+      "loss": 1.596,
+      "step": 1266
+    },
+    {
+      "epoch": 0.08828345469114726,
+      "grad_norm": 0.727639027077372,
+      "learning_rate": 6.879611979579343e-07,
+      "loss": 1.6517,
+      "step": 1267
+    },
+    {
+      "epoch": 0.08835313381876458,
+      "grad_norm": 0.7187124776898827,
+      "learning_rate": 6.879423090486051e-07,
+      "loss": 1.4588,
+      "step": 1268
+    },
+    {
+      "epoch": 0.0884228129463819,
+      "grad_norm": 0.714150308857557,
+      "learning_rate": 6.879234056216587e-07,
+      "loss": 1.5832,
+      "step": 1269
+    },
+    {
+      "epoch": 0.08849249207399923,
+      "grad_norm": 0.7638975002556709,
+      "learning_rate": 6.879044876780006e-07,
+      "loss": 1.6047,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08856217120161655,
+      "grad_norm": 0.7434466148742814,
+      "learning_rate": 6.878855552185377e-07,
+      "loss": 1.6107,
+      "step": 1271
+    },
+    {
+      "epoch": 0.08863185032923387,
+      "grad_norm": 0.7399244720941883,
+      "learning_rate": 6.878666082441772e-07,
+      "loss": 1.4261,
+      "step": 1272
+    },
+    {
+      "epoch": 0.0887015294568512,
+      "grad_norm": 0.6842006315513699,
+      "learning_rate": 6.87847646755827e-07,
+      "loss": 1.4954,
+      "step": 1273
+    },
+    {
+      "epoch": 0.08877120858446852,
+      "grad_norm": 0.7245096057643684,
+      "learning_rate": 6.878286707543958e-07,
+      "loss": 1.5579,
+      "step": 1274
+    },
+    {
+      "epoch": 0.08884088771208584,
+      "grad_norm": 0.7879892971492194,
+      "learning_rate": 6.878096802407931e-07,
+      "loss": 1.5088,
+      "step": 1275
+    },
+    {
+      "epoch": 0.08891056683970316,
+      "grad_norm": 0.6868473383580004,
+      "learning_rate": 6.877906752159289e-07,
+      "loss": 1.3988,
+      "step": 1276
+    },
+    {
+      "epoch": 0.08898024596732049,
+      "grad_norm": 0.7437043498439738,
+      "learning_rate": 6.877716556807138e-07,
+      "loss": 1.6203,
+      "step": 1277
+    },
+    {
+      "epoch": 0.08904992509493781,
+      "grad_norm": 0.7058403378765666,
+      "learning_rate": 6.877526216360595e-07,
+      "loss": 1.5629,
+      "step": 1278
+    },
+    {
+      "epoch": 0.08911960422255513,
+      "grad_norm": 0.7342911778185349,
+      "learning_rate": 6.87733573082878e-07,
+      "loss": 1.617,
+      "step": 1279
+    },
+    {
+      "epoch": 0.08918928335017245,
+      "grad_norm": 0.7737218614982889,
+      "learning_rate": 6.877145100220821e-07,
+      "loss": 1.703,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08925896247778978,
+      "grad_norm": 0.7115474712008767,
+      "learning_rate": 6.876954324545854e-07,
+      "loss": 1.5477,
+      "step": 1281
+    },
+    {
+      "epoch": 0.0893286416054071,
+      "grad_norm": 0.8068670560532931,
+      "learning_rate": 6.876763403813022e-07,
+      "loss": 1.6857,
+      "step": 1282
+    },
+    {
+      "epoch": 0.08939832073302442,
+      "grad_norm": 0.7093087007583438,
+      "learning_rate": 6.876572338031475e-07,
+      "loss": 1.4789,
+      "step": 1283
+    },
+    {
+      "epoch": 0.08946799986064174,
+      "grad_norm": 0.7420277841762087,
+      "learning_rate": 6.876381127210368e-07,
+      "loss": 1.4069,
+      "step": 1284
+    },
+    {
+      "epoch": 0.08953767898825907,
+      "grad_norm": 1.9990965003575498,
+      "learning_rate": 6.876189771358862e-07,
+      "loss": 1.4755,
+      "step": 1285
+    },
+    {
+      "epoch": 0.08960735811587639,
+      "grad_norm": 0.6977842602151991,
+      "learning_rate": 6.875998270486131e-07,
+      "loss": 1.6219,
+      "step": 1286
+    },
+    {
+      "epoch": 0.08967703724349371,
+      "grad_norm": 0.6590490730475049,
+      "learning_rate": 6.875806624601351e-07,
+      "loss": 1.5421,
+      "step": 1287
+    },
+    {
+      "epoch": 0.08974671637111103,
+      "grad_norm": 0.6887642541353808,
+      "learning_rate": 6.875614833713706e-07,
+      "loss": 1.4732,
+      "step": 1288
+    },
+    {
+      "epoch": 0.08981639549872836,
+      "grad_norm": 0.7376671506965188,
+      "learning_rate": 6.875422897832385e-07,
+      "loss": 1.3695,
+      "step": 1289
+    },
+    {
+      "epoch": 0.08988607462634568,
+      "grad_norm": 0.7259016277612667,
+      "learning_rate": 6.875230816966589e-07,
+      "loss": 1.5346,
+      "step": 1290
+    },
+    {
+      "epoch": 0.089955753753963,
+      "grad_norm": 0.724086003625605,
+      "learning_rate": 6.875038591125522e-07,
+      "loss": 1.6004,
+      "step": 1291
+    },
+    {
+      "epoch": 0.09002543288158033,
+      "grad_norm": 0.7178196848691761,
+      "learning_rate": 6.874846220318393e-07,
+      "loss": 1.6231,
+      "step": 1292
+    },
+    {
+      "epoch": 0.09009511200919765,
+      "grad_norm": 0.7373207554001411,
+      "learning_rate": 6.874653704554426e-07,
+      "loss": 1.4717,
+      "step": 1293
+    },
+    {
+      "epoch": 0.09016479113681497,
+      "grad_norm": 0.7134230514137152,
+      "learning_rate": 6.874461043842843e-07,
+      "loss": 1.6591,
+      "step": 1294
+    },
+    {
+      "epoch": 0.09023447026443229,
+      "grad_norm": 0.7376105614130277,
+      "learning_rate": 6.874268238192877e-07,
+      "loss": 1.6336,
+      "step": 1295
+    },
+    {
+      "epoch": 0.09030414939204962,
+      "grad_norm": 0.7450088115019615,
+      "learning_rate": 6.874075287613769e-07,
+      "loss": 1.6974,
+      "step": 1296
+    },
+    {
+      "epoch": 0.09037382851966694,
+      "grad_norm": 0.6893786584174302,
+      "learning_rate": 6.873882192114765e-07,
+      "loss": 1.5151,
+      "step": 1297
+    },
+    {
+      "epoch": 0.09044350764728426,
+      "grad_norm": 0.6715073587311645,
+      "learning_rate": 6.873688951705119e-07,
+      "loss": 1.4657,
+      "step": 1298
+    },
+    {
+      "epoch": 0.09051318677490158,
+      "grad_norm": 0.7566996663979243,
+      "learning_rate": 6.873495566394089e-07,
+      "loss": 1.5942,
+      "step": 1299
+    },
+    {
+      "epoch": 0.0905828659025189,
+      "grad_norm": 0.7357953944743499,
+      "learning_rate": 6.873302036190946e-07,
+      "loss": 1.6023,
+      "step": 1300
+    },
+    {
+      "epoch": 0.09065254503013623,
+      "grad_norm": 0.7581630163955304,
+      "learning_rate": 6.873108361104963e-07,
+      "loss": 1.7109,
+      "step": 1301
+    },
+    {
+      "epoch": 0.09072222415775355,
+      "grad_norm": 0.7341133536278991,
+      "learning_rate": 6.872914541145419e-07,
+      "loss": 1.6366,
+      "step": 1302
+    },
+    {
+      "epoch": 0.09079190328537087,
+      "grad_norm": 0.6626232009933386,
+      "learning_rate": 6.872720576321606e-07,
+      "loss": 1.5272,
+      "step": 1303
+    },
+    {
+      "epoch": 0.0908615824129882,
+      "grad_norm": 0.7910328700731327,
+      "learning_rate": 6.872526466642818e-07,
+      "loss": 1.7656,
+      "step": 1304
+    },
+    {
+      "epoch": 0.09093126154060552,
+      "grad_norm": 0.7662975077969607,
+      "learning_rate": 6.872332212118355e-07,
+      "loss": 1.635,
+      "step": 1305
+    },
+    {
+      "epoch": 0.09100094066822284,
+      "grad_norm": 0.6887721459909045,
+      "learning_rate": 6.872137812757528e-07,
+      "loss": 1.523,
+      "step": 1306
+    },
+    {
+      "epoch": 0.09107061979584015,
+      "grad_norm": 0.712920618786257,
+      "learning_rate": 6.871943268569653e-07,
+      "loss": 1.6317,
+      "step": 1307
+    },
+    {
+      "epoch": 0.09114029892345747,
+      "grad_norm": 0.7577558050195798,
+      "learning_rate": 6.871748579564053e-07,
+      "loss": 1.5641,
+      "step": 1308
+    },
+    {
+      "epoch": 0.0912099780510748,
+      "grad_norm": 0.6982117743762234,
+      "learning_rate": 6.871553745750056e-07,
+      "loss": 1.4785,
+      "step": 1309
+    },
+    {
+      "epoch": 0.09127965717869212,
+      "grad_norm": 0.6991004630390641,
+      "learning_rate": 6.871358767137003e-07,
+      "loss": 1.5554,
+      "step": 1310
+    },
+    {
+      "epoch": 0.09134933630630944,
+      "grad_norm": 0.7314176603239776,
+      "learning_rate": 6.871163643734233e-07,
+      "loss": 1.6083,
+      "step": 1311
+    },
+    {
+      "epoch": 0.09141901543392676,
+      "grad_norm": 0.6877124905657467,
+      "learning_rate": 6.870968375551098e-07,
+      "loss": 1.5961,
+      "step": 1312
+    },
+    {
+      "epoch": 0.09148869456154408,
+      "grad_norm": 0.6395093045596821,
+      "learning_rate": 6.870772962596959e-07,
+      "loss": 1.5792,
+      "step": 1313
+    },
+    {
+      "epoch": 0.09155837368916141,
+      "grad_norm": 0.7082289797343458,
+      "learning_rate": 6.870577404881177e-07,
+      "loss": 1.5305,
+      "step": 1314
+    },
+    {
+      "epoch": 0.09162805281677873,
+      "grad_norm": 0.7492565372235473,
+      "learning_rate": 6.870381702413124e-07,
+      "loss": 1.6021,
+      "step": 1315
+    },
+    {
+      "epoch": 0.09169773194439605,
+      "grad_norm": 0.7328857828061993,
+      "learning_rate": 6.87018585520218e-07,
+      "loss": 1.5604,
+      "step": 1316
+    },
+    {
+      "epoch": 0.09176741107201337,
+      "grad_norm": 0.9248051912076805,
+      "learning_rate": 6.86998986325773e-07,
+      "loss": 1.5056,
+      "step": 1317
+    },
+    {
+      "epoch": 0.0918370901996307,
+      "grad_norm": 0.7133062641617752,
+      "learning_rate": 6.869793726589165e-07,
+      "loss": 1.5892,
+      "step": 1318
+    },
+    {
+      "epoch": 0.09190676932724802,
+      "grad_norm": 0.7847077630418262,
+      "learning_rate": 6.869597445205885e-07,
+      "loss": 1.5176,
+      "step": 1319
+    },
+    {
+      "epoch": 0.09197644845486534,
+      "grad_norm": 0.6839226291668064,
+      "learning_rate": 6.869401019117297e-07,
+      "loss": 1.4888,
+      "step": 1320
+    },
+    {
+      "epoch": 0.09204612758248266,
+      "grad_norm": 0.721028875233305,
+      "learning_rate": 6.869204448332812e-07,
+      "loss": 1.5445,
+      "step": 1321
+    },
+    {
+      "epoch": 0.09211580671009999,
+      "grad_norm": 0.7977990236065278,
+      "learning_rate": 6.869007732861853e-07,
+      "loss": 1.6349,
+      "step": 1322
+    },
+    {
+      "epoch": 0.09218548583771731,
+      "grad_norm": 0.726214409816211,
+      "learning_rate": 6.868810872713846e-07,
+      "loss": 1.5163,
+      "step": 1323
+    },
+    {
+      "epoch": 0.09225516496533463,
+      "grad_norm": 0.7781168458751725,
+      "learning_rate": 6.868613867898223e-07,
+      "loss": 1.585,
+      "step": 1324
+    },
+    {
+      "epoch": 0.09232484409295195,
+      "grad_norm": 0.718082280962953,
+      "learning_rate": 6.868416718424427e-07,
+      "loss": 1.5275,
+      "step": 1325
+    },
+    {
+      "epoch": 0.09239452322056928,
+      "grad_norm": 0.752359718631759,
+      "learning_rate": 6.868219424301905e-07,
+      "loss": 1.7251,
+      "step": 1326
+    },
+    {
+      "epoch": 0.0924642023481866,
+      "grad_norm": 0.7757472708282024,
+      "learning_rate": 6.868021985540112e-07,
+      "loss": 1.5877,
+      "step": 1327
+    },
+    {
+      "epoch": 0.09253388147580392,
+      "grad_norm": 0.7188829520974873,
+      "learning_rate": 6.867824402148509e-07,
+      "loss": 1.6991,
+      "step": 1328
+    },
+    {
+      "epoch": 0.09260356060342124,
+      "grad_norm": 0.6927468915057636,
+      "learning_rate": 6.867626674136566e-07,
+      "loss": 1.4599,
+      "step": 1329
+    },
+    {
+      "epoch": 0.09267323973103857,
+      "grad_norm": 0.7793706193864617,
+      "learning_rate": 6.867428801513757e-07,
+      "loss": 1.8769,
+      "step": 1330
+    },
+    {
+      "epoch": 0.09274291885865589,
+      "grad_norm": 0.7050947771615956,
+      "learning_rate": 6.867230784289566e-07,
+      "loss": 1.5947,
+      "step": 1331
+    },
+    {
+      "epoch": 0.09281259798627321,
+      "grad_norm": 0.7363007364951878,
+      "learning_rate": 6.867032622473481e-07,
+      "loss": 1.4746,
+      "step": 1332
+    },
+    {
+      "epoch": 0.09288227711389054,
+      "grad_norm": 0.7210920566358469,
+      "learning_rate": 6.866834316074999e-07,
+      "loss": 1.4997,
+      "step": 1333
+    },
+    {
+      "epoch": 0.09295195624150786,
+      "grad_norm": 0.694334164410711,
+      "learning_rate": 6.866635865103623e-07,
+      "loss": 1.5192,
+      "step": 1334
+    },
+    {
+      "epoch": 0.09302163536912518,
+      "grad_norm": 0.6826381175796876,
+      "learning_rate": 6.866437269568864e-07,
+      "loss": 1.5416,
+      "step": 1335
+    },
+    {
+      "epoch": 0.0930913144967425,
+      "grad_norm": 0.7094518106250833,
+      "learning_rate": 6.866238529480238e-07,
+      "loss": 1.6305,
+      "step": 1336
+    },
+    {
+      "epoch": 0.09316099362435983,
+      "grad_norm": 0.7451153132227982,
+      "learning_rate": 6.86603964484727e-07,
+      "loss": 1.6427,
+      "step": 1337
+    },
+    {
+      "epoch": 0.09323067275197715,
+      "grad_norm": 0.6944135717548009,
+      "learning_rate": 6.865840615679489e-07,
+      "loss": 1.5994,
+      "step": 1338
+    },
+    {
+      "epoch": 0.09330035187959447,
+      "grad_norm": 0.7324834912449787,
+      "learning_rate": 6.865641441986436e-07,
+      "loss": 1.5387,
+      "step": 1339
+    },
+    {
+      "epoch": 0.09337003100721179,
+      "grad_norm": 0.6871232434674317,
+      "learning_rate": 6.865442123777652e-07,
+      "loss": 1.4626,
+      "step": 1340
+    },
+    {
+      "epoch": 0.09343971013482912,
+      "grad_norm": 0.75076436213832,
+      "learning_rate": 6.865242661062692e-07,
+      "loss": 1.4723,
+      "step": 1341
+    },
+    {
+      "epoch": 0.09350938926244644,
+      "grad_norm": 0.6986478206384793,
+      "learning_rate": 6.865043053851113e-07,
+      "loss": 1.546,
+      "step": 1342
+    },
+    {
+      "epoch": 0.09357906839006376,
+      "grad_norm": 0.6773347909600681,
+      "learning_rate": 6.864843302152483e-07,
+      "loss": 1.5233,
+      "step": 1343
+    },
+    {
+      "epoch": 0.09364874751768108,
+      "grad_norm": 0.7256916010451714,
+      "learning_rate": 6.864643405976371e-07,
+      "loss": 1.5341,
+      "step": 1344
+    },
+    {
+      "epoch": 0.0937184266452984,
+      "grad_norm": 0.7156842660159166,
+      "learning_rate": 6.864443365332357e-07,
+      "loss": 1.4524,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09378810577291573,
+      "grad_norm": 0.717844637914586,
+      "learning_rate": 6.86424318023003e-07,
+      "loss": 1.5503,
+      "step": 1346
+    },
+    {
+      "epoch": 0.09385778490053305,
+      "grad_norm": 0.7134994406384746,
+      "learning_rate": 6.864042850678981e-07,
+      "loss": 1.5557,
+      "step": 1347
+    },
+    {
+      "epoch": 0.09392746402815037,
+      "grad_norm": 0.7060719200241936,
+      "learning_rate": 6.863842376688812e-07,
+      "loss": 1.6098,
+      "step": 1348
+    },
+    {
+      "epoch": 0.0939971431557677,
+      "grad_norm": 0.7133840642791894,
+      "learning_rate": 6.863641758269128e-07,
+      "loss": 1.6069,
+      "step": 1349
+    },
+    {
+      "epoch": 0.09406682228338502,
+      "grad_norm": 0.7419267085106452,
+      "learning_rate": 6.863440995429543e-07,
+      "loss": 1.5277,
+      "step": 1350
+    },
+    {
+      "epoch": 0.09413650141100234,
+      "grad_norm": 0.7320435951110261,
+      "learning_rate": 6.863240088179681e-07,
+      "loss": 1.5051,
+      "step": 1351
+    },
+    {
+      "epoch": 0.09420618053861966,
+      "grad_norm": 0.7189557968560032,
+      "learning_rate": 6.863039036529167e-07,
+      "loss": 1.5563,
+      "step": 1352
+    },
+    {
+      "epoch": 0.09427585966623697,
+      "grad_norm": 0.7258479005816894,
+      "learning_rate": 6.862837840487637e-07,
+      "loss": 1.7108,
+      "step": 1353
+    },
+    {
+      "epoch": 0.0943455387938543,
+      "grad_norm": 0.7093271738297767,
+      "learning_rate": 6.862636500064733e-07,
+      "loss": 1.561,
+      "step": 1354
+    },
+    {
+      "epoch": 0.09441521792147162,
+      "grad_norm": 0.7289165862673944,
+      "learning_rate": 6.862435015270102e-07,
+      "loss": 1.5443,
+      "step": 1355
+    },
+    {
+      "epoch": 0.09448489704908894,
+      "grad_norm": 0.728731968183907,
+      "learning_rate": 6.862233386113402e-07,
+      "loss": 1.4719,
+      "step": 1356
+    },
+    {
+      "epoch": 0.09455457617670626,
+      "grad_norm": 0.7598985533740218,
+      "learning_rate": 6.862031612604292e-07,
+      "loss": 1.5874,
+      "step": 1357
+    },
+    {
+      "epoch": 0.09462425530432358,
+      "grad_norm": 0.7489312592826469,
+      "learning_rate": 6.861829694752445e-07,
+      "loss": 1.5196,
+      "step": 1358
+    },
+    {
+      "epoch": 0.09469393443194091,
+      "grad_norm": 0.7115801847246932,
+      "learning_rate": 6.861627632567536e-07,
+      "loss": 1.4183,
+      "step": 1359
+    },
+    {
+      "epoch": 0.09476361355955823,
+      "grad_norm": 0.7286465327425591,
+      "learning_rate": 6.861425426059247e-07,
+      "loss": 1.5136,
+      "step": 1360
+    },
+    {
+      "epoch": 0.09483329268717555,
+      "grad_norm": 0.7282399514777784,
+      "learning_rate": 6.861223075237272e-07,
+      "loss": 1.4448,
+      "step": 1361
+    },
+    {
+      "epoch": 0.09490297181479287,
+      "grad_norm": 0.6552566952542627,
+      "learning_rate": 6.861020580111302e-07,
+      "loss": 1.5461,
+      "step": 1362
+    },
+    {
+      "epoch": 0.0949726509424102,
+      "grad_norm": 0.6940211142791631,
+      "learning_rate": 6.860817940691046e-07,
+      "loss": 1.4516,
+      "step": 1363
+    },
+    {
+      "epoch": 0.09504233007002752,
+      "grad_norm": 0.7633128378880454,
+      "learning_rate": 6.860615156986212e-07,
+      "loss": 1.6427,
+      "step": 1364
+    },
+    {
+      "epoch": 0.09511200919764484,
+      "grad_norm": 0.6982614518760314,
+      "learning_rate": 6.86041222900652e-07,
+      "loss": 1.4,
+      "step": 1365
+    },
+    {
+      "epoch": 0.09518168832526216,
+      "grad_norm": 0.7177285924240916,
+      "learning_rate": 6.860209156761693e-07,
+      "loss": 1.4115,
+      "step": 1366
+    },
+    {
+      "epoch": 0.09525136745287949,
+      "grad_norm": 0.6952426491388587,
+      "learning_rate": 6.860005940261464e-07,
+      "loss": 1.4289,
+      "step": 1367
+    },
+    {
+      "epoch": 0.09532104658049681,
+      "grad_norm": 0.6700157416887091,
+      "learning_rate": 6.859802579515571e-07,
+      "loss": 1.4947,
+      "step": 1368
+    },
+    {
+      "epoch": 0.09539072570811413,
+      "grad_norm": 0.7096025037508805,
+      "learning_rate": 6.859599074533759e-07,
+      "loss": 1.5841,
+      "step": 1369
+    },
+    {
+      "epoch": 0.09546040483573145,
+      "grad_norm": 0.6999665684061958,
+      "learning_rate": 6.85939542532578e-07,
+      "loss": 1.5309,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09553008396334878,
+      "grad_norm": 0.7132517833298145,
+      "learning_rate": 6.859191631901395e-07,
+      "loss": 1.6011,
+      "step": 1371
+    },
+    {
+      "epoch": 0.0955997630909661,
+      "grad_norm": 0.6828272290576067,
+      "learning_rate": 6.858987694270371e-07,
+      "loss": 1.5264,
+      "step": 1372
+    },
+    {
+      "epoch": 0.09566944221858342,
+      "grad_norm": 0.7158171608458893,
+      "learning_rate": 6.858783612442477e-07,
+      "loss": 1.5748,
+      "step": 1373
+    },
+    {
+      "epoch": 0.09573912134620075,
+      "grad_norm": 0.7506877164228436,
+      "learning_rate": 6.858579386427496e-07,
+      "loss": 1.5836,
+      "step": 1374
+    },
+    {
+      "epoch": 0.09580880047381807,
+      "grad_norm": 0.7367438933731758,
+      "learning_rate": 6.858375016235214e-07,
+      "loss": 1.6368,
+      "step": 1375
+    },
+    {
+      "epoch": 0.09587847960143539,
+      "grad_norm": 0.7285757842868972,
+      "learning_rate": 6.858170501875426e-07,
+      "loss": 1.5309,
+      "step": 1376
+    },
+    {
+      "epoch": 0.09594815872905271,
+      "grad_norm": 0.73602276499145,
+      "learning_rate": 6.85796584335793e-07,
+      "loss": 1.5709,
+      "step": 1377
+    },
+    {
+      "epoch": 0.09601783785667004,
+      "grad_norm": 0.7170190034347591,
+      "learning_rate": 6.857761040692537e-07,
+      "loss": 1.5174,
+      "step": 1378
+    },
+    {
+      "epoch": 0.09608751698428736,
+      "grad_norm": 0.7080538784656087,
+      "learning_rate": 6.857556093889062e-07,
+      "loss": 1.5577,
+      "step": 1379
+    },
+    {
+      "epoch": 0.09615719611190468,
+      "grad_norm": 0.6842467145632883,
+      "learning_rate": 6.857351002957321e-07,
+      "loss": 1.5241,
+      "step": 1380
+    },
+    {
+      "epoch": 0.096226875239522,
+      "grad_norm": 0.7093581497996692,
+      "learning_rate": 6.857145767907148e-07,
+      "loss": 1.597,
+      "step": 1381
+    },
+    {
+      "epoch": 0.09629655436713933,
+      "grad_norm": 0.7003330075726656,
+      "learning_rate": 6.856940388748376e-07,
+      "loss": 1.5867,
+      "step": 1382
+    },
+    {
+      "epoch": 0.09636623349475665,
+      "grad_norm": 0.7464105256695647,
+      "learning_rate": 6.856734865490847e-07,
+      "loss": 1.6015,
+      "step": 1383
+    },
+    {
+      "epoch": 0.09643591262237397,
+      "grad_norm": 0.7417160276838659,
+      "learning_rate": 6.85652919814441e-07,
+      "loss": 1.564,
+      "step": 1384
+    },
+    {
+      "epoch": 0.09650559174999129,
+      "grad_norm": 0.7036341159599784,
+      "learning_rate": 6.856323386718923e-07,
+      "loss": 1.4837,
+      "step": 1385
+    },
+    {
+      "epoch": 0.09657527087760862,
+      "grad_norm": 0.7182081102121308,
+      "learning_rate": 6.856117431224246e-07,
+      "loss": 1.4602,
+      "step": 1386
+    },
+    {
+      "epoch": 0.09664495000522594,
+      "grad_norm": 0.7267870082254924,
+      "learning_rate": 6.855911331670251e-07,
+      "loss": 1.5879,
+      "step": 1387
+    },
+    {
+      "epoch": 0.09671462913284326,
+      "grad_norm": 0.7263497628374512,
+      "learning_rate": 6.855705088066814e-07,
+      "loss": 1.5322,
+      "step": 1388
+    },
+    {
+      "epoch": 0.09678430826046058,
+      "grad_norm": 0.7424830626660222,
+      "learning_rate": 6.855498700423819e-07,
+      "loss": 1.663,
+      "step": 1389
+    },
+    {
+      "epoch": 0.0968539873880779,
+      "grad_norm": 0.6792334902361514,
+      "learning_rate": 6.855292168751155e-07,
+      "loss": 1.6057,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09692366651569523,
+      "grad_norm": 0.6902998981753529,
+      "learning_rate": 6.855085493058721e-07,
+      "loss": 1.5234,
+      "step": 1391
+    },
+    {
+      "epoch": 0.09699334564331255,
+      "grad_norm": 0.6695799650473492,
+      "learning_rate": 6.854878673356421e-07,
+      "loss": 1.6354,
+      "step": 1392
+    },
+    {
+      "epoch": 0.09706302477092987,
+      "grad_norm": 0.707158420145716,
+      "learning_rate": 6.854671709654168e-07,
+      "loss": 1.6303,
+      "step": 1393
+    },
+    {
+      "epoch": 0.0971327038985472,
+      "grad_norm": 0.6919860585571154,
+      "learning_rate": 6.854464601961875e-07,
+      "loss": 1.5524,
+      "step": 1394
+    },
+    {
+      "epoch": 0.09720238302616452,
+      "grad_norm": 0.7174186095393233,
+      "learning_rate": 6.854257350289472e-07,
+      "loss": 1.5092,
+      "step": 1395
+    },
+    {
+      "epoch": 0.09727206215378184,
+      "grad_norm": 0.7194978937889569,
+      "learning_rate": 6.854049954646889e-07,
+      "loss": 1.6163,
+      "step": 1396
+    },
+    {
+      "epoch": 0.09734174128139916,
+      "grad_norm": 0.7669208281476789,
+      "learning_rate": 6.853842415044065e-07,
+      "loss": 1.4796,
+      "step": 1397
+    },
+    {
+      "epoch": 0.09741142040901649,
+      "grad_norm": 0.7872523249133949,
+      "learning_rate": 6.853634731490944e-07,
+      "loss": 1.5655,
+      "step": 1398
+    },
+    {
+      "epoch": 0.0974810995366338,
+      "grad_norm": 0.72278764585173,
+      "learning_rate": 6.853426903997482e-07,
+      "loss": 1.5231,
+      "step": 1399
+    },
+    {
+      "epoch": 0.09755077866425112,
+      "grad_norm": 0.6911638249630152,
+      "learning_rate": 6.853218932573636e-07,
+      "loss": 1.4571,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09762045779186844,
+      "grad_norm": 0.7409581114985209,
+      "learning_rate": 6.853010817229374e-07,
+      "loss": 1.5235,
+      "step": 1401
+    },
+    {
+      "epoch": 0.09769013691948576,
+      "grad_norm": 0.7512553391928395,
+      "learning_rate": 6.852802557974668e-07,
+      "loss": 1.6149,
+      "step": 1402
+    },
+    {
+      "epoch": 0.09775981604710308,
+      "grad_norm": 0.7542228208631525,
+      "learning_rate": 6.8525941548195e-07,
+      "loss": 1.6557,
+      "step": 1403
+    },
+    {
+      "epoch": 0.09782949517472041,
+      "grad_norm": 0.730933943838304,
+      "learning_rate": 6.852385607773855e-07,
+      "loss": 1.5203,
+      "step": 1404
+    },
+    {
+      "epoch": 0.09789917430233773,
+      "grad_norm": 0.7226965064395986,
+      "learning_rate": 6.852176916847728e-07,
+      "loss": 1.5156,
+      "step": 1405
+    },
+    {
+      "epoch": 0.09796885342995505,
+      "grad_norm": 0.7400386566230431,
+      "learning_rate": 6.851968082051119e-07,
+      "loss": 1.4812,
+      "step": 1406
+    },
+    {
+      "epoch": 0.09803853255757237,
+      "grad_norm": 0.7217069234943865,
+      "learning_rate": 6.851759103394038e-07,
+      "loss": 1.6011,
+      "step": 1407
+    },
+    {
+      "epoch": 0.0981082116851897,
+      "grad_norm": 0.7208248423892274,
+      "learning_rate": 6.851549980886498e-07,
+      "loss": 1.5482,
+      "step": 1408
+    },
+    {
+      "epoch": 0.09817789081280702,
+      "grad_norm": 0.73869785008963,
+      "learning_rate": 6.851340714538519e-07,
+      "loss": 1.5056,
+      "step": 1409
+    },
+    {
+      "epoch": 0.09824756994042434,
+      "grad_norm": 0.7205356992642813,
+      "learning_rate": 6.851131304360134e-07,
+      "loss": 1.6143,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09831724906804166,
+      "grad_norm": 0.7504594927506455,
+      "learning_rate": 6.850921750361374e-07,
+      "loss": 1.5684,
+      "step": 1411
+    },
+    {
+      "epoch": 0.09838692819565899,
+      "grad_norm": 0.7912872734883598,
+      "learning_rate": 6.850712052552282e-07,
+      "loss": 1.5114,
+      "step": 1412
+    },
+    {
+      "epoch": 0.09845660732327631,
+      "grad_norm": 0.6954212545087036,
+      "learning_rate": 6.85050221094291e-07,
+      "loss": 1.5148,
+      "step": 1413
+    },
+    {
+      "epoch": 0.09852628645089363,
+      "grad_norm": 0.7121224389351605,
+      "learning_rate": 6.850292225543312e-07,
+      "loss": 1.5543,
+      "step": 1414
+    },
+    {
+      "epoch": 0.09859596557851096,
+      "grad_norm": 0.7308811982222069,
+      "learning_rate": 6.850082096363551e-07,
+      "loss": 1.5675,
+      "step": 1415
+    },
+    {
+      "epoch": 0.09866564470612828,
+      "grad_norm": 0.6669232499820077,
+      "learning_rate": 6.849871823413696e-07,
+      "loss": 1.5222,
+      "step": 1416
+    },
+    {
+      "epoch": 0.0987353238337456,
+      "grad_norm": 0.7023704892031762,
+      "learning_rate": 6.849661406703825e-07,
+      "loss": 1.4323,
+      "step": 1417
+    },
+    {
+      "epoch": 0.09880500296136292,
+      "grad_norm": 0.7450305359211221,
+      "learning_rate": 6.849450846244022e-07,
+      "loss": 1.6129,
+      "step": 1418
+    },
+    {
+      "epoch": 0.09887468208898025,
+      "grad_norm": 0.7559519975027,
+      "learning_rate": 6.849240142044376e-07,
+      "loss": 1.543,
+      "step": 1419
+    },
+    {
+      "epoch": 0.09894436121659757,
+      "grad_norm": 0.7557039933040223,
+      "learning_rate": 6.849029294114985e-07,
+      "loss": 1.7281,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09901404034421489,
+      "grad_norm": 0.7535013429286772,
+      "learning_rate": 6.848818302465954e-07,
+      "loss": 1.5028,
+      "step": 1421
+    },
+    {
+      "epoch": 0.09908371947183221,
+      "grad_norm": 0.695752285288689,
+      "learning_rate": 6.848607167107393e-07,
+      "loss": 1.5552,
+      "step": 1422
+    },
+    {
+      "epoch": 0.09915339859944954,
+      "grad_norm": 0.7130712608592883,
+      "learning_rate": 6.848395888049421e-07,
+      "loss": 1.5192,
+      "step": 1423
+    },
+    {
+      "epoch": 0.09922307772706686,
+      "grad_norm": 0.7462396757316636,
+      "learning_rate": 6.84818446530216e-07,
+      "loss": 1.6576,
+      "step": 1424
+    },
+    {
+      "epoch": 0.09929275685468418,
+      "grad_norm": 0.6936681758464512,
+      "learning_rate": 6.847972898875747e-07,
+      "loss": 1.4941,
+      "step": 1425
+    },
+    {
+      "epoch": 0.0993624359823015,
+      "grad_norm": 0.7291394697109261,
+      "learning_rate": 6.847761188780318e-07,
+      "loss": 1.5412,
+      "step": 1426
+    },
+    {
+      "epoch": 0.09943211510991883,
+      "grad_norm": 0.8215857318532434,
+      "learning_rate": 6.847549335026017e-07,
+      "loss": 1.6619,
+      "step": 1427
+    },
+    {
+      "epoch": 0.09950179423753615,
+      "grad_norm": 0.7589146596740709,
+      "learning_rate": 6.847337337623e-07,
+      "loss": 1.5622,
+      "step": 1428
+    },
+    {
+      "epoch": 0.09957147336515347,
+      "grad_norm": 0.7000258820654407,
+      "learning_rate": 6.847125196581422e-07,
+      "loss": 1.6265,
+      "step": 1429
+    },
+    {
+      "epoch": 0.0996411524927708,
+      "grad_norm": 0.6653107210684309,
+      "learning_rate": 6.846912911911453e-07,
+      "loss": 1.4856,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09971083162038812,
+      "grad_norm": 0.6702010457721812,
+      "learning_rate": 6.846700483623265e-07,
+      "loss": 1.4824,
+      "step": 1431
+    },
+    {
+      "epoch": 0.09978051074800544,
+      "grad_norm": 0.7175711439850605,
+      "learning_rate": 6.846487911727036e-07,
+      "loss": 1.5737,
+      "step": 1432
+    },
+    {
+      "epoch": 0.09985018987562276,
+      "grad_norm": 0.729863487630728,
+      "learning_rate": 6.846275196232956e-07,
+      "loss": 1.4471,
+      "step": 1433
+    },
+    {
+      "epoch": 0.09991986900324008,
+      "grad_norm": 0.6618160691322408,
+      "learning_rate": 6.846062337151217e-07,
+      "loss": 1.4689,
+      "step": 1434
+    },
+    {
+      "epoch": 0.0999895481308574,
+      "grad_norm": 0.761821100047957,
+      "learning_rate": 6.845849334492021e-07,
+      "loss": 1.6445,
+      "step": 1435
+    },
+    {
+      "epoch": 0.10005922725847473,
+      "grad_norm": 0.7375583392439995,
+      "learning_rate": 6.845636188265573e-07,
+      "loss": 1.5686,
+      "step": 1436
+    },
+    {
+      "epoch": 0.10012890638609205,
+      "grad_norm": 0.7082128443723841,
+      "learning_rate": 6.845422898482089e-07,
+      "loss": 1.4708,
+      "step": 1437
+    },
+    {
+      "epoch": 0.10019858551370937,
+      "grad_norm": 0.7831935456199045,
+      "learning_rate": 6.845209465151791e-07,
+      "loss": 1.5965,
+      "step": 1438
+    },
+    {
+      "epoch": 0.1002682646413267,
+      "grad_norm": 0.6625792127601626,
+      "learning_rate": 6.844995888284906e-07,
+      "loss": 1.4301,
+      "step": 1439
+    },
+    {
+      "epoch": 0.10033794376894402,
+      "grad_norm": 0.7224274069117864,
+      "learning_rate": 6.84478216789167e-07,
+      "loss": 1.52,
+      "step": 1440
+    },
+    {
+      "epoch": 0.10040762289656134,
+      "grad_norm": 0.7306401875419833,
+      "learning_rate": 6.844568303982324e-07,
+      "loss": 1.6456,
+      "step": 1441
+    },
+    {
+      "epoch": 0.10047730202417866,
+      "grad_norm": 0.6991325586636551,
+      "learning_rate": 6.844354296567117e-07,
+      "loss": 1.5894,
+      "step": 1442
+    },
+    {
+      "epoch": 0.10054698115179599,
+      "grad_norm": 0.8232785004103219,
+      "learning_rate": 6.844140145656305e-07,
+      "loss": 1.7543,
+      "step": 1443
+    },
+    {
+      "epoch": 0.10061666027941331,
+      "grad_norm": 0.7215617175038418,
+      "learning_rate": 6.84392585126015e-07,
+      "loss": 1.5773,
+      "step": 1444
+    },
+    {
+      "epoch": 0.10068633940703062,
+      "grad_norm": 0.7374985957692638,
+      "learning_rate": 6.843711413388923e-07,
+      "loss": 1.5536,
+      "step": 1445
+    },
+    {
+      "epoch": 0.10075601853464794,
+      "grad_norm": 0.7104503086346701,
+      "learning_rate": 6.843496832052897e-07,
+      "loss": 1.5673,
+      "step": 1446
+    },
+    {
+      "epoch": 0.10082569766226526,
+      "grad_norm": 0.7036943520124541,
+      "learning_rate": 6.843282107262359e-07,
+      "loss": 1.4641,
+      "step": 1447
+    },
+    {
+      "epoch": 0.10089537678988258,
+      "grad_norm": 0.7682117601321635,
+      "learning_rate": 6.843067239027598e-07,
+      "loss": 1.4993,
+      "step": 1448
+    },
+    {
+      "epoch": 0.10096505591749991,
+      "grad_norm": 0.7987108957657223,
+      "learning_rate": 6.842852227358907e-07,
+      "loss": 1.6263,
+      "step": 1449
+    },
+    {
+      "epoch": 0.10103473504511723,
+      "grad_norm": 0.7182857850830695,
+      "learning_rate": 6.842637072266596e-07,
+      "loss": 1.53,
+      "step": 1450
+    },
+    {
+      "epoch": 0.10110441417273455,
+      "grad_norm": 0.7246888294850192,
+      "learning_rate": 6.842421773760972e-07,
+      "loss": 1.7401,
+      "step": 1451
+    },
+    {
+      "epoch": 0.10117409330035187,
+      "grad_norm": 0.7094639682875344,
+      "learning_rate": 6.842206331852352e-07,
+      "loss": 1.5597,
+      "step": 1452
+    },
+    {
+      "epoch": 0.1012437724279692,
+      "grad_norm": 0.744605882539227,
+      "learning_rate": 6.841990746551064e-07,
+      "loss": 1.6996,
+      "step": 1453
+    },
+    {
+      "epoch": 0.10131345155558652,
+      "grad_norm": 0.7732575296825245,
+      "learning_rate": 6.841775017867435e-07,
+      "loss": 1.6414,
+      "step": 1454
+    },
+    {
+      "epoch": 0.10138313068320384,
+      "grad_norm": 0.7219152072214178,
+      "learning_rate": 6.841559145811805e-07,
+      "loss": 1.6682,
+      "step": 1455
+    },
+    {
+      "epoch": 0.10145280981082117,
+      "grad_norm": 0.7089114420009685,
+      "learning_rate": 6.84134313039452e-07,
+      "loss": 1.6003,
+      "step": 1456
+    },
+    {
+      "epoch": 0.10152248893843849,
+      "grad_norm": 0.6822564437101662,
+      "learning_rate": 6.841126971625932e-07,
+      "loss": 1.5532,
+      "step": 1457
+    },
+    {
+      "epoch": 0.10159216806605581,
+      "grad_norm": 0.6908697206870166,
+      "learning_rate": 6.840910669516399e-07,
+      "loss": 1.551,
+      "step": 1458
+    },
+    {
+      "epoch": 0.10166184719367313,
+      "grad_norm": 0.749847037637897,
+      "learning_rate": 6.840694224076284e-07,
+      "loss": 1.511,
+      "step": 1459
+    },
+    {
+      "epoch": 0.10173152632129046,
+      "grad_norm": 0.6973808625903857,
+      "learning_rate": 6.840477635315965e-07,
+      "loss": 1.6447,
+      "step": 1460
+    },
+    {
+      "epoch": 0.10180120544890778,
+      "grad_norm": 0.6637049891036413,
+      "learning_rate": 6.840260903245816e-07,
+      "loss": 1.4964,
+      "step": 1461
+    },
+    {
+      "epoch": 0.1018708845765251,
+      "grad_norm": 0.7051328218519948,
+      "learning_rate": 6.840044027876226e-07,
+      "loss": 1.4516,
+      "step": 1462
+    },
+    {
+      "epoch": 0.10194056370414242,
+      "grad_norm": 0.6998029876564847,
+      "learning_rate": 6.839827009217589e-07,
+      "loss": 1.7098,
+      "step": 1463
+    },
+    {
+      "epoch": 0.10201024283175975,
+      "grad_norm": 0.8312269910488456,
+      "learning_rate": 6.839609847280303e-07,
+      "loss": 1.4785,
+      "step": 1464
+    },
+    {
+      "epoch": 0.10207992195937707,
+      "grad_norm": 0.692043398562744,
+      "learning_rate": 6.839392542074777e-07,
+      "loss": 1.3329,
+      "step": 1465
+    },
+    {
+      "epoch": 0.10214960108699439,
+      "grad_norm": 0.6905454250734092,
+      "learning_rate": 6.83917509361142e-07,
+      "loss": 1.6113,
+      "step": 1466
+    },
+    {
+      "epoch": 0.10221928021461171,
+      "grad_norm": 0.6692689047744544,
+      "learning_rate": 6.838957501900658e-07,
+      "loss": 1.5585,
+      "step": 1467
+    },
+    {
+      "epoch": 0.10228895934222904,
+      "grad_norm": 0.6866297965522439,
+      "learning_rate": 6.838739766952916e-07,
+      "loss": 1.491,
+      "step": 1468
+    },
+    {
+      "epoch": 0.10235863846984636,
+      "grad_norm": 0.713111200601478,
+      "learning_rate": 6.838521888778629e-07,
+      "loss": 1.4458,
+      "step": 1469
+    },
+    {
+      "epoch": 0.10242831759746368,
+      "grad_norm": 0.7596916282973877,
+      "learning_rate": 6.838303867388237e-07,
+      "loss": 1.5562,
+      "step": 1470
+    },
+    {
+      "epoch": 0.102497996725081,
+      "grad_norm": 0.7729511940587167,
+      "learning_rate": 6.83808570279219e-07,
+      "loss": 1.4941,
+      "step": 1471
+    },
+    {
+      "epoch": 0.10256767585269833,
+      "grad_norm": 0.6728813668527486,
+      "learning_rate": 6.83786739500094e-07,
+      "loss": 1.3944,
+      "step": 1472
+    },
+    {
+      "epoch": 0.10263735498031565,
+      "grad_norm": 0.7228001449563164,
+      "learning_rate": 6.837648944024951e-07,
+      "loss": 1.6582,
+      "step": 1473
+    },
+    {
+      "epoch": 0.10270703410793297,
+      "grad_norm": 0.7070507373346999,
+      "learning_rate": 6.837430349874693e-07,
+      "loss": 1.5626,
+      "step": 1474
+    },
+    {
+      "epoch": 0.1027767132355503,
+      "grad_norm": 0.6647571817731183,
+      "learning_rate": 6.837211612560636e-07,
+      "loss": 1.4195,
+      "step": 1475
+    },
+    {
+      "epoch": 0.10284639236316762,
+      "grad_norm": 0.7221508915094103,
+      "learning_rate": 6.836992732093267e-07,
+      "loss": 1.6211,
+      "step": 1476
+    },
+    {
+      "epoch": 0.10291607149078494,
+      "grad_norm": 0.7109611055734725,
+      "learning_rate": 6.836773708483076e-07,
+      "loss": 1.5389,
+      "step": 1477
+    },
+    {
+      "epoch": 0.10298575061840226,
+      "grad_norm": 0.6907251786674505,
+      "learning_rate": 6.836554541740556e-07,
+      "loss": 1.5164,
+      "step": 1478
+    },
+    {
+      "epoch": 0.10305542974601958,
+      "grad_norm": 0.699620393769317,
+      "learning_rate": 6.836335231876212e-07,
+      "loss": 1.7227,
+      "step": 1479
+    },
+    {
+      "epoch": 0.1031251088736369,
+      "grad_norm": 0.6940924493567484,
+      "learning_rate": 6.836115778900552e-07,
+      "loss": 1.5148,
+      "step": 1480
+    },
+    {
+      "epoch": 0.10319478800125423,
+      "grad_norm": 0.7238715870249767,
+      "learning_rate": 6.835896182824093e-07,
+      "loss": 1.5669,
+      "step": 1481
+    },
+    {
+      "epoch": 0.10326446712887155,
+      "grad_norm": 0.7088044163423685,
+      "learning_rate": 6.83567644365736e-07,
+      "loss": 1.4448,
+      "step": 1482
+    },
+    {
+      "epoch": 0.10333414625648887,
+      "grad_norm": 0.7447083571913127,
+      "learning_rate": 6.835456561410882e-07,
+      "loss": 1.6735,
+      "step": 1483
+    },
+    {
+      "epoch": 0.1034038253841062,
+      "grad_norm": 0.6910524206838882,
+      "learning_rate": 6.835236536095197e-07,
+      "loss": 1.5637,
+      "step": 1484
+    },
+    {
+      "epoch": 0.10347350451172352,
+      "grad_norm": 0.6593429324828214,
+      "learning_rate": 6.835016367720847e-07,
+      "loss": 1.4477,
+      "step": 1485
+    },
+    {
+      "epoch": 0.10354318363934084,
+      "grad_norm": 0.7235635260040205,
+      "learning_rate": 6.834796056298386e-07,
+      "loss": 1.5663,
+      "step": 1486
+    },
+    {
+      "epoch": 0.10361286276695816,
+      "grad_norm": 0.6955384974930857,
+      "learning_rate": 6.83457560183837e-07,
+      "loss": 1.5034,
+      "step": 1487
+    },
+    {
+      "epoch": 0.10368254189457549,
+      "grad_norm": 0.6869816822781529,
+      "learning_rate": 6.834355004351363e-07,
+      "loss": 1.5124,
+      "step": 1488
+    },
+    {
+      "epoch": 0.10375222102219281,
+      "grad_norm": 0.7239903773603513,
+      "learning_rate": 6.834134263847939e-07,
+      "loss": 1.5423,
+      "step": 1489
+    },
+    {
+      "epoch": 0.10382190014981013,
+      "grad_norm": 0.7255177486951737,
+      "learning_rate": 6.833913380338675e-07,
+      "loss": 1.468,
+      "step": 1490
+    },
+    {
+      "epoch": 0.10389157927742744,
+      "grad_norm": 0.6935427630827753,
+      "learning_rate": 6.833692353834154e-07,
+      "loss": 1.5646,
+      "step": 1491
+    },
+    {
+      "epoch": 0.10396125840504476,
+      "grad_norm": 0.78289308346883,
+      "learning_rate": 6.833471184344971e-07,
+      "loss": 1.61,
+      "step": 1492
+    },
+    {
+      "epoch": 0.10403093753266208,
+      "grad_norm": 0.7540578067091052,
+      "learning_rate": 6.833249871881725e-07,
+      "loss": 1.6918,
+      "step": 1493
+    },
+    {
+      "epoch": 0.10410061666027941,
+      "grad_norm": 0.6723185719673338,
+      "learning_rate": 6.833028416455019e-07,
+      "loss": 1.5178,
+      "step": 1494
+    },
+    {
+      "epoch": 0.10417029578789673,
+      "grad_norm": 0.7225132128946964,
+      "learning_rate": 6.832806818075467e-07,
+      "loss": 1.5849,
+      "step": 1495
+    },
+    {
+      "epoch": 0.10423997491551405,
+      "grad_norm": 0.7256132910891259,
+      "learning_rate": 6.83258507675369e-07,
+      "loss": 1.6121,
+      "step": 1496
+    },
+    {
+      "epoch": 0.10430965404313138,
+      "grad_norm": 0.7833180897554166,
+      "learning_rate": 6.832363192500312e-07,
+      "loss": 1.5221,
+      "step": 1497
+    },
+    {
+      "epoch": 0.1043793331707487,
+      "grad_norm": 0.7612167370203573,
+      "learning_rate": 6.832141165325967e-07,
+      "loss": 1.5844,
+      "step": 1498
+    },
+    {
+      "epoch": 0.10444901229836602,
+      "grad_norm": 0.7410871801404689,
+      "learning_rate": 6.831918995241296e-07,
+      "loss": 1.6095,
+      "step": 1499
+    },
+    {
+      "epoch": 0.10451869142598334,
+      "grad_norm": 0.7017298260754218,
+      "learning_rate": 6.831696682256944e-07,
+      "loss": 1.511,
+      "step": 1500
+    },
+    {
+      "epoch": 0.10458837055360067,
+      "grad_norm": 0.7339141096335122,
+      "learning_rate": 6.831474226383567e-07,
+      "loss": 1.6158,
+      "step": 1501
+    },
+    {
+      "epoch": 0.10465804968121799,
+      "grad_norm": 0.7315827206339433,
+      "learning_rate": 6.831251627631824e-07,
+      "loss": 1.5025,
+      "step": 1502
+    },
+    {
+      "epoch": 0.10472772880883531,
+      "grad_norm": 0.8214608688213216,
+      "learning_rate": 6.831028886012382e-07,
+      "loss": 1.4651,
+      "step": 1503
+    },
+    {
+      "epoch": 0.10479740793645263,
+      "grad_norm": 0.7050091177175483,
+      "learning_rate": 6.830806001535916e-07,
+      "loss": 1.5367,
+      "step": 1504
+    },
+    {
+      "epoch": 0.10486708706406996,
+      "grad_norm": 0.7055266701449938,
+      "learning_rate": 6.830582974213108e-07,
+      "loss": 1.5644,
+      "step": 1505
+    },
+    {
+      "epoch": 0.10493676619168728,
+      "grad_norm": 0.7094965831953183,
+      "learning_rate": 6.830359804054643e-07,
+      "loss": 1.6474,
+      "step": 1506
+    },
+    {
+      "epoch": 0.1050064453193046,
+      "grad_norm": 0.6640006713070633,
+      "learning_rate": 6.83013649107122e-07,
+      "loss": 1.3401,
+      "step": 1507
+    },
+    {
+      "epoch": 0.10507612444692192,
+      "grad_norm": 0.7105850301509256,
+      "learning_rate": 6.829913035273536e-07,
+      "loss": 1.455,
+      "step": 1508
+    },
+    {
+      "epoch": 0.10514580357453925,
+      "grad_norm": 0.6867742081585059,
+      "learning_rate": 6.829689436672302e-07,
+      "loss": 1.2886,
+      "step": 1509
+    },
+    {
+      "epoch": 0.10521548270215657,
+      "grad_norm": 0.6958585661416363,
+      "learning_rate": 6.829465695278233e-07,
+      "loss": 1.5809,
+      "step": 1510
+    },
+    {
+      "epoch": 0.10528516182977389,
+      "grad_norm": 0.7797334011110041,
+      "learning_rate": 6.829241811102052e-07,
+      "loss": 1.474,
+      "step": 1511
+    },
+    {
+      "epoch": 0.10535484095739121,
+      "grad_norm": 0.6753774832924908,
+      "learning_rate": 6.829017784154487e-07,
+      "loss": 1.5347,
+      "step": 1512
+    },
+    {
+      "epoch": 0.10542452008500854,
+      "grad_norm": 0.6657324139543032,
+      "learning_rate": 6.828793614446273e-07,
+      "loss": 1.4748,
+      "step": 1513
+    },
+    {
+      "epoch": 0.10549419921262586,
+      "grad_norm": 0.7217234617981039,
+      "learning_rate": 6.828569301988155e-07,
+      "loss": 1.5803,
+      "step": 1514
+    },
+    {
+      "epoch": 0.10556387834024318,
+      "grad_norm": 0.7292235276823041,
+      "learning_rate": 6.82834484679088e-07,
+      "loss": 1.4454,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1056335574678605,
+      "grad_norm": 0.6526371966253907,
+      "learning_rate": 6.828120248865206e-07,
+      "loss": 1.415,
+      "step": 1516
+    },
+    {
+      "epoch": 0.10570323659547783,
+      "grad_norm": 0.7198666759857738,
+      "learning_rate": 6.827895508221897e-07,
+      "loss": 1.5559,
+      "step": 1517
+    },
+    {
+      "epoch": 0.10577291572309515,
+      "grad_norm": 0.7628274611446516,
+      "learning_rate": 6.827670624871721e-07,
+      "loss": 1.5228,
+      "step": 1518
+    },
+    {
+      "epoch": 0.10584259485071247,
+      "grad_norm": 0.6985806196410942,
+      "learning_rate": 6.827445598825453e-07,
+      "loss": 1.5399,
+      "step": 1519
+    },
+    {
+      "epoch": 0.1059122739783298,
+      "grad_norm": 0.7268289166229313,
+      "learning_rate": 6.827220430093882e-07,
+      "loss": 1.5711,
+      "step": 1520
+    },
+    {
+      "epoch": 0.10598195310594712,
+      "grad_norm": 0.7159822726292894,
+      "learning_rate": 6.826995118687796e-07,
+      "loss": 1.3665,
+      "step": 1521
+    },
+    {
+      "epoch": 0.10605163223356444,
+      "grad_norm": 0.6799598821150186,
+      "learning_rate": 6.826769664617991e-07,
+      "loss": 1.5734,
+      "step": 1522
+    },
+    {
+      "epoch": 0.10612131136118176,
+      "grad_norm": 0.6759158480037062,
+      "learning_rate": 6.826544067895273e-07,
+      "loss": 1.5268,
+      "step": 1523
+    },
+    {
+      "epoch": 0.10619099048879908,
+      "grad_norm": 0.6764740596037916,
+      "learning_rate": 6.826318328530453e-07,
+      "loss": 1.5798,
+      "step": 1524
+    },
+    {
+      "epoch": 0.1062606696164164,
+      "grad_norm": 0.7544503094902575,
+      "learning_rate": 6.826092446534348e-07,
+      "loss": 1.6701,
+      "step": 1525
+    },
+    {
+      "epoch": 0.10633034874403373,
+      "grad_norm": 0.7599345785560571,
+      "learning_rate": 6.825866421917783e-07,
+      "loss": 1.5108,
+      "step": 1526
+    },
+    {
+      "epoch": 0.10640002787165105,
+      "grad_norm": 0.6887529792956971,
+      "learning_rate": 6.82564025469159e-07,
+      "loss": 1.4195,
+      "step": 1527
+    },
+    {
+      "epoch": 0.10646970699926837,
+      "grad_norm": 0.701570365983734,
+      "learning_rate": 6.825413944866607e-07,
+      "loss": 1.5902,
+      "step": 1528
+    },
+    {
+      "epoch": 0.1065393861268857,
+      "grad_norm": 0.6851192165134431,
+      "learning_rate": 6.825187492453679e-07,
+      "loss": 1.5846,
+      "step": 1529
+    },
+    {
+      "epoch": 0.10660906525450302,
+      "grad_norm": 0.6547795464181767,
+      "learning_rate": 6.82496089746366e-07,
+      "loss": 1.3953,
+      "step": 1530
+    },
+    {
+      "epoch": 0.10667874438212034,
+      "grad_norm": 0.7023921604836459,
+      "learning_rate": 6.824734159907405e-07,
+      "loss": 1.6031,
+      "step": 1531
+    },
+    {
+      "epoch": 0.10674842350973766,
+      "grad_norm": 0.7383875135991188,
+      "learning_rate": 6.824507279795784e-07,
+      "loss": 1.5285,
+      "step": 1532
+    },
+    {
+      "epoch": 0.10681810263735499,
+      "grad_norm": 0.7322590444258,
+      "learning_rate": 6.824280257139667e-07,
+      "loss": 1.4706,
+      "step": 1533
+    },
+    {
+      "epoch": 0.10688778176497231,
+      "grad_norm": 0.6595049248009317,
+      "learning_rate": 6.824053091949933e-07,
+      "loss": 1.4327,
+      "step": 1534
+    },
+    {
+      "epoch": 0.10695746089258963,
+      "grad_norm": 0.7165800718934179,
+      "learning_rate": 6.823825784237471e-07,
+      "loss": 1.5829,
+      "step": 1535
+    },
+    {
+      "epoch": 0.10702714002020695,
+      "grad_norm": 0.704731120016156,
+      "learning_rate": 6.82359833401317e-07,
+      "loss": 1.5255,
+      "step": 1536
+    },
+    {
+      "epoch": 0.10709681914782426,
+      "grad_norm": 0.6997247632243365,
+      "learning_rate": 6.823370741287933e-07,
+      "loss": 1.6407,
+      "step": 1537
+    },
+    {
+      "epoch": 0.10716649827544159,
+      "grad_norm": 0.7023753654540732,
+      "learning_rate": 6.823143006072667e-07,
+      "loss": 1.5886,
+      "step": 1538
+    },
+    {
+      "epoch": 0.10723617740305891,
+      "grad_norm": 0.6743001705870307,
+      "learning_rate": 6.822915128378284e-07,
+      "loss": 1.572,
+      "step": 1539
+    },
+    {
+      "epoch": 0.10730585653067623,
+      "grad_norm": 0.6750188030153429,
+      "learning_rate": 6.822687108215704e-07,
+      "loss": 1.5842,
+      "step": 1540
+    },
+    {
+      "epoch": 0.10737553565829355,
+      "grad_norm": 0.7262010807988295,
+      "learning_rate": 6.822458945595856e-07,
+      "loss": 1.5688,
+      "step": 1541
+    },
+    {
+      "epoch": 0.10744521478591088,
+      "grad_norm": 0.7212282946377836,
+      "learning_rate": 6.822230640529671e-07,
+      "loss": 1.5317,
+      "step": 1542
+    },
+    {
+      "epoch": 0.1075148939135282,
+      "grad_norm": 0.8885724857888603,
+      "learning_rate": 6.822002193028095e-07,
+      "loss": 1.6592,
+      "step": 1543
+    },
+    {
+      "epoch": 0.10758457304114552,
+      "grad_norm": 0.7197806691400269,
+      "learning_rate": 6.82177360310207e-07,
+      "loss": 1.5434,
+      "step": 1544
+    },
+    {
+      "epoch": 0.10765425216876284,
+      "grad_norm": 0.7558178313589539,
+      "learning_rate": 6.821544870762554e-07,
+      "loss": 1.6391,
+      "step": 1545
+    },
+    {
+      "epoch": 0.10772393129638017,
+      "grad_norm": 0.7987854415241106,
+      "learning_rate": 6.821315996020506e-07,
+      "loss": 1.5524,
+      "step": 1546
+    },
+    {
+      "epoch": 0.10779361042399749,
+      "grad_norm": 0.6686455155160611,
+      "learning_rate": 6.821086978886897e-07,
+      "loss": 1.4751,
+      "step": 1547
+    },
+    {
+      "epoch": 0.10786328955161481,
+      "grad_norm": 0.7341664427885704,
+      "learning_rate": 6.8208578193727e-07,
+      "loss": 1.5269,
+      "step": 1548
+    },
+    {
+      "epoch": 0.10793296867923213,
+      "grad_norm": 0.7003223328477466,
+      "learning_rate": 6.820628517488898e-07,
+      "loss": 1.5568,
+      "step": 1549
+    },
+    {
+      "epoch": 0.10800264780684946,
+      "grad_norm": 0.6864701685410314,
+      "learning_rate": 6.820399073246477e-07,
+      "loss": 1.4843,
+      "step": 1550
+    },
+    {
+      "epoch": 0.10807232693446678,
+      "grad_norm": 0.8687409341573422,
+      "learning_rate": 6.820169486656435e-07,
+      "loss": 1.5423,
+      "step": 1551
+    },
+    {
+      "epoch": 0.1081420060620841,
+      "grad_norm": 0.7607151628672985,
+      "learning_rate": 6.819939757729774e-07,
+      "loss": 1.5437,
+      "step": 1552
+    },
+    {
+      "epoch": 0.10821168518970142,
+      "grad_norm": 0.7048985366339954,
+      "learning_rate": 6.819709886477503e-07,
+      "loss": 1.5318,
+      "step": 1553
+    },
+    {
+      "epoch": 0.10828136431731875,
+      "grad_norm": 0.7314286735099996,
+      "learning_rate": 6.819479872910638e-07,
+      "loss": 1.5726,
+      "step": 1554
+    },
+    {
+      "epoch": 0.10835104344493607,
+      "grad_norm": 0.6962507248591363,
+      "learning_rate": 6.8192497170402e-07,
+      "loss": 1.4723,
+      "step": 1555
+    },
+    {
+      "epoch": 0.10842072257255339,
+      "grad_norm": 0.7785074386756621,
+      "learning_rate": 6.819019418877221e-07,
+      "loss": 1.6421,
+      "step": 1556
+    },
+    {
+      "epoch": 0.10849040170017071,
+      "grad_norm": 0.7954582117985666,
+      "learning_rate": 6.818788978432735e-07,
+      "loss": 1.664,
+      "step": 1557
+    },
+    {
+      "epoch": 0.10856008082778804,
+      "grad_norm": 0.7367228514312243,
+      "learning_rate": 6.818558395717786e-07,
+      "loss": 1.5606,
+      "step": 1558
+    },
+    {
+      "epoch": 0.10862975995540536,
+      "grad_norm": 0.745909368111922,
+      "learning_rate": 6.818327670743425e-07,
+      "loss": 1.4566,
+      "step": 1559
+    },
+    {
+      "epoch": 0.10869943908302268,
+      "grad_norm": 0.7600253418884132,
+      "learning_rate": 6.818096803520709e-07,
+      "loss": 1.6048,
+      "step": 1560
+    },
+    {
+      "epoch": 0.10876911821064,
+      "grad_norm": 0.7373741671607602,
+      "learning_rate": 6.817865794060699e-07,
+      "loss": 1.423,
+      "step": 1561
+    },
+    {
+      "epoch": 0.10883879733825733,
+      "grad_norm": 0.6896732729151899,
+      "learning_rate": 6.817634642374468e-07,
+      "loss": 1.5513,
+      "step": 1562
+    },
+    {
+      "epoch": 0.10890847646587465,
+      "grad_norm": 0.6983623496494458,
+      "learning_rate": 6.817403348473094e-07,
+      "loss": 1.5651,
+      "step": 1563
+    },
+    {
+      "epoch": 0.10897815559349197,
+      "grad_norm": 0.7525447676994738,
+      "learning_rate": 6.817171912367657e-07,
+      "loss": 1.6091,
+      "step": 1564
+    },
+    {
+      "epoch": 0.1090478347211093,
+      "grad_norm": 0.7532040188348436,
+      "learning_rate": 6.816940334069252e-07,
+      "loss": 1.5547,
+      "step": 1565
+    },
+    {
+      "epoch": 0.10911751384872662,
+      "grad_norm": 0.7377763746387574,
+      "learning_rate": 6.816708613588975e-07,
+      "loss": 1.6717,
+      "step": 1566
+    },
+    {
+      "epoch": 0.10918719297634394,
+      "grad_norm": 0.7404993105111815,
+      "learning_rate": 6.816476750937931e-07,
+      "loss": 1.5567,
+      "step": 1567
+    },
+    {
+      "epoch": 0.10925687210396126,
+      "grad_norm": 0.6607513401827204,
+      "learning_rate": 6.816244746127231e-07,
+      "loss": 1.4863,
+      "step": 1568
+    },
+    {
+      "epoch": 0.10932655123157858,
+      "grad_norm": 0.7327524646568696,
+      "learning_rate": 6.816012599167993e-07,
+      "loss": 1.4312,
+      "step": 1569
+    },
+    {
+      "epoch": 0.1093962303591959,
+      "grad_norm": 0.7166557657811868,
+      "learning_rate": 6.815780310071341e-07,
+      "loss": 1.5816,
+      "step": 1570
+    },
+    {
+      "epoch": 0.10946590948681323,
+      "grad_norm": 0.695293401858971,
+      "learning_rate": 6.81554787884841e-07,
+      "loss": 1.5191,
+      "step": 1571
+    },
+    {
+      "epoch": 0.10953558861443055,
+      "grad_norm": 0.7131918332055402,
+      "learning_rate": 6.815315305510336e-07,
+      "loss": 1.6284,
+      "step": 1572
+    },
+    {
+      "epoch": 0.10960526774204787,
+      "grad_norm": 0.7076583193113377,
+      "learning_rate": 6.815082590068264e-07,
+      "loss": 1.535,
+      "step": 1573
+    },
+    {
+      "epoch": 0.1096749468696652,
+      "grad_norm": 0.7135512952756091,
+      "learning_rate": 6.814849732533347e-07,
+      "loss": 1.5333,
+      "step": 1574
+    },
+    {
+      "epoch": 0.10974462599728252,
+      "grad_norm": 0.7538494298477586,
+      "learning_rate": 6.814616732916744e-07,
+      "loss": 1.4118,
+      "step": 1575
+    },
+    {
+      "epoch": 0.10981430512489984,
+      "grad_norm": 0.7178907661850447,
+      "learning_rate": 6.814383591229622e-07,
+      "loss": 1.3534,
+      "step": 1576
+    },
+    {
+      "epoch": 0.10988398425251716,
+      "grad_norm": 0.7308851126536335,
+      "learning_rate": 6.814150307483151e-07,
+      "loss": 1.5599,
+      "step": 1577
+    },
+    {
+      "epoch": 0.10995366338013449,
+      "grad_norm": 0.7669543415808457,
+      "learning_rate": 6.813916881688513e-07,
+      "loss": 1.54,
+      "step": 1578
+    },
+    {
+      "epoch": 0.11002334250775181,
+      "grad_norm": 0.8240784223589613,
+      "learning_rate": 6.813683313856894e-07,
+      "loss": 1.5358,
+      "step": 1579
+    },
+    {
+      "epoch": 0.11009302163536913,
+      "grad_norm": 0.7221931116839967,
+      "learning_rate": 6.813449603999485e-07,
+      "loss": 1.6244,
+      "step": 1580
+    },
+    {
+      "epoch": 0.11016270076298645,
+      "grad_norm": 0.7189519304364052,
+      "learning_rate": 6.813215752127488e-07,
+      "loss": 1.4958,
+      "step": 1581
+    },
+    {
+      "epoch": 0.11023237989060376,
+      "grad_norm": 0.7511549610377619,
+      "learning_rate": 6.812981758252108e-07,
+      "loss": 1.6661,
+      "step": 1582
+    },
+    {
+      "epoch": 0.11030205901822109,
+      "grad_norm": 0.7497943083182442,
+      "learning_rate": 6.81274762238456e-07,
+      "loss": 1.5513,
+      "step": 1583
+    },
+    {
+      "epoch": 0.11037173814583841,
+      "grad_norm": 0.7112454154563912,
+      "learning_rate": 6.812513344536063e-07,
+      "loss": 1.5281,
+      "step": 1584
+    },
+    {
+      "epoch": 0.11044141727345573,
+      "grad_norm": 0.7163312744702844,
+      "learning_rate": 6.812278924717844e-07,
+      "loss": 1.5135,
+      "step": 1585
+    },
+    {
+      "epoch": 0.11051109640107305,
+      "grad_norm": 0.7529791816288604,
+      "learning_rate": 6.812044362941139e-07,
+      "loss": 1.6745,
+      "step": 1586
+    },
+    {
+      "epoch": 0.11058077552869038,
+      "grad_norm": 0.6813105813519863,
+      "learning_rate": 6.811809659217186e-07,
+      "loss": 1.3124,
+      "step": 1587
+    },
+    {
+      "epoch": 0.1106504546563077,
+      "grad_norm": 0.6772757264130072,
+      "learning_rate": 6.811574813557234e-07,
+      "loss": 1.5298,
+      "step": 1588
+    },
+    {
+      "epoch": 0.11072013378392502,
+      "grad_norm": 0.6824509840298503,
+      "learning_rate": 6.811339825972538e-07,
+      "loss": 1.6369,
+      "step": 1589
+    },
+    {
+      "epoch": 0.11078981291154234,
+      "grad_norm": 0.7296519362331932,
+      "learning_rate": 6.811104696474356e-07,
+      "loss": 1.5423,
+      "step": 1590
+    },
+    {
+      "epoch": 0.11085949203915967,
+      "grad_norm": 0.718482315453647,
+      "learning_rate": 6.810869425073959e-07,
+      "loss": 1.6452,
+      "step": 1591
+    },
+    {
+      "epoch": 0.11092917116677699,
+      "grad_norm": 0.722378080045864,
+      "learning_rate": 6.81063401178262e-07,
+      "loss": 1.4309,
+      "step": 1592
+    },
+    {
+      "epoch": 0.11099885029439431,
+      "grad_norm": 0.6985374571080087,
+      "learning_rate": 6.810398456611623e-07,
+      "loss": 1.5046,
+      "step": 1593
+    },
+    {
+      "epoch": 0.11106852942201163,
+      "grad_norm": 0.7735897379532881,
+      "learning_rate": 6.810162759572252e-07,
+      "loss": 1.4598,
+      "step": 1594
+    },
+    {
+      "epoch": 0.11113820854962896,
+      "grad_norm": 0.7141646454049914,
+      "learning_rate": 6.809926920675806e-07,
+      "loss": 1.4788,
+      "step": 1595
+    },
+    {
+      "epoch": 0.11120788767724628,
+      "grad_norm": 0.7004030395772681,
+      "learning_rate": 6.809690939933585e-07,
+      "loss": 1.5618,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1112775668048636,
+      "grad_norm": 0.7195195106477383,
+      "learning_rate": 6.809454817356897e-07,
+      "loss": 1.5963,
+      "step": 1597
+    },
+    {
+      "epoch": 0.11134724593248092,
+      "grad_norm": 0.6591270162532643,
+      "learning_rate": 6.80921855295706e-07,
+      "loss": 1.4725,
+      "step": 1598
+    },
+    {
+      "epoch": 0.11141692506009825,
+      "grad_norm": 0.7034075202200033,
+      "learning_rate": 6.808982146745393e-07,
+      "loss": 1.5809,
+      "step": 1599
+    },
+    {
+      "epoch": 0.11148660418771557,
+      "grad_norm": 0.677382514036192,
+      "learning_rate": 6.808745598733229e-07,
+      "loss": 1.4277,
+      "step": 1600
+    },
+    {
+      "epoch": 0.11155628331533289,
+      "grad_norm": 0.6988131968542942,
+      "learning_rate": 6.8085089089319e-07,
+      "loss": 1.4795,
+      "step": 1601
+    },
+    {
+      "epoch": 0.11162596244295021,
+      "grad_norm": 0.6558486580469053,
+      "learning_rate": 6.808272077352751e-07,
+      "loss": 1.5151,
+      "step": 1602
+    },
+    {
+      "epoch": 0.11169564157056754,
+      "grad_norm": 0.7466317982625242,
+      "learning_rate": 6.808035104007131e-07,
+      "loss": 1.5656,
+      "step": 1603
+    },
+    {
+      "epoch": 0.11176532069818486,
+      "grad_norm": 0.7023736399262505,
+      "learning_rate": 6.807797988906397e-07,
+      "loss": 1.5659,
+      "step": 1604
+    },
+    {
+      "epoch": 0.11183499982580218,
+      "grad_norm": 0.6949112857794784,
+      "learning_rate": 6.807560732061909e-07,
+      "loss": 1.6362,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1119046789534195,
+      "grad_norm": 0.7026342000850446,
+      "learning_rate": 6.807323333485041e-07,
+      "loss": 1.4273,
+      "step": 1606
+    },
+    {
+      "epoch": 0.11197435808103683,
+      "grad_norm": 0.6892676901830094,
+      "learning_rate": 6.807085793187167e-07,
+      "loss": 1.6011,
+      "step": 1607
+    },
+    {
+      "epoch": 0.11204403720865415,
+      "grad_norm": 0.7217360369844035,
+      "learning_rate": 6.80684811117967e-07,
+      "loss": 1.5082,
+      "step": 1608
+    },
+    {
+      "epoch": 0.11211371633627147,
+      "grad_norm": 0.7305733477997847,
+      "learning_rate": 6.806610287473942e-07,
+      "loss": 1.4838,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1121833954638888,
+      "grad_norm": 0.7356577124965377,
+      "learning_rate": 6.806372322081379e-07,
+      "loss": 1.5469,
+      "step": 1610
+    },
+    {
+      "epoch": 0.11225307459150612,
+      "grad_norm": 0.678810220355509,
+      "learning_rate": 6.806134215013386e-07,
+      "loss": 1.4224,
+      "step": 1611
+    },
+    {
+      "epoch": 0.11232275371912344,
+      "grad_norm": 0.7015202486376189,
+      "learning_rate": 6.80589596628137e-07,
+      "loss": 1.5777,
+      "step": 1612
+    },
+    {
+      "epoch": 0.11239243284674076,
+      "grad_norm": 0.7235445572079848,
+      "learning_rate": 6.805657575896753e-07,
+      "loss": 1.5127,
+      "step": 1613
+    },
+    {
+      "epoch": 0.11246211197435808,
+      "grad_norm": 0.7302277793966115,
+      "learning_rate": 6.805419043870957e-07,
+      "loss": 1.5286,
+      "step": 1614
+    },
+    {
+      "epoch": 0.1125317911019754,
+      "grad_norm": 0.7486458593950519,
+      "learning_rate": 6.805180370215413e-07,
+      "loss": 1.6699,
+      "step": 1615
+    },
+    {
+      "epoch": 0.11260147022959273,
+      "grad_norm": 0.7586164782838338,
+      "learning_rate": 6.804941554941558e-07,
+      "loss": 1.5705,
+      "step": 1616
+    },
+    {
+      "epoch": 0.11267114935721005,
+      "grad_norm": 0.7399086186510679,
+      "learning_rate": 6.804702598060838e-07,
+      "loss": 1.6787,
+      "step": 1617
+    },
+    {
+      "epoch": 0.11274082848482737,
+      "grad_norm": 0.6983175095680024,
+      "learning_rate": 6.804463499584704e-07,
+      "loss": 1.7091,
+      "step": 1618
+    },
+    {
+      "epoch": 0.1128105076124447,
+      "grad_norm": 0.6907069923114252,
+      "learning_rate": 6.804224259524613e-07,
+      "loss": 1.6168,
+      "step": 1619
+    },
+    {
+      "epoch": 0.11288018674006202,
+      "grad_norm": 0.7384786246263818,
+      "learning_rate": 6.803984877892031e-07,
+      "loss": 1.516,
+      "step": 1620
+    },
+    {
+      "epoch": 0.11294986586767934,
+      "grad_norm": 0.7134317876278762,
+      "learning_rate": 6.80374535469843e-07,
+      "loss": 1.4856,
+      "step": 1621
+    },
+    {
+      "epoch": 0.11301954499529666,
+      "grad_norm": 0.7270747064059724,
+      "learning_rate": 6.803505689955286e-07,
+      "loss": 1.5908,
+      "step": 1622
+    },
+    {
+      "epoch": 0.11308922412291399,
+      "grad_norm": 0.7517895848474664,
+      "learning_rate": 6.803265883674087e-07,
+      "loss": 1.5437,
+      "step": 1623
+    },
+    {
+      "epoch": 0.11315890325053131,
+      "grad_norm": 0.6932504297112048,
+      "learning_rate": 6.803025935866324e-07,
+      "loss": 1.5541,
+      "step": 1624
+    },
+    {
+      "epoch": 0.11322858237814863,
+      "grad_norm": 0.7251577977979223,
+      "learning_rate": 6.802785846543495e-07,
+      "loss": 1.566,
+      "step": 1625
+    },
+    {
+      "epoch": 0.11329826150576595,
+      "grad_norm": 0.7127719443354646,
+      "learning_rate": 6.802545615717106e-07,
+      "loss": 1.6241,
+      "step": 1626
+    },
+    {
+      "epoch": 0.11336794063338328,
+      "grad_norm": 0.7287839552408284,
+      "learning_rate": 6.80230524339867e-07,
+      "loss": 1.6895,
+      "step": 1627
+    },
+    {
+      "epoch": 0.11343761976100059,
+      "grad_norm": 0.7003344394380433,
+      "learning_rate": 6.802064729599706e-07,
+      "loss": 1.5245,
+      "step": 1628
+    },
+    {
+      "epoch": 0.11350729888861791,
+      "grad_norm": 0.7174216722950388,
+      "learning_rate": 6.80182407433174e-07,
+      "loss": 1.6864,
+      "step": 1629
+    },
+    {
+      "epoch": 0.11357697801623523,
+      "grad_norm": 0.6850433524557609,
+      "learning_rate": 6.801583277606304e-07,
+      "loss": 1.5079,
+      "step": 1630
+    },
+    {
+      "epoch": 0.11364665714385255,
+      "grad_norm": 0.6871563613458431,
+      "learning_rate": 6.801342339434937e-07,
+      "loss": 1.4594,
+      "step": 1631
+    },
+    {
+      "epoch": 0.11371633627146988,
+      "grad_norm": 0.7170332340740067,
+      "learning_rate": 6.801101259829188e-07,
+      "loss": 1.656,
+      "step": 1632
+    },
+    {
+      "epoch": 0.1137860153990872,
+      "grad_norm": 0.7164452063926988,
+      "learning_rate": 6.800860038800607e-07,
+      "loss": 1.5006,
+      "step": 1633
+    },
+    {
+      "epoch": 0.11385569452670452,
+      "grad_norm": 0.6937900817795392,
+      "learning_rate": 6.800618676360755e-07,
+      "loss": 1.4742,
+      "step": 1634
+    },
+    {
+      "epoch": 0.11392537365432184,
+      "grad_norm": 0.7023556028757573,
+      "learning_rate": 6.800377172521199e-07,
+      "loss": 1.6291,
+      "step": 1635
+    },
+    {
+      "epoch": 0.11399505278193917,
+      "grad_norm": 0.7169038482009281,
+      "learning_rate": 6.800135527293511e-07,
+      "loss": 1.5947,
+      "step": 1636
+    },
+    {
+      "epoch": 0.11406473190955649,
+      "grad_norm": 0.7442205747389822,
+      "learning_rate": 6.799893740689272e-07,
+      "loss": 1.5385,
+      "step": 1637
+    },
+    {
+      "epoch": 0.11413441103717381,
+      "grad_norm": 0.6974269350674046,
+      "learning_rate": 6.79965181272007e-07,
+      "loss": 1.44,
+      "step": 1638
+    },
+    {
+      "epoch": 0.11420409016479113,
+      "grad_norm": 0.7202760860472499,
+      "learning_rate": 6.799409743397497e-07,
+      "loss": 1.5094,
+      "step": 1639
+    },
+    {
+      "epoch": 0.11427376929240846,
+      "grad_norm": 0.7228977897934166,
+      "learning_rate": 6.799167532733153e-07,
+      "loss": 1.7074,
+      "step": 1640
+    },
+    {
+      "epoch": 0.11434344842002578,
+      "grad_norm": 0.7422510209307608,
+      "learning_rate": 6.798925180738649e-07,
+      "loss": 1.4843,
+      "step": 1641
+    },
+    {
+      "epoch": 0.1144131275476431,
+      "grad_norm": 0.7648101870166034,
+      "learning_rate": 6.798682687425594e-07,
+      "loss": 1.6436,
+      "step": 1642
+    },
+    {
+      "epoch": 0.11448280667526042,
+      "grad_norm": 0.7794274991636997,
+      "learning_rate": 6.798440052805611e-07,
+      "loss": 1.6088,
+      "step": 1643
+    },
+    {
+      "epoch": 0.11455248580287775,
+      "grad_norm": 0.7468761912839483,
+      "learning_rate": 6.79819727689033e-07,
+      "loss": 1.6533,
+      "step": 1644
+    },
+    {
+      "epoch": 0.11462216493049507,
+      "grad_norm": 0.674230679158213,
+      "learning_rate": 6.79795435969138e-07,
+      "loss": 1.5058,
+      "step": 1645
+    },
+    {
+      "epoch": 0.11469184405811239,
+      "grad_norm": 0.7430350345094352,
+      "learning_rate": 6.797711301220406e-07,
+      "loss": 1.545,
+      "step": 1646
+    },
+    {
+      "epoch": 0.11476152318572971,
+      "grad_norm": 0.7276113930617449,
+      "learning_rate": 6.797468101489056e-07,
+      "loss": 1.5441,
+      "step": 1647
+    },
+    {
+      "epoch": 0.11483120231334704,
+      "grad_norm": 0.7398915067049099,
+      "learning_rate": 6.797224760508984e-07,
+      "loss": 1.6512,
+      "step": 1648
+    },
+    {
+      "epoch": 0.11490088144096436,
+      "grad_norm": 0.68940533642839,
+      "learning_rate": 6.796981278291849e-07,
+      "loss": 1.6151,
+      "step": 1649
+    },
+    {
+      "epoch": 0.11497056056858168,
+      "grad_norm": 0.7555918014240699,
+      "learning_rate": 6.796737654849322e-07,
+      "loss": 1.5668,
+      "step": 1650
+    },
+    {
+      "epoch": 0.115040239696199,
+      "grad_norm": 0.7099296673799076,
+      "learning_rate": 6.796493890193077e-07,
+      "loss": 1.4325,
+      "step": 1651
+    },
+    {
+      "epoch": 0.11510991882381633,
+      "grad_norm": 0.7595492751199421,
+      "learning_rate": 6.796249984334797e-07,
+      "loss": 1.6228,
+      "step": 1652
+    },
+    {
+      "epoch": 0.11517959795143365,
+      "grad_norm": 0.7632915208578668,
+      "learning_rate": 6.796005937286167e-07,
+      "loss": 1.5813,
+      "step": 1653
+    },
+    {
+      "epoch": 0.11524927707905097,
+      "grad_norm": 0.6922449395291023,
+      "learning_rate": 6.795761749058885e-07,
+      "loss": 1.5088,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1153189562066683,
+      "grad_norm": 0.7099657991630075,
+      "learning_rate": 6.795517419664653e-07,
+      "loss": 1.5136,
+      "step": 1655
+    },
+    {
+      "epoch": 0.11538863533428562,
+      "grad_norm": 0.7613853045497073,
+      "learning_rate": 6.795272949115179e-07,
+      "loss": 1.538,
+      "step": 1656
+    },
+    {
+      "epoch": 0.11545831446190294,
+      "grad_norm": 0.7195462148384297,
+      "learning_rate": 6.795028337422179e-07,
+      "loss": 1.6404,
+      "step": 1657
+    },
+    {
+      "epoch": 0.11552799358952026,
+      "grad_norm": 0.6747873760990837,
+      "learning_rate": 6.794783584597375e-07,
+      "loss": 1.5352,
+      "step": 1658
+    },
+    {
+      "epoch": 0.11559767271713758,
+      "grad_norm": 0.6519250279592536,
+      "learning_rate": 6.794538690652497e-07,
+      "loss": 1.5127,
+      "step": 1659
+    },
+    {
+      "epoch": 0.1156673518447549,
+      "grad_norm": 0.7162973540123703,
+      "learning_rate": 6.794293655599279e-07,
+      "loss": 1.4985,
+      "step": 1660
+    },
+    {
+      "epoch": 0.11573703097237223,
+      "grad_norm": 0.7358283562766068,
+      "learning_rate": 6.794048479449463e-07,
+      "loss": 1.4902,
+      "step": 1661
+    },
+    {
+      "epoch": 0.11580671009998955,
+      "grad_norm": 0.718594579591006,
+      "learning_rate": 6.793803162214801e-07,
+      "loss": 1.5252,
+      "step": 1662
+    },
+    {
+      "epoch": 0.11587638922760687,
+      "grad_norm": 0.6804222879849213,
+      "learning_rate": 6.793557703907049e-07,
+      "loss": 1.4665,
+      "step": 1663
+    },
+    {
+      "epoch": 0.1159460683552242,
+      "grad_norm": 0.6597234535748099,
+      "learning_rate": 6.793312104537968e-07,
+      "loss": 1.5345,
+      "step": 1664
+    },
+    {
+      "epoch": 0.11601574748284152,
+      "grad_norm": 0.6858789541075233,
+      "learning_rate": 6.793066364119327e-07,
+      "loss": 1.3725,
+      "step": 1665
+    },
+    {
+      "epoch": 0.11608542661045884,
+      "grad_norm": 0.7023154283430406,
+      "learning_rate": 6.792820482662906e-07,
+      "loss": 1.6363,
+      "step": 1666
+    },
+    {
+      "epoch": 0.11615510573807616,
+      "grad_norm": 0.6987195535137968,
+      "learning_rate": 6.792574460180486e-07,
+      "loss": 1.5657,
+      "step": 1667
+    },
+    {
+      "epoch": 0.11622478486569349,
+      "grad_norm": 0.6864641357349911,
+      "learning_rate": 6.792328296683856e-07,
+      "loss": 1.4052,
+      "step": 1668
+    },
+    {
+      "epoch": 0.11629446399331081,
+      "grad_norm": 0.7222713490050489,
+      "learning_rate": 6.792081992184813e-07,
+      "loss": 1.5152,
+      "step": 1669
+    },
+    {
+      "epoch": 0.11636414312092813,
+      "grad_norm": 0.7287387894977275,
+      "learning_rate": 6.791835546695162e-07,
+      "loss": 1.5372,
+      "step": 1670
+    },
+    {
+      "epoch": 0.11643382224854545,
+      "grad_norm": 0.7969883890704471,
+      "learning_rate": 6.791588960226712e-07,
+      "loss": 1.4504,
+      "step": 1671
+    },
+    {
+      "epoch": 0.11650350137616278,
+      "grad_norm": 0.6806299876221195,
+      "learning_rate": 6.79134223279128e-07,
+      "loss": 1.4165,
+      "step": 1672
+    },
+    {
+      "epoch": 0.1165731805037801,
+      "grad_norm": 0.7626172771312602,
+      "learning_rate": 6.791095364400689e-07,
+      "loss": 1.6344,
+      "step": 1673
+    },
+    {
+      "epoch": 0.11664285963139741,
+      "grad_norm": 0.7140725948962551,
+      "learning_rate": 6.790848355066771e-07,
+      "loss": 1.5458,
+      "step": 1674
+    },
+    {
+      "epoch": 0.11671253875901473,
+      "grad_norm": 0.7015644476735204,
+      "learning_rate": 6.790601204801361e-07,
+      "loss": 1.5793,
+      "step": 1675
+    },
+    {
+      "epoch": 0.11678221788663205,
+      "grad_norm": 0.7850028189623065,
+      "learning_rate": 6.790353913616307e-07,
+      "loss": 1.503,
+      "step": 1676
+    },
+    {
+      "epoch": 0.11685189701424938,
+      "grad_norm": 0.7580222452334741,
+      "learning_rate": 6.790106481523455e-07,
+      "loss": 1.5288,
+      "step": 1677
+    },
+    {
+      "epoch": 0.1169215761418667,
+      "grad_norm": 0.8206009490254131,
+      "learning_rate": 6.789858908534665e-07,
+      "loss": 1.5293,
+      "step": 1678
+    },
+    {
+      "epoch": 0.11699125526948402,
+      "grad_norm": 0.7647388384602489,
+      "learning_rate": 6.789611194661801e-07,
+      "loss": 1.5605,
+      "step": 1679
+    },
+    {
+      "epoch": 0.11706093439710134,
+      "grad_norm": 0.7159767052729963,
+      "learning_rate": 6.789363339916733e-07,
+      "loss": 1.6455,
+      "step": 1680
+    },
+    {
+      "epoch": 0.11713061352471867,
+      "grad_norm": 0.77035475571459,
+      "learning_rate": 6.78911534431134e-07,
+      "loss": 1.6127,
+      "step": 1681
+    },
+    {
+      "epoch": 0.11720029265233599,
+      "grad_norm": 0.6771402586909827,
+      "learning_rate": 6.788867207857505e-07,
+      "loss": 1.4831,
+      "step": 1682
+    },
+    {
+      "epoch": 0.11726997177995331,
+      "grad_norm": 0.7260329904784459,
+      "learning_rate": 6.78861893056712e-07,
+      "loss": 1.5792,
+      "step": 1683
+    },
+    {
+      "epoch": 0.11733965090757063,
+      "grad_norm": 0.6889949040174258,
+      "learning_rate": 6.788370512452083e-07,
+      "loss": 1.5853,
+      "step": 1684
+    },
+    {
+      "epoch": 0.11740933003518796,
+      "grad_norm": 0.7097739042353229,
+      "learning_rate": 6.7881219535243e-07,
+      "loss": 1.6495,
+      "step": 1685
+    },
+    {
+      "epoch": 0.11747900916280528,
+      "grad_norm": 0.76869404066403,
+      "learning_rate": 6.78787325379568e-07,
+      "loss": 1.5965,
+      "step": 1686
+    },
+    {
+      "epoch": 0.1175486882904226,
+      "grad_norm": 0.7519933050837854,
+      "learning_rate": 6.787624413278143e-07,
+      "loss": 1.536,
+      "step": 1687
+    },
+    {
+      "epoch": 0.11761836741803992,
+      "grad_norm": 0.6665676367181234,
+      "learning_rate": 6.787375431983613e-07,
+      "loss": 1.4836,
+      "step": 1688
+    },
+    {
+      "epoch": 0.11768804654565725,
+      "grad_norm": 0.6976889194202481,
+      "learning_rate": 6.787126309924023e-07,
+      "loss": 1.5132,
+      "step": 1689
+    },
+    {
+      "epoch": 0.11775772567327457,
+      "grad_norm": 0.6997377024825535,
+      "learning_rate": 6.786877047111309e-07,
+      "loss": 1.6681,
+      "step": 1690
+    },
+    {
+      "epoch": 0.11782740480089189,
+      "grad_norm": 0.7204082985724929,
+      "learning_rate": 6.786627643557416e-07,
+      "loss": 1.5786,
+      "step": 1691
+    },
+    {
+      "epoch": 0.11789708392850921,
+      "grad_norm": 0.6949032262013504,
+      "learning_rate": 6.7863780992743e-07,
+      "loss": 1.5075,
+      "step": 1692
+    },
+    {
+      "epoch": 0.11796676305612654,
+      "grad_norm": 0.7113507414950234,
+      "learning_rate": 6.786128414273917e-07,
+      "loss": 1.5558,
+      "step": 1693
+    },
+    {
+      "epoch": 0.11803644218374386,
+      "grad_norm": 0.7369337445932507,
+      "learning_rate": 6.785878588568232e-07,
+      "loss": 1.7454,
+      "step": 1694
+    },
+    {
+      "epoch": 0.11810612131136118,
+      "grad_norm": 0.7139653769172889,
+      "learning_rate": 6.785628622169219e-07,
+      "loss": 1.5648,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1181758004389785,
+      "grad_norm": 0.7636368697269599,
+      "learning_rate": 6.785378515088854e-07,
+      "loss": 1.6432,
+      "step": 1696
+    },
+    {
+      "epoch": 0.11824547956659583,
+      "grad_norm": 0.6979859259064816,
+      "learning_rate": 6.785128267339125e-07,
+      "loss": 1.6527,
+      "step": 1697
+    },
+    {
+      "epoch": 0.11831515869421315,
+      "grad_norm": 0.7449511856669205,
+      "learning_rate": 6.784877878932024e-07,
+      "loss": 1.5515,
+      "step": 1698
+    },
+    {
+      "epoch": 0.11838483782183047,
+      "grad_norm": 0.7035936657977012,
+      "learning_rate": 6.784627349879551e-07,
+      "loss": 1.476,
+      "step": 1699
+    },
+    {
+      "epoch": 0.1184545169494478,
+      "grad_norm": 0.7051512194691226,
+      "learning_rate": 6.784376680193709e-07,
+      "loss": 1.5173,
+      "step": 1700
+    },
+    {
+      "epoch": 0.11852419607706512,
+      "grad_norm": 0.6909050176678639,
+      "learning_rate": 6.784125869886512e-07,
+      "loss": 1.5744,
+      "step": 1701
+    },
+    {
+      "epoch": 0.11859387520468244,
+      "grad_norm": 0.7165979408686158,
+      "learning_rate": 6.78387491896998e-07,
+      "loss": 1.4485,
+      "step": 1702
+    },
+    {
+      "epoch": 0.11866355433229976,
+      "grad_norm": 0.7430454803926803,
+      "learning_rate": 6.783623827456139e-07,
+      "loss": 1.601,
+      "step": 1703
+    },
+    {
+      "epoch": 0.11873323345991708,
+      "grad_norm": 0.7117476169672308,
+      "learning_rate": 6.783372595357023e-07,
+      "loss": 1.3338,
+      "step": 1704
+    },
+    {
+      "epoch": 0.1188029125875344,
+      "grad_norm": 0.7970341950051524,
+      "learning_rate": 6.783121222684668e-07,
+      "loss": 1.5163,
+      "step": 1705
+    },
+    {
+      "epoch": 0.11887259171515173,
+      "grad_norm": 0.7299219462336711,
+      "learning_rate": 6.782869709451125e-07,
+      "loss": 1.527,
+      "step": 1706
+    },
+    {
+      "epoch": 0.11894227084276905,
+      "grad_norm": 0.7140113860452822,
+      "learning_rate": 6.782618055668442e-07,
+      "loss": 1.5946,
+      "step": 1707
+    },
+    {
+      "epoch": 0.11901194997038637,
+      "grad_norm": 0.8280777358865281,
+      "learning_rate": 6.782366261348682e-07,
+      "loss": 1.5883,
+      "step": 1708
+    },
+    {
+      "epoch": 0.1190816290980037,
+      "grad_norm": 0.761620219181658,
+      "learning_rate": 6.782114326503911e-07,
+      "loss": 1.5823,
+      "step": 1709
+    },
+    {
+      "epoch": 0.11915130822562102,
+      "grad_norm": 0.8154486412727795,
+      "learning_rate": 6.781862251146201e-07,
+      "loss": 1.639,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11922098735323834,
+      "grad_norm": 0.729402725558841,
+      "learning_rate": 6.781610035287634e-07,
+      "loss": 1.4666,
+      "step": 1711
+    },
+    {
+      "epoch": 0.11929066648085566,
+      "grad_norm": 0.7089244714070104,
+      "learning_rate": 6.781357678940296e-07,
+      "loss": 1.612,
+      "step": 1712
+    },
+    {
+      "epoch": 0.11936034560847299,
+      "grad_norm": 0.663561610229027,
+      "learning_rate": 6.781105182116277e-07,
+      "loss": 1.4386,
+      "step": 1713
+    },
+    {
+      "epoch": 0.11943002473609031,
+      "grad_norm": 0.7533233166790155,
+      "learning_rate": 6.780852544827683e-07,
+      "loss": 1.543,
+      "step": 1714
+    },
+    {
+      "epoch": 0.11949970386370763,
+      "grad_norm": 0.6484805021471591,
+      "learning_rate": 6.780599767086617e-07,
+      "loss": 1.4806,
+      "step": 1715
+    },
+    {
+      "epoch": 0.11956938299132495,
+      "grad_norm": 0.7670450412169437,
+      "learning_rate": 6.780346848905196e-07,
+      "loss": 1.569,
+      "step": 1716
+    },
+    {
+      "epoch": 0.11963906211894228,
+      "grad_norm": 0.6593231059169545,
+      "learning_rate": 6.780093790295537e-07,
+      "loss": 1.4938,
+      "step": 1717
+    },
+    {
+      "epoch": 0.1197087412465596,
+      "grad_norm": 0.7301933487162761,
+      "learning_rate": 6.779840591269766e-07,
+      "loss": 1.5712,
+      "step": 1718
+    },
+    {
+      "epoch": 0.11977842037417692,
+      "grad_norm": 0.703158568302166,
+      "learning_rate": 6.779587251840021e-07,
+      "loss": 1.6006,
+      "step": 1719
+    },
+    {
+      "epoch": 0.11984809950179423,
+      "grad_norm": 0.7672500156829497,
+      "learning_rate": 6.779333772018441e-07,
+      "loss": 1.5412,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11991777862941155,
+      "grad_norm": 0.7695414524452814,
+      "learning_rate": 6.779080151817172e-07,
+      "loss": 1.6895,
+      "step": 1721
+    },
+    {
+      "epoch": 0.11998745775702888,
+      "grad_norm": 0.677453250708521,
+      "learning_rate": 6.778826391248369e-07,
+      "loss": 1.5178,
+      "step": 1722
+    },
+    {
+      "epoch": 0.1200571368846462,
+      "grad_norm": 0.707782732326732,
+      "learning_rate": 6.778572490324192e-07,
+      "loss": 1.5283,
+      "step": 1723
+    },
+    {
+      "epoch": 0.12012681601226352,
+      "grad_norm": 0.6959585255791797,
+      "learning_rate": 6.778318449056811e-07,
+      "loss": 1.4923,
+      "step": 1724
+    },
+    {
+      "epoch": 0.12019649513988084,
+      "grad_norm": 0.7472007496068623,
+      "learning_rate": 6.778064267458396e-07,
+      "loss": 1.5479,
+      "step": 1725
+    },
+    {
+      "epoch": 0.12026617426749817,
+      "grad_norm": 0.7676470925470961,
+      "learning_rate": 6.77780994554113e-07,
+      "loss": 1.62,
+      "step": 1726
+    },
+    {
+      "epoch": 0.12033585339511549,
+      "grad_norm": 0.6976928226879325,
+      "learning_rate": 6.777555483317201e-07,
+      "loss": 1.5113,
+      "step": 1727
+    },
+    {
+      "epoch": 0.12040553252273281,
+      "grad_norm": 0.748334145125312,
+      "learning_rate": 6.777300880798806e-07,
+      "loss": 1.539,
+      "step": 1728
+    },
+    {
+      "epoch": 0.12047521165035013,
+      "grad_norm": 0.677784509849018,
+      "learning_rate": 6.777046137998139e-07,
+      "loss": 1.4158,
+      "step": 1729
+    },
+    {
+      "epoch": 0.12054489077796746,
+      "grad_norm": 0.6416320964645821,
+      "learning_rate": 6.776791254927415e-07,
+      "loss": 1.4458,
+      "step": 1730
+    },
+    {
+      "epoch": 0.12061456990558478,
+      "grad_norm": 0.7232355551723145,
+      "learning_rate": 6.776536231598843e-07,
+      "loss": 1.5007,
+      "step": 1731
+    },
+    {
+      "epoch": 0.1206842490332021,
+      "grad_norm": 0.6940114252253201,
+      "learning_rate": 6.776281068024648e-07,
+      "loss": 1.5597,
+      "step": 1732
+    },
+    {
+      "epoch": 0.12075392816081942,
+      "grad_norm": 0.766955235541318,
+      "learning_rate": 6.776025764217057e-07,
+      "loss": 1.6602,
+      "step": 1733
+    },
+    {
+      "epoch": 0.12082360728843675,
+      "grad_norm": 0.6672225734828413,
+      "learning_rate": 6.775770320188304e-07,
+      "loss": 1.5247,
+      "step": 1734
+    },
+    {
+      "epoch": 0.12089328641605407,
+      "grad_norm": 0.7381716051530299,
+      "learning_rate": 6.77551473595063e-07,
+      "loss": 1.5521,
+      "step": 1735
+    },
+    {
+      "epoch": 0.12096296554367139,
+      "grad_norm": 0.6818041083316355,
+      "learning_rate": 6.775259011516285e-07,
+      "loss": 1.5988,
+      "step": 1736
+    },
+    {
+      "epoch": 0.12103264467128871,
+      "grad_norm": 0.7532425489494298,
+      "learning_rate": 6.775003146897523e-07,
+      "loss": 1.5768,
+      "step": 1737
+    },
+    {
+      "epoch": 0.12110232379890604,
+      "grad_norm": 0.7004792129217193,
+      "learning_rate": 6.774747142106604e-07,
+      "loss": 1.4175,
+      "step": 1738
+    },
+    {
+      "epoch": 0.12117200292652336,
+      "grad_norm": 0.6979999270979104,
+      "learning_rate": 6.774490997155799e-07,
+      "loss": 1.5503,
+      "step": 1739
+    },
+    {
+      "epoch": 0.12124168205414068,
+      "grad_norm": 0.7309248475402028,
+      "learning_rate": 6.774234712057381e-07,
+      "loss": 1.5828,
+      "step": 1740
+    },
+    {
+      "epoch": 0.121311361181758,
+      "grad_norm": 0.7698670180342181,
+      "learning_rate": 6.773978286823632e-07,
+      "loss": 1.6536,
+      "step": 1741
+    },
+    {
+      "epoch": 0.12138104030937533,
+      "grad_norm": 0.751428111254726,
+      "learning_rate": 6.773721721466841e-07,
+      "loss": 1.7044,
+      "step": 1742
+    },
+    {
+      "epoch": 0.12145071943699265,
+      "grad_norm": 0.6960299296307737,
+      "learning_rate": 6.773465015999302e-07,
+      "loss": 1.5352,
+      "step": 1743
+    },
+    {
+      "epoch": 0.12152039856460997,
+      "grad_norm": 0.7043981294939846,
+      "learning_rate": 6.773208170433319e-07,
+      "loss": 1.5553,
+      "step": 1744
+    },
+    {
+      "epoch": 0.1215900776922273,
+      "grad_norm": 0.7442626020263271,
+      "learning_rate": 6.772951184781199e-07,
+      "loss": 1.4759,
+      "step": 1745
+    },
+    {
+      "epoch": 0.12165975681984462,
+      "grad_norm": 0.7031340655574903,
+      "learning_rate": 6.772694059055255e-07,
+      "loss": 1.5555,
+      "step": 1746
+    },
+    {
+      "epoch": 0.12172943594746194,
+      "grad_norm": 0.7117663767653726,
+      "learning_rate": 6.772436793267814e-07,
+      "loss": 1.5111,
+      "step": 1747
+    },
+    {
+      "epoch": 0.12179911507507926,
+      "grad_norm": 0.6919303229240988,
+      "learning_rate": 6.772179387431202e-07,
+      "loss": 1.5557,
+      "step": 1748
+    },
+    {
+      "epoch": 0.12186879420269658,
+      "grad_norm": 0.7151382209745435,
+      "learning_rate": 6.771921841557755e-07,
+      "loss": 1.5395,
+      "step": 1749
+    },
+    {
+      "epoch": 0.1219384733303139,
+      "grad_norm": 0.7159206817537367,
+      "learning_rate": 6.771664155659814e-07,
+      "loss": 1.5215,
+      "step": 1750
+    },
+    {
+      "epoch": 0.12200815245793123,
+      "grad_norm": 0.7647443834562503,
+      "learning_rate": 6.771406329749728e-07,
+      "loss": 1.5614,
+      "step": 1751
+    },
+    {
+      "epoch": 0.12207783158554855,
+      "grad_norm": 0.7242339466839115,
+      "learning_rate": 6.771148363839854e-07,
+      "loss": 1.4939,
+      "step": 1752
+    },
+    {
+      "epoch": 0.12214751071316587,
+      "grad_norm": 0.6616100019536842,
+      "learning_rate": 6.770890257942553e-07,
+      "loss": 1.461,
+      "step": 1753
+    },
+    {
+      "epoch": 0.1222171898407832,
+      "grad_norm": 0.688560043683067,
+      "learning_rate": 6.770632012070195e-07,
+      "loss": 1.5556,
+      "step": 1754
+    },
+    {
+      "epoch": 0.12228686896840052,
+      "grad_norm": 0.7160076848018615,
+      "learning_rate": 6.770373626235155e-07,
+      "loss": 1.5213,
+      "step": 1755
+    },
+    {
+      "epoch": 0.12235654809601784,
+      "grad_norm": 0.7547550213528791,
+      "learning_rate": 6.770115100449814e-07,
+      "loss": 1.6089,
+      "step": 1756
+    },
+    {
+      "epoch": 0.12242622722363516,
+      "grad_norm": 0.6926471414125789,
+      "learning_rate": 6.769856434726564e-07,
+      "loss": 1.6533,
+      "step": 1757
+    },
+    {
+      "epoch": 0.12249590635125249,
+      "grad_norm": 0.8418470820434377,
+      "learning_rate": 6.769597629077799e-07,
+      "loss": 1.616,
+      "step": 1758
+    },
+    {
+      "epoch": 0.12256558547886981,
+      "grad_norm": 0.7122046268206644,
+      "learning_rate": 6.769338683515921e-07,
+      "loss": 1.4934,
+      "step": 1759
+    },
+    {
+      "epoch": 0.12263526460648713,
+      "grad_norm": 0.6997019866605317,
+      "learning_rate": 6.76907959805334e-07,
+      "loss": 1.5801,
+      "step": 1760
+    },
+    {
+      "epoch": 0.12270494373410445,
+      "grad_norm": 0.7229950358350811,
+      "learning_rate": 6.768820372702473e-07,
+      "loss": 1.5317,
+      "step": 1761
+    },
+    {
+      "epoch": 0.12277462286172178,
+      "grad_norm": 0.6836420011968071,
+      "learning_rate": 6.768561007475743e-07,
+      "loss": 1.5504,
+      "step": 1762
+    },
+    {
+      "epoch": 0.1228443019893391,
+      "grad_norm": 0.729122942892889,
+      "learning_rate": 6.768301502385575e-07,
+      "loss": 1.544,
+      "step": 1763
+    },
+    {
+      "epoch": 0.12291398111695642,
+      "grad_norm": 0.7293140061308643,
+      "learning_rate": 6.768041857444408e-07,
+      "loss": 1.524,
+      "step": 1764
+    },
+    {
+      "epoch": 0.12298366024457374,
+      "grad_norm": 0.7587027647685078,
+      "learning_rate": 6.767782072664686e-07,
+      "loss": 1.6564,
+      "step": 1765
+    },
+    {
+      "epoch": 0.12305333937219105,
+      "grad_norm": 0.9828600493808015,
+      "learning_rate": 6.767522148058857e-07,
+      "loss": 1.7059,
+      "step": 1766
+    },
+    {
+      "epoch": 0.12312301849980838,
+      "grad_norm": 0.7534924840916932,
+      "learning_rate": 6.767262083639376e-07,
+      "loss": 1.5926,
+      "step": 1767
+    },
+    {
+      "epoch": 0.1231926976274257,
+      "grad_norm": 0.6925293310330989,
+      "learning_rate": 6.767001879418707e-07,
+      "loss": 1.6045,
+      "step": 1768
+    },
+    {
+      "epoch": 0.12326237675504302,
+      "grad_norm": 0.6414623823691792,
+      "learning_rate": 6.76674153540932e-07,
+      "loss": 1.458,
+      "step": 1769
+    },
+    {
+      "epoch": 0.12333205588266034,
+      "grad_norm": 0.6865414673148647,
+      "learning_rate": 6.766481051623689e-07,
+      "loss": 1.6308,
+      "step": 1770
+    },
+    {
+      "epoch": 0.12340173501027767,
+      "grad_norm": 0.7751892288650261,
+      "learning_rate": 6.766220428074302e-07,
+      "loss": 1.5195,
+      "step": 1771
+    },
+    {
+      "epoch": 0.12347141413789499,
+      "grad_norm": 0.757617472740173,
+      "learning_rate": 6.765959664773643e-07,
+      "loss": 1.5832,
+      "step": 1772
+    },
+    {
+      "epoch": 0.12354109326551231,
+      "grad_norm": 0.6845490688631454,
+      "learning_rate": 6.76569876173421e-07,
+      "loss": 1.4396,
+      "step": 1773
+    },
+    {
+      "epoch": 0.12361077239312963,
+      "grad_norm": 0.704065341495557,
+      "learning_rate": 6.765437718968508e-07,
+      "loss": 1.5786,
+      "step": 1774
+    },
+    {
+      "epoch": 0.12368045152074696,
+      "grad_norm": 0.7294914988615427,
+      "learning_rate": 6.765176536489044e-07,
+      "loss": 1.573,
+      "step": 1775
+    },
+    {
+      "epoch": 0.12375013064836428,
+      "grad_norm": 0.7297804629562296,
+      "learning_rate": 6.764915214308337e-07,
+      "loss": 1.4654,
+      "step": 1776
+    },
+    {
+      "epoch": 0.1238198097759816,
+      "grad_norm": 0.7017524980244302,
+      "learning_rate": 6.764653752438906e-07,
+      "loss": 1.6069,
+      "step": 1777
+    },
+    {
+      "epoch": 0.12388948890359892,
+      "grad_norm": 0.6905428204511646,
+      "learning_rate": 6.764392150893287e-07,
+      "loss": 1.5557,
+      "step": 1778
+    },
+    {
+      "epoch": 0.12395916803121625,
+      "grad_norm": 0.6926181291094297,
+      "learning_rate": 6.764130409684011e-07,
+      "loss": 1.6095,
+      "step": 1779
+    },
+    {
+      "epoch": 0.12402884715883357,
+      "grad_norm": 0.6707933571259431,
+      "learning_rate": 6.763868528823623e-07,
+      "loss": 1.4735,
+      "step": 1780
+    },
+    {
+      "epoch": 0.12409852628645089,
+      "grad_norm": 0.7107882072617481,
+      "learning_rate": 6.763606508324675e-07,
+      "loss": 1.5258,
+      "step": 1781
+    },
+    {
+      "epoch": 0.12416820541406821,
+      "grad_norm": 0.6739566888923207,
+      "learning_rate": 6.76334434819972e-07,
+      "loss": 1.4255,
+      "step": 1782
+    },
+    {
+      "epoch": 0.12423788454168554,
+      "grad_norm": 0.710791285317531,
+      "learning_rate": 6.763082048461322e-07,
+      "loss": 1.5633,
+      "step": 1783
+    },
+    {
+      "epoch": 0.12430756366930286,
+      "grad_norm": 0.7549679673837194,
+      "learning_rate": 6.762819609122052e-07,
+      "loss": 1.6048,
+      "step": 1784
+    },
+    {
+      "epoch": 0.12437724279692018,
+      "grad_norm": 0.6970338314567243,
+      "learning_rate": 6.762557030194489e-07,
+      "loss": 1.638,
+      "step": 1785
+    },
+    {
+      "epoch": 0.1244469219245375,
+      "grad_norm": 0.7329681304954031,
+      "learning_rate": 6.762294311691212e-07,
+      "loss": 1.6062,
+      "step": 1786
+    },
+    {
+      "epoch": 0.12451660105215483,
+      "grad_norm": 0.7518073218055812,
+      "learning_rate": 6.762031453624812e-07,
+      "loss": 1.5193,
+      "step": 1787
+    },
+    {
+      "epoch": 0.12458628017977215,
+      "grad_norm": 0.753125266754797,
+      "learning_rate": 6.761768456007888e-07,
+      "loss": 1.6375,
+      "step": 1788
+    },
+    {
+      "epoch": 0.12465595930738947,
+      "grad_norm": 0.7394329680594353,
+      "learning_rate": 6.76150531885304e-07,
+      "loss": 1.6436,
+      "step": 1789
+    },
+    {
+      "epoch": 0.1247256384350068,
+      "grad_norm": 0.6387955308763775,
+      "learning_rate": 6.761242042172882e-07,
+      "loss": 1.4788,
+      "step": 1790
+    },
+    {
+      "epoch": 0.12479531756262412,
+      "grad_norm": 0.7105153937432336,
+      "learning_rate": 6.760978625980027e-07,
+      "loss": 1.4578,
+      "step": 1791
+    },
+    {
+      "epoch": 0.12486499669024144,
+      "grad_norm": 0.7117812369399352,
+      "learning_rate": 6.760715070287101e-07,
+      "loss": 1.4701,
+      "step": 1792
+    },
+    {
+      "epoch": 0.12493467581785876,
+      "grad_norm": 0.6869069756247798,
+      "learning_rate": 6.760451375106733e-07,
+      "loss": 1.4755,
+      "step": 1793
+    },
+    {
+      "epoch": 0.12500435494547607,
+      "grad_norm": 0.6883046784262724,
+      "learning_rate": 6.76018754045156e-07,
+      "loss": 1.3836,
+      "step": 1794
+    },
+    {
+      "epoch": 0.1250740340730934,
+      "grad_norm": 0.7168411586962429,
+      "learning_rate": 6.759923566334225e-07,
+      "loss": 1.4577,
+      "step": 1795
+    },
+    {
+      "epoch": 0.12514371320071072,
+      "grad_norm": 0.753246753390711,
+      "learning_rate": 6.75965945276738e-07,
+      "loss": 1.4208,
+      "step": 1796
+    },
+    {
+      "epoch": 0.12521339232832804,
+      "grad_norm": 0.7194527123537333,
+      "learning_rate": 6.75939519976368e-07,
+      "loss": 1.5584,
+      "step": 1797
+    },
+    {
+      "epoch": 0.12528307145594536,
+      "grad_norm": 0.7148704507883208,
+      "learning_rate": 6.759130807335789e-07,
+      "loss": 1.5267,
+      "step": 1798
+    },
+    {
+      "epoch": 0.12535275058356268,
+      "grad_norm": 0.7706359160338018,
+      "learning_rate": 6.758866275496378e-07,
+      "loss": 1.622,
+      "step": 1799
+    },
+    {
+      "epoch": 0.12542242971118,
+      "grad_norm": 0.6968279687765971,
+      "learning_rate": 6.758601604258122e-07,
+      "loss": 1.5589,
+      "step": 1800
+    },
+    {
+      "epoch": 0.12549210883879733,
+      "grad_norm": 0.7075640465481938,
+      "learning_rate": 6.758336793633707e-07,
+      "loss": 1.5232,
+      "step": 1801
+    },
+    {
+      "epoch": 0.12556178796641465,
+      "grad_norm": 0.6986171636247963,
+      "learning_rate": 6.758071843635822e-07,
+      "loss": 1.5965,
+      "step": 1802
+    },
+    {
+      "epoch": 0.12563146709403197,
+      "grad_norm": 0.7080541368146622,
+      "learning_rate": 6.757806754277164e-07,
+      "loss": 1.6024,
+      "step": 1803
+    },
+    {
+      "epoch": 0.1257011462216493,
+      "grad_norm": 0.6954983639729273,
+      "learning_rate": 6.757541525570436e-07,
+      "loss": 1.4651,
+      "step": 1804
+    },
+    {
+      "epoch": 0.12577082534926662,
+      "grad_norm": 0.7513907877344518,
+      "learning_rate": 6.75727615752835e-07,
+      "loss": 1.6319,
+      "step": 1805
+    },
+    {
+      "epoch": 0.12584050447688394,
+      "grad_norm": 0.7377134006135804,
+      "learning_rate": 6.757010650163622e-07,
+      "loss": 1.4196,
+      "step": 1806
+    },
+    {
+      "epoch": 0.12591018360450126,
+      "grad_norm": 0.7379253766876244,
+      "learning_rate": 6.756745003488975e-07,
+      "loss": 1.6043,
+      "step": 1807
+    },
+    {
+      "epoch": 0.12597986273211859,
+      "grad_norm": 0.6780338111507962,
+      "learning_rate": 6.75647921751714e-07,
+      "loss": 1.5364,
+      "step": 1808
+    },
+    {
+      "epoch": 0.1260495418597359,
+      "grad_norm": 0.8363290539712473,
+      "learning_rate": 6.756213292260855e-07,
+      "loss": 1.4253,
+      "step": 1809
+    },
+    {
+      "epoch": 0.12611922098735323,
+      "grad_norm": 0.7171329233065877,
+      "learning_rate": 6.755947227732862e-07,
+      "loss": 1.6683,
+      "step": 1810
+    },
+    {
+      "epoch": 0.12618890011497055,
+      "grad_norm": 0.6956735267493139,
+      "learning_rate": 6.755681023945912e-07,
+      "loss": 1.5402,
+      "step": 1811
+    },
+    {
+      "epoch": 0.12625857924258788,
+      "grad_norm": 0.7265898886948043,
+      "learning_rate": 6.755414680912763e-07,
+      "loss": 1.5621,
+      "step": 1812
+    },
+    {
+      "epoch": 0.1263282583702052,
+      "grad_norm": 0.7832071974789533,
+      "learning_rate": 6.755148198646176e-07,
+      "loss": 1.6196,
+      "step": 1813
+    },
+    {
+      "epoch": 0.12639793749782252,
+      "grad_norm": 0.6980941771921907,
+      "learning_rate": 6.754881577158925e-07,
+      "loss": 1.469,
+      "step": 1814
+    },
+    {
+      "epoch": 0.12646761662543984,
+      "grad_norm": 0.7821845449425642,
+      "learning_rate": 6.754614816463783e-07,
+      "loss": 1.5984,
+      "step": 1815
+    },
+    {
+      "epoch": 0.12653729575305717,
+      "grad_norm": 0.7196655127524775,
+      "learning_rate": 6.754347916573539e-07,
+      "loss": 1.4687,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1266069748806745,
+      "grad_norm": 0.7132709263239155,
+      "learning_rate": 6.754080877500978e-07,
+      "loss": 1.5156,
+      "step": 1817
+    },
+    {
+      "epoch": 0.1266766540082918,
+      "grad_norm": 0.7073112967452371,
+      "learning_rate": 6.7538136992589e-07,
+      "loss": 1.5509,
+      "step": 1818
+    },
+    {
+      "epoch": 0.12674633313590913,
+      "grad_norm": 0.7151991872285213,
+      "learning_rate": 6.753546381860108e-07,
+      "loss": 1.5542,
+      "step": 1819
+    },
+    {
+      "epoch": 0.12681601226352646,
+      "grad_norm": 0.7242367578277304,
+      "learning_rate": 6.753278925317413e-07,
+      "loss": 1.5261,
+      "step": 1820
+    },
+    {
+      "epoch": 0.12688569139114378,
+      "grad_norm": 0.6651665644113366,
+      "learning_rate": 6.753011329643631e-07,
+      "loss": 1.4353,
+      "step": 1821
+    },
+    {
+      "epoch": 0.1269553705187611,
+      "grad_norm": 0.7588221525761546,
+      "learning_rate": 6.752743594851586e-07,
+      "loss": 1.4917,
+      "step": 1822
+    },
+    {
+      "epoch": 0.12702504964637842,
+      "grad_norm": 0.7726293303712665,
+      "learning_rate": 6.75247572095411e-07,
+      "loss": 1.6404,
+      "step": 1823
+    },
+    {
+      "epoch": 0.12709472877399575,
+      "grad_norm": 0.7105867827958886,
+      "learning_rate": 6.752207707964037e-07,
+      "loss": 1.484,
+      "step": 1824
+    },
+    {
+      "epoch": 0.12716440790161307,
+      "grad_norm": 0.6953351793350567,
+      "learning_rate": 6.751939555894213e-07,
+      "loss": 1.531,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1272340870292304,
+      "grad_norm": 0.6677024686574735,
+      "learning_rate": 6.75167126475749e-07,
+      "loss": 1.5333,
+      "step": 1826
+    },
+    {
+      "epoch": 0.12730376615684771,
+      "grad_norm": 0.7551124751250401,
+      "learning_rate": 6.751402834566721e-07,
+      "loss": 1.5653,
+      "step": 1827
+    },
+    {
+      "epoch": 0.12737344528446504,
+      "grad_norm": 0.7104248422758904,
+      "learning_rate": 6.751134265334772e-07,
+      "loss": 1.5667,
+      "step": 1828
+    },
+    {
+      "epoch": 0.12744312441208236,
+      "grad_norm": 0.6763604752203057,
+      "learning_rate": 6.750865557074514e-07,
+      "loss": 1.4499,
+      "step": 1829
+    },
+    {
+      "epoch": 0.12751280353969968,
+      "grad_norm": 0.6766323797495721,
+      "learning_rate": 6.750596709798822e-07,
+      "loss": 1.3995,
+      "step": 1830
+    },
+    {
+      "epoch": 0.127582482667317,
+      "grad_norm": 0.7736093226697133,
+      "learning_rate": 6.750327723520581e-07,
+      "loss": 1.5276,
+      "step": 1831
+    },
+    {
+      "epoch": 0.12765216179493433,
+      "grad_norm": 0.6719858842850682,
+      "learning_rate": 6.750058598252682e-07,
+      "loss": 1.4956,
+      "step": 1832
+    },
+    {
+      "epoch": 0.12772184092255165,
+      "grad_norm": 0.7346381580963557,
+      "learning_rate": 6.74978933400802e-07,
+      "loss": 1.5993,
+      "step": 1833
+    },
+    {
+      "epoch": 0.12779152005016897,
+      "grad_norm": 0.6808106123344357,
+      "learning_rate": 6.749519930799501e-07,
+      "loss": 1.5483,
+      "step": 1834
+    },
+    {
+      "epoch": 0.1278611991777863,
+      "grad_norm": 0.6727911376600302,
+      "learning_rate": 6.749250388640033e-07,
+      "loss": 1.5679,
+      "step": 1835
+    },
+    {
+      "epoch": 0.12793087830540362,
+      "grad_norm": 0.6908490917326462,
+      "learning_rate": 6.748980707542537e-07,
+      "loss": 1.4939,
+      "step": 1836
+    },
+    {
+      "epoch": 0.12800055743302094,
+      "grad_norm": 0.6991194309070923,
+      "learning_rate": 6.748710887519931e-07,
+      "loss": 1.5277,
+      "step": 1837
+    },
+    {
+      "epoch": 0.12807023656063826,
+      "grad_norm": 0.7535462584673673,
+      "learning_rate": 6.748440928585151e-07,
+      "loss": 1.5031,
+      "step": 1838
+    },
+    {
+      "epoch": 0.12813991568825558,
+      "grad_norm": 0.7257701505696328,
+      "learning_rate": 6.748170830751129e-07,
+      "loss": 1.6171,
+      "step": 1839
+    },
+    {
+      "epoch": 0.1282095948158729,
+      "grad_norm": 0.7197206514629374,
+      "learning_rate": 6.747900594030811e-07,
+      "loss": 1.5134,
+      "step": 1840
+    },
+    {
+      "epoch": 0.12827927394349023,
+      "grad_norm": 0.6863960162086155,
+      "learning_rate": 6.747630218437149e-07,
+      "loss": 1.6063,
+      "step": 1841
+    },
+    {
+      "epoch": 0.12834895307110755,
+      "grad_norm": 0.7451035417247105,
+      "learning_rate": 6.747359703983097e-07,
+      "loss": 1.5635,
+      "step": 1842
+    },
+    {
+      "epoch": 0.12841863219872487,
+      "grad_norm": 0.7056216450550601,
+      "learning_rate": 6.747089050681621e-07,
+      "loss": 1.4805,
+      "step": 1843
+    },
+    {
+      "epoch": 0.1284883113263422,
+      "grad_norm": 0.6961347423164467,
+      "learning_rate": 6.746818258545689e-07,
+      "loss": 1.4463,
+      "step": 1844
+    },
+    {
+      "epoch": 0.12855799045395952,
+      "grad_norm": 0.732709513808762,
+      "learning_rate": 6.746547327588279e-07,
+      "loss": 1.6429,
+      "step": 1845
+    },
+    {
+      "epoch": 0.12862766958157684,
+      "grad_norm": 0.7357288567627138,
+      "learning_rate": 6.746276257822375e-07,
+      "loss": 1.4988,
+      "step": 1846
+    },
+    {
+      "epoch": 0.12869734870919416,
+      "grad_norm": 0.6706891475934581,
+      "learning_rate": 6.746005049260967e-07,
+      "loss": 1.5309,
+      "step": 1847
+    },
+    {
+      "epoch": 0.1287670278368115,
+      "grad_norm": 0.7319525874852473,
+      "learning_rate": 6.745733701917052e-07,
+      "loss": 1.5278,
+      "step": 1848
+    },
+    {
+      "epoch": 0.1288367069644288,
+      "grad_norm": 0.6639104121057977,
+      "learning_rate": 6.745462215803632e-07,
+      "loss": 1.3677,
+      "step": 1849
+    },
+    {
+      "epoch": 0.12890638609204613,
+      "grad_norm": 0.7085480744877006,
+      "learning_rate": 6.745190590933719e-07,
+      "loss": 1.518,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12897606521966345,
+      "grad_norm": 0.7505374113469124,
+      "learning_rate": 6.744918827320328e-07,
+      "loss": 1.5333,
+      "step": 1851
+    },
+    {
+      "epoch": 0.12904574434728078,
+      "grad_norm": 0.7352795678485466,
+      "learning_rate": 6.744646924976485e-07,
+      "loss": 1.5475,
+      "step": 1852
+    },
+    {
+      "epoch": 0.1291154234748981,
+      "grad_norm": 0.7035551100291901,
+      "learning_rate": 6.744374883915218e-07,
+      "loss": 1.5256,
+      "step": 1853
+    },
+    {
+      "epoch": 0.12918510260251542,
+      "grad_norm": 0.7153574641045704,
+      "learning_rate": 6.744102704149565e-07,
+      "loss": 1.5735,
+      "step": 1854
+    },
+    {
+      "epoch": 0.12925478173013275,
+      "grad_norm": 0.6577428673649858,
+      "learning_rate": 6.743830385692569e-07,
+      "loss": 1.4839,
+      "step": 1855
+    },
+    {
+      "epoch": 0.12932446085775007,
+      "grad_norm": 0.7310828390027627,
+      "learning_rate": 6.743557928557279e-07,
+      "loss": 1.5818,
+      "step": 1856
+    },
+    {
+      "epoch": 0.1293941399853674,
+      "grad_norm": 0.6714264380020178,
+      "learning_rate": 6.743285332756753e-07,
+      "loss": 1.4134,
+      "step": 1857
+    },
+    {
+      "epoch": 0.1294638191129847,
+      "grad_norm": 0.7398456859492963,
+      "learning_rate": 6.743012598304055e-07,
+      "loss": 1.5288,
+      "step": 1858
+    },
+    {
+      "epoch": 0.12953349824060204,
+      "grad_norm": 0.7655971284671793,
+      "learning_rate": 6.742739725212255e-07,
+      "loss": 1.6144,
+      "step": 1859
+    },
+    {
+      "epoch": 0.12960317736821936,
+      "grad_norm": 0.7045382170108533,
+      "learning_rate": 6.742466713494427e-07,
+      "loss": 1.5161,
+      "step": 1860
+    },
+    {
+      "epoch": 0.12967285649583668,
+      "grad_norm": 0.7037565360640923,
+      "learning_rate": 6.742193563163656e-07,
+      "loss": 1.5649,
+      "step": 1861
+    },
+    {
+      "epoch": 0.129742535623454,
+      "grad_norm": 0.6883733361202018,
+      "learning_rate": 6.741920274233033e-07,
+      "loss": 1.5637,
+      "step": 1862
+    },
+    {
+      "epoch": 0.12981221475107133,
+      "grad_norm": 0.7285700573399307,
+      "learning_rate": 6.741646846715651e-07,
+      "loss": 1.6719,
+      "step": 1863
+    },
+    {
+      "epoch": 0.12988189387868865,
+      "grad_norm": 0.7301447510701851,
+      "learning_rate": 6.741373280624618e-07,
+      "loss": 1.5617,
+      "step": 1864
+    },
+    {
+      "epoch": 0.12995157300630597,
+      "grad_norm": 0.7031543215374925,
+      "learning_rate": 6.741099575973041e-07,
+      "loss": 1.4602,
+      "step": 1865
+    },
+    {
+      "epoch": 0.1300212521339233,
+      "grad_norm": 0.7094167448128131,
+      "learning_rate": 6.740825732774036e-07,
+      "loss": 1.616,
+      "step": 1866
+    },
+    {
+      "epoch": 0.13009093126154062,
+      "grad_norm": 0.7651565895226398,
+      "learning_rate": 6.740551751040729e-07,
+      "loss": 1.5392,
+      "step": 1867
+    },
+    {
+      "epoch": 0.13016061038915794,
+      "grad_norm": 0.7735065918268992,
+      "learning_rate": 6.740277630786246e-07,
+      "loss": 1.5778,
+      "step": 1868
+    },
+    {
+      "epoch": 0.13023028951677526,
+      "grad_norm": 0.7967705631098297,
+      "learning_rate": 6.740003372023727e-07,
+      "loss": 1.7259,
+      "step": 1869
+    },
+    {
+      "epoch": 0.13029996864439258,
+      "grad_norm": 0.6792582135508917,
+      "learning_rate": 6.739728974766312e-07,
+      "loss": 1.6366,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1303696477720099,
+      "grad_norm": 0.730703176577263,
+      "learning_rate": 6.739454439027153e-07,
+      "loss": 1.5665,
+      "step": 1871
+    },
+    {
+      "epoch": 0.13043932689962723,
+      "grad_norm": 0.6920045189349746,
+      "learning_rate": 6.739179764819405e-07,
+      "loss": 1.5654,
+      "step": 1872
+    },
+    {
+      "epoch": 0.13050900602724455,
+      "grad_norm": 0.7136204029101176,
+      "learning_rate": 6.738904952156231e-07,
+      "loss": 1.5337,
+      "step": 1873
+    },
+    {
+      "epoch": 0.13057868515486187,
+      "grad_norm": 0.6619804892238528,
+      "learning_rate": 6.738630001050801e-07,
+      "loss": 1.5429,
+      "step": 1874
+    },
+    {
+      "epoch": 0.1306483642824792,
+      "grad_norm": 0.7257253083228685,
+      "learning_rate": 6.738354911516292e-07,
+      "loss": 1.5279,
+      "step": 1875
+    },
+    {
+      "epoch": 0.13071804341009652,
+      "grad_norm": 0.7053782168665625,
+      "learning_rate": 6.738079683565885e-07,
+      "loss": 1.5864,
+      "step": 1876
+    },
+    {
+      "epoch": 0.13078772253771384,
+      "grad_norm": 0.763390428555883,
+      "learning_rate": 6.73780431721277e-07,
+      "loss": 1.5123,
+      "step": 1877
+    },
+    {
+      "epoch": 0.13085740166533116,
+      "grad_norm": 0.7091548271863509,
+      "learning_rate": 6.737528812470145e-07,
+      "loss": 1.4875,
+      "step": 1878
+    },
+    {
+      "epoch": 0.13092708079294849,
+      "grad_norm": 0.705405057687693,
+      "learning_rate": 6.737253169351209e-07,
+      "loss": 1.4578,
+      "step": 1879
+    },
+    {
+      "epoch": 0.13099675992056578,
+      "grad_norm": 0.6716338179893901,
+      "learning_rate": 6.736977387869176e-07,
+      "loss": 1.4275,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1310664390481831,
+      "grad_norm": 0.6687999558803702,
+      "learning_rate": 6.736701468037259e-07,
+      "loss": 1.4048,
+      "step": 1881
+    },
+    {
+      "epoch": 0.13113611817580043,
+      "grad_norm": 0.7167984046454882,
+      "learning_rate": 6.736425409868682e-07,
+      "loss": 1.7738,
+      "step": 1882
+    },
+    {
+      "epoch": 0.13120579730341775,
+      "grad_norm": 0.730282334768594,
+      "learning_rate": 6.736149213376672e-07,
+      "loss": 1.5536,
+      "step": 1883
+    },
+    {
+      "epoch": 0.13127547643103507,
+      "grad_norm": 0.7103695925549198,
+      "learning_rate": 6.735872878574467e-07,
+      "loss": 1.553,
+      "step": 1884
+    },
+    {
+      "epoch": 0.1313451555586524,
+      "grad_norm": 0.6854489314675175,
+      "learning_rate": 6.73559640547531e-07,
+      "loss": 1.5653,
+      "step": 1885
+    },
+    {
+      "epoch": 0.13141483468626972,
+      "grad_norm": 0.7505672021974592,
+      "learning_rate": 6.735319794092449e-07,
+      "loss": 1.5383,
+      "step": 1886
+    },
+    {
+      "epoch": 0.13148451381388704,
+      "grad_norm": 0.732423515927833,
+      "learning_rate": 6.73504304443914e-07,
+      "loss": 1.4525,
+      "step": 1887
+    },
+    {
+      "epoch": 0.13155419294150436,
+      "grad_norm": 0.685880905255038,
+      "learning_rate": 6.734766156528645e-07,
+      "loss": 1.616,
+      "step": 1888
+    },
+    {
+      "epoch": 0.13162387206912168,
+      "grad_norm": 0.6960458179247873,
+      "learning_rate": 6.734489130374234e-07,
+      "loss": 1.5326,
+      "step": 1889
+    },
+    {
+      "epoch": 0.131693551196739,
+      "grad_norm": 0.6866922904575083,
+      "learning_rate": 6.734211965989182e-07,
+      "loss": 1.6037,
+      "step": 1890
+    },
+    {
+      "epoch": 0.13176323032435633,
+      "grad_norm": 0.673778268440158,
+      "learning_rate": 6.73393466338677e-07,
+      "loss": 1.535,
+      "step": 1891
+    },
+    {
+      "epoch": 0.13183290945197365,
+      "grad_norm": 0.690886200558052,
+      "learning_rate": 6.73365722258029e-07,
+      "loss": 1.4375,
+      "step": 1892
+    },
+    {
+      "epoch": 0.13190258857959097,
+      "grad_norm": 0.6994952752280379,
+      "learning_rate": 6.733379643583036e-07,
+      "loss": 1.5044,
+      "step": 1893
+    },
+    {
+      "epoch": 0.1319722677072083,
+      "grad_norm": 0.7264263810645032,
+      "learning_rate": 6.733101926408308e-07,
+      "loss": 1.6547,
+      "step": 1894
+    },
+    {
+      "epoch": 0.13204194683482562,
+      "grad_norm": 0.7103588905968264,
+      "learning_rate": 6.732824071069419e-07,
+      "loss": 1.5439,
+      "step": 1895
+    },
+    {
+      "epoch": 0.13211162596244294,
+      "grad_norm": 0.672222207099807,
+      "learning_rate": 6.732546077579681e-07,
+      "loss": 1.491,
+      "step": 1896
+    },
+    {
+      "epoch": 0.13218130509006026,
+      "grad_norm": 0.751933233368762,
+      "learning_rate": 6.732267945952418e-07,
+      "loss": 1.6748,
+      "step": 1897
+    },
+    {
+      "epoch": 0.1322509842176776,
+      "grad_norm": 0.7135892877523559,
+      "learning_rate": 6.731989676200958e-07,
+      "loss": 1.6594,
+      "step": 1898
+    },
+    {
+      "epoch": 0.1323206633452949,
+      "grad_norm": 0.6841008584696798,
+      "learning_rate": 6.731711268338635e-07,
+      "loss": 1.555,
+      "step": 1899
+    },
+    {
+      "epoch": 0.13239034247291223,
+      "grad_norm": 0.7548485821820744,
+      "learning_rate": 6.731432722378794e-07,
+      "loss": 1.5304,
+      "step": 1900
+    },
+    {
+      "epoch": 0.13246002160052955,
+      "grad_norm": 0.7137250878688367,
+      "learning_rate": 6.73115403833478e-07,
+      "loss": 1.4984,
+      "step": 1901
+    },
+    {
+      "epoch": 0.13252970072814688,
+      "grad_norm": 0.6812981296452358,
+      "learning_rate": 6.730875216219948e-07,
+      "loss": 1.5644,
+      "step": 1902
+    },
+    {
+      "epoch": 0.1325993798557642,
+      "grad_norm": 0.7216405903372481,
+      "learning_rate": 6.730596256047663e-07,
+      "loss": 1.5853,
+      "step": 1903
+    },
+    {
+      "epoch": 0.13266905898338152,
+      "grad_norm": 0.7696458881454737,
+      "learning_rate": 6.730317157831293e-07,
+      "loss": 1.5555,
+      "step": 1904
+    },
+    {
+      "epoch": 0.13273873811099884,
+      "grad_norm": 0.7513839173466563,
+      "learning_rate": 6.730037921584209e-07,
+      "loss": 1.5812,
+      "step": 1905
+    },
+    {
+      "epoch": 0.13280841723861617,
+      "grad_norm": 0.7038080392696825,
+      "learning_rate": 6.729758547319796e-07,
+      "loss": 1.5747,
+      "step": 1906
+    },
+    {
+      "epoch": 0.1328780963662335,
+      "grad_norm": 0.7021482922132364,
+      "learning_rate": 6.729479035051443e-07,
+      "loss": 1.6157,
+      "step": 1907
+    },
+    {
+      "epoch": 0.1329477754938508,
+      "grad_norm": 0.7128968760411135,
+      "learning_rate": 6.729199384792542e-07,
+      "loss": 1.5565,
+      "step": 1908
+    },
+    {
+      "epoch": 0.13301745462146813,
+      "grad_norm": 0.7591474416194975,
+      "learning_rate": 6.728919596556496e-07,
+      "loss": 1.5124,
+      "step": 1909
+    },
+    {
+      "epoch": 0.13308713374908546,
+      "grad_norm": 0.731997601252689,
+      "learning_rate": 6.728639670356711e-07,
+      "loss": 1.5297,
+      "step": 1910
+    },
+    {
+      "epoch": 0.13315681287670278,
+      "grad_norm": 0.7360419831946642,
+      "learning_rate": 6.728359606206605e-07,
+      "loss": 1.7696,
+      "step": 1911
+    },
+    {
+      "epoch": 0.1332264920043201,
+      "grad_norm": 0.7143694573394039,
+      "learning_rate": 6.728079404119597e-07,
+      "loss": 1.4961,
+      "step": 1912
+    },
+    {
+      "epoch": 0.13329617113193742,
+      "grad_norm": 0.6871856694114608,
+      "learning_rate": 6.727799064109116e-07,
+      "loss": 1.5909,
+      "step": 1913
+    },
+    {
+      "epoch": 0.13336585025955475,
+      "grad_norm": 0.7120372611141601,
+      "learning_rate": 6.727518586188593e-07,
+      "loss": 1.6738,
+      "step": 1914
+    },
+    {
+      "epoch": 0.13343552938717207,
+      "grad_norm": 0.7250648448379915,
+      "learning_rate": 6.727237970371475e-07,
+      "loss": 1.4843,
+      "step": 1915
+    },
+    {
+      "epoch": 0.1335052085147894,
+      "grad_norm": 0.7316783552643227,
+      "learning_rate": 6.726957216671206e-07,
+      "loss": 1.537,
+      "step": 1916
+    },
+    {
+      "epoch": 0.13357488764240671,
+      "grad_norm": 0.7110209139763414,
+      "learning_rate": 6.72667632510124e-07,
+      "loss": 1.5645,
+      "step": 1917
+    },
+    {
+      "epoch": 0.13364456677002404,
+      "grad_norm": 0.7095671591176976,
+      "learning_rate": 6.72639529567504e-07,
+      "loss": 1.5458,
+      "step": 1918
+    },
+    {
+      "epoch": 0.13371424589764136,
+      "grad_norm": 0.7720388159693496,
+      "learning_rate": 6.726114128406072e-07,
+      "loss": 1.4622,
+      "step": 1919
+    },
+    {
+      "epoch": 0.13378392502525868,
+      "grad_norm": 0.7088699128516145,
+      "learning_rate": 6.72583282330781e-07,
+      "loss": 1.476,
+      "step": 1920
+    },
+    {
+      "epoch": 0.133853604152876,
+      "grad_norm": 0.6908728730889724,
+      "learning_rate": 6.725551380393735e-07,
+      "loss": 1.5533,
+      "step": 1921
+    },
+    {
+      "epoch": 0.13392328328049333,
+      "grad_norm": 0.6820942269733361,
+      "learning_rate": 6.725269799677335e-07,
+      "loss": 1.5603,
+      "step": 1922
+    },
+    {
+      "epoch": 0.13399296240811065,
+      "grad_norm": 0.6923947732216261,
+      "learning_rate": 6.724988081172102e-07,
+      "loss": 1.5315,
+      "step": 1923
+    },
+    {
+      "epoch": 0.13406264153572797,
+      "grad_norm": 0.7251093460202046,
+      "learning_rate": 6.72470622489154e-07,
+      "loss": 1.5488,
+      "step": 1924
+    },
+    {
+      "epoch": 0.1341323206633453,
+      "grad_norm": 0.7012512826060395,
+      "learning_rate": 6.724424230849153e-07,
+      "loss": 1.4948,
+      "step": 1925
+    },
+    {
+      "epoch": 0.13420199979096262,
+      "grad_norm": 0.7212684831182445,
+      "learning_rate": 6.724142099058455e-07,
+      "loss": 1.56,
+      "step": 1926
+    },
+    {
+      "epoch": 0.13427167891857994,
+      "grad_norm": 0.7048665231478228,
+      "learning_rate": 6.723859829532968e-07,
+      "loss": 1.5274,
+      "step": 1927
+    },
+    {
+      "epoch": 0.13434135804619726,
+      "grad_norm": 0.7718463262225225,
+      "learning_rate": 6.723577422286217e-07,
+      "loss": 1.7148,
+      "step": 1928
+    },
+    {
+      "epoch": 0.13441103717381458,
+      "grad_norm": 0.6792581126537035,
+      "learning_rate": 6.723294877331739e-07,
+      "loss": 1.4344,
+      "step": 1929
+    },
+    {
+      "epoch": 0.1344807163014319,
+      "grad_norm": 0.7081966504199131,
+      "learning_rate": 6.723012194683071e-07,
+      "loss": 1.5669,
+      "step": 1930
+    },
+    {
+      "epoch": 0.13455039542904923,
+      "grad_norm": 0.7534230117240622,
+      "learning_rate": 6.722729374353759e-07,
+      "loss": 1.4576,
+      "step": 1931
+    },
+    {
+      "epoch": 0.13462007455666655,
+      "grad_norm": 0.6962594541305029,
+      "learning_rate": 6.722446416357359e-07,
+      "loss": 1.5744,
+      "step": 1932
+    },
+    {
+      "epoch": 0.13468975368428387,
+      "grad_norm": 0.7580060974065307,
+      "learning_rate": 6.722163320707429e-07,
+      "loss": 1.6539,
+      "step": 1933
+    },
+    {
+      "epoch": 0.1347594328119012,
+      "grad_norm": 0.673830372373551,
+      "learning_rate": 6.721880087417536e-07,
+      "loss": 1.4916,
+      "step": 1934
+    },
+    {
+      "epoch": 0.13482911193951852,
+      "grad_norm": 0.7218253612280532,
+      "learning_rate": 6.721596716501253e-07,
+      "loss": 1.4981,
+      "step": 1935
+    },
+    {
+      "epoch": 0.13489879106713584,
+      "grad_norm": 0.7531583849102892,
+      "learning_rate": 6.721313207972162e-07,
+      "loss": 1.6348,
+      "step": 1936
+    },
+    {
+      "epoch": 0.13496847019475317,
+      "grad_norm": 0.7429401300806515,
+      "learning_rate": 6.721029561843847e-07,
+      "loss": 1.6212,
+      "step": 1937
+    },
+    {
+      "epoch": 0.1350381493223705,
+      "grad_norm": 0.6828835052353207,
+      "learning_rate": 6.720745778129899e-07,
+      "loss": 1.3943,
+      "step": 1938
+    },
+    {
+      "epoch": 0.1351078284499878,
+      "grad_norm": 0.7085091052785595,
+      "learning_rate": 6.720461856843922e-07,
+      "loss": 1.5354,
+      "step": 1939
+    },
+    {
+      "epoch": 0.13517750757760513,
+      "grad_norm": 0.7364242437633206,
+      "learning_rate": 6.720177797999519e-07,
+      "loss": 1.7678,
+      "step": 1940
+    },
+    {
+      "epoch": 0.13524718670522246,
+      "grad_norm": 0.6570866689949861,
+      "learning_rate": 6.719893601610304e-07,
+      "loss": 1.5072,
+      "step": 1941
+    },
+    {
+      "epoch": 0.13531686583283978,
+      "grad_norm": 0.7943585246747022,
+      "learning_rate": 6.719609267689896e-07,
+      "loss": 1.6721,
+      "step": 1942
+    },
+    {
+      "epoch": 0.1353865449604571,
+      "grad_norm": 0.7188461240489994,
+      "learning_rate": 6.71932479625192e-07,
+      "loss": 1.5127,
+      "step": 1943
+    },
+    {
+      "epoch": 0.13545622408807442,
+      "grad_norm": 0.7475198285216123,
+      "learning_rate": 6.719040187310009e-07,
+      "loss": 1.6445,
+      "step": 1944
+    },
+    {
+      "epoch": 0.13552590321569175,
+      "grad_norm": 0.7258450468992407,
+      "learning_rate": 6.718755440877802e-07,
+      "loss": 1.5219,
+      "step": 1945
+    },
+    {
+      "epoch": 0.13559558234330907,
+      "grad_norm": 0.7468264368633342,
+      "learning_rate": 6.718470556968946e-07,
+      "loss": 1.5828,
+      "step": 1946
+    },
+    {
+      "epoch": 0.1356652614709264,
+      "grad_norm": 0.7125015542157345,
+      "learning_rate": 6.718185535597091e-07,
+      "loss": 1.5189,
+      "step": 1947
+    },
+    {
+      "epoch": 0.1357349405985437,
+      "grad_norm": 0.7036505846596398,
+      "learning_rate": 6.717900376775899e-07,
+      "loss": 1.5386,
+      "step": 1948
+    },
+    {
+      "epoch": 0.13580461972616104,
+      "grad_norm": 0.7382895590391404,
+      "learning_rate": 6.71761508051903e-07,
+      "loss": 1.5655,
+      "step": 1949
+    },
+    {
+      "epoch": 0.13587429885377836,
+      "grad_norm": 0.7511916226599143,
+      "learning_rate": 6.717329646840162e-07,
+      "loss": 1.6689,
+      "step": 1950
+    },
+    {
+      "epoch": 0.13594397798139568,
+      "grad_norm": 0.7217544848044565,
+      "learning_rate": 6.717044075752969e-07,
+      "loss": 1.5589,
+      "step": 1951
+    },
+    {
+      "epoch": 0.136013657109013,
+      "grad_norm": 0.6802227867126514,
+      "learning_rate": 6.716758367271138e-07,
+      "loss": 1.4658,
+      "step": 1952
+    },
+    {
+      "epoch": 0.13608333623663033,
+      "grad_norm": 0.6972733413609539,
+      "learning_rate": 6.716472521408362e-07,
+      "loss": 1.5688,
+      "step": 1953
+    },
+    {
+      "epoch": 0.13615301536424765,
+      "grad_norm": 0.7289731126070428,
+      "learning_rate": 6.716186538178338e-07,
+      "loss": 1.572,
+      "step": 1954
+    },
+    {
+      "epoch": 0.13622269449186497,
+      "grad_norm": 0.7096236736018127,
+      "learning_rate": 6.715900417594769e-07,
+      "loss": 1.5759,
+      "step": 1955
+    },
+    {
+      "epoch": 0.1362923736194823,
+      "grad_norm": 0.7300802470602689,
+      "learning_rate": 6.71561415967137e-07,
+      "loss": 1.5944,
+      "step": 1956
+    },
+    {
+      "epoch": 0.13636205274709962,
+      "grad_norm": 0.7424757456231104,
+      "learning_rate": 6.715327764421858e-07,
+      "loss": 1.5659,
+      "step": 1957
+    },
+    {
+      "epoch": 0.13643173187471694,
+      "grad_norm": 0.6961374129664488,
+      "learning_rate": 6.715041231859956e-07,
+      "loss": 1.6426,
+      "step": 1958
+    },
+    {
+      "epoch": 0.13650141100233426,
+      "grad_norm": 0.7139600080889801,
+      "learning_rate": 6.714754561999395e-07,
+      "loss": 1.4646,
+      "step": 1959
+    },
+    {
+      "epoch": 0.13657109012995158,
+      "grad_norm": 0.737591170989514,
+      "learning_rate": 6.714467754853917e-07,
+      "loss": 1.5856,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1366407692575689,
+      "grad_norm": 0.7319158356469768,
+      "learning_rate": 6.714180810437263e-07,
+      "loss": 1.5179,
+      "step": 1961
+    },
+    {
+      "epoch": 0.13671044838518623,
+      "grad_norm": 0.8217089235096635,
+      "learning_rate": 6.713893728763184e-07,
+      "loss": 1.575,
+      "step": 1962
+    },
+    {
+      "epoch": 0.13678012751280355,
+      "grad_norm": 0.7349260888101283,
+      "learning_rate": 6.713606509845437e-07,
+      "loss": 1.4672,
+      "step": 1963
+    },
+    {
+      "epoch": 0.13684980664042087,
+      "grad_norm": 0.6870022968581666,
+      "learning_rate": 6.713319153697788e-07,
+      "loss": 1.611,
+      "step": 1964
+    },
+    {
+      "epoch": 0.1369194857680382,
+      "grad_norm": 0.6992649168712028,
+      "learning_rate": 6.713031660334007e-07,
+      "loss": 1.5757,
+      "step": 1965
+    },
+    {
+      "epoch": 0.13698916489565552,
+      "grad_norm": 0.7178246262660336,
+      "learning_rate": 6.712744029767871e-07,
+      "loss": 1.4761,
+      "step": 1966
+    },
+    {
+      "epoch": 0.13705884402327284,
+      "grad_norm": 0.7437799778592202,
+      "learning_rate": 6.712456262013164e-07,
+      "loss": 1.5656,
+      "step": 1967
+    },
+    {
+      "epoch": 0.13712852315089016,
+      "grad_norm": 0.7408121743553396,
+      "learning_rate": 6.712168357083677e-07,
+      "loss": 1.4855,
+      "step": 1968
+    },
+    {
+      "epoch": 0.1371982022785075,
+      "grad_norm": 0.7172543395443857,
+      "learning_rate": 6.711880314993205e-07,
+      "loss": 1.4648,
+      "step": 1969
+    },
+    {
+      "epoch": 0.1372678814061248,
+      "grad_norm": 0.7059469068330578,
+      "learning_rate": 6.711592135755555e-07,
+      "loss": 1.5729,
+      "step": 1970
+    },
+    {
+      "epoch": 0.13733756053374213,
+      "grad_norm": 0.7208503647318539,
+      "learning_rate": 6.711303819384533e-07,
+      "loss": 1.5759,
+      "step": 1971
+    },
+    {
+      "epoch": 0.13740723966135943,
+      "grad_norm": 0.7529947820776219,
+      "learning_rate": 6.711015365893959e-07,
+      "loss": 1.5245,
+      "step": 1972
+    },
+    {
+      "epoch": 0.13747691878897675,
+      "grad_norm": 0.7036672482791874,
+      "learning_rate": 6.710726775297655e-07,
+      "loss": 1.5135,
+      "step": 1973
+    },
+    {
+      "epoch": 0.13754659791659407,
+      "grad_norm": 0.7944363990655092,
+      "learning_rate": 6.710438047609452e-07,
+      "loss": 1.6168,
+      "step": 1974
+    },
+    {
+      "epoch": 0.1376162770442114,
+      "grad_norm": 0.7280065937367366,
+      "learning_rate": 6.710149182843183e-07,
+      "loss": 1.4074,
+      "step": 1975
+    },
+    {
+      "epoch": 0.13768595617182872,
+      "grad_norm": 0.727312215451071,
+      "learning_rate": 6.709860181012695e-07,
+      "loss": 1.6429,
+      "step": 1976
+    },
+    {
+      "epoch": 0.13775563529944604,
+      "grad_norm": 0.6953250639382146,
+      "learning_rate": 6.709571042131836e-07,
+      "loss": 1.596,
+      "step": 1977
+    },
+    {
+      "epoch": 0.13782531442706336,
+      "grad_norm": 0.7968221935128836,
+      "learning_rate": 6.709281766214462e-07,
+      "loss": 1.4812,
+      "step": 1978
+    },
+    {
+      "epoch": 0.13789499355468068,
+      "grad_norm": 0.731561061631931,
+      "learning_rate": 6.708992353274434e-07,
+      "loss": 1.5383,
+      "step": 1979
+    },
+    {
+      "epoch": 0.137964672682298,
+      "grad_norm": 0.697983322450153,
+      "learning_rate": 6.708702803325626e-07,
+      "loss": 1.4938,
+      "step": 1980
+    },
+    {
+      "epoch": 0.13803435180991533,
+      "grad_norm": 0.6999587134752484,
+      "learning_rate": 6.70841311638191e-07,
+      "loss": 1.5703,
+      "step": 1981
+    },
+    {
+      "epoch": 0.13810403093753265,
+      "grad_norm": 0.7522221971120099,
+      "learning_rate": 6.708123292457168e-07,
+      "loss": 1.4194,
+      "step": 1982
+    },
+    {
+      "epoch": 0.13817371006514997,
+      "grad_norm": 0.7996483404297469,
+      "learning_rate": 6.707833331565289e-07,
+      "loss": 1.3953,
+      "step": 1983
+    },
+    {
+      "epoch": 0.1382433891927673,
+      "grad_norm": 0.7338395677152775,
+      "learning_rate": 6.707543233720173e-07,
+      "loss": 1.6296,
+      "step": 1984
+    },
+    {
+      "epoch": 0.13831306832038462,
+      "grad_norm": 0.7426771851411896,
+      "learning_rate": 6.707252998935717e-07,
+      "loss": 1.5608,
+      "step": 1985
+    },
+    {
+      "epoch": 0.13838274744800194,
+      "grad_norm": 0.6530700467546554,
+      "learning_rate": 6.706962627225833e-07,
+      "loss": 1.5599,
+      "step": 1986
+    },
+    {
+      "epoch": 0.13845242657561926,
+      "grad_norm": 0.777687245902078,
+      "learning_rate": 6.706672118604433e-07,
+      "loss": 1.5721,
+      "step": 1987
+    },
+    {
+      "epoch": 0.1385221057032366,
+      "grad_norm": 0.6768719388702751,
+      "learning_rate": 6.706381473085441e-07,
+      "loss": 1.5406,
+      "step": 1988
+    },
+    {
+      "epoch": 0.1385917848308539,
+      "grad_norm": 0.7222488532422539,
+      "learning_rate": 6.706090690682784e-07,
+      "loss": 1.4633,
+      "step": 1989
+    },
+    {
+      "epoch": 0.13866146395847123,
+      "grad_norm": 0.723866634938114,
+      "learning_rate": 6.705799771410399e-07,
+      "loss": 1.5605,
+      "step": 1990
+    },
+    {
+      "epoch": 0.13873114308608855,
+      "grad_norm": 0.7870541821766887,
+      "learning_rate": 6.705508715282225e-07,
+      "loss": 1.5404,
+      "step": 1991
+    },
+    {
+      "epoch": 0.13880082221370588,
+      "grad_norm": 0.7334998345863415,
+      "learning_rate": 6.705217522312213e-07,
+      "loss": 1.6006,
+      "step": 1992
+    },
+    {
+      "epoch": 0.1388705013413232,
+      "grad_norm": 0.7960625479446524,
+      "learning_rate": 6.704926192514313e-07,
+      "loss": 1.6352,
+      "step": 1993
+    },
+    {
+      "epoch": 0.13894018046894052,
+      "grad_norm": 0.7140325267209116,
+      "learning_rate": 6.70463472590249e-07,
+      "loss": 1.5576,
+      "step": 1994
+    },
+    {
+      "epoch": 0.13900985959655784,
+      "grad_norm": 0.6927480097785106,
+      "learning_rate": 6.70434312249071e-07,
+      "loss": 1.4991,
+      "step": 1995
+    },
+    {
+      "epoch": 0.13907953872417517,
+      "grad_norm": 0.6642520213604098,
+      "learning_rate": 6.70405138229295e-07,
+      "loss": 1.4621,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1391492178517925,
+      "grad_norm": 0.6805635598441102,
+      "learning_rate": 6.703759505323186e-07,
+      "loss": 1.6133,
+      "step": 1997
+    },
+    {
+      "epoch": 0.1392188969794098,
+      "grad_norm": 0.7558783219492097,
+      "learning_rate": 6.703467491595409e-07,
+      "loss": 1.6302,
+      "step": 1998
+    },
+    {
+      "epoch": 0.13928857610702713,
+      "grad_norm": 0.7137225258702494,
+      "learning_rate": 6.703175341123611e-07,
+      "loss": 1.546,
+      "step": 1999
+    },
+    {
+      "epoch": 0.13935825523464446,
+      "grad_norm": 0.6896011638861811,
+      "learning_rate": 6.702883053921793e-07,
+      "loss": 1.5004,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13942793436226178,
+      "grad_norm": 0.7740936820333882,
+      "learning_rate": 6.702590630003963e-07,
+      "loss": 1.5578,
+      "step": 2001
+    },
+    {
+      "epoch": 0.1394976134898791,
+      "grad_norm": 0.7113997941083866,
+      "learning_rate": 6.702298069384134e-07,
+      "loss": 1.5332,
+      "step": 2002
+    },
+    {
+      "epoch": 0.13956729261749642,
+      "grad_norm": 0.6916645083605428,
+      "learning_rate": 6.702005372076325e-07,
+      "loss": 1.7124,
+      "step": 2003
+    },
+    {
+      "epoch": 0.13963697174511375,
+      "grad_norm": 0.8047028801866378,
+      "learning_rate": 6.701712538094564e-07,
+      "loss": 1.5595,
+      "step": 2004
+    },
+    {
+      "epoch": 0.13970665087273107,
+      "grad_norm": 0.6812523150401025,
+      "learning_rate": 6.701419567452884e-07,
+      "loss": 1.4507,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1397763300003484,
+      "grad_norm": 0.7103344035535895,
+      "learning_rate": 6.701126460165324e-07,
+      "loss": 1.5772,
+      "step": 2006
+    },
+    {
+      "epoch": 0.13984600912796571,
+      "grad_norm": 0.7102780550548473,
+      "learning_rate": 6.700833216245931e-07,
+      "loss": 1.5876,
+      "step": 2007
+    },
+    {
+      "epoch": 0.13991568825558304,
+      "grad_norm": 0.7369837252532143,
+      "learning_rate": 6.700539835708757e-07,
+      "loss": 1.5054,
+      "step": 2008
+    },
+    {
+      "epoch": 0.13998536738320036,
+      "grad_norm": 0.74091795158049,
+      "learning_rate": 6.700246318567862e-07,
+      "loss": 1.547,
+      "step": 2009
+    },
+    {
+      "epoch": 0.14005504651081768,
+      "grad_norm": 0.8230971961676586,
+      "learning_rate": 6.699952664837312e-07,
+      "loss": 1.6581,
+      "step": 2010
+    },
+    {
+      "epoch": 0.140124725638435,
+      "grad_norm": 0.6696729949760143,
+      "learning_rate": 6.699658874531181e-07,
+      "loss": 1.5002,
+      "step": 2011
+    },
+    {
+      "epoch": 0.14019440476605233,
+      "grad_norm": 0.6628325796270378,
+      "learning_rate": 6.699364947663546e-07,
+      "loss": 1.5461,
+      "step": 2012
+    },
+    {
+      "epoch": 0.14026408389366965,
+      "grad_norm": 0.7288767149466291,
+      "learning_rate": 6.699070884248492e-07,
+      "loss": 1.674,
+      "step": 2013
+    },
+    {
+      "epoch": 0.14033376302128697,
+      "grad_norm": 0.6900581440512029,
+      "learning_rate": 6.698776684300113e-07,
+      "loss": 1.539,
+      "step": 2014
+    },
+    {
+      "epoch": 0.1404034421489043,
+      "grad_norm": 0.7013138547776667,
+      "learning_rate": 6.698482347832506e-07,
+      "loss": 1.519,
+      "step": 2015
+    },
+    {
+      "epoch": 0.14047312127652162,
+      "grad_norm": 0.7269407616493614,
+      "learning_rate": 6.698187874859778e-07,
+      "loss": 1.4797,
+      "step": 2016
+    },
+    {
+      "epoch": 0.14054280040413894,
+      "grad_norm": 0.7810767647483735,
+      "learning_rate": 6.69789326539604e-07,
+      "loss": 1.6596,
+      "step": 2017
+    },
+    {
+      "epoch": 0.14061247953175626,
+      "grad_norm": 0.7286918267880533,
+      "learning_rate": 6.697598519455409e-07,
+      "loss": 1.4667,
+      "step": 2018
+    },
+    {
+      "epoch": 0.14068215865937359,
+      "grad_norm": 0.6977287876601965,
+      "learning_rate": 6.697303637052011e-07,
+      "loss": 1.3941,
+      "step": 2019
+    },
+    {
+      "epoch": 0.1407518377869909,
+      "grad_norm": 0.6633872121941339,
+      "learning_rate": 6.697008618199978e-07,
+      "loss": 1.4926,
+      "step": 2020
+    },
+    {
+      "epoch": 0.14082151691460823,
+      "grad_norm": 0.6908639722598806,
+      "learning_rate": 6.696713462913447e-07,
+      "loss": 1.5853,
+      "step": 2021
+    },
+    {
+      "epoch": 0.14089119604222555,
+      "grad_norm": 0.7069874421515819,
+      "learning_rate": 6.696418171206563e-07,
+      "loss": 1.4058,
+      "step": 2022
+    },
+    {
+      "epoch": 0.14096087516984288,
+      "grad_norm": 0.7027241913436704,
+      "learning_rate": 6.696122743093476e-07,
+      "loss": 1.5894,
+      "step": 2023
+    },
+    {
+      "epoch": 0.1410305542974602,
+      "grad_norm": 0.7560903773846185,
+      "learning_rate": 6.695827178588346e-07,
+      "loss": 1.6811,
+      "step": 2024
+    },
+    {
+      "epoch": 0.14110023342507752,
+      "grad_norm": 0.7511736752192385,
+      "learning_rate": 6.695531477705333e-07,
+      "loss": 1.6184,
+      "step": 2025
+    },
+    {
+      "epoch": 0.14116991255269484,
+      "grad_norm": 0.6690392393806831,
+      "learning_rate": 6.69523564045861e-07,
+      "loss": 1.4891,
+      "step": 2026
+    },
+    {
+      "epoch": 0.14123959168031217,
+      "grad_norm": 0.6912196783448034,
+      "learning_rate": 6.694939666862355e-07,
+      "loss": 1.4651,
+      "step": 2027
+    },
+    {
+      "epoch": 0.1413092708079295,
+      "grad_norm": 0.77137577829856,
+      "learning_rate": 6.69464355693075e-07,
+      "loss": 1.5403,
+      "step": 2028
+    },
+    {
+      "epoch": 0.1413789499355468,
+      "grad_norm": 0.6865381749824687,
+      "learning_rate": 6.694347310677985e-07,
+      "loss": 1.6508,
+      "step": 2029
+    },
+    {
+      "epoch": 0.14144862906316413,
+      "grad_norm": 0.7566844700374754,
+      "learning_rate": 6.69405092811826e-07,
+      "loss": 1.5167,
+      "step": 2030
+    },
+    {
+      "epoch": 0.14151830819078146,
+      "grad_norm": 0.6952455915837616,
+      "learning_rate": 6.693754409265774e-07,
+      "loss": 1.4989,
+      "step": 2031
+    },
+    {
+      "epoch": 0.14158798731839878,
+      "grad_norm": 0.7130424330099704,
+      "learning_rate": 6.693457754134739e-07,
+      "loss": 1.4692,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1416576664460161,
+      "grad_norm": 0.8873037536282933,
+      "learning_rate": 6.693160962739372e-07,
+      "loss": 1.4678,
+      "step": 2033
+    },
+    {
+      "epoch": 0.14172734557363342,
+      "grad_norm": 0.7036011332421183,
+      "learning_rate": 6.692864035093892e-07,
+      "loss": 1.5706,
+      "step": 2034
+    },
+    {
+      "epoch": 0.14179702470125075,
+      "grad_norm": 0.834179522399948,
+      "learning_rate": 6.692566971212533e-07,
+      "loss": 1.5409,
+      "step": 2035
+    },
+    {
+      "epoch": 0.14186670382886807,
+      "grad_norm": 0.6949574416710086,
+      "learning_rate": 6.69226977110953e-07,
+      "loss": 1.5529,
+      "step": 2036
+    },
+    {
+      "epoch": 0.1419363829564854,
+      "grad_norm": 0.7135623739474563,
+      "learning_rate": 6.691972434799122e-07,
+      "loss": 1.6741,
+      "step": 2037
+    },
+    {
+      "epoch": 0.1420060620841027,
+      "grad_norm": 0.6687369374009454,
+      "learning_rate": 6.691674962295562e-07,
+      "loss": 1.525,
+      "step": 2038
+    },
+    {
+      "epoch": 0.14207574121172004,
+      "grad_norm": 0.7433925112314416,
+      "learning_rate": 6.691377353613104e-07,
+      "loss": 1.4498,
+      "step": 2039
+    },
+    {
+      "epoch": 0.14214542033933736,
+      "grad_norm": 0.6745110392616911,
+      "learning_rate": 6.691079608766009e-07,
+      "loss": 1.6143,
+      "step": 2040
+    },
+    {
+      "epoch": 0.14221509946695468,
+      "grad_norm": 0.7382616845616087,
+      "learning_rate": 6.690781727768548e-07,
+      "loss": 1.541,
+      "step": 2041
+    },
+    {
+      "epoch": 0.142284778594572,
+      "grad_norm": 0.7045004156463536,
+      "learning_rate": 6.690483710634993e-07,
+      "loss": 1.4801,
+      "step": 2042
+    },
+    {
+      "epoch": 0.14235445772218933,
+      "grad_norm": 0.6965897989333067,
+      "learning_rate": 6.690185557379629e-07,
+      "loss": 1.489,
+      "step": 2043
+    },
+    {
+      "epoch": 0.14242413684980665,
+      "grad_norm": 0.7438196665174391,
+      "learning_rate": 6.689887268016741e-07,
+      "loss": 1.5633,
+      "step": 2044
+    },
+    {
+      "epoch": 0.14249381597742397,
+      "grad_norm": 0.7105715238696366,
+      "learning_rate": 6.689588842560625e-07,
+      "loss": 1.5686,
+      "step": 2045
+    },
+    {
+      "epoch": 0.1425634951050413,
+      "grad_norm": 0.7049215190610151,
+      "learning_rate": 6.689290281025582e-07,
+      "loss": 1.6438,
+      "step": 2046
+    },
+    {
+      "epoch": 0.14263317423265862,
+      "grad_norm": 0.6744774059408442,
+      "learning_rate": 6.68899158342592e-07,
+      "loss": 1.5095,
+      "step": 2047
+    },
+    {
+      "epoch": 0.14270285336027594,
+      "grad_norm": 0.680417356610093,
+      "learning_rate": 6.688692749775953e-07,
+      "loss": 1.5074,
+      "step": 2048
+    },
+    {
+      "epoch": 0.14277253248789326,
+      "grad_norm": 0.7151033904243572,
+      "learning_rate": 6.688393780090002e-07,
+      "loss": 1.604,
+      "step": 2049
+    },
+    {
+      "epoch": 0.14284221161551058,
+      "grad_norm": 0.724966418349942,
+      "learning_rate": 6.688094674382394e-07,
+      "loss": 1.5373,
+      "step": 2050
+    },
+    {
+      "epoch": 0.1429118907431279,
+      "grad_norm": 0.7360886176646285,
+      "learning_rate": 6.687795432667462e-07,
+      "loss": 1.5455,
+      "step": 2051
+    },
+    {
+      "epoch": 0.14298156987074523,
+      "grad_norm": 0.7172255708598817,
+      "learning_rate": 6.687496054959548e-07,
+      "loss": 1.6347,
+      "step": 2052
+    },
+    {
+      "epoch": 0.14305124899836255,
+      "grad_norm": 0.7699884407032171,
+      "learning_rate": 6.687196541272997e-07,
+      "loss": 1.6606,
+      "step": 2053
+    },
+    {
+      "epoch": 0.14312092812597987,
+      "grad_norm": 0.7014245175284832,
+      "learning_rate": 6.686896891622164e-07,
+      "loss": 1.6195,
+      "step": 2054
+    },
+    {
+      "epoch": 0.1431906072535972,
+      "grad_norm": 0.7050391085105184,
+      "learning_rate": 6.686597106021406e-07,
+      "loss": 1.4778,
+      "step": 2055
+    },
+    {
+      "epoch": 0.14326028638121452,
+      "grad_norm": 0.7475241606305587,
+      "learning_rate": 6.686297184485093e-07,
+      "loss": 1.522,
+      "step": 2056
+    },
+    {
+      "epoch": 0.14332996550883184,
+      "grad_norm": 0.7224955938632416,
+      "learning_rate": 6.685997127027597e-07,
+      "loss": 1.4863,
+      "step": 2057
+    },
+    {
+      "epoch": 0.14339964463644916,
+      "grad_norm": 0.7205299342896766,
+      "learning_rate": 6.685696933663295e-07,
+      "loss": 1.451,
+      "step": 2058
+    },
+    {
+      "epoch": 0.1434693237640665,
+      "grad_norm": 0.7375433453122535,
+      "learning_rate": 6.685396604406574e-07,
+      "loss": 1.5236,
+      "step": 2059
+    },
+    {
+      "epoch": 0.1435390028916838,
+      "grad_norm": 0.6683998772094839,
+      "learning_rate": 6.685096139271827e-07,
+      "loss": 1.519,
+      "step": 2060
+    },
+    {
+      "epoch": 0.14360868201930113,
+      "grad_norm": 0.7362126432095132,
+      "learning_rate": 6.684795538273452e-07,
+      "loss": 1.578,
+      "step": 2061
+    },
+    {
+      "epoch": 0.14367836114691845,
+      "grad_norm": 0.7395787426618212,
+      "learning_rate": 6.684494801425857e-07,
+      "loss": 1.5395,
+      "step": 2062
+    },
+    {
+      "epoch": 0.14374804027453578,
+      "grad_norm": 0.6894804607087299,
+      "learning_rate": 6.684193928743451e-07,
+      "loss": 1.5361,
+      "step": 2063
+    },
+    {
+      "epoch": 0.14381771940215307,
+      "grad_norm": 0.7135583941854905,
+      "learning_rate": 6.683892920240653e-07,
+      "loss": 1.4923,
+      "step": 2064
+    },
+    {
+      "epoch": 0.1438873985297704,
+      "grad_norm": 0.7829148162615202,
+      "learning_rate": 6.683591775931889e-07,
+      "loss": 1.5336,
+      "step": 2065
+    },
+    {
+      "epoch": 0.14395707765738772,
+      "grad_norm": 0.7303266441004318,
+      "learning_rate": 6.683290495831589e-07,
+      "loss": 1.6255,
+      "step": 2066
+    },
+    {
+      "epoch": 0.14402675678500504,
+      "grad_norm": 0.7839802506351182,
+      "learning_rate": 6.682989079954192e-07,
+      "loss": 1.5698,
+      "step": 2067
+    },
+    {
+      "epoch": 0.14409643591262236,
+      "grad_norm": 0.6629158355551916,
+      "learning_rate": 6.682687528314144e-07,
+      "loss": 1.5633,
+      "step": 2068
+    },
+    {
+      "epoch": 0.14416611504023968,
+      "grad_norm": 0.7241216398074919,
+      "learning_rate": 6.682385840925893e-07,
+      "loss": 1.545,
+      "step": 2069
+    },
+    {
+      "epoch": 0.144235794167857,
+      "grad_norm": 0.7164951634406063,
+      "learning_rate": 6.682084017803897e-07,
+      "loss": 1.4695,
+      "step": 2070
+    },
+    {
+      "epoch": 0.14430547329547433,
+      "grad_norm": 1.0271724115304646,
+      "learning_rate": 6.681782058962621e-07,
+      "loss": 1.4634,
+      "step": 2071
+    },
+    {
+      "epoch": 0.14437515242309165,
+      "grad_norm": 0.7823040178808669,
+      "learning_rate": 6.681479964416536e-07,
+      "loss": 1.5905,
+      "step": 2072
+    },
+    {
+      "epoch": 0.14444483155070897,
+      "grad_norm": 0.657817420957194,
+      "learning_rate": 6.681177734180118e-07,
+      "loss": 1.3752,
+      "step": 2073
+    },
+    {
+      "epoch": 0.1445145106783263,
+      "grad_norm": 0.6852262198461414,
+      "learning_rate": 6.68087536826785e-07,
+      "loss": 1.6397,
+      "step": 2074
+    },
+    {
+      "epoch": 0.14458418980594362,
+      "grad_norm": 0.756558074647055,
+      "learning_rate": 6.680572866694223e-07,
+      "loss": 1.635,
+      "step": 2075
+    },
+    {
+      "epoch": 0.14465386893356094,
+      "grad_norm": 0.7433246997372718,
+      "learning_rate": 6.680270229473733e-07,
+      "loss": 1.5365,
+      "step": 2076
+    },
+    {
+      "epoch": 0.14472354806117826,
+      "grad_norm": 0.78026047334549,
+      "learning_rate": 6.679967456620883e-07,
+      "loss": 1.5682,
+      "step": 2077
+    },
+    {
+      "epoch": 0.1447932271887956,
+      "grad_norm": 0.7072922446822634,
+      "learning_rate": 6.679664548150184e-07,
+      "loss": 1.4905,
+      "step": 2078
+    },
+    {
+      "epoch": 0.1448629063164129,
+      "grad_norm": 0.68509156320556,
+      "learning_rate": 6.67936150407615e-07,
+      "loss": 1.613,
+      "step": 2079
+    },
+    {
+      "epoch": 0.14493258544403023,
+      "grad_norm": 0.7336974814746486,
+      "learning_rate": 6.679058324413304e-07,
+      "loss": 1.6432,
+      "step": 2080
+    },
+    {
+      "epoch": 0.14500226457164755,
+      "grad_norm": 0.7801610741726437,
+      "learning_rate": 6.678755009176175e-07,
+      "loss": 1.7378,
+      "step": 2081
+    },
+    {
+      "epoch": 0.14507194369926488,
+      "grad_norm": 0.7463363382104677,
+      "learning_rate": 6.678451558379299e-07,
+      "loss": 1.5998,
+      "step": 2082
+    },
+    {
+      "epoch": 0.1451416228268822,
+      "grad_norm": 0.7695298804194249,
+      "learning_rate": 6.678147972037217e-07,
+      "loss": 1.4838,
+      "step": 2083
+    },
+    {
+      "epoch": 0.14521130195449952,
+      "grad_norm": 0.7298021201070521,
+      "learning_rate": 6.677844250164478e-07,
+      "loss": 1.7206,
+      "step": 2084
+    },
+    {
+      "epoch": 0.14528098108211684,
+      "grad_norm": 0.7045264264551108,
+      "learning_rate": 6.677540392775638e-07,
+      "loss": 1.6674,
+      "step": 2085
+    },
+    {
+      "epoch": 0.14535066020973417,
+      "grad_norm": 0.7080820049733552,
+      "learning_rate": 6.677236399885257e-07,
+      "loss": 1.3703,
+      "step": 2086
+    },
+    {
+      "epoch": 0.1454203393373515,
+      "grad_norm": 0.7405645499986991,
+      "learning_rate": 6.676932271507905e-07,
+      "loss": 1.6083,
+      "step": 2087
+    },
+    {
+      "epoch": 0.1454900184649688,
+      "grad_norm": 0.6755073635241531,
+      "learning_rate": 6.676628007658152e-07,
+      "loss": 1.4891,
+      "step": 2088
+    },
+    {
+      "epoch": 0.14555969759258613,
+      "grad_norm": 0.735196926318889,
+      "learning_rate": 6.676323608350584e-07,
+      "loss": 1.5538,
+      "step": 2089
+    },
+    {
+      "epoch": 0.14562937672020346,
+      "grad_norm": 0.7158832810121535,
+      "learning_rate": 6.676019073599786e-07,
+      "loss": 1.4395,
+      "step": 2090
+    },
+    {
+      "epoch": 0.14569905584782078,
+      "grad_norm": 0.8233995506467929,
+      "learning_rate": 6.675714403420352e-07,
+      "loss": 1.5909,
+      "step": 2091
+    },
+    {
+      "epoch": 0.1457687349754381,
+      "grad_norm": 0.6981047452018204,
+      "learning_rate": 6.675409597826883e-07,
+      "loss": 1.5379,
+      "step": 2092
+    },
+    {
+      "epoch": 0.14583841410305542,
+      "grad_norm": 0.7318049866634618,
+      "learning_rate": 6.675104656833985e-07,
+      "loss": 1.5167,
+      "step": 2093
+    },
+    {
+      "epoch": 0.14590809323067275,
+      "grad_norm": 0.6711256547636123,
+      "learning_rate": 6.674799580456273e-07,
+      "loss": 1.4706,
+      "step": 2094
+    },
+    {
+      "epoch": 0.14597777235829007,
+      "grad_norm": 0.7807090030494048,
+      "learning_rate": 6.674494368708365e-07,
+      "loss": 1.6408,
+      "step": 2095
+    },
+    {
+      "epoch": 0.1460474514859074,
+      "grad_norm": 0.7560459522384863,
+      "learning_rate": 6.674189021604889e-07,
+      "loss": 1.5771,
+      "step": 2096
+    },
+    {
+      "epoch": 0.14611713061352471,
+      "grad_norm": 0.7340658045296877,
+      "learning_rate": 6.673883539160478e-07,
+      "loss": 1.7691,
+      "step": 2097
+    },
+    {
+      "epoch": 0.14618680974114204,
+      "grad_norm": 0.7166206168364373,
+      "learning_rate": 6.673577921389768e-07,
+      "loss": 1.5528,
+      "step": 2098
+    },
+    {
+      "epoch": 0.14625648886875936,
+      "grad_norm": 0.7247772846536584,
+      "learning_rate": 6.673272168307408e-07,
+      "loss": 1.5739,
+      "step": 2099
+    },
+    {
+      "epoch": 0.14632616799637668,
+      "grad_norm": 0.7122755245520854,
+      "learning_rate": 6.672966279928051e-07,
+      "loss": 1.517,
+      "step": 2100
+    },
+    {
+      "epoch": 0.146395847123994,
+      "grad_norm": 0.709517658767935,
+      "learning_rate": 6.672660256266353e-07,
+      "loss": 1.5352,
+      "step": 2101
+    },
+    {
+      "epoch": 0.14646552625161133,
+      "grad_norm": 0.7526325805880144,
+      "learning_rate": 6.672354097336982e-07,
+      "loss": 1.4883,
+      "step": 2102
+    },
+    {
+      "epoch": 0.14653520537922865,
+      "grad_norm": 0.6888521999954886,
+      "learning_rate": 6.672047803154608e-07,
+      "loss": 1.6047,
+      "step": 2103
+    },
+    {
+      "epoch": 0.14660488450684597,
+      "grad_norm": 0.7384235333319291,
+      "learning_rate": 6.671741373733909e-07,
+      "loss": 1.5317,
+      "step": 2104
+    },
+    {
+      "epoch": 0.1466745636344633,
+      "grad_norm": 0.6736086119358556,
+      "learning_rate": 6.671434809089571e-07,
+      "loss": 1.5422,
+      "step": 2105
+    },
+    {
+      "epoch": 0.14674424276208062,
+      "grad_norm": 0.6661337493931547,
+      "learning_rate": 6.671128109236285e-07,
+      "loss": 1.501,
+      "step": 2106
+    },
+    {
+      "epoch": 0.14681392188969794,
+      "grad_norm": 0.7206096510228985,
+      "learning_rate": 6.670821274188747e-07,
+      "loss": 1.5317,
+      "step": 2107
+    },
+    {
+      "epoch": 0.14688360101731526,
+      "grad_norm": 0.6859249482438138,
+      "learning_rate": 6.670514303961664e-07,
+      "loss": 1.5287,
+      "step": 2108
+    },
+    {
+      "epoch": 0.14695328014493259,
+      "grad_norm": 0.7183963613016431,
+      "learning_rate": 6.670207198569745e-07,
+      "loss": 1.5517,
+      "step": 2109
+    },
+    {
+      "epoch": 0.1470229592725499,
+      "grad_norm": 0.7341898396673695,
+      "learning_rate": 6.669899958027707e-07,
+      "loss": 1.6989,
+      "step": 2110
+    },
+    {
+      "epoch": 0.14709263840016723,
+      "grad_norm": 0.7019688944440516,
+      "learning_rate": 6.669592582350273e-07,
+      "loss": 1.5366,
+      "step": 2111
+    },
+    {
+      "epoch": 0.14716231752778455,
+      "grad_norm": 0.7247432024634127,
+      "learning_rate": 6.669285071552174e-07,
+      "loss": 1.5756,
+      "step": 2112
+    },
+    {
+      "epoch": 0.14723199665540188,
+      "grad_norm": 0.7513673208989898,
+      "learning_rate": 6.668977425648149e-07,
+      "loss": 1.5718,
+      "step": 2113
+    },
+    {
+      "epoch": 0.1473016757830192,
+      "grad_norm": 0.7819729777404506,
+      "learning_rate": 6.668669644652937e-07,
+      "loss": 1.5458,
+      "step": 2114
+    },
+    {
+      "epoch": 0.14737135491063652,
+      "grad_norm": 0.750342648657007,
+      "learning_rate": 6.668361728581288e-07,
+      "loss": 1.4844,
+      "step": 2115
+    },
+    {
+      "epoch": 0.14744103403825384,
+      "grad_norm": 0.6908345407203063,
+      "learning_rate": 6.66805367744796e-07,
+      "loss": 1.6642,
+      "step": 2116
+    },
+    {
+      "epoch": 0.14751071316587117,
+      "grad_norm": 0.7344250712154198,
+      "learning_rate": 6.667745491267713e-07,
+      "loss": 1.6705,
+      "step": 2117
+    },
+    {
+      "epoch": 0.1475803922934885,
+      "grad_norm": 0.7510819008972629,
+      "learning_rate": 6.667437170055319e-07,
+      "loss": 1.5469,
+      "step": 2118
+    },
+    {
+      "epoch": 0.1476500714211058,
+      "grad_norm": 0.7555665446815104,
+      "learning_rate": 6.66712871382555e-07,
+      "loss": 1.6687,
+      "step": 2119
+    },
+    {
+      "epoch": 0.14771975054872313,
+      "grad_norm": 0.6604616741494789,
+      "learning_rate": 6.666820122593192e-07,
+      "loss": 1.5087,
+      "step": 2120
+    },
+    {
+      "epoch": 0.14778942967634046,
+      "grad_norm": 0.7348443193736736,
+      "learning_rate": 6.666511396373029e-07,
+      "loss": 1.5052,
+      "step": 2121
+    },
+    {
+      "epoch": 0.14785910880395778,
+      "grad_norm": 0.672958863560216,
+      "learning_rate": 6.666202535179859e-07,
+      "loss": 1.4961,
+      "step": 2122
+    },
+    {
+      "epoch": 0.1479287879315751,
+      "grad_norm": 0.7414274787253741,
+      "learning_rate": 6.665893539028481e-07,
+      "loss": 1.6306,
+      "step": 2123
+    },
+    {
+      "epoch": 0.14799846705919242,
+      "grad_norm": 0.727858108215168,
+      "learning_rate": 6.665584407933704e-07,
+      "loss": 1.5347,
+      "step": 2124
+    },
+    {
+      "epoch": 0.14806814618680975,
+      "grad_norm": 0.7294702840040114,
+      "learning_rate": 6.665275141910343e-07,
+      "loss": 1.3667,
+      "step": 2125
+    },
+    {
+      "epoch": 0.14813782531442707,
+      "grad_norm": 0.7349044745056671,
+      "learning_rate": 6.664965740973216e-07,
+      "loss": 1.7463,
+      "step": 2126
+    },
+    {
+      "epoch": 0.1482075044420444,
+      "grad_norm": 0.7758881380439382,
+      "learning_rate": 6.664656205137151e-07,
+      "loss": 1.6216,
+      "step": 2127
+    },
+    {
+      "epoch": 0.1482771835696617,
+      "grad_norm": 0.7495364666245434,
+      "learning_rate": 6.664346534416984e-07,
+      "loss": 1.6273,
+      "step": 2128
+    },
+    {
+      "epoch": 0.14834686269727904,
+      "grad_norm": 0.7040089705837504,
+      "learning_rate": 6.664036728827553e-07,
+      "loss": 1.5802,
+      "step": 2129
+    },
+    {
+      "epoch": 0.14841654182489636,
+      "grad_norm": 0.6667731454990864,
+      "learning_rate": 6.663726788383704e-07,
+      "loss": 1.5177,
+      "step": 2130
+    },
+    {
+      "epoch": 0.14848622095251368,
+      "grad_norm": 0.6997476564626672,
+      "learning_rate": 6.663416713100291e-07,
+      "loss": 1.5506,
+      "step": 2131
+    },
+    {
+      "epoch": 0.148555900080131,
+      "grad_norm": 0.6911327597765583,
+      "learning_rate": 6.663106502992175e-07,
+      "loss": 1.4742,
+      "step": 2132
+    },
+    {
+      "epoch": 0.14862557920774833,
+      "grad_norm": 0.7187628023471965,
+      "learning_rate": 6.662796158074218e-07,
+      "loss": 1.5424,
+      "step": 2133
+    },
+    {
+      "epoch": 0.14869525833536565,
+      "grad_norm": 0.8128369317563799,
+      "learning_rate": 6.662485678361296e-07,
+      "loss": 1.5732,
+      "step": 2134
+    },
+    {
+      "epoch": 0.14876493746298297,
+      "grad_norm": 0.782738410518543,
+      "learning_rate": 6.662175063868286e-07,
+      "loss": 1.6439,
+      "step": 2135
+    },
+    {
+      "epoch": 0.1488346165906003,
+      "grad_norm": 0.7095268270329291,
+      "learning_rate": 6.661864314610073e-07,
+      "loss": 1.5901,
+      "step": 2136
+    },
+    {
+      "epoch": 0.14890429571821762,
+      "grad_norm": 0.6888819253469545,
+      "learning_rate": 6.66155343060155e-07,
+      "loss": 1.4262,
+      "step": 2137
+    },
+    {
+      "epoch": 0.14897397484583494,
+      "grad_norm": 0.698034927176739,
+      "learning_rate": 6.661242411857614e-07,
+      "loss": 1.4935,
+      "step": 2138
+    },
+    {
+      "epoch": 0.14904365397345226,
+      "grad_norm": 0.7034207813846144,
+      "learning_rate": 6.660931258393171e-07,
+      "loss": 1.6056,
+      "step": 2139
+    },
+    {
+      "epoch": 0.14911333310106958,
+      "grad_norm": 0.714720004406207,
+      "learning_rate": 6.66061997022313e-07,
+      "loss": 1.4297,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1491830122286869,
+      "grad_norm": 0.70720621104776,
+      "learning_rate": 6.660308547362412e-07,
+      "loss": 1.5657,
+      "step": 2141
+    },
+    {
+      "epoch": 0.14925269135630423,
+      "grad_norm": 0.6818808254535282,
+      "learning_rate": 6.659996989825938e-07,
+      "loss": 1.4919,
+      "step": 2142
+    },
+    {
+      "epoch": 0.14932237048392155,
+      "grad_norm": 1.2047689714013203,
+      "learning_rate": 6.65968529762864e-07,
+      "loss": 1.6194,
+      "step": 2143
+    },
+    {
+      "epoch": 0.14939204961153887,
+      "grad_norm": 0.7234531463009618,
+      "learning_rate": 6.659373470785454e-07,
+      "loss": 1.6053,
+      "step": 2144
+    },
+    {
+      "epoch": 0.1494617287391562,
+      "grad_norm": 0.8116627746519786,
+      "learning_rate": 6.659061509311322e-07,
+      "loss": 1.5547,
+      "step": 2145
+    },
+    {
+      "epoch": 0.14953140786677352,
+      "grad_norm": 0.7010744160731812,
+      "learning_rate": 6.658749413221197e-07,
+      "loss": 1.6412,
+      "step": 2146
+    },
+    {
+      "epoch": 0.14960108699439084,
+      "grad_norm": 0.7281161646233438,
+      "learning_rate": 6.658437182530034e-07,
+      "loss": 1.6422,
+      "step": 2147
+    },
+    {
+      "epoch": 0.14967076612200816,
+      "grad_norm": 0.7123633183041211,
+      "learning_rate": 6.658124817252796e-07,
+      "loss": 1.6012,
+      "step": 2148
+    },
+    {
+      "epoch": 0.1497404452496255,
+      "grad_norm": 0.7779085968751687,
+      "learning_rate": 6.657812317404451e-07,
+      "loss": 1.5819,
+      "step": 2149
+    },
+    {
+      "epoch": 0.1498101243772428,
+      "grad_norm": 0.7245364414178849,
+      "learning_rate": 6.657499682999976e-07,
+      "loss": 1.6811,
+      "step": 2150
+    },
+    {
+      "epoch": 0.14987980350486013,
+      "grad_norm": 0.7291548999158717,
+      "learning_rate": 6.657186914054353e-07,
+      "loss": 1.5303,
+      "step": 2151
+    },
+    {
+      "epoch": 0.14994948263247745,
+      "grad_norm": 0.6850518007923204,
+      "learning_rate": 6.656874010582568e-07,
+      "loss": 1.6017,
+      "step": 2152
+    },
+    {
+      "epoch": 0.15001916176009478,
+      "grad_norm": 0.7636515207105723,
+      "learning_rate": 6.65656097259962e-07,
+      "loss": 1.6767,
+      "step": 2153
+    },
+    {
+      "epoch": 0.1500888408877121,
+      "grad_norm": 0.7095907892985733,
+      "learning_rate": 6.656247800120508e-07,
+      "loss": 1.5909,
+      "step": 2154
+    },
+    {
+      "epoch": 0.15015852001532942,
+      "grad_norm": 0.8017043567634776,
+      "learning_rate": 6.65593449316024e-07,
+      "loss": 1.6075,
+      "step": 2155
+    },
+    {
+      "epoch": 0.15022819914294672,
+      "grad_norm": 0.7075063766905534,
+      "learning_rate": 6.655621051733831e-07,
+      "loss": 1.5024,
+      "step": 2156
+    },
+    {
+      "epoch": 0.15029787827056404,
+      "grad_norm": 0.7010902957564245,
+      "learning_rate": 6.6553074758563e-07,
+      "loss": 1.5442,
+      "step": 2157
+    },
+    {
+      "epoch": 0.15036755739818136,
+      "grad_norm": 0.6911690892841524,
+      "learning_rate": 6.654993765542677e-07,
+      "loss": 1.539,
+      "step": 2158
+    },
+    {
+      "epoch": 0.15043723652579868,
+      "grad_norm": 0.707345315596661,
+      "learning_rate": 6.654679920807994e-07,
+      "loss": 1.4807,
+      "step": 2159
+    },
+    {
+      "epoch": 0.150506915653416,
+      "grad_norm": 0.7415392921869378,
+      "learning_rate": 6.654365941667291e-07,
+      "loss": 1.5474,
+      "step": 2160
+    },
+    {
+      "epoch": 0.15057659478103333,
+      "grad_norm": 0.706173361060631,
+      "learning_rate": 6.654051828135615e-07,
+      "loss": 1.4426,
+      "step": 2161
+    },
+    {
+      "epoch": 0.15064627390865065,
+      "grad_norm": 0.7303397511035342,
+      "learning_rate": 6.653737580228018e-07,
+      "loss": 1.5252,
+      "step": 2162
+    },
+    {
+      "epoch": 0.15071595303626797,
+      "grad_norm": 0.6801700281353229,
+      "learning_rate": 6.653423197959561e-07,
+      "loss": 1.6305,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1507856321638853,
+      "grad_norm": 0.7319612660717305,
+      "learning_rate": 6.653108681345308e-07,
+      "loss": 1.5535,
+      "step": 2164
+    },
+    {
+      "epoch": 0.15085531129150262,
+      "grad_norm": 0.739665723589392,
+      "learning_rate": 6.652794030400334e-07,
+      "loss": 1.4885,
+      "step": 2165
+    },
+    {
+      "epoch": 0.15092499041911994,
+      "grad_norm": 0.7300207068686901,
+      "learning_rate": 6.652479245139715e-07,
+      "loss": 1.556,
+      "step": 2166
+    },
+    {
+      "epoch": 0.15099466954673726,
+      "grad_norm": 0.7478877415893503,
+      "learning_rate": 6.652164325578538e-07,
+      "loss": 1.5387,
+      "step": 2167
+    },
+    {
+      "epoch": 0.1510643486743546,
+      "grad_norm": 0.7583353301827843,
+      "learning_rate": 6.651849271731893e-07,
+      "loss": 1.5861,
+      "step": 2168
+    },
+    {
+      "epoch": 0.1511340278019719,
+      "grad_norm": 0.7393233041700598,
+      "learning_rate": 6.651534083614879e-07,
+      "loss": 1.6991,
+      "step": 2169
+    },
+    {
+      "epoch": 0.15120370692958923,
+      "grad_norm": 0.6878120430029538,
+      "learning_rate": 6.6512187612426e-07,
+      "loss": 1.5472,
+      "step": 2170
+    },
+    {
+      "epoch": 0.15127338605720655,
+      "grad_norm": 0.7431748916767359,
+      "learning_rate": 6.650903304630168e-07,
+      "loss": 1.5066,
+      "step": 2171
+    },
+    {
+      "epoch": 0.15134306518482388,
+      "grad_norm": 0.7058577241054244,
+      "learning_rate": 6.650587713792698e-07,
+      "loss": 1.4571,
+      "step": 2172
+    },
+    {
+      "epoch": 0.1514127443124412,
+      "grad_norm": 0.7446431893400042,
+      "learning_rate": 6.650271988745317e-07,
+      "loss": 1.4761,
+      "step": 2173
+    },
+    {
+      "epoch": 0.15148242344005852,
+      "grad_norm": 0.7038821399942174,
+      "learning_rate": 6.649956129503152e-07,
+      "loss": 1.4445,
+      "step": 2174
+    },
+    {
+      "epoch": 0.15155210256767584,
+      "grad_norm": 0.8191464308394847,
+      "learning_rate": 6.649640136081343e-07,
+      "loss": 1.5595,
+      "step": 2175
+    },
+    {
+      "epoch": 0.15162178169529317,
+      "grad_norm": 0.7133519258464575,
+      "learning_rate": 6.64932400849503e-07,
+      "loss": 1.5781,
+      "step": 2176
+    },
+    {
+      "epoch": 0.1516914608229105,
+      "grad_norm": 0.6795663922248834,
+      "learning_rate": 6.649007746759362e-07,
+      "loss": 1.6141,
+      "step": 2177
+    },
+    {
+      "epoch": 0.1517611399505278,
+      "grad_norm": 0.69403858097068,
+      "learning_rate": 6.648691350889498e-07,
+      "loss": 1.4282,
+      "step": 2178
+    },
+    {
+      "epoch": 0.15183081907814514,
+      "grad_norm": 0.7119608330094656,
+      "learning_rate": 6.6483748209006e-07,
+      "loss": 1.3755,
+      "step": 2179
+    },
+    {
+      "epoch": 0.15190049820576246,
+      "grad_norm": 0.7564469568638418,
+      "learning_rate": 6.648058156807836e-07,
+      "loss": 1.5567,
+      "step": 2180
+    },
+    {
+      "epoch": 0.15197017733337978,
+      "grad_norm": 0.7486844475397654,
+      "learning_rate": 6.647741358626378e-07,
+      "loss": 1.5791,
+      "step": 2181
+    },
+    {
+      "epoch": 0.1520398564609971,
+      "grad_norm": 0.7055998936778343,
+      "learning_rate": 6.647424426371411e-07,
+      "loss": 1.5442,
+      "step": 2182
+    },
+    {
+      "epoch": 0.15210953558861443,
+      "grad_norm": 0.7470502207739791,
+      "learning_rate": 6.647107360058124e-07,
+      "loss": 1.5793,
+      "step": 2183
+    },
+    {
+      "epoch": 0.15217921471623175,
+      "grad_norm": 0.7341460938876649,
+      "learning_rate": 6.64679015970171e-07,
+      "loss": 1.5151,
+      "step": 2184
+    },
+    {
+      "epoch": 0.15224889384384907,
+      "grad_norm": 0.7346400065000422,
+      "learning_rate": 6.646472825317368e-07,
+      "loss": 1.5545,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1523185729714664,
+      "grad_norm": 0.7201065762380195,
+      "learning_rate": 6.646155356920309e-07,
+      "loss": 1.5169,
+      "step": 2186
+    },
+    {
+      "epoch": 0.15238825209908372,
+      "grad_norm": 0.7393498689538193,
+      "learning_rate": 6.645837754525743e-07,
+      "loss": 1.5674,
+      "step": 2187
+    },
+    {
+      "epoch": 0.15245793122670104,
+      "grad_norm": 0.7785664711016976,
+      "learning_rate": 6.645520018148894e-07,
+      "loss": 1.5922,
+      "step": 2188
+    },
+    {
+      "epoch": 0.15252761035431836,
+      "grad_norm": 0.7293345596385744,
+      "learning_rate": 6.645202147804986e-07,
+      "loss": 1.5481,
+      "step": 2189
+    },
+    {
+      "epoch": 0.15259728948193568,
+      "grad_norm": 0.6894072142217631,
+      "learning_rate": 6.644884143509253e-07,
+      "loss": 1.389,
+      "step": 2190
+    },
+    {
+      "epoch": 0.152666968609553,
+      "grad_norm": 0.6740333835131167,
+      "learning_rate": 6.644566005276932e-07,
+      "loss": 1.572,
+      "step": 2191
+    },
+    {
+      "epoch": 0.15273664773717033,
+      "grad_norm": 0.6956811615086338,
+      "learning_rate": 6.644247733123273e-07,
+      "loss": 1.5829,
+      "step": 2192
+    },
+    {
+      "epoch": 0.15280632686478765,
+      "grad_norm": 0.7320923923783303,
+      "learning_rate": 6.643929327063526e-07,
+      "loss": 1.628,
+      "step": 2193
+    },
+    {
+      "epoch": 0.15287600599240497,
+      "grad_norm": 0.7024646754235628,
+      "learning_rate": 6.643610787112949e-07,
+      "loss": 1.4922,
+      "step": 2194
+    },
+    {
+      "epoch": 0.1529456851200223,
+      "grad_norm": 0.7097415544166261,
+      "learning_rate": 6.643292113286809e-07,
+      "loss": 1.5992,
+      "step": 2195
+    },
+    {
+      "epoch": 0.15301536424763962,
+      "grad_norm": 0.6985595433662738,
+      "learning_rate": 6.642973305600375e-07,
+      "loss": 1.4852,
+      "step": 2196
+    },
+    {
+      "epoch": 0.15308504337525694,
+      "grad_norm": 0.6991483533810473,
+      "learning_rate": 6.642654364068927e-07,
+      "loss": 1.4809,
+      "step": 2197
+    },
+    {
+      "epoch": 0.15315472250287426,
+      "grad_norm": 0.7908856281336831,
+      "learning_rate": 6.642335288707749e-07,
+      "loss": 1.5996,
+      "step": 2198
+    },
+    {
+      "epoch": 0.15322440163049159,
+      "grad_norm": 0.696976024748579,
+      "learning_rate": 6.642016079532131e-07,
+      "loss": 1.5017,
+      "step": 2199
+    },
+    {
+      "epoch": 0.1532940807581089,
+      "grad_norm": 0.6691993666994807,
+      "learning_rate": 6.64169673655737e-07,
+      "loss": 1.5449,
+      "step": 2200
+    },
+    {
+      "epoch": 0.15336375988572623,
+      "grad_norm": 0.738678893957917,
+      "learning_rate": 6.641377259798771e-07,
+      "loss": 1.5894,
+      "step": 2201
+    },
+    {
+      "epoch": 0.15343343901334355,
+      "grad_norm": 0.6999886972654056,
+      "learning_rate": 6.641057649271644e-07,
+      "loss": 1.5552,
+      "step": 2202
+    },
+    {
+      "epoch": 0.15350311814096088,
+      "grad_norm": 0.7839373696668602,
+      "learning_rate": 6.640737904991303e-07,
+      "loss": 1.628,
+      "step": 2203
+    },
+    {
+      "epoch": 0.1535727972685782,
+      "grad_norm": 0.6958760832939552,
+      "learning_rate": 6.640418026973073e-07,
+      "loss": 1.5806,
+      "step": 2204
+    },
+    {
+      "epoch": 0.15364247639619552,
+      "grad_norm": 0.6935626553735474,
+      "learning_rate": 6.640098015232282e-07,
+      "loss": 1.5275,
+      "step": 2205
+    },
+    {
+      "epoch": 0.15371215552381284,
+      "grad_norm": 0.7681046260488631,
+      "learning_rate": 6.639777869784266e-07,
+      "loss": 1.6231,
+      "step": 2206
+    },
+    {
+      "epoch": 0.15378183465143017,
+      "grad_norm": 0.6894945589980458,
+      "learning_rate": 6.639457590644367e-07,
+      "loss": 1.4775,
+      "step": 2207
+    },
+    {
+      "epoch": 0.1538515137790475,
+      "grad_norm": 0.7261850091840004,
+      "learning_rate": 6.639137177827935e-07,
+      "loss": 1.5407,
+      "step": 2208
+    },
+    {
+      "epoch": 0.1539211929066648,
+      "grad_norm": 0.6606501584232484,
+      "learning_rate": 6.638816631350324e-07,
+      "loss": 1.4989,
+      "step": 2209
+    },
+    {
+      "epoch": 0.15399087203428213,
+      "grad_norm": 0.7152784548111163,
+      "learning_rate": 6.638495951226891e-07,
+      "loss": 1.4855,
+      "step": 2210
+    },
+    {
+      "epoch": 0.15406055116189946,
+      "grad_norm": 0.7522656609768341,
+      "learning_rate": 6.638175137473011e-07,
+      "loss": 1.5111,
+      "step": 2211
+    },
+    {
+      "epoch": 0.15413023028951678,
+      "grad_norm": 0.7745867847921484,
+      "learning_rate": 6.637854190104053e-07,
+      "loss": 1.534,
+      "step": 2212
+    },
+    {
+      "epoch": 0.1541999094171341,
+      "grad_norm": 0.7233835267335126,
+      "learning_rate": 6.637533109135399e-07,
+      "loss": 1.623,
+      "step": 2213
+    },
+    {
+      "epoch": 0.15426958854475142,
+      "grad_norm": 0.7270410330528998,
+      "learning_rate": 6.637211894582435e-07,
+      "loss": 1.6425,
+      "step": 2214
+    },
+    {
+      "epoch": 0.15433926767236875,
+      "grad_norm": 0.7697491418177266,
+      "learning_rate": 6.636890546460556e-07,
+      "loss": 1.5283,
+      "step": 2215
+    },
+    {
+      "epoch": 0.15440894679998607,
+      "grad_norm": 0.7117412043799898,
+      "learning_rate": 6.636569064785162e-07,
+      "loss": 1.5912,
+      "step": 2216
+    },
+    {
+      "epoch": 0.1544786259276034,
+      "grad_norm": 0.7505816377254368,
+      "learning_rate": 6.636247449571654e-07,
+      "loss": 1.458,
+      "step": 2217
+    },
+    {
+      "epoch": 0.1545483050552207,
+      "grad_norm": 0.6735296173568921,
+      "learning_rate": 6.63592570083545e-07,
+      "loss": 1.5044,
+      "step": 2218
+    },
+    {
+      "epoch": 0.15461798418283804,
+      "grad_norm": 0.6846941209657414,
+      "learning_rate": 6.635603818591967e-07,
+      "loss": 1.5567,
+      "step": 2219
+    },
+    {
+      "epoch": 0.15468766331045536,
+      "grad_norm": 0.6929598112898522,
+      "learning_rate": 6.635281802856632e-07,
+      "loss": 1.5509,
+      "step": 2220
+    },
+    {
+      "epoch": 0.15475734243807268,
+      "grad_norm": 0.6792500496086945,
+      "learning_rate": 6.634959653644873e-07,
+      "loss": 1.5822,
+      "step": 2221
+    },
+    {
+      "epoch": 0.15482702156569,
+      "grad_norm": 0.6842858576959354,
+      "learning_rate": 6.634637370972131e-07,
+      "loss": 1.5215,
+      "step": 2222
+    },
+    {
+      "epoch": 0.15489670069330733,
+      "grad_norm": 0.6924953687899594,
+      "learning_rate": 6.634314954853847e-07,
+      "loss": 1.5317,
+      "step": 2223
+    },
+    {
+      "epoch": 0.15496637982092465,
+      "grad_norm": 0.7730565766794356,
+      "learning_rate": 6.633992405305477e-07,
+      "loss": 1.5001,
+      "step": 2224
+    },
+    {
+      "epoch": 0.15503605894854197,
+      "grad_norm": 0.7204812325831732,
+      "learning_rate": 6.633669722342475e-07,
+      "loss": 1.5332,
+      "step": 2225
+    },
+    {
+      "epoch": 0.1551057380761593,
+      "grad_norm": 0.6750704593792517,
+      "learning_rate": 6.633346905980304e-07,
+      "loss": 1.5395,
+      "step": 2226
+    },
+    {
+      "epoch": 0.15517541720377662,
+      "grad_norm": 0.7399175686766039,
+      "learning_rate": 6.633023956234436e-07,
+      "loss": 1.6439,
+      "step": 2227
+    },
+    {
+      "epoch": 0.15524509633139394,
+      "grad_norm": 0.6596536831519212,
+      "learning_rate": 6.632700873120346e-07,
+      "loss": 1.4794,
+      "step": 2228
+    },
+    {
+      "epoch": 0.15531477545901126,
+      "grad_norm": 0.6992076941173299,
+      "learning_rate": 6.632377656653518e-07,
+      "loss": 1.5523,
+      "step": 2229
+    },
+    {
+      "epoch": 0.15538445458662858,
+      "grad_norm": 0.680191070215852,
+      "learning_rate": 6.63205430684944e-07,
+      "loss": 1.7234,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1554541337142459,
+      "grad_norm": 0.7405456874943854,
+      "learning_rate": 6.631730823723609e-07,
+      "loss": 1.6782,
+      "step": 2231
+    },
+    {
+      "epoch": 0.15552381284186323,
+      "grad_norm": 0.7827715409835366,
+      "learning_rate": 6.631407207291526e-07,
+      "loss": 1.5043,
+      "step": 2232
+    },
+    {
+      "epoch": 0.15559349196948055,
+      "grad_norm": 0.7125663974576547,
+      "learning_rate": 6.631083457568699e-07,
+      "loss": 1.5413,
+      "step": 2233
+    },
+    {
+      "epoch": 0.15566317109709787,
+      "grad_norm": 0.7187003784037695,
+      "learning_rate": 6.630759574570644e-07,
+      "loss": 1.4771,
+      "step": 2234
+    },
+    {
+      "epoch": 0.1557328502247152,
+      "grad_norm": 0.6781974858369967,
+      "learning_rate": 6.63043555831288e-07,
+      "loss": 1.523,
+      "step": 2235
+    },
+    {
+      "epoch": 0.15580252935233252,
+      "grad_norm": 0.6951594427619806,
+      "learning_rate": 6.630111408810937e-07,
+      "loss": 1.6086,
+      "step": 2236
+    },
+    {
+      "epoch": 0.15587220847994984,
+      "grad_norm": 0.7338407170382629,
+      "learning_rate": 6.629787126080348e-07,
+      "loss": 1.6122,
+      "step": 2237
+    },
+    {
+      "epoch": 0.15594188760756716,
+      "grad_norm": 0.7159255072884011,
+      "learning_rate": 6.629462710136653e-07,
+      "loss": 1.5316,
+      "step": 2238
+    },
+    {
+      "epoch": 0.1560115667351845,
+      "grad_norm": 0.7532205783926076,
+      "learning_rate": 6.629138160995399e-07,
+      "loss": 1.7628,
+      "step": 2239
+    },
+    {
+      "epoch": 0.1560812458628018,
+      "grad_norm": 0.6981258987523404,
+      "learning_rate": 6.628813478672139e-07,
+      "loss": 1.4858,
+      "step": 2240
+    },
+    {
+      "epoch": 0.15615092499041913,
+      "grad_norm": 0.6879762245284705,
+      "learning_rate": 6.628488663182431e-07,
+      "loss": 1.5243,
+      "step": 2241
+    },
+    {
+      "epoch": 0.15622060411803645,
+      "grad_norm": 0.7106885629753927,
+      "learning_rate": 6.628163714541844e-07,
+      "loss": 1.4806,
+      "step": 2242
+    },
+    {
+      "epoch": 0.15629028324565378,
+      "grad_norm": 0.774813261180234,
+      "learning_rate": 6.627838632765946e-07,
+      "loss": 1.5369,
+      "step": 2243
+    },
+    {
+      "epoch": 0.1563599623732711,
+      "grad_norm": 0.7211700093562131,
+      "learning_rate": 6.627513417870319e-07,
+      "loss": 1.5611,
+      "step": 2244
+    },
+    {
+      "epoch": 0.15642964150088842,
+      "grad_norm": 0.6932231497601744,
+      "learning_rate": 6.627188069870548e-07,
+      "loss": 1.4099,
+      "step": 2245
+    },
+    {
+      "epoch": 0.15649932062850574,
+      "grad_norm": 0.7036873950506464,
+      "learning_rate": 6.626862588782221e-07,
+      "loss": 1.5305,
+      "step": 2246
+    },
+    {
+      "epoch": 0.15656899975612304,
+      "grad_norm": 0.7089690133313291,
+      "learning_rate": 6.626536974620939e-07,
+      "loss": 1.4183,
+      "step": 2247
+    },
+    {
+      "epoch": 0.15663867888374036,
+      "grad_norm": 0.7105684077832143,
+      "learning_rate": 6.626211227402306e-07,
+      "loss": 1.5593,
+      "step": 2248
+    },
+    {
+      "epoch": 0.15670835801135768,
+      "grad_norm": 0.6832997394422372,
+      "learning_rate": 6.625885347141931e-07,
+      "loss": 1.4509,
+      "step": 2249
+    },
+    {
+      "epoch": 0.156778037138975,
+      "grad_norm": 0.6977345495829199,
+      "learning_rate": 6.625559333855431e-07,
+      "loss": 1.6193,
+      "step": 2250
+    },
+    {
+      "epoch": 0.15684771626659233,
+      "grad_norm": 0.6908184277057992,
+      "learning_rate": 6.625233187558431e-07,
+      "loss": 1.5806,
+      "step": 2251
+    },
+    {
+      "epoch": 0.15691739539420965,
+      "grad_norm": 0.7326095761597486,
+      "learning_rate": 6.624906908266557e-07,
+      "loss": 1.6088,
+      "step": 2252
+    },
+    {
+      "epoch": 0.15698707452182697,
+      "grad_norm": 0.650882245462365,
+      "learning_rate": 6.624580495995447e-07,
+      "loss": 1.4173,
+      "step": 2253
+    },
+    {
+      "epoch": 0.1570567536494443,
+      "grad_norm": 0.7035964201999952,
+      "learning_rate": 6.624253950760746e-07,
+      "loss": 1.6664,
+      "step": 2254
+    },
+    {
+      "epoch": 0.15712643277706162,
+      "grad_norm": 0.7171947278524486,
+      "learning_rate": 6.623927272578098e-07,
+      "loss": 1.5436,
+      "step": 2255
+    },
+    {
+      "epoch": 0.15719611190467894,
+      "grad_norm": 0.7582978366124182,
+      "learning_rate": 6.623600461463162e-07,
+      "loss": 1.5061,
+      "step": 2256
+    },
+    {
+      "epoch": 0.15726579103229626,
+      "grad_norm": 0.7197080632735179,
+      "learning_rate": 6.623273517431597e-07,
+      "loss": 1.4875,
+      "step": 2257
+    },
+    {
+      "epoch": 0.1573354701599136,
+      "grad_norm": 0.6975155150989609,
+      "learning_rate": 6.622946440499072e-07,
+      "loss": 1.375,
+      "step": 2258
+    },
+    {
+      "epoch": 0.1574051492875309,
+      "grad_norm": 0.7211246058188371,
+      "learning_rate": 6.62261923068126e-07,
+      "loss": 1.5137,
+      "step": 2259
+    },
+    {
+      "epoch": 0.15747482841514823,
+      "grad_norm": 0.6786300156106961,
+      "learning_rate": 6.622291887993843e-07,
+      "loss": 1.4291,
+      "step": 2260
+    },
+    {
+      "epoch": 0.15754450754276556,
+      "grad_norm": 0.7286101497808452,
+      "learning_rate": 6.621964412452507e-07,
+      "loss": 1.4559,
+      "step": 2261
+    },
+    {
+      "epoch": 0.15761418667038288,
+      "grad_norm": 0.6679841831373213,
+      "learning_rate": 6.621636804072947e-07,
+      "loss": 1.4893,
+      "step": 2262
+    },
+    {
+      "epoch": 0.1576838657980002,
+      "grad_norm": 0.7239015658745319,
+      "learning_rate": 6.621309062870859e-07,
+      "loss": 1.5985,
+      "step": 2263
+    },
+    {
+      "epoch": 0.15775354492561752,
+      "grad_norm": 0.8033552658475741,
+      "learning_rate": 6.620981188861952e-07,
+      "loss": 1.5316,
+      "step": 2264
+    },
+    {
+      "epoch": 0.15782322405323485,
+      "grad_norm": 0.8488918942742487,
+      "learning_rate": 6.620653182061938e-07,
+      "loss": 1.6772,
+      "step": 2265
+    },
+    {
+      "epoch": 0.15789290318085217,
+      "grad_norm": 0.7367837946918524,
+      "learning_rate": 6.620325042486537e-07,
+      "loss": 1.5063,
+      "step": 2266
+    },
+    {
+      "epoch": 0.1579625823084695,
+      "grad_norm": 0.7017635363309525,
+      "learning_rate": 6.619996770151471e-07,
+      "loss": 1.3885,
+      "step": 2267
+    },
+    {
+      "epoch": 0.1580322614360868,
+      "grad_norm": 0.7708473358379301,
+      "learning_rate": 6.619668365072472e-07,
+      "loss": 1.5045,
+      "step": 2268
+    },
+    {
+      "epoch": 0.15810194056370414,
+      "grad_norm": 0.6922310385911277,
+      "learning_rate": 6.619339827265281e-07,
+      "loss": 1.445,
+      "step": 2269
+    },
+    {
+      "epoch": 0.15817161969132146,
+      "grad_norm": 0.7504965139466662,
+      "learning_rate": 6.619011156745639e-07,
+      "loss": 1.5939,
+      "step": 2270
+    },
+    {
+      "epoch": 0.15824129881893878,
+      "grad_norm": 0.7091372028940393,
+      "learning_rate": 6.618682353529299e-07,
+      "loss": 1.5027,
+      "step": 2271
+    },
+    {
+      "epoch": 0.1583109779465561,
+      "grad_norm": 0.6747431666388362,
+      "learning_rate": 6.618353417632015e-07,
+      "loss": 1.5093,
+      "step": 2272
+    },
+    {
+      "epoch": 0.15838065707417343,
+      "grad_norm": 0.7506009039438517,
+      "learning_rate": 6.618024349069553e-07,
+      "loss": 1.5443,
+      "step": 2273
+    },
+    {
+      "epoch": 0.15845033620179075,
+      "grad_norm": 0.7136774347676139,
+      "learning_rate": 6.617695147857681e-07,
+      "loss": 1.4922,
+      "step": 2274
+    },
+    {
+      "epoch": 0.15852001532940807,
+      "grad_norm": 0.6607762672674673,
+      "learning_rate": 6.617365814012176e-07,
+      "loss": 1.4119,
+      "step": 2275
+    },
+    {
+      "epoch": 0.1585896944570254,
+      "grad_norm": 0.7219437275521969,
+      "learning_rate": 6.61703634754882e-07,
+      "loss": 1.5683,
+      "step": 2276
+    },
+    {
+      "epoch": 0.15865937358464272,
+      "grad_norm": 0.7557107319667966,
+      "learning_rate": 6.616706748483401e-07,
+      "loss": 1.6325,
+      "step": 2277
+    },
+    {
+      "epoch": 0.15872905271226004,
+      "grad_norm": 0.7296816006462254,
+      "learning_rate": 6.616377016831715e-07,
+      "loss": 1.6051,
+      "step": 2278
+    },
+    {
+      "epoch": 0.15879873183987736,
+      "grad_norm": 0.6952674966414605,
+      "learning_rate": 6.616047152609562e-07,
+      "loss": 1.484,
+      "step": 2279
+    },
+    {
+      "epoch": 0.15886841096749468,
+      "grad_norm": 0.7239655613929008,
+      "learning_rate": 6.615717155832753e-07,
+      "loss": 1.4567,
+      "step": 2280
+    },
+    {
+      "epoch": 0.158938090095112,
+      "grad_norm": 0.6928217629736545,
+      "learning_rate": 6.615387026517097e-07,
+      "loss": 1.6081,
+      "step": 2281
+    },
+    {
+      "epoch": 0.15900776922272933,
+      "grad_norm": 0.6737837113780031,
+      "learning_rate": 6.61505676467842e-07,
+      "loss": 1.5322,
+      "step": 2282
+    },
+    {
+      "epoch": 0.15907744835034665,
+      "grad_norm": 0.7578826835208271,
+      "learning_rate": 6.614726370332546e-07,
+      "loss": 1.5239,
+      "step": 2283
+    },
+    {
+      "epoch": 0.15914712747796397,
+      "grad_norm": 0.707541021895692,
+      "learning_rate": 6.614395843495308e-07,
+      "loss": 1.4852,
+      "step": 2284
+    },
+    {
+      "epoch": 0.1592168066055813,
+      "grad_norm": 0.7077574571137548,
+      "learning_rate": 6.614065184182546e-07,
+      "loss": 1.6096,
+      "step": 2285
+    },
+    {
+      "epoch": 0.15928648573319862,
+      "grad_norm": 0.7182478632727132,
+      "learning_rate": 6.613734392410106e-07,
+      "loss": 1.5395,
+      "step": 2286
+    },
+    {
+      "epoch": 0.15935616486081594,
+      "grad_norm": 0.7367290616820319,
+      "learning_rate": 6.61340346819384e-07,
+      "loss": 1.5181,
+      "step": 2287
+    },
+    {
+      "epoch": 0.15942584398843326,
+      "grad_norm": 0.7042290216704357,
+      "learning_rate": 6.613072411549606e-07,
+      "loss": 1.5792,
+      "step": 2288
+    },
+    {
+      "epoch": 0.15949552311605059,
+      "grad_norm": 0.7080735049459942,
+      "learning_rate": 6.61274122249327e-07,
+      "loss": 1.5474,
+      "step": 2289
+    },
+    {
+      "epoch": 0.1595652022436679,
+      "grad_norm": 0.8896374246651081,
+      "learning_rate": 6.612409901040703e-07,
+      "loss": 1.697,
+      "step": 2290
+    },
+    {
+      "epoch": 0.15963488137128523,
+      "grad_norm": 0.7105735026902991,
+      "learning_rate": 6.612078447207782e-07,
+      "loss": 1.5946,
+      "step": 2291
+    },
+    {
+      "epoch": 0.15970456049890255,
+      "grad_norm": 0.6961245254676526,
+      "learning_rate": 6.611746861010392e-07,
+      "loss": 1.4702,
+      "step": 2292
+    },
+    {
+      "epoch": 0.15977423962651988,
+      "grad_norm": 0.7613831462750842,
+      "learning_rate": 6.611415142464423e-07,
+      "loss": 1.4611,
+      "step": 2293
+    },
+    {
+      "epoch": 0.1598439187541372,
+      "grad_norm": 0.7766601055554404,
+      "learning_rate": 6.61108329158577e-07,
+      "loss": 1.5536,
+      "step": 2294
+    },
+    {
+      "epoch": 0.15991359788175452,
+      "grad_norm": 0.7216761614181849,
+      "learning_rate": 6.610751308390338e-07,
+      "loss": 1.62,
+      "step": 2295
+    },
+    {
+      "epoch": 0.15998327700937184,
+      "grad_norm": 0.7602123159942696,
+      "learning_rate": 6.610419192894035e-07,
+      "loss": 1.5354,
+      "step": 2296
+    },
+    {
+      "epoch": 0.16005295613698917,
+      "grad_norm": 0.7208680007068906,
+      "learning_rate": 6.610086945112779e-07,
+      "loss": 1.507,
+      "step": 2297
+    },
+    {
+      "epoch": 0.1601226352646065,
+      "grad_norm": 0.7271009203166914,
+      "learning_rate": 6.609754565062488e-07,
+      "loss": 1.5454,
+      "step": 2298
+    },
+    {
+      "epoch": 0.1601923143922238,
+      "grad_norm": 0.7155591566547312,
+      "learning_rate": 6.609422052759093e-07,
+      "loss": 1.5072,
+      "step": 2299
+    },
+    {
+      "epoch": 0.16026199351984113,
+      "grad_norm": 0.7646577990143016,
+      "learning_rate": 6.60908940821853e-07,
+      "loss": 1.5958,
+      "step": 2300
+    },
+    {
+      "epoch": 0.16033167264745846,
+      "grad_norm": 0.8432643486292445,
+      "learning_rate": 6.608756631456737e-07,
+      "loss": 1.5313,
+      "step": 2301
+    },
+    {
+      "epoch": 0.16040135177507578,
+      "grad_norm": 0.7165983860712775,
+      "learning_rate": 6.608423722489663e-07,
+      "loss": 1.5866,
+      "step": 2302
+    },
+    {
+      "epoch": 0.1604710309026931,
+      "grad_norm": 0.6948385437059152,
+      "learning_rate": 6.608090681333261e-07,
+      "loss": 1.4906,
+      "step": 2303
+    },
+    {
+      "epoch": 0.16054071003031042,
+      "grad_norm": 0.6881623948883927,
+      "learning_rate": 6.607757508003492e-07,
+      "loss": 1.6308,
+      "step": 2304
+    },
+    {
+      "epoch": 0.16061038915792775,
+      "grad_norm": 0.7202743271110196,
+      "learning_rate": 6.60742420251632e-07,
+      "loss": 1.5586,
+      "step": 2305
+    },
+    {
+      "epoch": 0.16068006828554507,
+      "grad_norm": 0.728905182320733,
+      "learning_rate": 6.607090764887721e-07,
+      "loss": 1.382,
+      "step": 2306
+    },
+    {
+      "epoch": 0.1607497474131624,
+      "grad_norm": 0.7111857908738513,
+      "learning_rate": 6.606757195133672e-07,
+      "loss": 1.5649,
+      "step": 2307
+    },
+    {
+      "epoch": 0.16081942654077971,
+      "grad_norm": 0.7302731077054829,
+      "learning_rate": 6.606423493270158e-07,
+      "loss": 1.5731,
+      "step": 2308
+    },
+    {
+      "epoch": 0.16088910566839704,
+      "grad_norm": 0.6801391548735751,
+      "learning_rate": 6.606089659313172e-07,
+      "loss": 1.5673,
+      "step": 2309
+    },
+    {
+      "epoch": 0.16095878479601436,
+      "grad_norm": 0.7827230409718977,
+      "learning_rate": 6.605755693278711e-07,
+      "loss": 1.502,
+      "step": 2310
+    },
+    {
+      "epoch": 0.16102846392363168,
+      "grad_norm": 0.7035852984193754,
+      "learning_rate": 6.605421595182779e-07,
+      "loss": 1.5612,
+      "step": 2311
+    },
+    {
+      "epoch": 0.161098143051249,
+      "grad_norm": 0.6825492690800449,
+      "learning_rate": 6.605087365041389e-07,
+      "loss": 1.5236,
+      "step": 2312
+    },
+    {
+      "epoch": 0.16116782217886633,
+      "grad_norm": 0.6942928057626722,
+      "learning_rate": 6.604753002870555e-07,
+      "loss": 1.5146,
+      "step": 2313
+    },
+    {
+      "epoch": 0.16123750130648365,
+      "grad_norm": 0.8494657143717437,
+      "learning_rate": 6.604418508686302e-07,
+      "loss": 1.7396,
+      "step": 2314
+    },
+    {
+      "epoch": 0.16130718043410097,
+      "grad_norm": 0.6673471103730584,
+      "learning_rate": 6.604083882504659e-07,
+      "loss": 1.5284,
+      "step": 2315
+    },
+    {
+      "epoch": 0.1613768595617183,
+      "grad_norm": 0.7528593628272984,
+      "learning_rate": 6.603749124341663e-07,
+      "loss": 1.6444,
+      "step": 2316
+    },
+    {
+      "epoch": 0.16144653868933562,
+      "grad_norm": 0.7210025295412614,
+      "learning_rate": 6.603414234213357e-07,
+      "loss": 1.6882,
+      "step": 2317
+    },
+    {
+      "epoch": 0.16151621781695294,
+      "grad_norm": 0.7069438472104946,
+      "learning_rate": 6.603079212135785e-07,
+      "loss": 1.6581,
+      "step": 2318
+    },
+    {
+      "epoch": 0.16158589694457026,
+      "grad_norm": 0.7138732731517777,
+      "learning_rate": 6.602744058125009e-07,
+      "loss": 1.5197,
+      "step": 2319
+    },
+    {
+      "epoch": 0.16165557607218758,
+      "grad_norm": 0.7168132849247152,
+      "learning_rate": 6.602408772197084e-07,
+      "loss": 1.4548,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1617252551998049,
+      "grad_norm": 0.7326189178797672,
+      "learning_rate": 6.60207335436808e-07,
+      "loss": 1.6001,
+      "step": 2321
+    },
+    {
+      "epoch": 0.16179493432742223,
+      "grad_norm": 0.6927914280769315,
+      "learning_rate": 6.601737804654071e-07,
+      "loss": 1.6246,
+      "step": 2322
+    },
+    {
+      "epoch": 0.16186461345503955,
+      "grad_norm": 0.7548539239584744,
+      "learning_rate": 6.601402123071138e-07,
+      "loss": 1.7135,
+      "step": 2323
+    },
+    {
+      "epoch": 0.16193429258265687,
+      "grad_norm": 0.8231932734792815,
+      "learning_rate": 6.601066309635366e-07,
+      "loss": 1.6542,
+      "step": 2324
+    },
+    {
+      "epoch": 0.1620039717102742,
+      "grad_norm": 0.7076281356457971,
+      "learning_rate": 6.600730364362849e-07,
+      "loss": 1.6272,
+      "step": 2325
+    },
+    {
+      "epoch": 0.16207365083789152,
+      "grad_norm": 0.7152242206236852,
+      "learning_rate": 6.600394287269687e-07,
+      "loss": 1.4146,
+      "step": 2326
+    },
+    {
+      "epoch": 0.16214332996550884,
+      "grad_norm": 0.7079726966566244,
+      "learning_rate": 6.600058078371983e-07,
+      "loss": 1.4889,
+      "step": 2327
+    },
+    {
+      "epoch": 0.16221300909312616,
+      "grad_norm": 0.7254514558992783,
+      "learning_rate": 6.599721737685849e-07,
+      "loss": 1.502,
+      "step": 2328
+    },
+    {
+      "epoch": 0.1622826882207435,
+      "grad_norm": 0.6421208663306851,
+      "learning_rate": 6.599385265227405e-07,
+      "loss": 1.4715,
+      "step": 2329
+    },
+    {
+      "epoch": 0.1623523673483608,
+      "grad_norm": 0.7134578161355969,
+      "learning_rate": 6.599048661012776e-07,
+      "loss": 1.4987,
+      "step": 2330
+    },
+    {
+      "epoch": 0.16242204647597813,
+      "grad_norm": 0.6944470963507515,
+      "learning_rate": 6.598711925058091e-07,
+      "loss": 1.518,
+      "step": 2331
+    },
+    {
+      "epoch": 0.16249172560359545,
+      "grad_norm": 0.7024672276158266,
+      "learning_rate": 6.598375057379487e-07,
+      "loss": 1.6774,
+      "step": 2332
+    },
+    {
+      "epoch": 0.16256140473121278,
+      "grad_norm": 0.7529700252137836,
+      "learning_rate": 6.598038057993109e-07,
+      "loss": 1.5766,
+      "step": 2333
+    },
+    {
+      "epoch": 0.1626310838588301,
+      "grad_norm": 0.7632965108002485,
+      "learning_rate": 6.597700926915103e-07,
+      "loss": 1.6279,
+      "step": 2334
+    },
+    {
+      "epoch": 0.16270076298644742,
+      "grad_norm": 0.7201730099281007,
+      "learning_rate": 6.59736366416163e-07,
+      "loss": 1.5469,
+      "step": 2335
+    },
+    {
+      "epoch": 0.16277044211406475,
+      "grad_norm": 0.6996857758852515,
+      "learning_rate": 6.59702626974885e-07,
+      "loss": 1.5442,
+      "step": 2336
+    },
+    {
+      "epoch": 0.16284012124168207,
+      "grad_norm": 0.6917112721568993,
+      "learning_rate": 6.596688743692931e-07,
+      "loss": 1.563,
+      "step": 2337
+    },
+    {
+      "epoch": 0.1629098003692994,
+      "grad_norm": 0.666913569996407,
+      "learning_rate": 6.596351086010048e-07,
+      "loss": 1.638,
+      "step": 2338
+    },
+    {
+      "epoch": 0.16297947949691668,
+      "grad_norm": 0.6802046539952517,
+      "learning_rate": 6.596013296716384e-07,
+      "loss": 1.5565,
+      "step": 2339
+    },
+    {
+      "epoch": 0.163049158624534,
+      "grad_norm": 0.7063793504298344,
+      "learning_rate": 6.595675375828124e-07,
+      "loss": 1.5484,
+      "step": 2340
+    },
+    {
+      "epoch": 0.16311883775215133,
+      "grad_norm": 0.7332190383926808,
+      "learning_rate": 6.595337323361465e-07,
+      "loss": 1.5243,
+      "step": 2341
+    },
+    {
+      "epoch": 0.16318851687976865,
+      "grad_norm": 0.7329461018523288,
+      "learning_rate": 6.594999139332605e-07,
+      "loss": 1.7221,
+      "step": 2342
+    },
+    {
+      "epoch": 0.16325819600738598,
+      "grad_norm": 0.7419199688652642,
+      "learning_rate": 6.594660823757749e-07,
+      "loss": 1.5138,
+      "step": 2343
+    },
+    {
+      "epoch": 0.1633278751350033,
+      "grad_norm": 0.7639302342058242,
+      "learning_rate": 6.594322376653114e-07,
+      "loss": 1.6771,
+      "step": 2344
+    },
+    {
+      "epoch": 0.16339755426262062,
+      "grad_norm": 0.7472337265405039,
+      "learning_rate": 6.593983798034915e-07,
+      "loss": 1.518,
+      "step": 2345
+    },
+    {
+      "epoch": 0.16346723339023794,
+      "grad_norm": 0.7281142288983273,
+      "learning_rate": 6.59364508791938e-07,
+      "loss": 1.5627,
+      "step": 2346
+    },
+    {
+      "epoch": 0.16353691251785527,
+      "grad_norm": 0.7110147549449994,
+      "learning_rate": 6.593306246322739e-07,
+      "loss": 1.5052,
+      "step": 2347
+    },
+    {
+      "epoch": 0.1636065916454726,
+      "grad_norm": 0.7074364175708336,
+      "learning_rate": 6.592967273261232e-07,
+      "loss": 1.5836,
+      "step": 2348
+    },
+    {
+      "epoch": 0.1636762707730899,
+      "grad_norm": 0.7421192254211886,
+      "learning_rate": 6.592628168751102e-07,
+      "loss": 1.5003,
+      "step": 2349
+    },
+    {
+      "epoch": 0.16374594990070723,
+      "grad_norm": 0.7168279539068398,
+      "learning_rate": 6.592288932808598e-07,
+      "loss": 1.5503,
+      "step": 2350
+    },
+    {
+      "epoch": 0.16381562902832456,
+      "grad_norm": 0.7188929197436819,
+      "learning_rate": 6.591949565449979e-07,
+      "loss": 1.5518,
+      "step": 2351
+    },
+    {
+      "epoch": 0.16388530815594188,
+      "grad_norm": 0.762876427581947,
+      "learning_rate": 6.591610066691508e-07,
+      "loss": 1.5161,
+      "step": 2352
+    },
+    {
+      "epoch": 0.1639549872835592,
+      "grad_norm": 0.764966904869516,
+      "learning_rate": 6.591270436549456e-07,
+      "loss": 1.5519,
+      "step": 2353
+    },
+    {
+      "epoch": 0.16402466641117652,
+      "grad_norm": 0.7005514577699263,
+      "learning_rate": 6.590930675040095e-07,
+      "loss": 1.5517,
+      "step": 2354
+    },
+    {
+      "epoch": 0.16409434553879385,
+      "grad_norm": 0.6770300769048003,
+      "learning_rate": 6.59059078217971e-07,
+      "loss": 1.4233,
+      "step": 2355
+    },
+    {
+      "epoch": 0.16416402466641117,
+      "grad_norm": 0.7150779632036103,
+      "learning_rate": 6.590250757984588e-07,
+      "loss": 1.5995,
+      "step": 2356
+    },
+    {
+      "epoch": 0.1642337037940285,
+      "grad_norm": 0.6993402330838724,
+      "learning_rate": 6.589910602471023e-07,
+      "loss": 1.4289,
+      "step": 2357
+    },
+    {
+      "epoch": 0.1643033829216458,
+      "grad_norm": 0.7129939019203483,
+      "learning_rate": 6.589570315655318e-07,
+      "loss": 1.6153,
+      "step": 2358
+    },
+    {
+      "epoch": 0.16437306204926314,
+      "grad_norm": 0.7046647737874796,
+      "learning_rate": 6.589229897553779e-07,
+      "loss": 1.582,
+      "step": 2359
+    },
+    {
+      "epoch": 0.16444274117688046,
+      "grad_norm": 0.7290829227675538,
+      "learning_rate": 6.58888934818272e-07,
+      "loss": 1.4972,
+      "step": 2360
+    },
+    {
+      "epoch": 0.16451242030449778,
+      "grad_norm": 0.7072365195859416,
+      "learning_rate": 6.588548667558461e-07,
+      "loss": 1.4581,
+      "step": 2361
+    },
+    {
+      "epoch": 0.1645820994321151,
+      "grad_norm": 0.735679978353686,
+      "learning_rate": 6.588207855697326e-07,
+      "loss": 1.5999,
+      "step": 2362
+    },
+    {
+      "epoch": 0.16465177855973243,
+      "grad_norm": 0.7120299813416344,
+      "learning_rate": 6.587866912615651e-07,
+      "loss": 1.5704,
+      "step": 2363
+    },
+    {
+      "epoch": 0.16472145768734975,
+      "grad_norm": 0.6820387705988862,
+      "learning_rate": 6.587525838329772e-07,
+      "loss": 1.4342,
+      "step": 2364
+    },
+    {
+      "epoch": 0.16479113681496707,
+      "grad_norm": 0.7107393107714317,
+      "learning_rate": 6.587184632856035e-07,
+      "loss": 1.544,
+      "step": 2365
+    },
+    {
+      "epoch": 0.1648608159425844,
+      "grad_norm": 0.7535777040315426,
+      "learning_rate": 6.58684329621079e-07,
+      "loss": 1.6942,
+      "step": 2366
+    },
+    {
+      "epoch": 0.16493049507020172,
+      "grad_norm": 0.6771116719963444,
+      "learning_rate": 6.586501828410397e-07,
+      "loss": 1.5631,
+      "step": 2367
+    },
+    {
+      "epoch": 0.16500017419781904,
+      "grad_norm": 0.6617367548227207,
+      "learning_rate": 6.586160229471219e-07,
+      "loss": 1.4529,
+      "step": 2368
+    },
+    {
+      "epoch": 0.16506985332543636,
+      "grad_norm": 0.756471565795342,
+      "learning_rate": 6.585818499409624e-07,
+      "loss": 1.6085,
+      "step": 2369
+    },
+    {
+      "epoch": 0.16513953245305368,
+      "grad_norm": 0.7215067440748905,
+      "learning_rate": 6.585476638241991e-07,
+      "loss": 1.5597,
+      "step": 2370
+    },
+    {
+      "epoch": 0.165209211580671,
+      "grad_norm": 0.6715639533999088,
+      "learning_rate": 6.585134645984701e-07,
+      "loss": 1.5579,
+      "step": 2371
+    },
+    {
+      "epoch": 0.16527889070828833,
+      "grad_norm": 0.7528494030685173,
+      "learning_rate": 6.584792522654144e-07,
+      "loss": 1.613,
+      "step": 2372
+    },
+    {
+      "epoch": 0.16534856983590565,
+      "grad_norm": 0.9070409370704179,
+      "learning_rate": 6.584450268266715e-07,
+      "loss": 1.5739,
+      "step": 2373
+    },
+    {
+      "epoch": 0.16541824896352297,
+      "grad_norm": 0.7127053800060373,
+      "learning_rate": 6.584107882838815e-07,
+      "loss": 1.5985,
+      "step": 2374
+    },
+    {
+      "epoch": 0.1654879280911403,
+      "grad_norm": 0.7285186039275199,
+      "learning_rate": 6.583765366386853e-07,
+      "loss": 1.5165,
+      "step": 2375
+    },
+    {
+      "epoch": 0.16555760721875762,
+      "grad_norm": 0.7571340368115554,
+      "learning_rate": 6.583422718927242e-07,
+      "loss": 1.4696,
+      "step": 2376
+    },
+    {
+      "epoch": 0.16562728634637494,
+      "grad_norm": 0.6658658814069529,
+      "learning_rate": 6.583079940476402e-07,
+      "loss": 1.4661,
+      "step": 2377
+    },
+    {
+      "epoch": 0.16569696547399226,
+      "grad_norm": 0.8030332451545158,
+      "learning_rate": 6.582737031050761e-07,
+      "loss": 1.8509,
+      "step": 2378
+    },
+    {
+      "epoch": 0.1657666446016096,
+      "grad_norm": 0.6966252141405037,
+      "learning_rate": 6.582393990666751e-07,
+      "loss": 1.5828,
+      "step": 2379
+    },
+    {
+      "epoch": 0.1658363237292269,
+      "grad_norm": 0.7364229221914593,
+      "learning_rate": 6.582050819340812e-07,
+      "loss": 1.6218,
+      "step": 2380
+    },
+    {
+      "epoch": 0.16590600285684423,
+      "grad_norm": 0.7196295616760052,
+      "learning_rate": 6.581707517089389e-07,
+      "loss": 1.5562,
+      "step": 2381
+    },
+    {
+      "epoch": 0.16597568198446155,
+      "grad_norm": 0.7378354060426576,
+      "learning_rate": 6.581364083928932e-07,
+      "loss": 1.6569,
+      "step": 2382
+    },
+    {
+      "epoch": 0.16604536111207888,
+      "grad_norm": 0.7020021368288268,
+      "learning_rate": 6.581020519875903e-07,
+      "loss": 1.4947,
+      "step": 2383
+    },
+    {
+      "epoch": 0.1661150402396962,
+      "grad_norm": 23.011605268027264,
+      "learning_rate": 6.580676824946763e-07,
+      "loss": 1.4042,
+      "step": 2384
+    },
+    {
+      "epoch": 0.16618471936731352,
+      "grad_norm": 0.7040826666782872,
+      "learning_rate": 6.580332999157983e-07,
+      "loss": 1.469,
+      "step": 2385
+    },
+    {
+      "epoch": 0.16625439849493084,
+      "grad_norm": 0.7218140697486265,
+      "learning_rate": 6.579989042526041e-07,
+      "loss": 1.5228,
+      "step": 2386
+    },
+    {
+      "epoch": 0.16632407762254817,
+      "grad_norm": 0.7123392833850649,
+      "learning_rate": 6.579644955067421e-07,
+      "loss": 1.5075,
+      "step": 2387
+    },
+    {
+      "epoch": 0.1663937567501655,
+      "grad_norm": 0.742613976751209,
+      "learning_rate": 6.579300736798608e-07,
+      "loss": 1.6328,
+      "step": 2388
+    },
+    {
+      "epoch": 0.1664634358777828,
+      "grad_norm": 0.71508688345372,
+      "learning_rate": 6.578956387736101e-07,
+      "loss": 1.5266,
+      "step": 2389
+    },
+    {
+      "epoch": 0.16653311500540013,
+      "grad_norm": 0.7778468846082288,
+      "learning_rate": 6.578611907896403e-07,
+      "loss": 1.6486,
+      "step": 2390
+    },
+    {
+      "epoch": 0.16660279413301746,
+      "grad_norm": 0.6972662449499131,
+      "learning_rate": 6.57826729729602e-07,
+      "loss": 1.5245,
+      "step": 2391
+    },
+    {
+      "epoch": 0.16667247326063478,
+      "grad_norm": 0.6742424513816168,
+      "learning_rate": 6.577922555951467e-07,
+      "loss": 1.4659,
+      "step": 2392
+    },
+    {
+      "epoch": 0.1667421523882521,
+      "grad_norm": 0.7191566959015561,
+      "learning_rate": 6.577577683879266e-07,
+      "loss": 1.6706,
+      "step": 2393
+    },
+    {
+      "epoch": 0.16681183151586942,
+      "grad_norm": 0.742215586917538,
+      "learning_rate": 6.577232681095941e-07,
+      "loss": 1.6183,
+      "step": 2394
+    },
+    {
+      "epoch": 0.16688151064348675,
+      "grad_norm": 0.7459211762633696,
+      "learning_rate": 6.576887547618028e-07,
+      "loss": 1.6543,
+      "step": 2395
+    },
+    {
+      "epoch": 0.16695118977110407,
+      "grad_norm": 0.7406229483744687,
+      "learning_rate": 6.576542283462065e-07,
+      "loss": 1.6511,
+      "step": 2396
+    },
+    {
+      "epoch": 0.1670208688987214,
+      "grad_norm": 0.7041449161282237,
+      "learning_rate": 6.576196888644599e-07,
+      "loss": 1.4184,
+      "step": 2397
+    },
+    {
+      "epoch": 0.16709054802633871,
+      "grad_norm": 0.705938722848296,
+      "learning_rate": 6.57585136318218e-07,
+      "loss": 1.5844,
+      "step": 2398
+    },
+    {
+      "epoch": 0.16716022715395604,
+      "grad_norm": 0.7396198963679226,
+      "learning_rate": 6.575505707091368e-07,
+      "loss": 1.5891,
+      "step": 2399
+    },
+    {
+      "epoch": 0.16722990628157336,
+      "grad_norm": 0.7350092724605711,
+      "learning_rate": 6.575159920388728e-07,
+      "loss": 1.6338,
+      "step": 2400
+    },
+    {
+      "epoch": 0.16729958540919068,
+      "grad_norm": 0.7163445839743655,
+      "learning_rate": 6.57481400309083e-07,
+      "loss": 1.5543,
+      "step": 2401
+    },
+    {
+      "epoch": 0.167369264536808,
+      "grad_norm": 0.7149076387617239,
+      "learning_rate": 6.574467955214251e-07,
+      "loss": 1.5448,
+      "step": 2402
+    },
+    {
+      "epoch": 0.16743894366442533,
+      "grad_norm": 0.7125466696460487,
+      "learning_rate": 6.574121776775573e-07,
+      "loss": 1.3838,
+      "step": 2403
+    },
+    {
+      "epoch": 0.16750862279204265,
+      "grad_norm": 0.7253541486554131,
+      "learning_rate": 6.573775467791388e-07,
+      "loss": 1.6541,
+      "step": 2404
+    },
+    {
+      "epoch": 0.16757830191965997,
+      "grad_norm": 0.6888729373269321,
+      "learning_rate": 6.57342902827829e-07,
+      "loss": 1.4966,
+      "step": 2405
+    },
+    {
+      "epoch": 0.1676479810472773,
+      "grad_norm": 0.7242788546055735,
+      "learning_rate": 6.573082458252883e-07,
+      "loss": 1.4608,
+      "step": 2406
+    },
+    {
+      "epoch": 0.16771766017489462,
+      "grad_norm": 0.719095591694848,
+      "learning_rate": 6.572735757731774e-07,
+      "loss": 1.5231,
+      "step": 2407
+    },
+    {
+      "epoch": 0.16778733930251194,
+      "grad_norm": 0.7062431253120159,
+      "learning_rate": 6.572388926731578e-07,
+      "loss": 1.5036,
+      "step": 2408
+    },
+    {
+      "epoch": 0.16785701843012926,
+      "grad_norm": 0.7380879998828161,
+      "learning_rate": 6.572041965268916e-07,
+      "loss": 1.6707,
+      "step": 2409
+    },
+    {
+      "epoch": 0.16792669755774658,
+      "grad_norm": 0.6863769124791635,
+      "learning_rate": 6.571694873360414e-07,
+      "loss": 1.5437,
+      "step": 2410
+    },
+    {
+      "epoch": 0.1679963766853639,
+      "grad_norm": 0.708848587877779,
+      "learning_rate": 6.571347651022706e-07,
+      "loss": 1.5714,
+      "step": 2411
+    },
+    {
+      "epoch": 0.16806605581298123,
+      "grad_norm": 0.7038223757104983,
+      "learning_rate": 6.571000298272432e-07,
+      "loss": 1.5171,
+      "step": 2412
+    },
+    {
+      "epoch": 0.16813573494059855,
+      "grad_norm": 0.760760658766724,
+      "learning_rate": 6.570652815126238e-07,
+      "loss": 1.5314,
+      "step": 2413
+    },
+    {
+      "epoch": 0.16820541406821587,
+      "grad_norm": 0.7801390262080484,
+      "learning_rate": 6.570305201600774e-07,
+      "loss": 1.6455,
+      "step": 2414
+    },
+    {
+      "epoch": 0.1682750931958332,
+      "grad_norm": 0.6874636309530708,
+      "learning_rate": 6.569957457712702e-07,
+      "loss": 1.4671,
+      "step": 2415
+    },
+    {
+      "epoch": 0.16834477232345052,
+      "grad_norm": 0.7394775403102596,
+      "learning_rate": 6.569609583478684e-07,
+      "loss": 1.5806,
+      "step": 2416
+    },
+    {
+      "epoch": 0.16841445145106784,
+      "grad_norm": 0.7229398082360742,
+      "learning_rate": 6.56926157891539e-07,
+      "loss": 1.6871,
+      "step": 2417
+    },
+    {
+      "epoch": 0.16848413057868517,
+      "grad_norm": 0.6976349597214885,
+      "learning_rate": 6.568913444039502e-07,
+      "loss": 1.4284,
+      "step": 2418
+    },
+    {
+      "epoch": 0.1685538097063025,
+      "grad_norm": 0.7426082913327527,
+      "learning_rate": 6.568565178867696e-07,
+      "loss": 1.6139,
+      "step": 2419
+    },
+    {
+      "epoch": 0.1686234888339198,
+      "grad_norm": 0.7408162730120007,
+      "learning_rate": 6.568216783416669e-07,
+      "loss": 1.4382,
+      "step": 2420
+    },
+    {
+      "epoch": 0.16869316796153713,
+      "grad_norm": 0.7843645939651729,
+      "learning_rate": 6.567868257703112e-07,
+      "loss": 1.5234,
+      "step": 2421
+    },
+    {
+      "epoch": 0.16876284708915446,
+      "grad_norm": 0.7106920389232463,
+      "learning_rate": 6.567519601743728e-07,
+      "loss": 1.5792,
+      "step": 2422
+    },
+    {
+      "epoch": 0.16883252621677178,
+      "grad_norm": 0.7380262499470066,
+      "learning_rate": 6.567170815555226e-07,
+      "loss": 1.5791,
+      "step": 2423
+    },
+    {
+      "epoch": 0.1689022053443891,
+      "grad_norm": 0.7456240829771935,
+      "learning_rate": 6.56682189915432e-07,
+      "loss": 1.5671,
+      "step": 2424
+    },
+    {
+      "epoch": 0.16897188447200642,
+      "grad_norm": 0.7274486962987895,
+      "learning_rate": 6.566472852557731e-07,
+      "loss": 1.5414,
+      "step": 2425
+    },
+    {
+      "epoch": 0.16904156359962375,
+      "grad_norm": 0.7512202796475355,
+      "learning_rate": 6.566123675782187e-07,
+      "loss": 1.5159,
+      "step": 2426
+    },
+    {
+      "epoch": 0.16911124272724107,
+      "grad_norm": 0.709986504523135,
+      "learning_rate": 6.56577436884442e-07,
+      "loss": 1.5121,
+      "step": 2427
+    },
+    {
+      "epoch": 0.1691809218548584,
+      "grad_norm": 0.714474528207041,
+      "learning_rate": 6.56542493176117e-07,
+      "loss": 1.4527,
+      "step": 2428
+    },
+    {
+      "epoch": 0.1692506009824757,
+      "grad_norm": 0.6975956737888491,
+      "learning_rate": 6.565075364549182e-07,
+      "loss": 1.5378,
+      "step": 2429
+    },
+    {
+      "epoch": 0.16932028011009304,
+      "grad_norm": 0.7889336946448815,
+      "learning_rate": 6.564725667225209e-07,
+      "loss": 1.5498,
+      "step": 2430
+    },
+    {
+      "epoch": 0.16938995923771033,
+      "grad_norm": 0.7181576447687493,
+      "learning_rate": 6.564375839806009e-07,
+      "loss": 1.6174,
+      "step": 2431
+    },
+    {
+      "epoch": 0.16945963836532765,
+      "grad_norm": 0.7157683881150451,
+      "learning_rate": 6.564025882308345e-07,
+      "loss": 1.5108,
+      "step": 2432
+    },
+    {
+      "epoch": 0.16952931749294498,
+      "grad_norm": 0.719495796453952,
+      "learning_rate": 6.56367579474899e-07,
+      "loss": 1.5718,
+      "step": 2433
+    },
+    {
+      "epoch": 0.1695989966205623,
+      "grad_norm": 0.7061695225090542,
+      "learning_rate": 6.56332557714472e-07,
+      "loss": 1.4915,
+      "step": 2434
+    },
+    {
+      "epoch": 0.16966867574817962,
+      "grad_norm": 0.7577425760099268,
+      "learning_rate": 6.562975229512317e-07,
+      "loss": 1.5643,
+      "step": 2435
+    },
+    {
+      "epoch": 0.16973835487579694,
+      "grad_norm": 0.6952801720982502,
+      "learning_rate": 6.562624751868573e-07,
+      "loss": 1.49,
+      "step": 2436
+    },
+    {
+      "epoch": 0.16980803400341427,
+      "grad_norm": 0.7197087596785499,
+      "learning_rate": 6.562274144230281e-07,
+      "loss": 1.6057,
+      "step": 2437
+    },
+    {
+      "epoch": 0.1698777131310316,
+      "grad_norm": 0.6958464094196392,
+      "learning_rate": 6.561923406614243e-07,
+      "loss": 1.5574,
+      "step": 2438
+    },
+    {
+      "epoch": 0.1699473922586489,
+      "grad_norm": 0.7366661054000561,
+      "learning_rate": 6.56157253903727e-07,
+      "loss": 1.5172,
+      "step": 2439
+    },
+    {
+      "epoch": 0.17001707138626623,
+      "grad_norm": 0.6770931017295103,
+      "learning_rate": 6.561221541516173e-07,
+      "loss": 1.5862,
+      "step": 2440
+    },
+    {
+      "epoch": 0.17008675051388356,
+      "grad_norm": 0.7110420060229062,
+      "learning_rate": 6.560870414067773e-07,
+      "loss": 1.597,
+      "step": 2441
+    },
+    {
+      "epoch": 0.17015642964150088,
+      "grad_norm": 0.7620049791494978,
+      "learning_rate": 6.560519156708898e-07,
+      "loss": 1.6246,
+      "step": 2442
+    },
+    {
+      "epoch": 0.1702261087691182,
+      "grad_norm": 0.7140259867968641,
+      "learning_rate": 6.560167769456381e-07,
+      "loss": 1.5858,
+      "step": 2443
+    },
+    {
+      "epoch": 0.17029578789673552,
+      "grad_norm": 0.7617281906746146,
+      "learning_rate": 6.559816252327059e-07,
+      "loss": 1.6849,
+      "step": 2444
+    },
+    {
+      "epoch": 0.17036546702435285,
+      "grad_norm": 0.7099678487031629,
+      "learning_rate": 6.55946460533778e-07,
+      "loss": 1.5603,
+      "step": 2445
+    },
+    {
+      "epoch": 0.17043514615197017,
+      "grad_norm": 0.8066709045974286,
+      "learning_rate": 6.559112828505397e-07,
+      "loss": 1.6058,
+      "step": 2446
+    },
+    {
+      "epoch": 0.1705048252795875,
+      "grad_norm": 0.7135349848853826,
+      "learning_rate": 6.558760921846762e-07,
+      "loss": 1.6007,
+      "step": 2447
+    },
+    {
+      "epoch": 0.1705745044072048,
+      "grad_norm": 0.7441499343871727,
+      "learning_rate": 6.558408885378744e-07,
+      "loss": 1.5822,
+      "step": 2448
+    },
+    {
+      "epoch": 0.17064418353482214,
+      "grad_norm": 0.7321585891025582,
+      "learning_rate": 6.558056719118212e-07,
+      "loss": 1.6208,
+      "step": 2449
+    },
+    {
+      "epoch": 0.17071386266243946,
+      "grad_norm": 0.7988542772381881,
+      "learning_rate": 6.557704423082042e-07,
+      "loss": 1.6834,
+      "step": 2450
+    },
+    {
+      "epoch": 0.17078354179005678,
+      "grad_norm": 0.7160620851253273,
+      "learning_rate": 6.557351997287118e-07,
+      "loss": 1.5184,
+      "step": 2451
+    },
+    {
+      "epoch": 0.1708532209176741,
+      "grad_norm": 0.6947991412978013,
+      "learning_rate": 6.556999441750328e-07,
+      "loss": 1.547,
+      "step": 2452
+    },
+    {
+      "epoch": 0.17092290004529143,
+      "grad_norm": 0.7031959563327964,
+      "learning_rate": 6.556646756488567e-07,
+      "loss": 1.4695,
+      "step": 2453
+    },
+    {
+      "epoch": 0.17099257917290875,
+      "grad_norm": 0.7041369358140038,
+      "learning_rate": 6.556293941518737e-07,
+      "loss": 1.5845,
+      "step": 2454
+    },
+    {
+      "epoch": 0.17106225830052607,
+      "grad_norm": 0.7228048984189448,
+      "learning_rate": 6.555940996857746e-07,
+      "loss": 1.5521,
+      "step": 2455
+    },
+    {
+      "epoch": 0.1711319374281434,
+      "grad_norm": 0.763823425561266,
+      "learning_rate": 6.555587922522507e-07,
+      "loss": 1.514,
+      "step": 2456
+    },
+    {
+      "epoch": 0.17120161655576072,
+      "grad_norm": 0.7060298331702053,
+      "learning_rate": 6.55523471852994e-07,
+      "loss": 1.6084,
+      "step": 2457
+    },
+    {
+      "epoch": 0.17127129568337804,
+      "grad_norm": 0.7791863111355747,
+      "learning_rate": 6.554881384896971e-07,
+      "loss": 1.6451,
+      "step": 2458
+    },
+    {
+      "epoch": 0.17134097481099536,
+      "grad_norm": 0.7548442713753598,
+      "learning_rate": 6.554527921640534e-07,
+      "loss": 1.5738,
+      "step": 2459
+    },
+    {
+      "epoch": 0.17141065393861268,
+      "grad_norm": 0.7318943333159449,
+      "learning_rate": 6.554174328777566e-07,
+      "loss": 1.5405,
+      "step": 2460
+    },
+    {
+      "epoch": 0.17148033306623,
+      "grad_norm": 0.7107952467450837,
+      "learning_rate": 6.553820606325013e-07,
+      "loss": 1.5816,
+      "step": 2461
+    },
+    {
+      "epoch": 0.17155001219384733,
+      "grad_norm": 0.7148920719538931,
+      "learning_rate": 6.553466754299825e-07,
+      "loss": 1.4621,
+      "step": 2462
+    },
+    {
+      "epoch": 0.17161969132146465,
+      "grad_norm": 15.284258988235981,
+      "learning_rate": 6.553112772718961e-07,
+      "loss": 1.4982,
+      "step": 2463
+    },
+    {
+      "epoch": 0.17168937044908197,
+      "grad_norm": 0.7606998847724221,
+      "learning_rate": 6.552758661599384e-07,
+      "loss": 1.505,
+      "step": 2464
+    },
+    {
+      "epoch": 0.1717590495766993,
+      "grad_norm": 0.737717860072526,
+      "learning_rate": 6.552404420958061e-07,
+      "loss": 1.6274,
+      "step": 2465
+    },
+    {
+      "epoch": 0.17182872870431662,
+      "grad_norm": 0.6936147594438409,
+      "learning_rate": 6.552050050811973e-07,
+      "loss": 1.4483,
+      "step": 2466
+    },
+    {
+      "epoch": 0.17189840783193394,
+      "grad_norm": 0.6821739277267969,
+      "learning_rate": 6.551695551178097e-07,
+      "loss": 1.5747,
+      "step": 2467
+    },
+    {
+      "epoch": 0.17196808695955126,
+      "grad_norm": 0.7209446081125167,
+      "learning_rate": 6.551340922073425e-07,
+      "loss": 1.6935,
+      "step": 2468
+    },
+    {
+      "epoch": 0.1720377660871686,
+      "grad_norm": 0.706917031810206,
+      "learning_rate": 6.55098616351495e-07,
+      "loss": 1.5429,
+      "step": 2469
+    },
+    {
+      "epoch": 0.1721074452147859,
+      "grad_norm": 0.6657369029274698,
+      "learning_rate": 6.550631275519674e-07,
+      "loss": 1.6344,
+      "step": 2470
+    },
+    {
+      "epoch": 0.17217712434240323,
+      "grad_norm": 0.7300297982250075,
+      "learning_rate": 6.550276258104601e-07,
+      "loss": 1.7434,
+      "step": 2471
+    },
+    {
+      "epoch": 0.17224680347002055,
+      "grad_norm": 0.7218093987525189,
+      "learning_rate": 6.549921111286748e-07,
+      "loss": 1.5561,
+      "step": 2472
+    },
+    {
+      "epoch": 0.17231648259763788,
+      "grad_norm": 0.7178449444982237,
+      "learning_rate": 6.549565835083131e-07,
+      "loss": 1.4582,
+      "step": 2473
+    },
+    {
+      "epoch": 0.1723861617252552,
+      "grad_norm": 0.7341473296783434,
+      "learning_rate": 6.549210429510778e-07,
+      "loss": 1.5608,
+      "step": 2474
+    },
+    {
+      "epoch": 0.17245584085287252,
+      "grad_norm": 0.7524545461753286,
+      "learning_rate": 6.548854894586719e-07,
+      "loss": 1.5571,
+      "step": 2475
+    },
+    {
+      "epoch": 0.17252551998048984,
+      "grad_norm": 0.7265542621338041,
+      "learning_rate": 6.548499230327993e-07,
+      "loss": 1.524,
+      "step": 2476
+    },
+    {
+      "epoch": 0.17259519910810717,
+      "grad_norm": 0.7105786157504762,
+      "learning_rate": 6.548143436751646e-07,
+      "loss": 1.4915,
+      "step": 2477
+    },
+    {
+      "epoch": 0.1726648782357245,
+      "grad_norm": 0.7174301051303112,
+      "learning_rate": 6.547787513874723e-07,
+      "loss": 1.4659,
+      "step": 2478
+    },
+    {
+      "epoch": 0.1727345573633418,
+      "grad_norm": 0.7503192380533867,
+      "learning_rate": 6.547431461714286e-07,
+      "loss": 1.5658,
+      "step": 2479
+    },
+    {
+      "epoch": 0.17280423649095913,
+      "grad_norm": 0.6907154911256304,
+      "learning_rate": 6.547075280287396e-07,
+      "loss": 1.5539,
+      "step": 2480
+    },
+    {
+      "epoch": 0.17287391561857646,
+      "grad_norm": 0.7794515823710214,
+      "learning_rate": 6.546718969611121e-07,
+      "loss": 1.7881,
+      "step": 2481
+    },
+    {
+      "epoch": 0.17294359474619378,
+      "grad_norm": 0.6603868909619415,
+      "learning_rate": 6.546362529702536e-07,
+      "loss": 1.4557,
+      "step": 2482
+    },
+    {
+      "epoch": 0.1730132738738111,
+      "grad_norm": 0.7586159790450311,
+      "learning_rate": 6.546005960578724e-07,
+      "loss": 1.5587,
+      "step": 2483
+    },
+    {
+      "epoch": 0.17308295300142842,
+      "grad_norm": 0.7404883917977331,
+      "learning_rate": 6.545649262256771e-07,
+      "loss": 1.483,
+      "step": 2484
+    },
+    {
+      "epoch": 0.17315263212904575,
+      "grad_norm": 0.7526067704071594,
+      "learning_rate": 6.545292434753772e-07,
+      "loss": 1.4999,
+      "step": 2485
+    },
+    {
+      "epoch": 0.17322231125666307,
+      "grad_norm": 0.799844700462213,
+      "learning_rate": 6.544935478086825e-07,
+      "loss": 1.6473,
+      "step": 2486
+    },
+    {
+      "epoch": 0.1732919903842804,
+      "grad_norm": 0.7884391841826116,
+      "learning_rate": 6.544578392273038e-07,
+      "loss": 1.5843,
+      "step": 2487
+    },
+    {
+      "epoch": 0.17336166951189771,
+      "grad_norm": 0.722546423895997,
+      "learning_rate": 6.544221177329522e-07,
+      "loss": 1.5647,
+      "step": 2488
+    },
+    {
+      "epoch": 0.17343134863951504,
+      "grad_norm": 0.6995711761145036,
+      "learning_rate": 6.543863833273397e-07,
+      "loss": 1.4946,
+      "step": 2489
+    },
+    {
+      "epoch": 0.17350102776713236,
+      "grad_norm": 0.7608256105047294,
+      "learning_rate": 6.543506360121787e-07,
+      "loss": 1.5278,
+      "step": 2490
+    },
+    {
+      "epoch": 0.17357070689474968,
+      "grad_norm": 0.7060114399131842,
+      "learning_rate": 6.543148757891821e-07,
+      "loss": 1.4342,
+      "step": 2491
+    },
+    {
+      "epoch": 0.173640386022367,
+      "grad_norm": 0.7038526017797239,
+      "learning_rate": 6.542791026600639e-07,
+      "loss": 1.5103,
+      "step": 2492
+    },
+    {
+      "epoch": 0.17371006514998433,
+      "grad_norm": 0.7454708545350585,
+      "learning_rate": 6.542433166265382e-07,
+      "loss": 1.5699,
+      "step": 2493
+    },
+    {
+      "epoch": 0.17377974427760165,
+      "grad_norm": 0.7241434175996339,
+      "learning_rate": 6.5420751769032e-07,
+      "loss": 1.5849,
+      "step": 2494
+    },
+    {
+      "epoch": 0.17384942340521897,
+      "grad_norm": 0.6828214180093903,
+      "learning_rate": 6.54171705853125e-07,
+      "loss": 1.5575,
+      "step": 2495
+    },
+    {
+      "epoch": 0.1739191025328363,
+      "grad_norm": 0.7645534413737075,
+      "learning_rate": 6.541358811166691e-07,
+      "loss": 1.7047,
+      "step": 2496
+    },
+    {
+      "epoch": 0.17398878166045362,
+      "grad_norm": 0.7599757807395299,
+      "learning_rate": 6.541000434826694e-07,
+      "loss": 1.5909,
+      "step": 2497
+    },
+    {
+      "epoch": 0.17405846078807094,
+      "grad_norm": 0.8077318342213113,
+      "learning_rate": 6.540641929528431e-07,
+      "loss": 1.6113,
+      "step": 2498
+    },
+    {
+      "epoch": 0.17412813991568826,
+      "grad_norm": 0.7421508001687054,
+      "learning_rate": 6.540283295289082e-07,
+      "loss": 1.5622,
+      "step": 2499
+    },
+    {
+      "epoch": 0.17419781904330559,
+      "grad_norm": 0.7088253814000094,
+      "learning_rate": 6.539924532125835e-07,
+      "loss": 1.6426,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1742674981709229,
+      "grad_norm": 0.7474674019877895,
+      "learning_rate": 6.539565640055884e-07,
+      "loss": 1.5854,
+      "step": 2501
+    },
+    {
+      "epoch": 0.17433717729854023,
+      "grad_norm": 0.7365714423846182,
+      "learning_rate": 6.539206619096424e-07,
+      "loss": 1.6054,
+      "step": 2502
+    },
+    {
+      "epoch": 0.17440685642615755,
+      "grad_norm": 0.7123103017402581,
+      "learning_rate": 6.538847469264663e-07,
+      "loss": 1.6297,
+      "step": 2503
+    },
+    {
+      "epoch": 0.17447653555377488,
+      "grad_norm": 0.7224845735928211,
+      "learning_rate": 6.53848819057781e-07,
+      "loss": 1.6559,
+      "step": 2504
+    },
+    {
+      "epoch": 0.1745462146813922,
+      "grad_norm": 0.6926473364257885,
+      "learning_rate": 6.538128783053084e-07,
+      "loss": 1.6328,
+      "step": 2505
+    },
+    {
+      "epoch": 0.17461589380900952,
+      "grad_norm": 0.7662335923901936,
+      "learning_rate": 6.537769246707708e-07,
+      "loss": 1.405,
+      "step": 2506
+    },
+    {
+      "epoch": 0.17468557293662684,
+      "grad_norm": 0.6971301204534728,
+      "learning_rate": 6.537409581558911e-07,
+      "loss": 1.4508,
+      "step": 2507
+    },
+    {
+      "epoch": 0.17475525206424417,
+      "grad_norm": 0.7439320230916382,
+      "learning_rate": 6.537049787623931e-07,
+      "loss": 1.5743,
+      "step": 2508
+    },
+    {
+      "epoch": 0.1748249311918615,
+      "grad_norm": 0.7336682024026748,
+      "learning_rate": 6.536689864920009e-07,
+      "loss": 1.6968,
+      "step": 2509
+    },
+    {
+      "epoch": 0.1748946103194788,
+      "grad_norm": 0.7449360351785614,
+      "learning_rate": 6.536329813464393e-07,
+      "loss": 1.6851,
+      "step": 2510
+    },
+    {
+      "epoch": 0.17496428944709613,
+      "grad_norm": 0.7593962246800922,
+      "learning_rate": 6.535969633274337e-07,
+      "loss": 1.5848,
+      "step": 2511
+    },
+    {
+      "epoch": 0.17503396857471346,
+      "grad_norm": 0.6892811934428639,
+      "learning_rate": 6.535609324367102e-07,
+      "loss": 1.5004,
+      "step": 2512
+    },
+    {
+      "epoch": 0.17510364770233078,
+      "grad_norm": 0.7357891038175344,
+      "learning_rate": 6.535248886759954e-07,
+      "loss": 1.5088,
+      "step": 2513
+    },
+    {
+      "epoch": 0.1751733268299481,
+      "grad_norm": 0.7200528123000481,
+      "learning_rate": 6.534888320470168e-07,
+      "loss": 1.5738,
+      "step": 2514
+    },
+    {
+      "epoch": 0.17524300595756542,
+      "grad_norm": 0.6873780049343139,
+      "learning_rate": 6.53452762551502e-07,
+      "loss": 1.4716,
+      "step": 2515
+    },
+    {
+      "epoch": 0.17531268508518275,
+      "grad_norm": 0.724561950813103,
+      "learning_rate": 6.534166801911799e-07,
+      "loss": 1.434,
+      "step": 2516
+    },
+    {
+      "epoch": 0.17538236421280007,
+      "grad_norm": 0.7489153459676924,
+      "learning_rate": 6.533805849677792e-07,
+      "loss": 1.4961,
+      "step": 2517
+    },
+    {
+      "epoch": 0.1754520433404174,
+      "grad_norm": 0.7401673701928653,
+      "learning_rate": 6.533444768830302e-07,
+      "loss": 1.5407,
+      "step": 2518
+    },
+    {
+      "epoch": 0.1755217224680347,
+      "grad_norm": 0.6934330185732609,
+      "learning_rate": 6.533083559386627e-07,
+      "loss": 1.4262,
+      "step": 2519
+    },
+    {
+      "epoch": 0.17559140159565204,
+      "grad_norm": 0.7294097563407264,
+      "learning_rate": 6.532722221364081e-07,
+      "loss": 1.6563,
+      "step": 2520
+    },
+    {
+      "epoch": 0.17566108072326936,
+      "grad_norm": 0.6607520492198753,
+      "learning_rate": 6.532360754779977e-07,
+      "loss": 1.4277,
+      "step": 2521
+    },
+    {
+      "epoch": 0.17573075985088665,
+      "grad_norm": 0.734449690451649,
+      "learning_rate": 6.53199915965164e-07,
+      "loss": 1.5041,
+      "step": 2522
+    },
+    {
+      "epoch": 0.17580043897850398,
+      "grad_norm": 0.7278895608033406,
+      "learning_rate": 6.531637435996397e-07,
+      "loss": 1.5348,
+      "step": 2523
+    },
+    {
+      "epoch": 0.1758701181061213,
+      "grad_norm": 0.6880733498662125,
+      "learning_rate": 6.531275583831584e-07,
+      "loss": 1.5598,
+      "step": 2524
+    },
+    {
+      "epoch": 0.17593979723373862,
+      "grad_norm": 0.7219542430932584,
+      "learning_rate": 6.530913603174539e-07,
+      "loss": 1.6169,
+      "step": 2525
+    },
+    {
+      "epoch": 0.17600947636135594,
+      "grad_norm": 0.6768504677014932,
+      "learning_rate": 6.530551494042611e-07,
+      "loss": 1.3468,
+      "step": 2526
+    },
+    {
+      "epoch": 0.17607915548897327,
+      "grad_norm": 0.7628140093864075,
+      "learning_rate": 6.530189256453151e-07,
+      "loss": 1.4331,
+      "step": 2527
+    },
+    {
+      "epoch": 0.1761488346165906,
+      "grad_norm": 0.6937139477500073,
+      "learning_rate": 6.529826890423521e-07,
+      "loss": 1.5771,
+      "step": 2528
+    },
+    {
+      "epoch": 0.1762185137442079,
+      "grad_norm": 0.7446668334736429,
+      "learning_rate": 6.529464395971085e-07,
+      "loss": 1.5671,
+      "step": 2529
+    },
+    {
+      "epoch": 0.17628819287182523,
+      "grad_norm": 0.8951513009342587,
+      "learning_rate": 6.529101773113213e-07,
+      "loss": 1.6593,
+      "step": 2530
+    },
+    {
+      "epoch": 0.17635787199944256,
+      "grad_norm": 0.7501370933617739,
+      "learning_rate": 6.528739021867285e-07,
+      "loss": 1.5608,
+      "step": 2531
+    },
+    {
+      "epoch": 0.17642755112705988,
+      "grad_norm": 0.7259327370195917,
+      "learning_rate": 6.528376142250684e-07,
+      "loss": 1.6285,
+      "step": 2532
+    },
+    {
+      "epoch": 0.1764972302546772,
+      "grad_norm": 0.6513213101252865,
+      "learning_rate": 6.528013134280799e-07,
+      "loss": 1.4204,
+      "step": 2533
+    },
+    {
+      "epoch": 0.17656690938229452,
+      "grad_norm": 0.6998756742851133,
+      "learning_rate": 6.527649997975026e-07,
+      "loss": 1.495,
+      "step": 2534
+    },
+    {
+      "epoch": 0.17663658850991185,
+      "grad_norm": 0.7206725602051455,
+      "learning_rate": 6.52728673335077e-07,
+      "loss": 1.6582,
+      "step": 2535
+    },
+    {
+      "epoch": 0.17670626763752917,
+      "grad_norm": 0.725602973875328,
+      "learning_rate": 6.526923340425437e-07,
+      "loss": 1.5383,
+      "step": 2536
+    },
+    {
+      "epoch": 0.1767759467651465,
+      "grad_norm": 0.7314479526711624,
+      "learning_rate": 6.52655981921644e-07,
+      "loss": 1.378,
+      "step": 2537
+    },
+    {
+      "epoch": 0.1768456258927638,
+      "grad_norm": 0.7116441177034473,
+      "learning_rate": 6.526196169741202e-07,
+      "loss": 1.4627,
+      "step": 2538
+    },
+    {
+      "epoch": 0.17691530502038114,
+      "grad_norm": 0.7290784007685396,
+      "learning_rate": 6.525832392017151e-07,
+      "loss": 1.5596,
+      "step": 2539
+    },
+    {
+      "epoch": 0.17698498414799846,
+      "grad_norm": 0.6934353435568296,
+      "learning_rate": 6.525468486061717e-07,
+      "loss": 1.4273,
+      "step": 2540
+    },
+    {
+      "epoch": 0.17705466327561578,
+      "grad_norm": 0.7232881714758217,
+      "learning_rate": 6.525104451892341e-07,
+      "loss": 1.5556,
+      "step": 2541
+    },
+    {
+      "epoch": 0.1771243424032331,
+      "grad_norm": 0.7330226056875261,
+      "learning_rate": 6.524740289526469e-07,
+      "loss": 1.5676,
+      "step": 2542
+    },
+    {
+      "epoch": 0.17719402153085043,
+      "grad_norm": 0.6942435985096442,
+      "learning_rate": 6.524375998981549e-07,
+      "loss": 1.4992,
+      "step": 2543
+    },
+    {
+      "epoch": 0.17726370065846775,
+      "grad_norm": 0.7087192076326799,
+      "learning_rate": 6.524011580275041e-07,
+      "loss": 1.5234,
+      "step": 2544
+    },
+    {
+      "epoch": 0.17733337978608507,
+      "grad_norm": 0.6906658016292048,
+      "learning_rate": 6.52364703342441e-07,
+      "loss": 1.5446,
+      "step": 2545
+    },
+    {
+      "epoch": 0.1774030589137024,
+      "grad_norm": 0.6590162253016592,
+      "learning_rate": 6.523282358447123e-07,
+      "loss": 1.4942,
+      "step": 2546
+    },
+    {
+      "epoch": 0.17747273804131972,
+      "grad_norm": 0.7374886336643569,
+      "learning_rate": 6.522917555360658e-07,
+      "loss": 1.6685,
+      "step": 2547
+    },
+    {
+      "epoch": 0.17754241716893704,
+      "grad_norm": 0.7288316061604521,
+      "learning_rate": 6.522552624182495e-07,
+      "loss": 1.5612,
+      "step": 2548
+    },
+    {
+      "epoch": 0.17761209629655436,
+      "grad_norm": 0.7986604299216298,
+      "learning_rate": 6.522187564930125e-07,
+      "loss": 1.5161,
+      "step": 2549
+    },
+    {
+      "epoch": 0.17768177542417168,
+      "grad_norm": 0.7490611352894104,
+      "learning_rate": 6.521822377621042e-07,
+      "loss": 1.571,
+      "step": 2550
+    },
+    {
+      "epoch": 0.177751454551789,
+      "grad_norm": 0.7096302346445801,
+      "learning_rate": 6.521457062272743e-07,
+      "loss": 1.4839,
+      "step": 2551
+    },
+    {
+      "epoch": 0.17782113367940633,
+      "grad_norm": 0.7300140825304802,
+      "learning_rate": 6.521091618902738e-07,
+      "loss": 1.615,
+      "step": 2552
+    },
+    {
+      "epoch": 0.17789081280702365,
+      "grad_norm": 0.6966168480177054,
+      "learning_rate": 6.520726047528539e-07,
+      "loss": 1.5791,
+      "step": 2553
+    },
+    {
+      "epoch": 0.17796049193464097,
+      "grad_norm": 0.7468143622329642,
+      "learning_rate": 6.520360348167666e-07,
+      "loss": 1.5228,
+      "step": 2554
+    },
+    {
+      "epoch": 0.1780301710622583,
+      "grad_norm": 0.6872146413398952,
+      "learning_rate": 6.519994520837641e-07,
+      "loss": 1.4354,
+      "step": 2555
+    },
+    {
+      "epoch": 0.17809985018987562,
+      "grad_norm": 0.7369132879346765,
+      "learning_rate": 6.519628565555998e-07,
+      "loss": 1.564,
+      "step": 2556
+    },
+    {
+      "epoch": 0.17816952931749294,
+      "grad_norm": 0.7484987827082581,
+      "learning_rate": 6.519262482340275e-07,
+      "loss": 1.4987,
+      "step": 2557
+    },
+    {
+      "epoch": 0.17823920844511026,
+      "grad_norm": 0.7590198399272611,
+      "learning_rate": 6.518896271208012e-07,
+      "loss": 1.7448,
+      "step": 2558
+    },
+    {
+      "epoch": 0.1783088875727276,
+      "grad_norm": 0.6911442734510137,
+      "learning_rate": 6.518529932176761e-07,
+      "loss": 1.6123,
+      "step": 2559
+    },
+    {
+      "epoch": 0.1783785667003449,
+      "grad_norm": 0.6705004696348974,
+      "learning_rate": 6.518163465264078e-07,
+      "loss": 1.5614,
+      "step": 2560
+    },
+    {
+      "epoch": 0.17844824582796223,
+      "grad_norm": 0.750693147635284,
+      "learning_rate": 6.517796870487524e-07,
+      "loss": 1.5328,
+      "step": 2561
+    },
+    {
+      "epoch": 0.17851792495557955,
+      "grad_norm": 0.6856127610822711,
+      "learning_rate": 6.517430147864667e-07,
+      "loss": 1.48,
+      "step": 2562
+    },
+    {
+      "epoch": 0.17858760408319688,
+      "grad_norm": 0.7275820774071743,
+      "learning_rate": 6.51706329741308e-07,
+      "loss": 1.5447,
+      "step": 2563
+    },
+    {
+      "epoch": 0.1786572832108142,
+      "grad_norm": 0.6711195393492667,
+      "learning_rate": 6.516696319150345e-07,
+      "loss": 1.5349,
+      "step": 2564
+    },
+    {
+      "epoch": 0.17872696233843152,
+      "grad_norm": 0.7017299635771622,
+      "learning_rate": 6.51632921309405e-07,
+      "loss": 1.4836,
+      "step": 2565
+    },
+    {
+      "epoch": 0.17879664146604884,
+      "grad_norm": 0.6998012714366909,
+      "learning_rate": 6.515961979261782e-07,
+      "loss": 1.4723,
+      "step": 2566
+    },
+    {
+      "epoch": 0.17886632059366617,
+      "grad_norm": 0.7104918945327922,
+      "learning_rate": 6.515594617671145e-07,
+      "loss": 1.582,
+      "step": 2567
+    },
+    {
+      "epoch": 0.1789359997212835,
+      "grad_norm": 0.6731653151031656,
+      "learning_rate": 6.515227128339741e-07,
+      "loss": 1.5919,
+      "step": 2568
+    },
+    {
+      "epoch": 0.1790056788489008,
+      "grad_norm": 0.703887895548082,
+      "learning_rate": 6.51485951128518e-07,
+      "loss": 1.5629,
+      "step": 2569
+    },
+    {
+      "epoch": 0.17907535797651813,
+      "grad_norm": 0.7025048472660292,
+      "learning_rate": 6.514491766525082e-07,
+      "loss": 1.5729,
+      "step": 2570
+    },
+    {
+      "epoch": 0.17914503710413546,
+      "grad_norm": 0.6551133693668119,
+      "learning_rate": 6.514123894077068e-07,
+      "loss": 1.5205,
+      "step": 2571
+    },
+    {
+      "epoch": 0.17921471623175278,
+      "grad_norm": 0.7821354033326472,
+      "learning_rate": 6.513755893958768e-07,
+      "loss": 1.5875,
+      "step": 2572
+    },
+    {
+      "epoch": 0.1792843953593701,
+      "grad_norm": 0.7426558766874091,
+      "learning_rate": 6.513387766187816e-07,
+      "loss": 1.5528,
+      "step": 2573
+    },
+    {
+      "epoch": 0.17935407448698742,
+      "grad_norm": 0.7372215085776161,
+      "learning_rate": 6.513019510781856e-07,
+      "loss": 1.6203,
+      "step": 2574
+    },
+    {
+      "epoch": 0.17942375361460475,
+      "grad_norm": 0.7242908564459493,
+      "learning_rate": 6.512651127758533e-07,
+      "loss": 1.564,
+      "step": 2575
+    },
+    {
+      "epoch": 0.17949343274222207,
+      "grad_norm": 0.7324473423283194,
+      "learning_rate": 6.512282617135501e-07,
+      "loss": 1.5448,
+      "step": 2576
+    },
+    {
+      "epoch": 0.1795631118698394,
+      "grad_norm": 0.6708021000359229,
+      "learning_rate": 6.511913978930421e-07,
+      "loss": 1.4548,
+      "step": 2577
+    },
+    {
+      "epoch": 0.17963279099745671,
+      "grad_norm": 0.7622906166024526,
+      "learning_rate": 6.511545213160959e-07,
+      "loss": 1.72,
+      "step": 2578
+    },
+    {
+      "epoch": 0.17970247012507404,
+      "grad_norm": 0.8097203764903231,
+      "learning_rate": 6.511176319844786e-07,
+      "loss": 1.5922,
+      "step": 2579
+    },
+    {
+      "epoch": 0.17977214925269136,
+      "grad_norm": 0.7167045481979863,
+      "learning_rate": 6.51080729899958e-07,
+      "loss": 1.5534,
+      "step": 2580
+    },
+    {
+      "epoch": 0.17984182838030868,
+      "grad_norm": 0.7130025941415435,
+      "learning_rate": 6.510438150643025e-07,
+      "loss": 1.4931,
+      "step": 2581
+    },
+    {
+      "epoch": 0.179911507507926,
+      "grad_norm": 0.7206872658927816,
+      "learning_rate": 6.510068874792813e-07,
+      "loss": 1.549,
+      "step": 2582
+    },
+    {
+      "epoch": 0.17998118663554333,
+      "grad_norm": 0.6855445976689892,
+      "learning_rate": 6.50969947146664e-07,
+      "loss": 1.5097,
+      "step": 2583
+    },
+    {
+      "epoch": 0.18005086576316065,
+      "grad_norm": 0.8358535909689165,
+      "learning_rate": 6.509329940682207e-07,
+      "loss": 1.6702,
+      "step": 2584
+    },
+    {
+      "epoch": 0.18012054489077797,
+      "grad_norm": 0.7251461113964899,
+      "learning_rate": 6.508960282457224e-07,
+      "loss": 1.5584,
+      "step": 2585
+    },
+    {
+      "epoch": 0.1801902240183953,
+      "grad_norm": 0.731726399487568,
+      "learning_rate": 6.508590496809407e-07,
+      "loss": 1.6408,
+      "step": 2586
+    },
+    {
+      "epoch": 0.18025990314601262,
+      "grad_norm": 0.771355055088224,
+      "learning_rate": 6.508220583756474e-07,
+      "loss": 1.6582,
+      "step": 2587
+    },
+    {
+      "epoch": 0.18032958227362994,
+      "grad_norm": 0.6991176945437412,
+      "learning_rate": 6.507850543316153e-07,
+      "loss": 1.5338,
+      "step": 2588
+    },
+    {
+      "epoch": 0.18039926140124726,
+      "grad_norm": 0.6942418119782533,
+      "learning_rate": 6.507480375506177e-07,
+      "loss": 1.6089,
+      "step": 2589
+    },
+    {
+      "epoch": 0.18046894052886459,
+      "grad_norm": 0.7744690223635904,
+      "learning_rate": 6.507110080344287e-07,
+      "loss": 1.6037,
+      "step": 2590
+    },
+    {
+      "epoch": 0.1805386196564819,
+      "grad_norm": 0.748844004383975,
+      "learning_rate": 6.506739657848227e-07,
+      "loss": 1.5866,
+      "step": 2591
+    },
+    {
+      "epoch": 0.18060829878409923,
+      "grad_norm": 0.7649898548673842,
+      "learning_rate": 6.506369108035749e-07,
+      "loss": 1.5733,
+      "step": 2592
+    },
+    {
+      "epoch": 0.18067797791171655,
+      "grad_norm": 0.7134369595094384,
+      "learning_rate": 6.505998430924608e-07,
+      "loss": 1.515,
+      "step": 2593
+    },
+    {
+      "epoch": 0.18074765703933388,
+      "grad_norm": 0.678241174097191,
+      "learning_rate": 6.505627626532572e-07,
+      "loss": 1.5061,
+      "step": 2594
+    },
+    {
+      "epoch": 0.1808173361669512,
+      "grad_norm": 0.7642942419987621,
+      "learning_rate": 6.505256694877406e-07,
+      "loss": 1.5778,
+      "step": 2595
+    },
+    {
+      "epoch": 0.18088701529456852,
+      "grad_norm": 0.7179287288029527,
+      "learning_rate": 6.50488563597689e-07,
+      "loss": 1.6228,
+      "step": 2596
+    },
+    {
+      "epoch": 0.18095669442218584,
+      "grad_norm": 0.7435897191353903,
+      "learning_rate": 6.504514449848803e-07,
+      "loss": 1.5228,
+      "step": 2597
+    },
+    {
+      "epoch": 0.18102637354980317,
+      "grad_norm": 0.7270413279809006,
+      "learning_rate": 6.504143136510932e-07,
+      "loss": 1.5702,
+      "step": 2598
+    },
+    {
+      "epoch": 0.1810960526774205,
+      "grad_norm": 0.6872507120511585,
+      "learning_rate": 6.503771695981076e-07,
+      "loss": 1.4811,
+      "step": 2599
+    },
+    {
+      "epoch": 0.1811657318050378,
+      "grad_norm": 0.7803447095575899,
+      "learning_rate": 6.503400128277032e-07,
+      "loss": 1.4736,
+      "step": 2600
+    },
+    {
+      "epoch": 0.18123541093265513,
+      "grad_norm": 0.7177509621225837,
+      "learning_rate": 6.503028433416605e-07,
+      "loss": 1.5244,
+      "step": 2601
+    },
+    {
+      "epoch": 0.18130509006027246,
+      "grad_norm": 0.7965152976759453,
+      "learning_rate": 6.502656611417609e-07,
+      "loss": 1.4269,
+      "step": 2602
+    },
+    {
+      "epoch": 0.18137476918788978,
+      "grad_norm": 0.7101628004895346,
+      "learning_rate": 6.502284662297863e-07,
+      "loss": 1.5922,
+      "step": 2603
+    },
+    {
+      "epoch": 0.1814444483155071,
+      "grad_norm": 0.6982929845687866,
+      "learning_rate": 6.50191258607519e-07,
+      "loss": 1.5254,
+      "step": 2604
+    },
+    {
+      "epoch": 0.18151412744312442,
+      "grad_norm": 0.6950322548952821,
+      "learning_rate": 6.501540382767424e-07,
+      "loss": 1.4629,
+      "step": 2605
+    },
+    {
+      "epoch": 0.18158380657074175,
+      "grad_norm": 0.7202886551001328,
+      "learning_rate": 6.501168052392397e-07,
+      "loss": 1.552,
+      "step": 2606
+    },
+    {
+      "epoch": 0.18165348569835907,
+      "grad_norm": 0.7161505166062673,
+      "learning_rate": 6.500795594967954e-07,
+      "loss": 1.6631,
+      "step": 2607
+    },
+    {
+      "epoch": 0.1817231648259764,
+      "grad_norm": 0.6878116644197528,
+      "learning_rate": 6.500423010511946e-07,
+      "loss": 1.4918,
+      "step": 2608
+    },
+    {
+      "epoch": 0.1817928439535937,
+      "grad_norm": 0.7353866773063502,
+      "learning_rate": 6.500050299042224e-07,
+      "loss": 1.554,
+      "step": 2609
+    },
+    {
+      "epoch": 0.18186252308121104,
+      "grad_norm": 0.7164396573550172,
+      "learning_rate": 6.499677460576653e-07,
+      "loss": 1.4531,
+      "step": 2610
+    },
+    {
+      "epoch": 0.18193220220882836,
+      "grad_norm": 0.7912418932841199,
+      "learning_rate": 6.499304495133098e-07,
+      "loss": 1.5789,
+      "step": 2611
+    },
+    {
+      "epoch": 0.18200188133644568,
+      "grad_norm": 0.763240467980432,
+      "learning_rate": 6.498931402729434e-07,
+      "loss": 1.5829,
+      "step": 2612
+    },
+    {
+      "epoch": 0.182071560464063,
+      "grad_norm": 0.7439100105793851,
+      "learning_rate": 6.498558183383537e-07,
+      "loss": 1.5418,
+      "step": 2613
+    },
+    {
+      "epoch": 0.1821412395916803,
+      "grad_norm": 0.6859069824967786,
+      "learning_rate": 6.498184837113295e-07,
+      "loss": 1.6184,
+      "step": 2614
+    },
+    {
+      "epoch": 0.18221091871929762,
+      "grad_norm": 0.7132179337585111,
+      "learning_rate": 6.497811363936598e-07,
+      "loss": 1.4808,
+      "step": 2615
+    },
+    {
+      "epoch": 0.18228059784691494,
+      "grad_norm": 0.7409771501660511,
+      "learning_rate": 6.497437763871346e-07,
+      "loss": 1.4906,
+      "step": 2616
+    },
+    {
+      "epoch": 0.18235027697453227,
+      "grad_norm": 0.6944759248214315,
+      "learning_rate": 6.497064036935441e-07,
+      "loss": 1.5897,
+      "step": 2617
+    },
+    {
+      "epoch": 0.1824199561021496,
+      "grad_norm": 0.7015003873915269,
+      "learning_rate": 6.496690183146794e-07,
+      "loss": 1.509,
+      "step": 2618
+    },
+    {
+      "epoch": 0.1824896352297669,
+      "grad_norm": 0.6690613285222292,
+      "learning_rate": 6.496316202523318e-07,
+      "loss": 1.4293,
+      "step": 2619
+    },
+    {
+      "epoch": 0.18255931435738423,
+      "grad_norm": 0.749788872274255,
+      "learning_rate": 6.495942095082938e-07,
+      "loss": 1.5291,
+      "step": 2620
+    },
+    {
+      "epoch": 0.18262899348500156,
+      "grad_norm": 0.7473142725831187,
+      "learning_rate": 6.49556786084358e-07,
+      "loss": 1.6343,
+      "step": 2621
+    },
+    {
+      "epoch": 0.18269867261261888,
+      "grad_norm": 0.7249410808106185,
+      "learning_rate": 6.49519349982318e-07,
+      "loss": 1.4134,
+      "step": 2622
+    },
+    {
+      "epoch": 0.1827683517402362,
+      "grad_norm": 0.6833719220191706,
+      "learning_rate": 6.494819012039675e-07,
+      "loss": 1.5008,
+      "step": 2623
+    },
+    {
+      "epoch": 0.18283803086785352,
+      "grad_norm": 0.7862174459510464,
+      "learning_rate": 6.494444397511014e-07,
+      "loss": 1.6386,
+      "step": 2624
+    },
+    {
+      "epoch": 0.18290770999547085,
+      "grad_norm": 0.730100169454004,
+      "learning_rate": 6.49406965625515e-07,
+      "loss": 1.6254,
+      "step": 2625
+    },
+    {
+      "epoch": 0.18297738912308817,
+      "grad_norm": 0.7313275186062451,
+      "learning_rate": 6.493694788290038e-07,
+      "loss": 1.5602,
+      "step": 2626
+    },
+    {
+      "epoch": 0.1830470682507055,
+      "grad_norm": 0.8129828741881585,
+      "learning_rate": 6.493319793633646e-07,
+      "loss": 1.594,
+      "step": 2627
+    },
+    {
+      "epoch": 0.18311674737832281,
+      "grad_norm": 0.6898176815693812,
+      "learning_rate": 6.492944672303941e-07,
+      "loss": 1.634,
+      "step": 2628
+    },
+    {
+      "epoch": 0.18318642650594014,
+      "grad_norm": 0.708461442459301,
+      "learning_rate": 6.492569424318904e-07,
+      "loss": 1.5954,
+      "step": 2629
+    },
+    {
+      "epoch": 0.18325610563355746,
+      "grad_norm": 0.7568109034106955,
+      "learning_rate": 6.492194049696513e-07,
+      "loss": 1.5439,
+      "step": 2630
+    },
+    {
+      "epoch": 0.18332578476117478,
+      "grad_norm": 0.6771063848093974,
+      "learning_rate": 6.49181854845476e-07,
+      "loss": 1.5704,
+      "step": 2631
+    },
+    {
+      "epoch": 0.1833954638887921,
+      "grad_norm": 0.676580159279378,
+      "learning_rate": 6.491442920611637e-07,
+      "loss": 1.5426,
+      "step": 2632
+    },
+    {
+      "epoch": 0.18346514301640943,
+      "grad_norm": 0.6600845607953997,
+      "learning_rate": 6.491067166185148e-07,
+      "loss": 1.4584,
+      "step": 2633
+    },
+    {
+      "epoch": 0.18353482214402675,
+      "grad_norm": 0.724526501155734,
+      "learning_rate": 6.490691285193298e-07,
+      "loss": 1.4994,
+      "step": 2634
+    },
+    {
+      "epoch": 0.18360450127164407,
+      "grad_norm": 0.7652557824216829,
+      "learning_rate": 6.490315277654101e-07,
+      "loss": 1.5651,
+      "step": 2635
+    },
+    {
+      "epoch": 0.1836741803992614,
+      "grad_norm": 0.7187368483648745,
+      "learning_rate": 6.489939143585574e-07,
+      "loss": 1.5492,
+      "step": 2636
+    },
+    {
+      "epoch": 0.18374385952687872,
+      "grad_norm": 0.7394868496695424,
+      "learning_rate": 6.489562883005744e-07,
+      "loss": 1.7474,
+      "step": 2637
+    },
+    {
+      "epoch": 0.18381353865449604,
+      "grad_norm": 0.7707547565878867,
+      "learning_rate": 6.489186495932642e-07,
+      "loss": 1.6094,
+      "step": 2638
+    },
+    {
+      "epoch": 0.18388321778211336,
+      "grad_norm": 0.6817021323889921,
+      "learning_rate": 6.488809982384305e-07,
+      "loss": 1.5874,
+      "step": 2639
+    },
+    {
+      "epoch": 0.18395289690973068,
+      "grad_norm": 0.6530574602289403,
+      "learning_rate": 6.488433342378776e-07,
+      "loss": 1.3532,
+      "step": 2640
+    },
+    {
+      "epoch": 0.184022576037348,
+      "grad_norm": 0.6891115609332592,
+      "learning_rate": 6.488056575934104e-07,
+      "loss": 1.5033,
+      "step": 2641
+    },
+    {
+      "epoch": 0.18409225516496533,
+      "grad_norm": 0.7075654971214155,
+      "learning_rate": 6.487679683068348e-07,
+      "loss": 1.5582,
+      "step": 2642
+    },
+    {
+      "epoch": 0.18416193429258265,
+      "grad_norm": 0.6979922146733539,
+      "learning_rate": 6.487302663799565e-07,
+      "loss": 1.5725,
+      "step": 2643
+    },
+    {
+      "epoch": 0.18423161342019997,
+      "grad_norm": 0.7038692584176877,
+      "learning_rate": 6.486925518145822e-07,
+      "loss": 1.5976,
+      "step": 2644
+    },
+    {
+      "epoch": 0.1843012925478173,
+      "grad_norm": 0.7598408079706039,
+      "learning_rate": 6.486548246125197e-07,
+      "loss": 1.5574,
+      "step": 2645
+    },
+    {
+      "epoch": 0.18437097167543462,
+      "grad_norm": 0.6855757317942535,
+      "learning_rate": 6.486170847755767e-07,
+      "loss": 1.4375,
+      "step": 2646
+    },
+    {
+      "epoch": 0.18444065080305194,
+      "grad_norm": 0.7631668431699284,
+      "learning_rate": 6.485793323055616e-07,
+      "loss": 1.616,
+      "step": 2647
+    },
+    {
+      "epoch": 0.18451032993066926,
+      "grad_norm": 0.7368144191558978,
+      "learning_rate": 6.485415672042839e-07,
+      "loss": 1.5148,
+      "step": 2648
+    },
+    {
+      "epoch": 0.1845800090582866,
+      "grad_norm": 0.7685902842120601,
+      "learning_rate": 6.485037894735534e-07,
+      "loss": 1.7105,
+      "step": 2649
+    },
+    {
+      "epoch": 0.1846496881859039,
+      "grad_norm": 0.6976902801252264,
+      "learning_rate": 6.484659991151802e-07,
+      "loss": 1.4219,
+      "step": 2650
+    },
+    {
+      "epoch": 0.18471936731352123,
+      "grad_norm": 0.7580061537305252,
+      "learning_rate": 6.484281961309754e-07,
+      "loss": 1.5707,
+      "step": 2651
+    },
+    {
+      "epoch": 0.18478904644113855,
+      "grad_norm": 0.7500641059710291,
+      "learning_rate": 6.483903805227508e-07,
+      "loss": 1.5827,
+      "step": 2652
+    },
+    {
+      "epoch": 0.18485872556875588,
+      "grad_norm": 0.6649438106538588,
+      "learning_rate": 6.483525522923183e-07,
+      "loss": 1.4317,
+      "step": 2653
+    },
+    {
+      "epoch": 0.1849284046963732,
+      "grad_norm": 0.7351706654236744,
+      "learning_rate": 6.483147114414908e-07,
+      "loss": 1.5674,
+      "step": 2654
+    },
+    {
+      "epoch": 0.18499808382399052,
+      "grad_norm": 0.6680546391703736,
+      "learning_rate": 6.482768579720819e-07,
+      "loss": 1.5953,
+      "step": 2655
+    },
+    {
+      "epoch": 0.18506776295160784,
+      "grad_norm": 0.7247126042568591,
+      "learning_rate": 6.482389918859054e-07,
+      "loss": 1.4191,
+      "step": 2656
+    },
+    {
+      "epoch": 0.18513744207922517,
+      "grad_norm": 0.6911414934075633,
+      "learning_rate": 6.48201113184776e-07,
+      "loss": 1.438,
+      "step": 2657
+    },
+    {
+      "epoch": 0.1852071212068425,
+      "grad_norm": 0.734603021831186,
+      "learning_rate": 6.481632218705089e-07,
+      "loss": 1.5527,
+      "step": 2658
+    },
+    {
+      "epoch": 0.1852768003344598,
+      "grad_norm": 0.7504027564292364,
+      "learning_rate": 6.4812531794492e-07,
+      "loss": 1.6085,
+      "step": 2659
+    },
+    {
+      "epoch": 0.18534647946207713,
+      "grad_norm": 0.6875098502184053,
+      "learning_rate": 6.480874014098256e-07,
+      "loss": 1.4462,
+      "step": 2660
+    },
+    {
+      "epoch": 0.18541615858969446,
+      "grad_norm": 0.7557531183356848,
+      "learning_rate": 6.480494722670429e-07,
+      "loss": 1.5846,
+      "step": 2661
+    },
+    {
+      "epoch": 0.18548583771731178,
+      "grad_norm": 0.7337273902395194,
+      "learning_rate": 6.480115305183894e-07,
+      "loss": 1.4645,
+      "step": 2662
+    },
+    {
+      "epoch": 0.1855555168449291,
+      "grad_norm": 0.7159247755309575,
+      "learning_rate": 6.479735761656834e-07,
+      "loss": 1.5828,
+      "step": 2663
+    },
+    {
+      "epoch": 0.18562519597254643,
+      "grad_norm": 0.7390118659196949,
+      "learning_rate": 6.479356092107438e-07,
+      "loss": 1.62,
+      "step": 2664
+    },
+    {
+      "epoch": 0.18569487510016375,
+      "grad_norm": 0.758500507595704,
+      "learning_rate": 6.478976296553899e-07,
+      "loss": 1.5198,
+      "step": 2665
+    },
+    {
+      "epoch": 0.18576455422778107,
+      "grad_norm": 0.7302673339657927,
+      "learning_rate": 6.478596375014419e-07,
+      "loss": 1.6104,
+      "step": 2666
+    },
+    {
+      "epoch": 0.1858342333553984,
+      "grad_norm": 0.7771031453203399,
+      "learning_rate": 6.478216327507206e-07,
+      "loss": 1.6206,
+      "step": 2667
+    },
+    {
+      "epoch": 0.18590391248301572,
+      "grad_norm": 0.7260781373547577,
+      "learning_rate": 6.477836154050468e-07,
+      "loss": 1.5566,
+      "step": 2668
+    },
+    {
+      "epoch": 0.18597359161063304,
+      "grad_norm": 0.7006581746625344,
+      "learning_rate": 6.47745585466243e-07,
+      "loss": 1.3939,
+      "step": 2669
+    },
+    {
+      "epoch": 0.18604327073825036,
+      "grad_norm": 0.7801900637952093,
+      "learning_rate": 6.477075429361312e-07,
+      "loss": 1.5328,
+      "step": 2670
+    },
+    {
+      "epoch": 0.18611294986586768,
+      "grad_norm": 0.7423656847875956,
+      "learning_rate": 6.476694878165345e-07,
+      "loss": 1.6035,
+      "step": 2671
+    },
+    {
+      "epoch": 0.186182628993485,
+      "grad_norm": 0.7195463532032887,
+      "learning_rate": 6.476314201092768e-07,
+      "loss": 1.5965,
+      "step": 2672
+    },
+    {
+      "epoch": 0.18625230812110233,
+      "grad_norm": 0.7532581893821502,
+      "learning_rate": 6.475933398161822e-07,
+      "loss": 1.6735,
+      "step": 2673
+    },
+    {
+      "epoch": 0.18632198724871965,
+      "grad_norm": 0.7769211162227893,
+      "learning_rate": 6.475552469390758e-07,
+      "loss": 1.4756,
+      "step": 2674
+    },
+    {
+      "epoch": 0.18639166637633697,
+      "grad_norm": 0.7550947624847407,
+      "learning_rate": 6.475171414797828e-07,
+      "loss": 1.6108,
+      "step": 2675
+    },
+    {
+      "epoch": 0.1864613455039543,
+      "grad_norm": 0.6690296420086265,
+      "learning_rate": 6.474790234401296e-07,
+      "loss": 1.4959,
+      "step": 2676
+    },
+    {
+      "epoch": 0.18653102463157162,
+      "grad_norm": 0.6832451741796913,
+      "learning_rate": 6.474408928219426e-07,
+      "loss": 1.498,
+      "step": 2677
+    },
+    {
+      "epoch": 0.18660070375918894,
+      "grad_norm": 0.747393796002889,
+      "learning_rate": 6.474027496270493e-07,
+      "loss": 1.562,
+      "step": 2678
+    },
+    {
+      "epoch": 0.18667038288680626,
+      "grad_norm": 0.7611818415803452,
+      "learning_rate": 6.473645938572774e-07,
+      "loss": 1.6005,
+      "step": 2679
+    },
+    {
+      "epoch": 0.18674006201442359,
+      "grad_norm": 0.732431807713295,
+      "learning_rate": 6.473264255144558e-07,
+      "loss": 1.6562,
+      "step": 2680
+    },
+    {
+      "epoch": 0.1868097411420409,
+      "grad_norm": 0.8045845891090851,
+      "learning_rate": 6.472882446004132e-07,
+      "loss": 1.5858,
+      "step": 2681
+    },
+    {
+      "epoch": 0.18687942026965823,
+      "grad_norm": 0.6645879405992747,
+      "learning_rate": 6.472500511169794e-07,
+      "loss": 1.4494,
+      "step": 2682
+    },
+    {
+      "epoch": 0.18694909939727555,
+      "grad_norm": 0.7534293667028573,
+      "learning_rate": 6.472118450659848e-07,
+      "loss": 1.4723,
+      "step": 2683
+    },
+    {
+      "epoch": 0.18701877852489288,
+      "grad_norm": 0.7352604989735883,
+      "learning_rate": 6.471736264492604e-07,
+      "loss": 1.431,
+      "step": 2684
+    },
+    {
+      "epoch": 0.1870884576525102,
+      "grad_norm": 0.7933364996076561,
+      "learning_rate": 6.471353952686373e-07,
+      "loss": 1.524,
+      "step": 2685
+    },
+    {
+      "epoch": 0.18715813678012752,
+      "grad_norm": 0.7600850522766464,
+      "learning_rate": 6.470971515259481e-07,
+      "loss": 1.46,
+      "step": 2686
+    },
+    {
+      "epoch": 0.18722781590774484,
+      "grad_norm": 0.7938840091949416,
+      "learning_rate": 6.470588952230254e-07,
+      "loss": 1.5313,
+      "step": 2687
+    },
+    {
+      "epoch": 0.18729749503536217,
+      "grad_norm": 0.7195140084978836,
+      "learning_rate": 6.470206263617024e-07,
+      "loss": 1.4697,
+      "step": 2688
+    },
+    {
+      "epoch": 0.1873671741629795,
+      "grad_norm": 0.7005601442693249,
+      "learning_rate": 6.469823449438129e-07,
+      "loss": 1.5984,
+      "step": 2689
+    },
+    {
+      "epoch": 0.1874368532905968,
+      "grad_norm": 0.7700719550441382,
+      "learning_rate": 6.469440509711918e-07,
+      "loss": 1.5505,
+      "step": 2690
+    },
+    {
+      "epoch": 0.18750653241821413,
+      "grad_norm": 0.6955391533447876,
+      "learning_rate": 6.469057444456739e-07,
+      "loss": 1.5746,
+      "step": 2691
+    },
+    {
+      "epoch": 0.18757621154583146,
+      "grad_norm": 0.7533355959862432,
+      "learning_rate": 6.46867425369095e-07,
+      "loss": 1.4649,
+      "step": 2692
+    },
+    {
+      "epoch": 0.18764589067344878,
+      "grad_norm": 0.6861651730926349,
+      "learning_rate": 6.468290937432916e-07,
+      "loss": 1.556,
+      "step": 2693
+    },
+    {
+      "epoch": 0.1877155698010661,
+      "grad_norm": 0.6969686824642832,
+      "learning_rate": 6.467907495701004e-07,
+      "loss": 1.4597,
+      "step": 2694
+    },
+    {
+      "epoch": 0.18778524892868342,
+      "grad_norm": 0.7780056460087894,
+      "learning_rate": 6.467523928513592e-07,
+      "loss": 1.584,
+      "step": 2695
+    },
+    {
+      "epoch": 0.18785492805630075,
+      "grad_norm": 0.6971351165916029,
+      "learning_rate": 6.467140235889058e-07,
+      "loss": 1.3735,
+      "step": 2696
+    },
+    {
+      "epoch": 0.18792460718391807,
+      "grad_norm": 0.7640130727336821,
+      "learning_rate": 6.466756417845792e-07,
+      "loss": 1.6152,
+      "step": 2697
+    },
+    {
+      "epoch": 0.1879942863115354,
+      "grad_norm": 0.7174718505159229,
+      "learning_rate": 6.466372474402185e-07,
+      "loss": 1.5299,
+      "step": 2698
+    },
+    {
+      "epoch": 0.1880639654391527,
+      "grad_norm": 0.705437048563619,
+      "learning_rate": 6.465988405576638e-07,
+      "loss": 1.5884,
+      "step": 2699
+    },
+    {
+      "epoch": 0.18813364456677004,
+      "grad_norm": 0.8496788834128441,
+      "learning_rate": 6.465604211387557e-07,
+      "loss": 1.6326,
+      "step": 2700
+    },
+    {
+      "epoch": 0.18820332369438736,
+      "grad_norm": 0.7183308884189702,
+      "learning_rate": 6.465219891853353e-07,
+      "loss": 1.4403,
+      "step": 2701
+    },
+    {
+      "epoch": 0.18827300282200468,
+      "grad_norm": 0.7363655214283799,
+      "learning_rate": 6.464835446992441e-07,
+      "loss": 1.502,
+      "step": 2702
+    },
+    {
+      "epoch": 0.188342681949622,
+      "grad_norm": 0.7137203184429629,
+      "learning_rate": 6.464450876823248e-07,
+      "loss": 1.5203,
+      "step": 2703
+    },
+    {
+      "epoch": 0.18841236107723933,
+      "grad_norm": 0.7252300369350988,
+      "learning_rate": 6.464066181364201e-07,
+      "loss": 1.5046,
+      "step": 2704
+    },
+    {
+      "epoch": 0.18848204020485665,
+      "grad_norm": 0.7117919170248129,
+      "learning_rate": 6.463681360633735e-07,
+      "loss": 1.6556,
+      "step": 2705
+    },
+    {
+      "epoch": 0.18855171933247394,
+      "grad_norm": 0.720869879698167,
+      "learning_rate": 6.463296414650294e-07,
+      "loss": 1.5998,
+      "step": 2706
+    },
+    {
+      "epoch": 0.18862139846009127,
+      "grad_norm": 0.7449188955522009,
+      "learning_rate": 6.462911343432322e-07,
+      "loss": 1.4649,
+      "step": 2707
+    },
+    {
+      "epoch": 0.1886910775877086,
+      "grad_norm": 0.7439890979647477,
+      "learning_rate": 6.462526146998275e-07,
+      "loss": 1.6382,
+      "step": 2708
+    },
+    {
+      "epoch": 0.1887607567153259,
+      "grad_norm": 0.7126331244413916,
+      "learning_rate": 6.462140825366612e-07,
+      "loss": 1.5312,
+      "step": 2709
+    },
+    {
+      "epoch": 0.18883043584294323,
+      "grad_norm": 0.7735940649273119,
+      "learning_rate": 6.461755378555798e-07,
+      "loss": 1.5014,
+      "step": 2710
+    },
+    {
+      "epoch": 0.18890011497056056,
+      "grad_norm": 0.7440600101186636,
+      "learning_rate": 6.461369806584305e-07,
+      "loss": 1.6966,
+      "step": 2711
+    },
+    {
+      "epoch": 0.18896979409817788,
+      "grad_norm": 0.6759130137405256,
+      "learning_rate": 6.460984109470609e-07,
+      "loss": 1.5484,
+      "step": 2712
+    },
+    {
+      "epoch": 0.1890394732257952,
+      "grad_norm": 0.8264161895423117,
+      "learning_rate": 6.460598287233195e-07,
+      "loss": 1.5834,
+      "step": 2713
+    },
+    {
+      "epoch": 0.18910915235341252,
+      "grad_norm": 0.7491457334304705,
+      "learning_rate": 6.46021233989055e-07,
+      "loss": 1.5877,
+      "step": 2714
+    },
+    {
+      "epoch": 0.18917883148102985,
+      "grad_norm": 0.6864192440850518,
+      "learning_rate": 6.459826267461172e-07,
+      "loss": 1.5166,
+      "step": 2715
+    },
+    {
+      "epoch": 0.18924851060864717,
+      "grad_norm": 0.7259227539963868,
+      "learning_rate": 6.459440069963562e-07,
+      "loss": 1.5007,
+      "step": 2716
+    },
+    {
+      "epoch": 0.1893181897362645,
+      "grad_norm": 0.7092268754966882,
+      "learning_rate": 6.459053747416225e-07,
+      "loss": 1.5473,
+      "step": 2717
+    },
+    {
+      "epoch": 0.18938786886388181,
+      "grad_norm": 0.6882956952095184,
+      "learning_rate": 6.458667299837677e-07,
+      "loss": 1.5654,
+      "step": 2718
+    },
+    {
+      "epoch": 0.18945754799149914,
+      "grad_norm": 0.8355331625656581,
+      "learning_rate": 6.458280727246438e-07,
+      "loss": 1.5154,
+      "step": 2719
+    },
+    {
+      "epoch": 0.18952722711911646,
+      "grad_norm": 0.7352037316311917,
+      "learning_rate": 6.45789402966103e-07,
+      "loss": 1.6327,
+      "step": 2720
+    },
+    {
+      "epoch": 0.18959690624673378,
+      "grad_norm": 0.6941993455689607,
+      "learning_rate": 6.457507207099988e-07,
+      "loss": 1.4324,
+      "step": 2721
+    },
+    {
+      "epoch": 0.1896665853743511,
+      "grad_norm": 0.7649191465845473,
+      "learning_rate": 6.457120259581845e-07,
+      "loss": 1.6697,
+      "step": 2722
+    },
+    {
+      "epoch": 0.18973626450196843,
+      "grad_norm": 0.7466998683828353,
+      "learning_rate": 6.456733187125149e-07,
+      "loss": 1.659,
+      "step": 2723
+    },
+    {
+      "epoch": 0.18980594362958575,
+      "grad_norm": 0.7588182856390634,
+      "learning_rate": 6.456345989748445e-07,
+      "loss": 1.5306,
+      "step": 2724
+    },
+    {
+      "epoch": 0.18987562275720307,
+      "grad_norm": 0.6838571523104907,
+      "learning_rate": 6.455958667470293e-07,
+      "loss": 1.6308,
+      "step": 2725
+    },
+    {
+      "epoch": 0.1899453018848204,
+      "grad_norm": 0.7125291705245504,
+      "learning_rate": 6.455571220309251e-07,
+      "loss": 1.5757,
+      "step": 2726
+    },
+    {
+      "epoch": 0.19001498101243772,
+      "grad_norm": 0.7950404350336909,
+      "learning_rate": 6.455183648283886e-07,
+      "loss": 1.5175,
+      "step": 2727
+    },
+    {
+      "epoch": 0.19008466014005504,
+      "grad_norm": 0.6832770487165002,
+      "learning_rate": 6.454795951412773e-07,
+      "loss": 1.5996,
+      "step": 2728
+    },
+    {
+      "epoch": 0.19015433926767236,
+      "grad_norm": 0.723433406934566,
+      "learning_rate": 6.45440812971449e-07,
+      "loss": 1.5254,
+      "step": 2729
+    },
+    {
+      "epoch": 0.19022401839528968,
+      "grad_norm": 0.7054317859702796,
+      "learning_rate": 6.454020183207623e-07,
+      "loss": 1.5587,
+      "step": 2730
+    },
+    {
+      "epoch": 0.190293697522907,
+      "grad_norm": 0.7909451789471312,
+      "learning_rate": 6.453632111910763e-07,
+      "loss": 1.6194,
+      "step": 2731
+    },
+    {
+      "epoch": 0.19036337665052433,
+      "grad_norm": 0.6953562567380657,
+      "learning_rate": 6.453243915842507e-07,
+      "loss": 1.503,
+      "step": 2732
+    },
+    {
+      "epoch": 0.19043305577814165,
+      "grad_norm": 0.7637019576203848,
+      "learning_rate": 6.452855595021457e-07,
+      "loss": 1.6011,
+      "step": 2733
+    },
+    {
+      "epoch": 0.19050273490575897,
+      "grad_norm": 0.6822582597425052,
+      "learning_rate": 6.452467149466225e-07,
+      "loss": 1.3143,
+      "step": 2734
+    },
+    {
+      "epoch": 0.1905724140333763,
+      "grad_norm": 0.7178449079461839,
+      "learning_rate": 6.452078579195424e-07,
+      "loss": 1.5419,
+      "step": 2735
+    },
+    {
+      "epoch": 0.19064209316099362,
+      "grad_norm": 0.7138948499344089,
+      "learning_rate": 6.451689884227674e-07,
+      "loss": 1.4824,
+      "step": 2736
+    },
+    {
+      "epoch": 0.19071177228861094,
+      "grad_norm": 0.7322828001630005,
+      "learning_rate": 6.451301064581605e-07,
+      "loss": 1.506,
+      "step": 2737
+    },
+    {
+      "epoch": 0.19078145141622826,
+      "grad_norm": 0.811654935475173,
+      "learning_rate": 6.45091212027585e-07,
+      "loss": 1.5939,
+      "step": 2738
+    },
+    {
+      "epoch": 0.1908511305438456,
+      "grad_norm": 0.6802507343095897,
+      "learning_rate": 6.450523051329044e-07,
+      "loss": 1.5206,
+      "step": 2739
+    },
+    {
+      "epoch": 0.1909208096714629,
+      "grad_norm": 0.7181069784358606,
+      "learning_rate": 6.450133857759836e-07,
+      "loss": 1.566,
+      "step": 2740
+    },
+    {
+      "epoch": 0.19099048879908023,
+      "grad_norm": 0.7119050567279736,
+      "learning_rate": 6.449744539586875e-07,
+      "loss": 1.4723,
+      "step": 2741
+    },
+    {
+      "epoch": 0.19106016792669755,
+      "grad_norm": 0.7359612711039888,
+      "learning_rate": 6.449355096828818e-07,
+      "loss": 1.4862,
+      "step": 2742
+    },
+    {
+      "epoch": 0.19112984705431488,
+      "grad_norm": 0.7097328511153883,
+      "learning_rate": 6.44896552950433e-07,
+      "loss": 1.5173,
+      "step": 2743
+    },
+    {
+      "epoch": 0.1911995261819322,
+      "grad_norm": 0.6929533047668857,
+      "learning_rate": 6.448575837632079e-07,
+      "loss": 1.5577,
+      "step": 2744
+    },
+    {
+      "epoch": 0.19126920530954952,
+      "grad_norm": 0.7041347408340864,
+      "learning_rate": 6.448186021230737e-07,
+      "loss": 1.452,
+      "step": 2745
+    },
+    {
+      "epoch": 0.19133888443716685,
+      "grad_norm": 0.72942054200097,
+      "learning_rate": 6.447796080318986e-07,
+      "loss": 1.5466,
+      "step": 2746
+    },
+    {
+      "epoch": 0.19140856356478417,
+      "grad_norm": 0.7194334581281807,
+      "learning_rate": 6.447406014915515e-07,
+      "loss": 1.5971,
+      "step": 2747
+    },
+    {
+      "epoch": 0.1914782426924015,
+      "grad_norm": 0.746922395656889,
+      "learning_rate": 6.447015825039016e-07,
+      "loss": 1.6671,
+      "step": 2748
+    },
+    {
+      "epoch": 0.1915479218200188,
+      "grad_norm": 0.6766416878379267,
+      "learning_rate": 6.446625510708187e-07,
+      "loss": 1.5329,
+      "step": 2749
+    },
+    {
+      "epoch": 0.19161760094763614,
+      "grad_norm": 0.7661740967695121,
+      "learning_rate": 6.446235071941732e-07,
+      "loss": 1.6282,
+      "step": 2750
+    },
+    {
+      "epoch": 0.19168728007525346,
+      "grad_norm": 0.7270021181227755,
+      "learning_rate": 6.445844508758363e-07,
+      "loss": 1.4649,
+      "step": 2751
+    },
+    {
+      "epoch": 0.19175695920287078,
+      "grad_norm": 0.6904466633241558,
+      "learning_rate": 6.445453821176795e-07,
+      "loss": 1.4696,
+      "step": 2752
+    },
+    {
+      "epoch": 0.1918266383304881,
+      "grad_norm": 0.7594226179740771,
+      "learning_rate": 6.445063009215751e-07,
+      "loss": 1.5229,
+      "step": 2753
+    },
+    {
+      "epoch": 0.19189631745810543,
+      "grad_norm": 0.7340481003476917,
+      "learning_rate": 6.444672072893962e-07,
+      "loss": 1.5785,
+      "step": 2754
+    },
+    {
+      "epoch": 0.19196599658572275,
+      "grad_norm": 0.7553905406477419,
+      "learning_rate": 6.444281012230159e-07,
+      "loss": 1.5664,
+      "step": 2755
+    },
+    {
+      "epoch": 0.19203567571334007,
+      "grad_norm": 0.7126604908145795,
+      "learning_rate": 6.443889827243085e-07,
+      "loss": 1.5051,
+      "step": 2756
+    },
+    {
+      "epoch": 0.1921053548409574,
+      "grad_norm": 0.7123761884060794,
+      "learning_rate": 6.443498517951485e-07,
+      "loss": 1.4743,
+      "step": 2757
+    },
+    {
+      "epoch": 0.19217503396857472,
+      "grad_norm": 0.7877452122261493,
+      "learning_rate": 6.443107084374112e-07,
+      "loss": 1.5403,
+      "step": 2758
+    },
+    {
+      "epoch": 0.19224471309619204,
+      "grad_norm": 0.6537712332432835,
+      "learning_rate": 6.442715526529724e-07,
+      "loss": 1.3929,
+      "step": 2759
+    },
+    {
+      "epoch": 0.19231439222380936,
+      "grad_norm": 0.6982635432356034,
+      "learning_rate": 6.442323844437085e-07,
+      "loss": 1.5458,
+      "step": 2760
+    },
+    {
+      "epoch": 0.19238407135142668,
+      "grad_norm": 0.7007628799980677,
+      "learning_rate": 6.441932038114964e-07,
+      "loss": 1.4499,
+      "step": 2761
+    },
+    {
+      "epoch": 0.192453750479044,
+      "grad_norm": 0.6998289345209359,
+      "learning_rate": 6.441540107582141e-07,
+      "loss": 1.527,
+      "step": 2762
+    },
+    {
+      "epoch": 0.19252342960666133,
+      "grad_norm": 0.7224707964363176,
+      "learning_rate": 6.441148052857394e-07,
+      "loss": 1.5791,
+      "step": 2763
+    },
+    {
+      "epoch": 0.19259310873427865,
+      "grad_norm": 0.7549588551015194,
+      "learning_rate": 6.440755873959513e-07,
+      "loss": 1.5293,
+      "step": 2764
+    },
+    {
+      "epoch": 0.19266278786189597,
+      "grad_norm": 0.8271235962565269,
+      "learning_rate": 6.440363570907294e-07,
+      "loss": 1.6131,
+      "step": 2765
+    },
+    {
+      "epoch": 0.1927324669895133,
+      "grad_norm": 0.7332975981376336,
+      "learning_rate": 6.439971143719531e-07,
+      "loss": 1.5622,
+      "step": 2766
+    },
+    {
+      "epoch": 0.19280214611713062,
+      "grad_norm": 0.7431066568250002,
+      "learning_rate": 6.439578592415036e-07,
+      "loss": 1.4984,
+      "step": 2767
+    },
+    {
+      "epoch": 0.19287182524474794,
+      "grad_norm": 0.7062334001251804,
+      "learning_rate": 6.43918591701262e-07,
+      "loss": 1.5469,
+      "step": 2768
+    },
+    {
+      "epoch": 0.19294150437236526,
+      "grad_norm": 0.8259805932604753,
+      "learning_rate": 6.438793117531097e-07,
+      "loss": 1.813,
+      "step": 2769
+    },
+    {
+      "epoch": 0.19301118349998259,
+      "grad_norm": 0.7265022203713082,
+      "learning_rate": 6.438400193989292e-07,
+      "loss": 1.5044,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1930808626275999,
+      "grad_norm": 0.7800710904673711,
+      "learning_rate": 6.438007146406037e-07,
+      "loss": 1.5255,
+      "step": 2771
+    },
+    {
+      "epoch": 0.19315054175521723,
+      "grad_norm": 0.6915373563261974,
+      "learning_rate": 6.437613974800168e-07,
+      "loss": 1.4791,
+      "step": 2772
+    },
+    {
+      "epoch": 0.19322022088283455,
+      "grad_norm": 0.7455382855726259,
+      "learning_rate": 6.437220679190524e-07,
+      "loss": 1.6197,
+      "step": 2773
+    },
+    {
+      "epoch": 0.19328990001045188,
+      "grad_norm": 0.741388638904269,
+      "learning_rate": 6.436827259595954e-07,
+      "loss": 1.5326,
+      "step": 2774
+    },
+    {
+      "epoch": 0.1933595791380692,
+      "grad_norm": 0.7274532124482767,
+      "learning_rate": 6.436433716035309e-07,
+      "loss": 1.6042,
+      "step": 2775
+    },
+    {
+      "epoch": 0.19342925826568652,
+      "grad_norm": 0.6613153079163728,
+      "learning_rate": 6.436040048527453e-07,
+      "loss": 1.423,
+      "step": 2776
+    },
+    {
+      "epoch": 0.19349893739330384,
+      "grad_norm": 0.7526717406315007,
+      "learning_rate": 6.435646257091248e-07,
+      "loss": 1.6477,
+      "step": 2777
+    },
+    {
+      "epoch": 0.19356861652092117,
+      "grad_norm": 0.7457469971419062,
+      "learning_rate": 6.435252341745566e-07,
+      "loss": 1.7472,
+      "step": 2778
+    },
+    {
+      "epoch": 0.1936382956485385,
+      "grad_norm": 0.7283337615791947,
+      "learning_rate": 6.434858302509284e-07,
+      "loss": 1.6787,
+      "step": 2779
+    },
+    {
+      "epoch": 0.1937079747761558,
+      "grad_norm": 0.7479235036434875,
+      "learning_rate": 6.434464139401287e-07,
+      "loss": 1.8808,
+      "step": 2780
+    },
+    {
+      "epoch": 0.19377765390377313,
+      "grad_norm": 0.6914895910578951,
+      "learning_rate": 6.434069852440461e-07,
+      "loss": 1.5673,
+      "step": 2781
+    },
+    {
+      "epoch": 0.19384733303139046,
+      "grad_norm": 0.7254723360609027,
+      "learning_rate": 6.433675441645703e-07,
+      "loss": 1.539,
+      "step": 2782
+    },
+    {
+      "epoch": 0.19391701215900778,
+      "grad_norm": 0.6833639219823902,
+      "learning_rate": 6.433280907035914e-07,
+      "loss": 1.5052,
+      "step": 2783
+    },
+    {
+      "epoch": 0.1939866912866251,
+      "grad_norm": 0.7420464590441822,
+      "learning_rate": 6.432886248630001e-07,
+      "loss": 1.5594,
+      "step": 2784
+    },
+    {
+      "epoch": 0.19405637041424242,
+      "grad_norm": 0.6652408829333618,
+      "learning_rate": 6.432491466446876e-07,
+      "loss": 1.2722,
+      "step": 2785
+    },
+    {
+      "epoch": 0.19412604954185975,
+      "grad_norm": 0.6626300147201509,
+      "learning_rate": 6.432096560505458e-07,
+      "loss": 1.398,
+      "step": 2786
+    },
+    {
+      "epoch": 0.19419572866947707,
+      "grad_norm": 0.7465600044204658,
+      "learning_rate": 6.431701530824671e-07,
+      "loss": 1.5446,
+      "step": 2787
+    },
+    {
+      "epoch": 0.1942654077970944,
+      "grad_norm": 0.7559501299835806,
+      "learning_rate": 6.431306377423449e-07,
+      "loss": 1.4106,
+      "step": 2788
+    },
+    {
+      "epoch": 0.19433508692471171,
+      "grad_norm": 0.7012209472341381,
+      "learning_rate": 6.430911100320723e-07,
+      "loss": 1.492,
+      "step": 2789
+    },
+    {
+      "epoch": 0.19440476605232904,
+      "grad_norm": 0.711451679800713,
+      "learning_rate": 6.430515699535441e-07,
+      "loss": 1.516,
+      "step": 2790
+    },
+    {
+      "epoch": 0.19447444517994636,
+      "grad_norm": 0.7340907534836132,
+      "learning_rate": 6.430120175086548e-07,
+      "loss": 1.5429,
+      "step": 2791
+    },
+    {
+      "epoch": 0.19454412430756368,
+      "grad_norm": 0.7246947540430131,
+      "learning_rate": 6.429724526993e-07,
+      "loss": 1.4757,
+      "step": 2792
+    },
+    {
+      "epoch": 0.194613803435181,
+      "grad_norm": 0.6721144443775302,
+      "learning_rate": 6.429328755273755e-07,
+      "loss": 1.4574,
+      "step": 2793
+    },
+    {
+      "epoch": 0.19468348256279833,
+      "grad_norm": 0.7162298227034134,
+      "learning_rate": 6.428932859947782e-07,
+      "loss": 1.5135,
+      "step": 2794
+    },
+    {
+      "epoch": 0.19475316169041565,
+      "grad_norm": 0.6931706305166699,
+      "learning_rate": 6.428536841034052e-07,
+      "loss": 1.6258,
+      "step": 2795
+    },
+    {
+      "epoch": 0.19482284081803297,
+      "grad_norm": 0.6696857466893444,
+      "learning_rate": 6.428140698551542e-07,
+      "loss": 1.4967,
+      "step": 2796
+    },
+    {
+      "epoch": 0.19489251994565027,
+      "grad_norm": 0.7079042834337671,
+      "learning_rate": 6.427744432519237e-07,
+      "loss": 1.4456,
+      "step": 2797
+    },
+    {
+      "epoch": 0.1949621990732676,
+      "grad_norm": 0.7021521523597029,
+      "learning_rate": 6.427348042956129e-07,
+      "loss": 1.6123,
+      "step": 2798
+    },
+    {
+      "epoch": 0.1950318782008849,
+      "grad_norm": 0.7125268105966331,
+      "learning_rate": 6.426951529881209e-07,
+      "loss": 1.4722,
+      "step": 2799
+    },
+    {
+      "epoch": 0.19510155732850223,
+      "grad_norm": 0.6802108240355292,
+      "learning_rate": 6.426554893313482e-07,
+      "loss": 1.5581,
+      "step": 2800
+    },
+    {
+      "epoch": 0.19517123645611956,
+      "grad_norm": 0.7984301225935575,
+      "learning_rate": 6.426158133271956e-07,
+      "loss": 1.7121,
+      "step": 2801
+    },
+    {
+      "epoch": 0.19524091558373688,
+      "grad_norm": 0.6879133160196618,
+      "learning_rate": 6.425761249775642e-07,
+      "loss": 1.5569,
+      "step": 2802
+    },
+    {
+      "epoch": 0.1953105947113542,
+      "grad_norm": 0.6831021054413613,
+      "learning_rate": 6.425364242843561e-07,
+      "loss": 1.6133,
+      "step": 2803
+    },
+    {
+      "epoch": 0.19538027383897152,
+      "grad_norm": 0.7194943179719568,
+      "learning_rate": 6.42496711249474e-07,
+      "loss": 1.6279,
+      "step": 2804
+    },
+    {
+      "epoch": 0.19544995296658885,
+      "grad_norm": 0.7155004699065798,
+      "learning_rate": 6.424569858748207e-07,
+      "loss": 1.5022,
+      "step": 2805
+    },
+    {
+      "epoch": 0.19551963209420617,
+      "grad_norm": 0.7372121801210776,
+      "learning_rate": 6.424172481623002e-07,
+      "loss": 1.4252,
+      "step": 2806
+    },
+    {
+      "epoch": 0.1955893112218235,
+      "grad_norm": 0.7147709147935137,
+      "learning_rate": 6.423774981138167e-07,
+      "loss": 1.6929,
+      "step": 2807
+    },
+    {
+      "epoch": 0.19565899034944081,
+      "grad_norm": 0.7087472475940669,
+      "learning_rate": 6.423377357312749e-07,
+      "loss": 1.6249,
+      "step": 2808
+    },
+    {
+      "epoch": 0.19572866947705814,
+      "grad_norm": 0.6918451314684043,
+      "learning_rate": 6.422979610165808e-07,
+      "loss": 1.5191,
+      "step": 2809
+    },
+    {
+      "epoch": 0.19579834860467546,
+      "grad_norm": 0.7084086537536417,
+      "learning_rate": 6.422581739716399e-07,
+      "loss": 1.5604,
+      "step": 2810
+    },
+    {
+      "epoch": 0.19586802773229278,
+      "grad_norm": 0.703091479351792,
+      "learning_rate": 6.422183745983593e-07,
+      "loss": 1.6192,
+      "step": 2811
+    },
+    {
+      "epoch": 0.1959377068599101,
+      "grad_norm": 0.7999940834598073,
+      "learning_rate": 6.421785628986463e-07,
+      "loss": 1.6106,
+      "step": 2812
+    },
+    {
+      "epoch": 0.19600738598752743,
+      "grad_norm": 0.7439364790535018,
+      "learning_rate": 6.421387388744083e-07,
+      "loss": 1.4405,
+      "step": 2813
+    },
+    {
+      "epoch": 0.19607706511514475,
+      "grad_norm": 0.7334064002022052,
+      "learning_rate": 6.420989025275542e-07,
+      "loss": 1.2867,
+      "step": 2814
+    },
+    {
+      "epoch": 0.19614674424276207,
+      "grad_norm": 0.746628498475421,
+      "learning_rate": 6.420590538599928e-07,
+      "loss": 1.4626,
+      "step": 2815
+    },
+    {
+      "epoch": 0.1962164233703794,
+      "grad_norm": 0.6954573088788314,
+      "learning_rate": 6.420191928736339e-07,
+      "loss": 1.5406,
+      "step": 2816
+    },
+    {
+      "epoch": 0.19628610249799672,
+      "grad_norm": 0.6996811523016951,
+      "learning_rate": 6.419793195703875e-07,
+      "loss": 1.6512,
+      "step": 2817
+    },
+    {
+      "epoch": 0.19635578162561404,
+      "grad_norm": 0.7488194125821913,
+      "learning_rate": 6.419394339521647e-07,
+      "loss": 1.5551,
+      "step": 2818
+    },
+    {
+      "epoch": 0.19642546075323136,
+      "grad_norm": 0.6858185427092187,
+      "learning_rate": 6.418995360208766e-07,
+      "loss": 1.5719,
+      "step": 2819
+    },
+    {
+      "epoch": 0.19649513988084868,
+      "grad_norm": 0.8012233152006635,
+      "learning_rate": 6.418596257784354e-07,
+      "loss": 1.5205,
+      "step": 2820
+    },
+    {
+      "epoch": 0.196564819008466,
+      "grad_norm": 0.6869950460759762,
+      "learning_rate": 6.418197032267536e-07,
+      "loss": 1.4701,
+      "step": 2821
+    },
+    {
+      "epoch": 0.19663449813608333,
+      "grad_norm": 0.7558716905890024,
+      "learning_rate": 6.417797683677444e-07,
+      "loss": 1.5402,
+      "step": 2822
+    },
+    {
+      "epoch": 0.19670417726370065,
+      "grad_norm": 0.6739914935874416,
+      "learning_rate": 6.417398212033214e-07,
+      "loss": 1.5758,
+      "step": 2823
+    },
+    {
+      "epoch": 0.19677385639131798,
+      "grad_norm": 0.7447417175089804,
+      "learning_rate": 6.416998617353993e-07,
+      "loss": 1.6096,
+      "step": 2824
+    },
+    {
+      "epoch": 0.1968435355189353,
+      "grad_norm": 0.6751254016745756,
+      "learning_rate": 6.416598899658926e-07,
+      "loss": 1.4143,
+      "step": 2825
+    },
+    {
+      "epoch": 0.19691321464655262,
+      "grad_norm": 0.7045687553583755,
+      "learning_rate": 6.416199058967173e-07,
+      "loss": 1.6254,
+      "step": 2826
+    },
+    {
+      "epoch": 0.19698289377416994,
+      "grad_norm": 0.7201577315390923,
+      "learning_rate": 6.41579909529789e-07,
+      "loss": 1.561,
+      "step": 2827
+    },
+    {
+      "epoch": 0.19705257290178727,
+      "grad_norm": 0.6772138874740977,
+      "learning_rate": 6.41539900867025e-07,
+      "loss": 1.6012,
+      "step": 2828
+    },
+    {
+      "epoch": 0.1971222520294046,
+      "grad_norm": 0.7547835224780833,
+      "learning_rate": 6.414998799103421e-07,
+      "loss": 1.6242,
+      "step": 2829
+    },
+    {
+      "epoch": 0.1971919311570219,
+      "grad_norm": 0.7656065190946523,
+      "learning_rate": 6.414598466616585e-07,
+      "loss": 1.6035,
+      "step": 2830
+    },
+    {
+      "epoch": 0.19726161028463923,
+      "grad_norm": 0.7244976383852213,
+      "learning_rate": 6.414198011228923e-07,
+      "loss": 1.6974,
+      "step": 2831
+    },
+    {
+      "epoch": 0.19733128941225656,
+      "grad_norm": 0.7259803282047949,
+      "learning_rate": 6.413797432959631e-07,
+      "loss": 1.5899,
+      "step": 2832
+    },
+    {
+      "epoch": 0.19740096853987388,
+      "grad_norm": 0.7657140131144465,
+      "learning_rate": 6.413396731827901e-07,
+      "loss": 1.5469,
+      "step": 2833
+    },
+    {
+      "epoch": 0.1974706476674912,
+      "grad_norm": 0.7295398215507435,
+      "learning_rate": 6.412995907852937e-07,
+      "loss": 1.5622,
+      "step": 2834
+    },
+    {
+      "epoch": 0.19754032679510852,
+      "grad_norm": 0.7318433019876475,
+      "learning_rate": 6.412594961053948e-07,
+      "loss": 1.5309,
+      "step": 2835
+    },
+    {
+      "epoch": 0.19761000592272585,
+      "grad_norm": 0.71192020542836,
+      "learning_rate": 6.412193891450147e-07,
+      "loss": 1.5897,
+      "step": 2836
+    },
+    {
+      "epoch": 0.19767968505034317,
+      "grad_norm": 0.7021056662939148,
+      "learning_rate": 6.411792699060755e-07,
+      "loss": 1.5287,
+      "step": 2837
+    },
+    {
+      "epoch": 0.1977493641779605,
+      "grad_norm": 0.7681891994041536,
+      "learning_rate": 6.411391383904998e-07,
+      "loss": 1.5876,
+      "step": 2838
+    },
+    {
+      "epoch": 0.1978190433055778,
+      "grad_norm": 0.7493583117634097,
+      "learning_rate": 6.410989946002107e-07,
+      "loss": 1.5165,
+      "step": 2839
+    },
+    {
+      "epoch": 0.19788872243319514,
+      "grad_norm": 0.7378067482423463,
+      "learning_rate": 6.41058838537132e-07,
+      "loss": 1.4552,
+      "step": 2840
+    },
+    {
+      "epoch": 0.19795840156081246,
+      "grad_norm": 0.7445427788569985,
+      "learning_rate": 6.41018670203188e-07,
+      "loss": 1.5561,
+      "step": 2841
+    },
+    {
+      "epoch": 0.19802808068842978,
+      "grad_norm": 0.7622629977423278,
+      "learning_rate": 6.409784896003039e-07,
+      "loss": 1.5894,
+      "step": 2842
+    },
+    {
+      "epoch": 0.1980977598160471,
+      "grad_norm": 0.7665002701824735,
+      "learning_rate": 6.40938296730405e-07,
+      "loss": 1.5411,
+      "step": 2843
+    },
+    {
+      "epoch": 0.19816743894366443,
+      "grad_norm": 0.7254647373761564,
+      "learning_rate": 6.408980915954175e-07,
+      "loss": 1.464,
+      "step": 2844
+    },
+    {
+      "epoch": 0.19823711807128175,
+      "grad_norm": 0.7271343977692023,
+      "learning_rate": 6.40857874197268e-07,
+      "loss": 1.6614,
+      "step": 2845
+    },
+    {
+      "epoch": 0.19830679719889907,
+      "grad_norm": 0.7139000862151373,
+      "learning_rate": 6.40817644537884e-07,
+      "loss": 1.5599,
+      "step": 2846
+    },
+    {
+      "epoch": 0.1983764763265164,
+      "grad_norm": 0.6681638182102779,
+      "learning_rate": 6.407774026191933e-07,
+      "loss": 1.6665,
+      "step": 2847
+    },
+    {
+      "epoch": 0.19844615545413372,
+      "grad_norm": 0.7868727777360315,
+      "learning_rate": 6.407371484431243e-07,
+      "loss": 1.5568,
+      "step": 2848
+    },
+    {
+      "epoch": 0.19851583458175104,
+      "grad_norm": 0.7553355394120074,
+      "learning_rate": 6.406968820116061e-07,
+      "loss": 1.551,
+      "step": 2849
+    },
+    {
+      "epoch": 0.19858551370936836,
+      "grad_norm": 0.745667391752033,
+      "learning_rate": 6.406566033265686e-07,
+      "loss": 1.467,
+      "step": 2850
+    },
+    {
+      "epoch": 0.19865519283698568,
+      "grad_norm": 0.7214889502095606,
+      "learning_rate": 6.406163123899416e-07,
+      "loss": 1.5626,
+      "step": 2851
+    },
+    {
+      "epoch": 0.198724871964603,
+      "grad_norm": 0.7039722794357991,
+      "learning_rate": 6.405760092036561e-07,
+      "loss": 1.4811,
+      "step": 2852
+    },
+    {
+      "epoch": 0.19879455109222033,
+      "grad_norm": 0.7028228796727578,
+      "learning_rate": 6.405356937696437e-07,
+      "loss": 1.5365,
+      "step": 2853
+    },
+    {
+      "epoch": 0.19886423021983765,
+      "grad_norm": 0.7327098572778575,
+      "learning_rate": 6.404953660898362e-07,
+      "loss": 1.6272,
+      "step": 2854
+    },
+    {
+      "epoch": 0.19893390934745497,
+      "grad_norm": 0.6640798875637417,
+      "learning_rate": 6.404550261661662e-07,
+      "loss": 1.4853,
+      "step": 2855
+    },
+    {
+      "epoch": 0.1990035884750723,
+      "grad_norm": 0.6993347936293075,
+      "learning_rate": 6.40414674000567e-07,
+      "loss": 1.565,
+      "step": 2856
+    },
+    {
+      "epoch": 0.19907326760268962,
+      "grad_norm": 0.68104277668191,
+      "learning_rate": 6.403743095949722e-07,
+      "loss": 1.4548,
+      "step": 2857
+    },
+    {
+      "epoch": 0.19914294673030694,
+      "grad_norm": 0.7617610054585428,
+      "learning_rate": 6.403339329513161e-07,
+      "loss": 1.6688,
+      "step": 2858
+    },
+    {
+      "epoch": 0.19921262585792426,
+      "grad_norm": 0.7131057683123566,
+      "learning_rate": 6.402935440715339e-07,
+      "loss": 1.6218,
+      "step": 2859
+    },
+    {
+      "epoch": 0.1992823049855416,
+      "grad_norm": 0.6968176358706246,
+      "learning_rate": 6.402531429575609e-07,
+      "loss": 1.4128,
+      "step": 2860
+    },
+    {
+      "epoch": 0.1993519841131589,
+      "grad_norm": 0.718108124628976,
+      "learning_rate": 6.402127296113334e-07,
+      "loss": 1.6234,
+      "step": 2861
+    },
+    {
+      "epoch": 0.19942166324077623,
+      "grad_norm": 0.7445450879925423,
+      "learning_rate": 6.401723040347878e-07,
+      "loss": 1.4594,
+      "step": 2862
+    },
+    {
+      "epoch": 0.19949134236839355,
+      "grad_norm": 0.7261122364370047,
+      "learning_rate": 6.401318662298615e-07,
+      "loss": 1.5517,
+      "step": 2863
+    },
+    {
+      "epoch": 0.19956102149601088,
+      "grad_norm": 0.7220852520629888,
+      "learning_rate": 6.400914161984925e-07,
+      "loss": 1.5947,
+      "step": 2864
+    },
+    {
+      "epoch": 0.1996307006236282,
+      "grad_norm": 0.6652206452519076,
+      "learning_rate": 6.400509539426191e-07,
+      "loss": 1.4174,
+      "step": 2865
+    },
+    {
+      "epoch": 0.19970037975124552,
+      "grad_norm": 0.7005410182147795,
+      "learning_rate": 6.400104794641804e-07,
+      "loss": 1.6145,
+      "step": 2866
+    },
+    {
+      "epoch": 0.19977005887886284,
+      "grad_norm": 0.7164367875663481,
+      "learning_rate": 6.39969992765116e-07,
+      "loss": 1.6245,
+      "step": 2867
+    },
+    {
+      "epoch": 0.19983973800648017,
+      "grad_norm": 0.7140554642769725,
+      "learning_rate": 6.399294938473661e-07,
+      "loss": 1.5329,
+      "step": 2868
+    },
+    {
+      "epoch": 0.1999094171340975,
+      "grad_norm": 0.7436557780926554,
+      "learning_rate": 6.398889827128715e-07,
+      "loss": 1.6935,
+      "step": 2869
+    },
+    {
+      "epoch": 0.1999790962617148,
+      "grad_norm": 0.7231366816956215,
+      "learning_rate": 6.398484593635735e-07,
+      "loss": 1.5485,
+      "step": 2870
+    },
+    {
+      "epoch": 0.20004877538933213,
+      "grad_norm": 0.66648456054266,
+      "learning_rate": 6.398079238014141e-07,
+      "loss": 1.4717,
+      "step": 2871
+    },
+    {
+      "epoch": 0.20011845451694946,
+      "grad_norm": 0.7575185945561026,
+      "learning_rate": 6.39767376028336e-07,
+      "loss": 1.6266,
+      "step": 2872
+    },
+    {
+      "epoch": 0.20018813364456678,
+      "grad_norm": 0.7215300676555968,
+      "learning_rate": 6.397268160462822e-07,
+      "loss": 1.5363,
+      "step": 2873
+    },
+    {
+      "epoch": 0.2002578127721841,
+      "grad_norm": 0.7612344619035825,
+      "learning_rate": 6.396862438571965e-07,
+      "loss": 1.5018,
+      "step": 2874
+    },
+    {
+      "epoch": 0.20032749189980142,
+      "grad_norm": 0.7581568365524771,
+      "learning_rate": 6.39645659463023e-07,
+      "loss": 1.5225,
+      "step": 2875
+    },
+    {
+      "epoch": 0.20039717102741875,
+      "grad_norm": 0.7129849329155351,
+      "learning_rate": 6.396050628657068e-07,
+      "loss": 1.5566,
+      "step": 2876
+    },
+    {
+      "epoch": 0.20046685015503607,
+      "grad_norm": 0.7277241962059409,
+      "learning_rate": 6.395644540671932e-07,
+      "loss": 1.5693,
+      "step": 2877
+    },
+    {
+      "epoch": 0.2005365292826534,
+      "grad_norm": 0.7268033481810099,
+      "learning_rate": 6.395238330694284e-07,
+      "loss": 1.7033,
+      "step": 2878
+    },
+    {
+      "epoch": 0.20060620841027071,
+      "grad_norm": 0.7578464136236305,
+      "learning_rate": 6.39483199874359e-07,
+      "loss": 1.5682,
+      "step": 2879
+    },
+    {
+      "epoch": 0.20067588753788804,
+      "grad_norm": 0.685464342009214,
+      "learning_rate": 6.394425544839323e-07,
+      "loss": 1.5303,
+      "step": 2880
+    },
+    {
+      "epoch": 0.20074556666550536,
+      "grad_norm": 0.7182341780133301,
+      "learning_rate": 6.394018969000958e-07,
+      "loss": 1.592,
+      "step": 2881
+    },
+    {
+      "epoch": 0.20081524579312268,
+      "grad_norm": 0.6970355774318064,
+      "learning_rate": 6.393612271247984e-07,
+      "loss": 1.5237,
+      "step": 2882
+    },
+    {
+      "epoch": 0.20088492492074,
+      "grad_norm": 0.692437523420572,
+      "learning_rate": 6.393205451599887e-07,
+      "loss": 1.5745,
+      "step": 2883
+    },
+    {
+      "epoch": 0.20095460404835733,
+      "grad_norm": 0.7421810031357134,
+      "learning_rate": 6.392798510076162e-07,
+      "loss": 1.5122,
+      "step": 2884
+    },
+    {
+      "epoch": 0.20102428317597465,
+      "grad_norm": 0.7247718274256905,
+      "learning_rate": 6.392391446696315e-07,
+      "loss": 1.5325,
+      "step": 2885
+    },
+    {
+      "epoch": 0.20109396230359197,
+      "grad_norm": 0.7015610226696939,
+      "learning_rate": 6.391984261479848e-07,
+      "loss": 1.4387,
+      "step": 2886
+    },
+    {
+      "epoch": 0.2011636414312093,
+      "grad_norm": 0.7136527685788008,
+      "learning_rate": 6.391576954446278e-07,
+      "loss": 1.5067,
+      "step": 2887
+    },
+    {
+      "epoch": 0.20123332055882662,
+      "grad_norm": 0.751014147824683,
+      "learning_rate": 6.391169525615121e-07,
+      "loss": 1.6104,
+      "step": 2888
+    },
+    {
+      "epoch": 0.2013029996864439,
+      "grad_norm": 0.8270177555880811,
+      "learning_rate": 6.390761975005905e-07,
+      "loss": 1.6917,
+      "step": 2889
+    },
+    {
+      "epoch": 0.20137267881406123,
+      "grad_norm": 0.7686620792228451,
+      "learning_rate": 6.39035430263816e-07,
+      "loss": 1.4402,
+      "step": 2890
+    },
+    {
+      "epoch": 0.20144235794167856,
+      "grad_norm": 0.6852734847807493,
+      "learning_rate": 6.38994650853142e-07,
+      "loss": 1.4847,
+      "step": 2891
+    },
+    {
+      "epoch": 0.20151203706929588,
+      "grad_norm": 0.6861463442143995,
+      "learning_rate": 6.389538592705229e-07,
+      "loss": 1.56,
+      "step": 2892
+    },
+    {
+      "epoch": 0.2015817161969132,
+      "grad_norm": 0.7185857414585152,
+      "learning_rate": 6.389130555179134e-07,
+      "loss": 1.4896,
+      "step": 2893
+    },
+    {
+      "epoch": 0.20165139532453052,
+      "grad_norm": 0.7038209840042945,
+      "learning_rate": 6.388722395972692e-07,
+      "loss": 1.4935,
+      "step": 2894
+    },
+    {
+      "epoch": 0.20172107445214785,
+      "grad_norm": 0.7107114427154652,
+      "learning_rate": 6.38831411510546e-07,
+      "loss": 1.6031,
+      "step": 2895
+    },
+    {
+      "epoch": 0.20179075357976517,
+      "grad_norm": 0.6991671503850788,
+      "learning_rate": 6.387905712597004e-07,
+      "loss": 1.5549,
+      "step": 2896
+    },
+    {
+      "epoch": 0.2018604327073825,
+      "grad_norm": 0.7530770913699526,
+      "learning_rate": 6.387497188466897e-07,
+      "loss": 1.4468,
+      "step": 2897
+    },
+    {
+      "epoch": 0.20193011183499981,
+      "grad_norm": 0.6670243465782024,
+      "learning_rate": 6.387088542734715e-07,
+      "loss": 1.6758,
+      "step": 2898
+    },
+    {
+      "epoch": 0.20199979096261714,
+      "grad_norm": 0.7092271431291487,
+      "learning_rate": 6.386679775420042e-07,
+      "loss": 1.4933,
+      "step": 2899
+    },
+    {
+      "epoch": 0.20206947009023446,
+      "grad_norm": 0.736826431597258,
+      "learning_rate": 6.386270886542466e-07,
+      "loss": 1.4696,
+      "step": 2900
+    },
+    {
+      "epoch": 0.20213914921785178,
+      "grad_norm": 0.6884689205263212,
+      "learning_rate": 6.385861876121582e-07,
+      "loss": 1.5714,
+      "step": 2901
+    },
+    {
+      "epoch": 0.2022088283454691,
+      "grad_norm": 0.7191280188616926,
+      "learning_rate": 6.385452744176991e-07,
+      "loss": 1.4575,
+      "step": 2902
+    },
+    {
+      "epoch": 0.20227850747308643,
+      "grad_norm": 0.7246993662960771,
+      "learning_rate": 6.385043490728299e-07,
+      "loss": 1.5546,
+      "step": 2903
+    },
+    {
+      "epoch": 0.20234818660070375,
+      "grad_norm": 0.7135160908956892,
+      "learning_rate": 6.384634115795119e-07,
+      "loss": 1.5166,
+      "step": 2904
+    },
+    {
+      "epoch": 0.20241786572832107,
+      "grad_norm": 0.7142107711582788,
+      "learning_rate": 6.384224619397069e-07,
+      "loss": 1.4995,
+      "step": 2905
+    },
+    {
+      "epoch": 0.2024875448559384,
+      "grad_norm": 0.7530747506999542,
+      "learning_rate": 6.383815001553772e-07,
+      "loss": 1.5028,
+      "step": 2906
+    },
+    {
+      "epoch": 0.20255722398355572,
+      "grad_norm": 0.6930300478756766,
+      "learning_rate": 6.383405262284859e-07,
+      "loss": 1.6004,
+      "step": 2907
+    },
+    {
+      "epoch": 0.20262690311117304,
+      "grad_norm": 0.8030841056649302,
+      "learning_rate": 6.382995401609964e-07,
+      "loss": 1.5417,
+      "step": 2908
+    },
+    {
+      "epoch": 0.20269658223879036,
+      "grad_norm": 0.7682765721959441,
+      "learning_rate": 6.38258541954873e-07,
+      "loss": 1.6121,
+      "step": 2909
+    },
+    {
+      "epoch": 0.20276626136640769,
+      "grad_norm": 0.7125907867109845,
+      "learning_rate": 6.382175316120802e-07,
+      "loss": 1.4901,
+      "step": 2910
+    },
+    {
+      "epoch": 0.202835940494025,
+      "grad_norm": 0.7415556985743339,
+      "learning_rate": 6.381765091345836e-07,
+      "loss": 1.624,
+      "step": 2911
+    },
+    {
+      "epoch": 0.20290561962164233,
+      "grad_norm": 0.7215852980031595,
+      "learning_rate": 6.381354745243489e-07,
+      "loss": 1.5852,
+      "step": 2912
+    },
+    {
+      "epoch": 0.20297529874925965,
+      "grad_norm": 0.6972105029569049,
+      "learning_rate": 6.380944277833425e-07,
+      "loss": 1.4699,
+      "step": 2913
+    },
+    {
+      "epoch": 0.20304497787687698,
+      "grad_norm": 0.7509877613104287,
+      "learning_rate": 6.380533689135316e-07,
+      "loss": 1.5867,
+      "step": 2914
+    },
+    {
+      "epoch": 0.2031146570044943,
+      "grad_norm": 0.7736493844250907,
+      "learning_rate": 6.380122979168837e-07,
+      "loss": 1.5804,
+      "step": 2915
+    },
+    {
+      "epoch": 0.20318433613211162,
+      "grad_norm": 0.689680312040028,
+      "learning_rate": 6.379712147953671e-07,
+      "loss": 1.5172,
+      "step": 2916
+    },
+    {
+      "epoch": 0.20325401525972894,
+      "grad_norm": 0.7474879041731053,
+      "learning_rate": 6.379301195509505e-07,
+      "loss": 1.4659,
+      "step": 2917
+    },
+    {
+      "epoch": 0.20332369438734627,
+      "grad_norm": 0.7394318317131775,
+      "learning_rate": 6.378890121856034e-07,
+      "loss": 1.5584,
+      "step": 2918
+    },
+    {
+      "epoch": 0.2033933735149636,
+      "grad_norm": 0.6932945340619577,
+      "learning_rate": 6.378478927012956e-07,
+      "loss": 1.5079,
+      "step": 2919
+    },
+    {
+      "epoch": 0.2034630526425809,
+      "grad_norm": 0.7306806137581724,
+      "learning_rate": 6.378067610999978e-07,
+      "loss": 1.6069,
+      "step": 2920
+    },
+    {
+      "epoch": 0.20353273177019823,
+      "grad_norm": 0.7651741778475234,
+      "learning_rate": 6.377656173836808e-07,
+      "loss": 1.4645,
+      "step": 2921
+    },
+    {
+      "epoch": 0.20360241089781556,
+      "grad_norm": 0.7220245915728065,
+      "learning_rate": 6.377244615543167e-07,
+      "loss": 1.5917,
+      "step": 2922
+    },
+    {
+      "epoch": 0.20367209002543288,
+      "grad_norm": 0.7149824274459259,
+      "learning_rate": 6.376832936138774e-07,
+      "loss": 1.6134,
+      "step": 2923
+    },
+    {
+      "epoch": 0.2037417691530502,
+      "grad_norm": 0.7180373040033633,
+      "learning_rate": 6.37642113564336e-07,
+      "loss": 1.4994,
+      "step": 2924
+    },
+    {
+      "epoch": 0.20381144828066752,
+      "grad_norm": 0.7100774209110662,
+      "learning_rate": 6.376009214076658e-07,
+      "loss": 1.4923,
+      "step": 2925
+    },
+    {
+      "epoch": 0.20388112740828485,
+      "grad_norm": 0.6672838981463166,
+      "learning_rate": 6.37559717145841e-07,
+      "loss": 1.5192,
+      "step": 2926
+    },
+    {
+      "epoch": 0.20395080653590217,
+      "grad_norm": 0.7585267274723554,
+      "learning_rate": 6.37518500780836e-07,
+      "loss": 1.556,
+      "step": 2927
+    },
+    {
+      "epoch": 0.2040204856635195,
+      "grad_norm": 0.7175807603218348,
+      "learning_rate": 6.374772723146259e-07,
+      "loss": 1.592,
+      "step": 2928
+    },
+    {
+      "epoch": 0.2040901647911368,
+      "grad_norm": 0.7037496395350076,
+      "learning_rate": 6.374360317491867e-07,
+      "loss": 1.5354,
+      "step": 2929
+    },
+    {
+      "epoch": 0.20415984391875414,
+      "grad_norm": 0.691830573824489,
+      "learning_rate": 6.373947790864947e-07,
+      "loss": 1.5469,
+      "step": 2930
+    },
+    {
+      "epoch": 0.20422952304637146,
+      "grad_norm": 0.7167616804294645,
+      "learning_rate": 6.373535143285267e-07,
+      "loss": 1.6573,
+      "step": 2931
+    },
+    {
+      "epoch": 0.20429920217398878,
+      "grad_norm": 0.6779235031573106,
+      "learning_rate": 6.373122374772601e-07,
+      "loss": 1.5974,
+      "step": 2932
+    },
+    {
+      "epoch": 0.2043688813016061,
+      "grad_norm": 0.6919311558080987,
+      "learning_rate": 6.372709485346731e-07,
+      "loss": 1.5194,
+      "step": 2933
+    },
+    {
+      "epoch": 0.20443856042922343,
+      "grad_norm": 0.6804788103429456,
+      "learning_rate": 6.372296475027444e-07,
+      "loss": 1.4618,
+      "step": 2934
+    },
+    {
+      "epoch": 0.20450823955684075,
+      "grad_norm": 0.7378684110679562,
+      "learning_rate": 6.371883343834532e-07,
+      "loss": 1.591,
+      "step": 2935
+    },
+    {
+      "epoch": 0.20457791868445807,
+      "grad_norm": 0.7033761641976377,
+      "learning_rate": 6.371470091787792e-07,
+      "loss": 1.4276,
+      "step": 2936
+    },
+    {
+      "epoch": 0.2046475978120754,
+      "grad_norm": 0.7226159322335646,
+      "learning_rate": 6.371056718907029e-07,
+      "loss": 1.4797,
+      "step": 2937
+    },
+    {
+      "epoch": 0.20471727693969272,
+      "grad_norm": 0.7733144725121075,
+      "learning_rate": 6.370643225212054e-07,
+      "loss": 1.6505,
+      "step": 2938
+    },
+    {
+      "epoch": 0.20478695606731004,
+      "grad_norm": 0.7425310543293385,
+      "learning_rate": 6.370229610722679e-07,
+      "loss": 1.5771,
+      "step": 2939
+    },
+    {
+      "epoch": 0.20485663519492736,
+      "grad_norm": 0.6828119499786712,
+      "learning_rate": 6.369815875458727e-07,
+      "loss": 1.5115,
+      "step": 2940
+    },
+    {
+      "epoch": 0.20492631432254468,
+      "grad_norm": 0.7025329013795072,
+      "learning_rate": 6.369402019440027e-07,
+      "loss": 1.5258,
+      "step": 2941
+    },
+    {
+      "epoch": 0.204995993450162,
+      "grad_norm": 0.7426863118585209,
+      "learning_rate": 6.368988042686408e-07,
+      "loss": 1.4511,
+      "step": 2942
+    },
+    {
+      "epoch": 0.20506567257777933,
+      "grad_norm": 0.6975590020208378,
+      "learning_rate": 6.368573945217712e-07,
+      "loss": 1.5915,
+      "step": 2943
+    },
+    {
+      "epoch": 0.20513535170539665,
+      "grad_norm": 0.6555378771055111,
+      "learning_rate": 6.368159727053781e-07,
+      "loss": 1.53,
+      "step": 2944
+    },
+    {
+      "epoch": 0.20520503083301397,
+      "grad_norm": 0.730334550632792,
+      "learning_rate": 6.367745388214467e-07,
+      "loss": 1.5767,
+      "step": 2945
+    },
+    {
+      "epoch": 0.2052747099606313,
+      "grad_norm": 0.6796967601212938,
+      "learning_rate": 6.367330928719625e-07,
+      "loss": 1.5181,
+      "step": 2946
+    },
+    {
+      "epoch": 0.20534438908824862,
+      "grad_norm": 0.7077642966530233,
+      "learning_rate": 6.366916348589116e-07,
+      "loss": 1.3541,
+      "step": 2947
+    },
+    {
+      "epoch": 0.20541406821586594,
+      "grad_norm": 0.739817028672939,
+      "learning_rate": 6.36650164784281e-07,
+      "loss": 1.5237,
+      "step": 2948
+    },
+    {
+      "epoch": 0.20548374734348326,
+      "grad_norm": 0.7232713294038303,
+      "learning_rate": 6.366086826500578e-07,
+      "loss": 1.642,
+      "step": 2949
+    },
+    {
+      "epoch": 0.2055534264711006,
+      "grad_norm": 0.7529785742217403,
+      "learning_rate": 6.3656718845823e-07,
+      "loss": 1.5518,
+      "step": 2950
+    },
+    {
+      "epoch": 0.2056231055987179,
+      "grad_norm": 0.7295660194953272,
+      "learning_rate": 6.365256822107859e-07,
+      "loss": 1.4517,
+      "step": 2951
+    },
+    {
+      "epoch": 0.20569278472633523,
+      "grad_norm": 0.6964765639169348,
+      "learning_rate": 6.36484163909715e-07,
+      "loss": 1.6233,
+      "step": 2952
+    },
+    {
+      "epoch": 0.20576246385395255,
+      "grad_norm": 0.710909846668233,
+      "learning_rate": 6.364426335570066e-07,
+      "loss": 1.6418,
+      "step": 2953
+    },
+    {
+      "epoch": 0.20583214298156988,
+      "grad_norm": 0.7652461920615087,
+      "learning_rate": 6.364010911546508e-07,
+      "loss": 1.5322,
+      "step": 2954
+    },
+    {
+      "epoch": 0.2059018221091872,
+      "grad_norm": 0.7045875156326461,
+      "learning_rate": 6.363595367046389e-07,
+      "loss": 1.5305,
+      "step": 2955
+    },
+    {
+      "epoch": 0.20597150123680452,
+      "grad_norm": 0.7563366210538794,
+      "learning_rate": 6.363179702089618e-07,
+      "loss": 1.5893,
+      "step": 2956
+    },
+    {
+      "epoch": 0.20604118036442184,
+      "grad_norm": 0.7511868496946549,
+      "learning_rate": 6.362763916696117e-07,
+      "loss": 1.732,
+      "step": 2957
+    },
+    {
+      "epoch": 0.20611085949203917,
+      "grad_norm": 0.6697718587186111,
+      "learning_rate": 6.362348010885809e-07,
+      "loss": 1.5072,
+      "step": 2958
+    },
+    {
+      "epoch": 0.2061805386196565,
+      "grad_norm": 0.7530888567410823,
+      "learning_rate": 6.361931984678628e-07,
+      "loss": 1.5611,
+      "step": 2959
+    },
+    {
+      "epoch": 0.2062502177472738,
+      "grad_norm": 0.7185438842375362,
+      "learning_rate": 6.361515838094509e-07,
+      "loss": 1.6223,
+      "step": 2960
+    },
+    {
+      "epoch": 0.20631989687489113,
+      "grad_norm": 0.7036427853275619,
+      "learning_rate": 6.361099571153395e-07,
+      "loss": 1.4914,
+      "step": 2961
+    },
+    {
+      "epoch": 0.20638957600250846,
+      "grad_norm": 0.7408387503191232,
+      "learning_rate": 6.360683183875235e-07,
+      "loss": 1.5372,
+      "step": 2962
+    },
+    {
+      "epoch": 0.20645925513012578,
+      "grad_norm": 0.7178825080999842,
+      "learning_rate": 6.360266676279981e-07,
+      "loss": 1.6065,
+      "step": 2963
+    },
+    {
+      "epoch": 0.2065289342577431,
+      "grad_norm": 0.7429861155708916,
+      "learning_rate": 6.359850048387594e-07,
+      "loss": 1.4552,
+      "step": 2964
+    },
+    {
+      "epoch": 0.20659861338536042,
+      "grad_norm": 0.7202664843056981,
+      "learning_rate": 6.359433300218042e-07,
+      "loss": 1.594,
+      "step": 2965
+    },
+    {
+      "epoch": 0.20666829251297775,
+      "grad_norm": 0.7222321118467498,
+      "learning_rate": 6.359016431791294e-07,
+      "loss": 1.5255,
+      "step": 2966
+    },
+    {
+      "epoch": 0.20673797164059507,
+      "grad_norm": 0.734344864117076,
+      "learning_rate": 6.358599443127328e-07,
+      "loss": 1.5508,
+      "step": 2967
+    },
+    {
+      "epoch": 0.2068076507682124,
+      "grad_norm": 0.7206494532948653,
+      "learning_rate": 6.358182334246125e-07,
+      "loss": 1.4947,
+      "step": 2968
+    },
+    {
+      "epoch": 0.20687732989582971,
+      "grad_norm": 0.7140087905352821,
+      "learning_rate": 6.357765105167676e-07,
+      "loss": 1.5102,
+      "step": 2969
+    },
+    {
+      "epoch": 0.20694700902344704,
+      "grad_norm": 0.7852221694206425,
+      "learning_rate": 6.357347755911976e-07,
+      "loss": 1.6091,
+      "step": 2970
+    },
+    {
+      "epoch": 0.20701668815106436,
+      "grad_norm": 0.7048526642376749,
+      "learning_rate": 6.356930286499024e-07,
+      "loss": 1.5422,
+      "step": 2971
+    },
+    {
+      "epoch": 0.20708636727868168,
+      "grad_norm": 0.8007350555344261,
+      "learning_rate": 6.356512696948825e-07,
+      "loss": 1.5874,
+      "step": 2972
+    },
+    {
+      "epoch": 0.207156046406299,
+      "grad_norm": 0.7615620997332188,
+      "learning_rate": 6.356094987281391e-07,
+      "loss": 1.4809,
+      "step": 2973
+    },
+    {
+      "epoch": 0.20722572553391633,
+      "grad_norm": 0.7434804810145779,
+      "learning_rate": 6.355677157516743e-07,
+      "loss": 1.5758,
+      "step": 2974
+    },
+    {
+      "epoch": 0.20729540466153365,
+      "grad_norm": 0.7457304333301299,
+      "learning_rate": 6.355259207674899e-07,
+      "loss": 1.4783,
+      "step": 2975
+    },
+    {
+      "epoch": 0.20736508378915097,
+      "grad_norm": 0.706990856124059,
+      "learning_rate": 6.354841137775891e-07,
+      "loss": 1.4879,
+      "step": 2976
+    },
+    {
+      "epoch": 0.2074347629167683,
+      "grad_norm": 0.6955401431087637,
+      "learning_rate": 6.354422947839755e-07,
+      "loss": 1.5454,
+      "step": 2977
+    },
+    {
+      "epoch": 0.20750444204438562,
+      "grad_norm": 0.7384175680602987,
+      "learning_rate": 6.354004637886528e-07,
+      "loss": 1.5092,
+      "step": 2978
+    },
+    {
+      "epoch": 0.20757412117200294,
+      "grad_norm": 0.685234971809416,
+      "learning_rate": 6.35358620793626e-07,
+      "loss": 1.4753,
+      "step": 2979
+    },
+    {
+      "epoch": 0.20764380029962026,
+      "grad_norm": 0.70711394306406,
+      "learning_rate": 6.353167658009e-07,
+      "loss": 1.4555,
+      "step": 2980
+    },
+    {
+      "epoch": 0.20771347942723756,
+      "grad_norm": 0.7962597663635689,
+      "learning_rate": 6.352748988124805e-07,
+      "loss": 1.4569,
+      "step": 2981
+    },
+    {
+      "epoch": 0.20778315855485488,
+      "grad_norm": 0.6868913171560058,
+      "learning_rate": 6.352330198303742e-07,
+      "loss": 1.4446,
+      "step": 2982
+    },
+    {
+      "epoch": 0.2078528376824722,
+      "grad_norm": 0.7735566971642617,
+      "learning_rate": 6.35191128856588e-07,
+      "loss": 1.4883,
+      "step": 2983
+    },
+    {
+      "epoch": 0.20792251681008952,
+      "grad_norm": 0.7079499524333197,
+      "learning_rate": 6.351492258931292e-07,
+      "loss": 1.5655,
+      "step": 2984
+    },
+    {
+      "epoch": 0.20799219593770685,
+      "grad_norm": 0.7512965571925899,
+      "learning_rate": 6.351073109420059e-07,
+      "loss": 1.4857,
+      "step": 2985
+    },
+    {
+      "epoch": 0.20806187506532417,
+      "grad_norm": 0.7695609293964266,
+      "learning_rate": 6.350653840052269e-07,
+      "loss": 1.5553,
+      "step": 2986
+    },
+    {
+      "epoch": 0.2081315541929415,
+      "grad_norm": 0.7377077462933005,
+      "learning_rate": 6.350234450848013e-07,
+      "loss": 1.5265,
+      "step": 2987
+    },
+    {
+      "epoch": 0.20820123332055882,
+      "grad_norm": 0.7476217154283284,
+      "learning_rate": 6.349814941827387e-07,
+      "loss": 1.6797,
+      "step": 2988
+    },
+    {
+      "epoch": 0.20827091244817614,
+      "grad_norm": 0.7456407850803368,
+      "learning_rate": 6.3493953130105e-07,
+      "loss": 1.4848,
+      "step": 2989
+    },
+    {
+      "epoch": 0.20834059157579346,
+      "grad_norm": 0.7976077275272218,
+      "learning_rate": 6.348975564417456e-07,
+      "loss": 1.583,
+      "step": 2990
+    },
+    {
+      "epoch": 0.20841027070341078,
+      "grad_norm": 0.6623779094432293,
+      "learning_rate": 6.348555696068374e-07,
+      "loss": 1.4389,
+      "step": 2991
+    },
+    {
+      "epoch": 0.2084799498310281,
+      "grad_norm": 0.7602196162699404,
+      "learning_rate": 6.348135707983374e-07,
+      "loss": 1.6182,
+      "step": 2992
+    },
+    {
+      "epoch": 0.20854962895864543,
+      "grad_norm": 0.7172420272508857,
+      "learning_rate": 6.347715600182582e-07,
+      "loss": 1.5597,
+      "step": 2993
+    },
+    {
+      "epoch": 0.20861930808626275,
+      "grad_norm": 0.7716121956224239,
+      "learning_rate": 6.347295372686129e-07,
+      "loss": 1.574,
+      "step": 2994
+    },
+    {
+      "epoch": 0.20868898721388007,
+      "grad_norm": 0.6863157372578493,
+      "learning_rate": 6.346875025514157e-07,
+      "loss": 1.5149,
+      "step": 2995
+    },
+    {
+      "epoch": 0.2087586663414974,
+      "grad_norm": 0.7010956595448394,
+      "learning_rate": 6.346454558686807e-07,
+      "loss": 1.491,
+      "step": 2996
+    },
+    {
+      "epoch": 0.20882834546911472,
+      "grad_norm": 0.7242536207410032,
+      "learning_rate": 6.34603397222423e-07,
+      "loss": 1.4427,
+      "step": 2997
+    },
+    {
+      "epoch": 0.20889802459673204,
+      "grad_norm": 0.7040783271824286,
+      "learning_rate": 6.345613266146581e-07,
+      "loss": 1.4716,
+      "step": 2998
+    },
+    {
+      "epoch": 0.20896770372434936,
+      "grad_norm": 0.7624020600820641,
+      "learning_rate": 6.34519244047402e-07,
+      "loss": 1.7078,
+      "step": 2999
+    },
+    {
+      "epoch": 0.20903738285196669,
+      "grad_norm": 0.7175278942824543,
+      "learning_rate": 6.344771495226715e-07,
+      "loss": 1.5841,
+      "step": 3000
+    },
+    {
+      "epoch": 0.209107061979584,
+      "grad_norm": 0.7130038274068573,
+      "learning_rate": 6.344350430424838e-07,
+      "loss": 1.5632,
+      "step": 3001
+    },
+    {
+      "epoch": 0.20917674110720133,
+      "grad_norm": 0.7453698321614705,
+      "learning_rate": 6.343929246088567e-07,
+      "loss": 1.6019,
+      "step": 3002
+    },
+    {
+      "epoch": 0.20924642023481865,
+      "grad_norm": 0.7151850979286338,
+      "learning_rate": 6.343507942238088e-07,
+      "loss": 1.6559,
+      "step": 3003
+    },
+    {
+      "epoch": 0.20931609936243598,
+      "grad_norm": 0.6936345747851341,
+      "learning_rate": 6.343086518893588e-07,
+      "loss": 1.4821,
+      "step": 3004
+    },
+    {
+      "epoch": 0.2093857784900533,
+      "grad_norm": 0.7956254510612235,
+      "learning_rate": 6.342664976075265e-07,
+      "loss": 1.5881,
+      "step": 3005
+    },
+    {
+      "epoch": 0.20945545761767062,
+      "grad_norm": 0.7429121066816823,
+      "learning_rate": 6.342243313803317e-07,
+      "loss": 1.5633,
+      "step": 3006
+    },
+    {
+      "epoch": 0.20952513674528794,
+      "grad_norm": 0.8037190129898167,
+      "learning_rate": 6.341821532097956e-07,
+      "loss": 1.5909,
+      "step": 3007
+    },
+    {
+      "epoch": 0.20959481587290527,
+      "grad_norm": 0.6477652783679634,
+      "learning_rate": 6.34139963097939e-07,
+      "loss": 1.488,
+      "step": 3008
+    },
+    {
+      "epoch": 0.2096644950005226,
+      "grad_norm": 0.7000979494433776,
+      "learning_rate": 6.340977610467839e-07,
+      "loss": 1.4087,
+      "step": 3009
+    },
+    {
+      "epoch": 0.2097341741281399,
+      "grad_norm": 0.7648185849056666,
+      "learning_rate": 6.340555470583526e-07,
+      "loss": 1.5636,
+      "step": 3010
+    },
+    {
+      "epoch": 0.20980385325575723,
+      "grad_norm": 0.6881245272099351,
+      "learning_rate": 6.340133211346684e-07,
+      "loss": 1.4254,
+      "step": 3011
+    },
+    {
+      "epoch": 0.20987353238337456,
+      "grad_norm": 0.7170054857436351,
+      "learning_rate": 6.339710832777545e-07,
+      "loss": 1.5242,
+      "step": 3012
+    },
+    {
+      "epoch": 0.20994321151099188,
+      "grad_norm": 0.7091403344071238,
+      "learning_rate": 6.339288334896352e-07,
+      "loss": 1.5227,
+      "step": 3013
+    },
+    {
+      "epoch": 0.2100128906386092,
+      "grad_norm": 0.7188824298263891,
+      "learning_rate": 6.338865717723351e-07,
+      "loss": 1.5106,
+      "step": 3014
+    },
+    {
+      "epoch": 0.21008256976622652,
+      "grad_norm": 0.7123102173208373,
+      "learning_rate": 6.338442981278796e-07,
+      "loss": 1.6579,
+      "step": 3015
+    },
+    {
+      "epoch": 0.21015224889384385,
+      "grad_norm": 0.7421057665043513,
+      "learning_rate": 6.338020125582944e-07,
+      "loss": 1.5459,
+      "step": 3016
+    },
+    {
+      "epoch": 0.21022192802146117,
+      "grad_norm": 0.726883902227082,
+      "learning_rate": 6.33759715065606e-07,
+      "loss": 1.493,
+      "step": 3017
+    },
+    {
+      "epoch": 0.2102916071490785,
+      "grad_norm": 0.7538909742546808,
+      "learning_rate": 6.337174056518413e-07,
+      "loss": 1.6548,
+      "step": 3018
+    },
+    {
+      "epoch": 0.2103612862766958,
+      "grad_norm": 0.6874902631466682,
+      "learning_rate": 6.336750843190281e-07,
+      "loss": 1.5101,
+      "step": 3019
+    },
+    {
+      "epoch": 0.21043096540431314,
+      "grad_norm": 0.6991743258217981,
+      "learning_rate": 6.336327510691941e-07,
+      "loss": 1.4517,
+      "step": 3020
+    },
+    {
+      "epoch": 0.21050064453193046,
+      "grad_norm": 0.7216212073939949,
+      "learning_rate": 6.335904059043684e-07,
+      "loss": 1.5649,
+      "step": 3021
+    },
+    {
+      "epoch": 0.21057032365954778,
+      "grad_norm": 0.6953455603901181,
+      "learning_rate": 6.335480488265801e-07,
+      "loss": 1.5648,
+      "step": 3022
+    },
+    {
+      "epoch": 0.2106400027871651,
+      "grad_norm": 0.7269944245240746,
+      "learning_rate": 6.33505679837859e-07,
+      "loss": 1.5551,
+      "step": 3023
+    },
+    {
+      "epoch": 0.21070968191478243,
+      "grad_norm": 0.7521226273898199,
+      "learning_rate": 6.334632989402356e-07,
+      "loss": 1.5345,
+      "step": 3024
+    },
+    {
+      "epoch": 0.21077936104239975,
+      "grad_norm": 0.7288849333564393,
+      "learning_rate": 6.334209061357407e-07,
+      "loss": 1.5542,
+      "step": 3025
+    },
+    {
+      "epoch": 0.21084904017001707,
+      "grad_norm": 0.723684896541463,
+      "learning_rate": 6.333785014264061e-07,
+      "loss": 1.4942,
+      "step": 3026
+    },
+    {
+      "epoch": 0.2109187192976344,
+      "grad_norm": 0.8354732057230474,
+      "learning_rate": 6.333360848142637e-07,
+      "loss": 1.5765,
+      "step": 3027
+    },
+    {
+      "epoch": 0.21098839842525172,
+      "grad_norm": 0.7266486063179814,
+      "learning_rate": 6.332936563013465e-07,
+      "loss": 1.5522,
+      "step": 3028
+    },
+    {
+      "epoch": 0.21105807755286904,
+      "grad_norm": 0.7620578829901353,
+      "learning_rate": 6.332512158896873e-07,
+      "loss": 1.6742,
+      "step": 3029
+    },
+    {
+      "epoch": 0.21112775668048636,
+      "grad_norm": 0.6453483203498609,
+      "learning_rate": 6.332087635813202e-07,
+      "loss": 1.4459,
+      "step": 3030
+    },
+    {
+      "epoch": 0.21119743580810368,
+      "grad_norm": 0.7671250472225403,
+      "learning_rate": 6.331662993782797e-07,
+      "loss": 1.608,
+      "step": 3031
+    },
+    {
+      "epoch": 0.211267114935721,
+      "grad_norm": 0.7297600261040783,
+      "learning_rate": 6.331238232826006e-07,
+      "loss": 1.5218,
+      "step": 3032
+    },
+    {
+      "epoch": 0.21133679406333833,
+      "grad_norm": 0.6976240359205483,
+      "learning_rate": 6.330813352963186e-07,
+      "loss": 1.5028,
+      "step": 3033
+    },
+    {
+      "epoch": 0.21140647319095565,
+      "grad_norm": 0.7366466923541604,
+      "learning_rate": 6.330388354214697e-07,
+      "loss": 1.5963,
+      "step": 3034
+    },
+    {
+      "epoch": 0.21147615231857297,
+      "grad_norm": 0.743045908625905,
+      "learning_rate": 6.329963236600905e-07,
+      "loss": 1.5822,
+      "step": 3035
+    },
+    {
+      "epoch": 0.2115458314461903,
+      "grad_norm": 0.7694559232108855,
+      "learning_rate": 6.329538000142183e-07,
+      "loss": 1.5745,
+      "step": 3036
+    },
+    {
+      "epoch": 0.21161551057380762,
+      "grad_norm": 0.7136032068936781,
+      "learning_rate": 6.329112644858911e-07,
+      "loss": 1.6823,
+      "step": 3037
+    },
+    {
+      "epoch": 0.21168518970142494,
+      "grad_norm": 0.7209064494933916,
+      "learning_rate": 6.328687170771472e-07,
+      "loss": 1.5415,
+      "step": 3038
+    },
+    {
+      "epoch": 0.21175486882904226,
+      "grad_norm": 0.7601206403956654,
+      "learning_rate": 6.328261577900253e-07,
+      "loss": 1.531,
+      "step": 3039
+    },
+    {
+      "epoch": 0.2118245479566596,
+      "grad_norm": 0.6959998472222233,
+      "learning_rate": 6.327835866265652e-07,
+      "loss": 1.4477,
+      "step": 3040
+    },
+    {
+      "epoch": 0.2118942270842769,
+      "grad_norm": 0.6953710184416528,
+      "learning_rate": 6.327410035888071e-07,
+      "loss": 1.58,
+      "step": 3041
+    },
+    {
+      "epoch": 0.21196390621189423,
+      "grad_norm": 0.6943533580699583,
+      "learning_rate": 6.326984086787913e-07,
+      "loss": 1.5223,
+      "step": 3042
+    },
+    {
+      "epoch": 0.21203358533951155,
+      "grad_norm": 0.7408013816517577,
+      "learning_rate": 6.326558018985593e-07,
+      "loss": 1.5385,
+      "step": 3043
+    },
+    {
+      "epoch": 0.21210326446712888,
+      "grad_norm": 0.7128173593393061,
+      "learning_rate": 6.32613183250153e-07,
+      "loss": 1.5008,
+      "step": 3044
+    },
+    {
+      "epoch": 0.2121729435947462,
+      "grad_norm": 0.7578937862434474,
+      "learning_rate": 6.325705527356144e-07,
+      "loss": 1.5314,
+      "step": 3045
+    },
+    {
+      "epoch": 0.21224262272236352,
+      "grad_norm": 0.7242230650402078,
+      "learning_rate": 6.325279103569868e-07,
+      "loss": 1.5105,
+      "step": 3046
+    },
+    {
+      "epoch": 0.21231230184998084,
+      "grad_norm": 0.7271388530049001,
+      "learning_rate": 6.324852561163136e-07,
+      "loss": 1.56,
+      "step": 3047
+    },
+    {
+      "epoch": 0.21238198097759817,
+      "grad_norm": 0.7251757959446895,
+      "learning_rate": 6.324425900156387e-07,
+      "loss": 1.6189,
+      "step": 3048
+    },
+    {
+      "epoch": 0.2124516601052155,
+      "grad_norm": 0.6949230910117429,
+      "learning_rate": 6.32399912057007e-07,
+      "loss": 1.4936,
+      "step": 3049
+    },
+    {
+      "epoch": 0.2125213392328328,
+      "grad_norm": 0.7083185802548545,
+      "learning_rate": 6.323572222424636e-07,
+      "loss": 1.6193,
+      "step": 3050
+    },
+    {
+      "epoch": 0.21259101836045013,
+      "grad_norm": 0.7188362586038636,
+      "learning_rate": 6.323145205740543e-07,
+      "loss": 1.5033,
+      "step": 3051
+    },
+    {
+      "epoch": 0.21266069748806746,
+      "grad_norm": 0.6807201669919434,
+      "learning_rate": 6.322718070538254e-07,
+      "loss": 1.4721,
+      "step": 3052
+    },
+    {
+      "epoch": 0.21273037661568478,
+      "grad_norm": 0.7290543207491141,
+      "learning_rate": 6.32229081683824e-07,
+      "loss": 1.464,
+      "step": 3053
+    },
+    {
+      "epoch": 0.2128000557433021,
+      "grad_norm": 0.7049188321594511,
+      "learning_rate": 6.321863444660972e-07,
+      "loss": 1.5481,
+      "step": 3054
+    },
+    {
+      "epoch": 0.21286973487091942,
+      "grad_norm": 0.7531590172268701,
+      "learning_rate": 6.321435954026935e-07,
+      "loss": 1.5502,
+      "step": 3055
+    },
+    {
+      "epoch": 0.21293941399853675,
+      "grad_norm": 0.7204791442591438,
+      "learning_rate": 6.321008344956612e-07,
+      "loss": 1.5385,
+      "step": 3056
+    },
+    {
+      "epoch": 0.21300909312615407,
+      "grad_norm": 0.766737235051107,
+      "learning_rate": 6.320580617470498e-07,
+      "loss": 1.5535,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2130787722537714,
+      "grad_norm": 0.6949576898262453,
+      "learning_rate": 6.320152771589089e-07,
+      "loss": 1.5787,
+      "step": 3058
+    },
+    {
+      "epoch": 0.21314845138138871,
+      "grad_norm": 0.7854280947188981,
+      "learning_rate": 6.319724807332886e-07,
+      "loss": 1.5751,
+      "step": 3059
+    },
+    {
+      "epoch": 0.21321813050900604,
+      "grad_norm": 0.7310738000268694,
+      "learning_rate": 6.319296724722401e-07,
+      "loss": 1.4912,
+      "step": 3060
+    },
+    {
+      "epoch": 0.21328780963662336,
+      "grad_norm": 0.710144969632082,
+      "learning_rate": 6.318868523778147e-07,
+      "loss": 1.501,
+      "step": 3061
+    },
+    {
+      "epoch": 0.21335748876424068,
+      "grad_norm": 0.7624755707474923,
+      "learning_rate": 6.318440204520646e-07,
+      "loss": 1.5925,
+      "step": 3062
+    },
+    {
+      "epoch": 0.213427167891858,
+      "grad_norm": 0.6779707671850002,
+      "learning_rate": 6.318011766970422e-07,
+      "loss": 1.566,
+      "step": 3063
+    },
+    {
+      "epoch": 0.21349684701947533,
+      "grad_norm": 0.7282347315793222,
+      "learning_rate": 6.317583211148007e-07,
+      "loss": 1.637,
+      "step": 3064
+    },
+    {
+      "epoch": 0.21356652614709265,
+      "grad_norm": 0.6777857214243174,
+      "learning_rate": 6.31715453707394e-07,
+      "loss": 1.5141,
+      "step": 3065
+    },
+    {
+      "epoch": 0.21363620527470997,
+      "grad_norm": 0.7409112879146345,
+      "learning_rate": 6.316725744768762e-07,
+      "loss": 1.5923,
+      "step": 3066
+    },
+    {
+      "epoch": 0.2137058844023273,
+      "grad_norm": 0.7024204744147935,
+      "learning_rate": 6.316296834253022e-07,
+      "loss": 1.5464,
+      "step": 3067
+    },
+    {
+      "epoch": 0.21377556352994462,
+      "grad_norm": 0.6877856023309812,
+      "learning_rate": 6.315867805547275e-07,
+      "loss": 1.5618,
+      "step": 3068
+    },
+    {
+      "epoch": 0.21384524265756194,
+      "grad_norm": 0.7212486356758446,
+      "learning_rate": 6.31543865867208e-07,
+      "loss": 1.5316,
+      "step": 3069
+    },
+    {
+      "epoch": 0.21391492178517926,
+      "grad_norm": 0.776152740751751,
+      "learning_rate": 6.315009393648003e-07,
+      "loss": 1.6051,
+      "step": 3070
+    },
+    {
+      "epoch": 0.21398460091279659,
+      "grad_norm": 0.7071509959023832,
+      "learning_rate": 6.314580010495615e-07,
+      "loss": 1.5317,
+      "step": 3071
+    },
+    {
+      "epoch": 0.2140542800404139,
+      "grad_norm": 0.6884549731762428,
+      "learning_rate": 6.314150509235494e-07,
+      "loss": 1.4914,
+      "step": 3072
+    },
+    {
+      "epoch": 0.2141239591680312,
+      "grad_norm": 0.7408015065712721,
+      "learning_rate": 6.313720889888221e-07,
+      "loss": 1.5193,
+      "step": 3073
+    },
+    {
+      "epoch": 0.21419363829564853,
+      "grad_norm": 0.7518489520556682,
+      "learning_rate": 6.313291152474385e-07,
+      "loss": 1.5984,
+      "step": 3074
+    },
+    {
+      "epoch": 0.21426331742326585,
+      "grad_norm": 0.6931740419432002,
+      "learning_rate": 6.312861297014581e-07,
+      "loss": 1.5873,
+      "step": 3075
+    },
+    {
+      "epoch": 0.21433299655088317,
+      "grad_norm": 0.7460738164111779,
+      "learning_rate": 6.312431323529407e-07,
+      "loss": 1.5159,
+      "step": 3076
+    },
+    {
+      "epoch": 0.2144026756785005,
+      "grad_norm": 0.7066855472272406,
+      "learning_rate": 6.312001232039468e-07,
+      "loss": 1.4636,
+      "step": 3077
+    },
+    {
+      "epoch": 0.21447235480611782,
+      "grad_norm": 0.7111733820215189,
+      "learning_rate": 6.311571022565376e-07,
+      "loss": 1.6072,
+      "step": 3078
+    },
+    {
+      "epoch": 0.21454203393373514,
+      "grad_norm": 0.7146743692003119,
+      "learning_rate": 6.311140695127748e-07,
+      "loss": 1.5797,
+      "step": 3079
+    },
+    {
+      "epoch": 0.21461171306135246,
+      "grad_norm": 0.7428280015586619,
+      "learning_rate": 6.310710249747204e-07,
+      "loss": 1.5731,
+      "step": 3080
+    },
+    {
+      "epoch": 0.21468139218896978,
+      "grad_norm": 0.7142132227243223,
+      "learning_rate": 6.310279686444374e-07,
+      "loss": 1.5772,
+      "step": 3081
+    },
+    {
+      "epoch": 0.2147510713165871,
+      "grad_norm": 0.7367688211537811,
+      "learning_rate": 6.30984900523989e-07,
+      "loss": 1.5302,
+      "step": 3082
+    },
+    {
+      "epoch": 0.21482075044420443,
+      "grad_norm": 0.7051161284853323,
+      "learning_rate": 6.309418206154392e-07,
+      "loss": 1.3608,
+      "step": 3083
+    },
+    {
+      "epoch": 0.21489042957182175,
+      "grad_norm": 0.6561827725117855,
+      "learning_rate": 6.308987289208523e-07,
+      "loss": 1.442,
+      "step": 3084
+    },
+    {
+      "epoch": 0.21496010869943907,
+      "grad_norm": 0.7060953414362706,
+      "learning_rate": 6.308556254422937e-07,
+      "loss": 1.5811,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2150297878270564,
+      "grad_norm": 0.712986152298592,
+      "learning_rate": 6.308125101818287e-07,
+      "loss": 1.4907,
+      "step": 3086
+    },
+    {
+      "epoch": 0.21509946695467372,
+      "grad_norm": 0.7064297350621125,
+      "learning_rate": 6.307693831415236e-07,
+      "loss": 1.4074,
+      "step": 3087
+    },
+    {
+      "epoch": 0.21516914608229104,
+      "grad_norm": 0.6978571997252455,
+      "learning_rate": 6.307262443234451e-07,
+      "loss": 1.5763,
+      "step": 3088
+    },
+    {
+      "epoch": 0.21523882520990836,
+      "grad_norm": 0.7204214153527864,
+      "learning_rate": 6.306830937296605e-07,
+      "loss": 1.5953,
+      "step": 3089
+    },
+    {
+      "epoch": 0.21530850433752569,
+      "grad_norm": 0.7251712825238252,
+      "learning_rate": 6.306399313622376e-07,
+      "loss": 1.4922,
+      "step": 3090
+    },
+    {
+      "epoch": 0.215378183465143,
+      "grad_norm": 0.789317477528938,
+      "learning_rate": 6.30596757223245e-07,
+      "loss": 1.5323,
+      "step": 3091
+    },
+    {
+      "epoch": 0.21544786259276033,
+      "grad_norm": 0.7018313808348219,
+      "learning_rate": 6.305535713147515e-07,
+      "loss": 1.5525,
+      "step": 3092
+    },
+    {
+      "epoch": 0.21551754172037765,
+      "grad_norm": 0.6641796971823418,
+      "learning_rate": 6.305103736388268e-07,
+      "loss": 1.4296,
+      "step": 3093
+    },
+    {
+      "epoch": 0.21558722084799498,
+      "grad_norm": 0.7333376762555791,
+      "learning_rate": 6.30467164197541e-07,
+      "loss": 1.5683,
+      "step": 3094
+    },
+    {
+      "epoch": 0.2156568999756123,
+      "grad_norm": 0.715162682097904,
+      "learning_rate": 6.304239429929647e-07,
+      "loss": 1.5447,
+      "step": 3095
+    },
+    {
+      "epoch": 0.21572657910322962,
+      "grad_norm": 0.705125764957599,
+      "learning_rate": 6.303807100271694e-07,
+      "loss": 1.5388,
+      "step": 3096
+    },
+    {
+      "epoch": 0.21579625823084694,
+      "grad_norm": 0.6932910886448272,
+      "learning_rate": 6.303374653022264e-07,
+      "loss": 1.3299,
+      "step": 3097
+    },
+    {
+      "epoch": 0.21586593735846427,
+      "grad_norm": 0.7285862258034365,
+      "learning_rate": 6.302942088202086e-07,
+      "loss": 1.5698,
+      "step": 3098
+    },
+    {
+      "epoch": 0.2159356164860816,
+      "grad_norm": 0.6525649404815276,
+      "learning_rate": 6.302509405831887e-07,
+      "loss": 1.4322,
+      "step": 3099
+    },
+    {
+      "epoch": 0.2160052956136989,
+      "grad_norm": 0.7069636227599213,
+      "learning_rate": 6.302076605932402e-07,
+      "loss": 1.5997,
+      "step": 3100
+    },
+    {
+      "epoch": 0.21607497474131623,
+      "grad_norm": 0.7333395774155352,
+      "learning_rate": 6.301643688524372e-07,
+      "loss": 1.5292,
+      "step": 3101
+    },
+    {
+      "epoch": 0.21614465386893356,
+      "grad_norm": 0.7868825512640599,
+      "learning_rate": 6.301210653628545e-07,
+      "loss": 1.5164,
+      "step": 3102
+    },
+    {
+      "epoch": 0.21621433299655088,
+      "grad_norm": 0.6477530449996382,
+      "learning_rate": 6.300777501265669e-07,
+      "loss": 1.4932,
+      "step": 3103
+    },
+    {
+      "epoch": 0.2162840121241682,
+      "grad_norm": 0.7382703043519941,
+      "learning_rate": 6.300344231456505e-07,
+      "loss": 1.5096,
+      "step": 3104
+    },
+    {
+      "epoch": 0.21635369125178552,
+      "grad_norm": 0.7197826816393428,
+      "learning_rate": 6.299910844221815e-07,
+      "loss": 1.5127,
+      "step": 3105
+    },
+    {
+      "epoch": 0.21642337037940285,
+      "grad_norm": 0.7162601287647885,
+      "learning_rate": 6.299477339582367e-07,
+      "loss": 1.5613,
+      "step": 3106
+    },
+    {
+      "epoch": 0.21649304950702017,
+      "grad_norm": 0.6993308172485544,
+      "learning_rate": 6.299043717558937e-07,
+      "loss": 1.3961,
+      "step": 3107
+    },
+    {
+      "epoch": 0.2165627286346375,
+      "grad_norm": 0.7050100563001422,
+      "learning_rate": 6.298609978172303e-07,
+      "loss": 1.522,
+      "step": 3108
+    },
+    {
+      "epoch": 0.2166324077622548,
+      "grad_norm": 0.7030621104722599,
+      "learning_rate": 6.298176121443252e-07,
+      "loss": 1.5093,
+      "step": 3109
+    },
+    {
+      "epoch": 0.21670208688987214,
+      "grad_norm": 0.6658177312948771,
+      "learning_rate": 6.297742147392577e-07,
+      "loss": 1.5186,
+      "step": 3110
+    },
+    {
+      "epoch": 0.21677176601748946,
+      "grad_norm": 0.6695545637176513,
+      "learning_rate": 6.297308056041072e-07,
+      "loss": 1.4912,
+      "step": 3111
+    },
+    {
+      "epoch": 0.21684144514510678,
+      "grad_norm": 0.6955784740818475,
+      "learning_rate": 6.296873847409542e-07,
+      "loss": 1.5496,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2169111242727241,
+      "grad_norm": 0.6878524383136772,
+      "learning_rate": 6.296439521518792e-07,
+      "loss": 1.5043,
+      "step": 3113
+    },
+    {
+      "epoch": 0.21698080340034143,
+      "grad_norm": 0.7220128624196103,
+      "learning_rate": 6.296005078389639e-07,
+      "loss": 1.5356,
+      "step": 3114
+    },
+    {
+      "epoch": 0.21705048252795875,
+      "grad_norm": 0.6693205323609417,
+      "learning_rate": 6.2955705180429e-07,
+      "loss": 1.3807,
+      "step": 3115
+    },
+    {
+      "epoch": 0.21712016165557607,
+      "grad_norm": 0.68769141490313,
+      "learning_rate": 6.295135840499401e-07,
+      "loss": 1.6205,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2171898407831934,
+      "grad_norm": 0.7197082361841207,
+      "learning_rate": 6.294701045779974e-07,
+      "loss": 1.4412,
+      "step": 3117
+    },
+    {
+      "epoch": 0.21725951991081072,
+      "grad_norm": 0.7310208796664356,
+      "learning_rate": 6.294266133905453e-07,
+      "loss": 1.612,
+      "step": 3118
+    },
+    {
+      "epoch": 0.21732919903842804,
+      "grad_norm": 0.7665599518177661,
+      "learning_rate": 6.293831104896682e-07,
+      "loss": 1.5117,
+      "step": 3119
+    },
+    {
+      "epoch": 0.21739887816604536,
+      "grad_norm": 0.7444580585825434,
+      "learning_rate": 6.293395958774507e-07,
+      "loss": 1.6397,
+      "step": 3120
+    },
+    {
+      "epoch": 0.21746855729366268,
+      "grad_norm": 0.7523077640557353,
+      "learning_rate": 6.29296069555978e-07,
+      "loss": 1.5666,
+      "step": 3121
+    },
+    {
+      "epoch": 0.21753823642128,
+      "grad_norm": 0.7124509029071153,
+      "learning_rate": 6.292525315273363e-07,
+      "loss": 1.3989,
+      "step": 3122
+    },
+    {
+      "epoch": 0.21760791554889733,
+      "grad_norm": 0.6772900926590302,
+      "learning_rate": 6.292089817936119e-07,
+      "loss": 1.4556,
+      "step": 3123
+    },
+    {
+      "epoch": 0.21767759467651465,
+      "grad_norm": 0.6820042166343526,
+      "learning_rate": 6.291654203568915e-07,
+      "loss": 1.4178,
+      "step": 3124
+    },
+    {
+      "epoch": 0.21774727380413197,
+      "grad_norm": 0.7656657737383707,
+      "learning_rate": 6.29121847219263e-07,
+      "loss": 1.5057,
+      "step": 3125
+    },
+    {
+      "epoch": 0.2178169529317493,
+      "grad_norm": 0.7702721728012195,
+      "learning_rate": 6.290782623828146e-07,
+      "loss": 1.7169,
+      "step": 3126
+    },
+    {
+      "epoch": 0.21788663205936662,
+      "grad_norm": 0.7139401052066671,
+      "learning_rate": 6.290346658496345e-07,
+      "loss": 1.4959,
+      "step": 3127
+    },
+    {
+      "epoch": 0.21795631118698394,
+      "grad_norm": 0.6650157900893546,
+      "learning_rate": 6.289910576218124e-07,
+      "loss": 1.5168,
+      "step": 3128
+    },
+    {
+      "epoch": 0.21802599031460126,
+      "grad_norm": 0.7229572770264344,
+      "learning_rate": 6.289474377014378e-07,
+      "loss": 1.5193,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2180956694422186,
+      "grad_norm": 0.7516357460383902,
+      "learning_rate": 6.289038060906011e-07,
+      "loss": 1.6042,
+      "step": 3130
+    },
+    {
+      "epoch": 0.2181653485698359,
+      "grad_norm": 0.6990707991385572,
+      "learning_rate": 6.288601627913935e-07,
+      "loss": 1.5118,
+      "step": 3131
+    },
+    {
+      "epoch": 0.21823502769745323,
+      "grad_norm": 0.7402279645798079,
+      "learning_rate": 6.288165078059062e-07,
+      "loss": 1.5159,
+      "step": 3132
+    },
+    {
+      "epoch": 0.21830470682507055,
+      "grad_norm": 0.7071125829282526,
+      "learning_rate": 6.287728411362312e-07,
+      "loss": 1.4857,
+      "step": 3133
+    },
+    {
+      "epoch": 0.21837438595268788,
+      "grad_norm": 0.7145022706649667,
+      "learning_rate": 6.287291627844613e-07,
+      "loss": 1.4423,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2184440650803052,
+      "grad_norm": 0.7153095213061711,
+      "learning_rate": 6.286854727526895e-07,
+      "loss": 1.5401,
+      "step": 3135
+    },
+    {
+      "epoch": 0.21851374420792252,
+      "grad_norm": 0.7388320724188436,
+      "learning_rate": 6.286417710430096e-07,
+      "loss": 1.5551,
+      "step": 3136
+    },
+    {
+      "epoch": 0.21858342333553984,
+      "grad_norm": 0.7351605398058645,
+      "learning_rate": 6.285980576575158e-07,
+      "loss": 1.5743,
+      "step": 3137
+    },
+    {
+      "epoch": 0.21865310246315717,
+      "grad_norm": 0.7103834284273812,
+      "learning_rate": 6.28554332598303e-07,
+      "loss": 1.62,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2187227815907745,
+      "grad_norm": 0.7124917000500621,
+      "learning_rate": 6.285105958674667e-07,
+      "loss": 1.5462,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2187924607183918,
+      "grad_norm": 0.6719455946928833,
+      "learning_rate": 6.284668474671026e-07,
+      "loss": 1.4454,
+      "step": 3140
+    },
+    {
+      "epoch": 0.21886213984600913,
+      "grad_norm": 0.7342723618991475,
+      "learning_rate": 6.284230873993073e-07,
+      "loss": 1.5297,
+      "step": 3141
+    },
+    {
+      "epoch": 0.21893181897362646,
+      "grad_norm": 0.7050821767435831,
+      "learning_rate": 6.28379315666178e-07,
+      "loss": 1.4949,
+      "step": 3142
+    },
+    {
+      "epoch": 0.21900149810124378,
+      "grad_norm": 0.7305929985653432,
+      "learning_rate": 6.283355322698121e-07,
+      "loss": 1.5449,
+      "step": 3143
+    },
+    {
+      "epoch": 0.2190711772288611,
+      "grad_norm": 0.7355604484531012,
+      "learning_rate": 6.282917372123081e-07,
+      "loss": 1.631,
+      "step": 3144
+    },
+    {
+      "epoch": 0.21914085635647843,
+      "grad_norm": 0.6723855105370937,
+      "learning_rate": 6.282479304957646e-07,
+      "loss": 1.521,
+      "step": 3145
+    },
+    {
+      "epoch": 0.21921053548409575,
+      "grad_norm": 0.7500049197897934,
+      "learning_rate": 6.282041121222808e-07,
+      "loss": 1.4776,
+      "step": 3146
+    },
+    {
+      "epoch": 0.21928021461171307,
+      "grad_norm": 0.705067231005872,
+      "learning_rate": 6.281602820939566e-07,
+      "loss": 1.4577,
+      "step": 3147
+    },
+    {
+      "epoch": 0.2193498937393304,
+      "grad_norm": 0.7447843195035967,
+      "learning_rate": 6.281164404128927e-07,
+      "loss": 1.464,
+      "step": 3148
+    },
+    {
+      "epoch": 0.21941957286694772,
+      "grad_norm": 0.6669666094624056,
+      "learning_rate": 6.280725870811896e-07,
+      "loss": 1.4811,
+      "step": 3149
+    },
+    {
+      "epoch": 0.21948925199456504,
+      "grad_norm": 0.7372591631824635,
+      "learning_rate": 6.280287221009493e-07,
+      "loss": 1.4874,
+      "step": 3150
+    },
+    {
+      "epoch": 0.21955893112218236,
+      "grad_norm": 0.7400195760354873,
+      "learning_rate": 6.279848454742736e-07,
+      "loss": 1.5233,
+      "step": 3151
+    },
+    {
+      "epoch": 0.21962861024979968,
+      "grad_norm": 0.7100152026287823,
+      "learning_rate": 6.279409572032653e-07,
+      "loss": 1.5391,
+      "step": 3152
+    },
+    {
+      "epoch": 0.219698289377417,
+      "grad_norm": 0.6998698777309963,
+      "learning_rate": 6.278970572900277e-07,
+      "loss": 1.5045,
+      "step": 3153
+    },
+    {
+      "epoch": 0.21976796850503433,
+      "grad_norm": 0.8130378158487237,
+      "learning_rate": 6.278531457366644e-07,
+      "loss": 1.5623,
+      "step": 3154
+    },
+    {
+      "epoch": 0.21983764763265165,
+      "grad_norm": 0.7189054680097751,
+      "learning_rate": 6.278092225452796e-07,
+      "loss": 1.4549,
+      "step": 3155
+    },
+    {
+      "epoch": 0.21990732676026897,
+      "grad_norm": 0.7060714272932562,
+      "learning_rate": 6.277652877179786e-07,
+      "loss": 1.6297,
+      "step": 3156
+    },
+    {
+      "epoch": 0.2199770058878863,
+      "grad_norm": 0.7851668596549601,
+      "learning_rate": 6.277213412568665e-07,
+      "loss": 1.5901,
+      "step": 3157
+    },
+    {
+      "epoch": 0.22004668501550362,
+      "grad_norm": 0.721129749016274,
+      "learning_rate": 6.276773831640495e-07,
+      "loss": 1.5617,
+      "step": 3158
+    },
+    {
+      "epoch": 0.22011636414312094,
+      "grad_norm": 0.7291404792430866,
+      "learning_rate": 6.276334134416341e-07,
+      "loss": 1.6845,
+      "step": 3159
+    },
+    {
+      "epoch": 0.22018604327073826,
+      "grad_norm": 0.6621504202587892,
+      "learning_rate": 6.275894320917273e-07,
+      "loss": 1.5213,
+      "step": 3160
+    },
+    {
+      "epoch": 0.22025572239835559,
+      "grad_norm": 0.7520204029846614,
+      "learning_rate": 6.27545439116437e-07,
+      "loss": 1.5872,
+      "step": 3161
+    },
+    {
+      "epoch": 0.2203254015259729,
+      "grad_norm": 0.7445862369015398,
+      "learning_rate": 6.275014345178713e-07,
+      "loss": 1.5763,
+      "step": 3162
+    },
+    {
+      "epoch": 0.22039508065359023,
+      "grad_norm": 0.7290777915425307,
+      "learning_rate": 6.27457418298139e-07,
+      "loss": 1.6284,
+      "step": 3163
+    },
+    {
+      "epoch": 0.22046475978120753,
+      "grad_norm": 0.752459227931593,
+      "learning_rate": 6.274133904593496e-07,
+      "loss": 1.6027,
+      "step": 3164
+    },
+    {
+      "epoch": 0.22053443890882485,
+      "grad_norm": 0.6618546415584925,
+      "learning_rate": 6.273693510036126e-07,
+      "loss": 1.5229,
+      "step": 3165
+    },
+    {
+      "epoch": 0.22060411803644217,
+      "grad_norm": 0.6864959676253173,
+      "learning_rate": 6.273252999330389e-07,
+      "loss": 1.6194,
+      "step": 3166
+    },
+    {
+      "epoch": 0.2206737971640595,
+      "grad_norm": 0.8145056745835206,
+      "learning_rate": 6.272812372497392e-07,
+      "loss": 1.3994,
+      "step": 3167
+    },
+    {
+      "epoch": 0.22074347629167682,
+      "grad_norm": 0.7224351871692737,
+      "learning_rate": 6.272371629558254e-07,
+      "loss": 1.6291,
+      "step": 3168
+    },
+    {
+      "epoch": 0.22081315541929414,
+      "grad_norm": 0.752897078662678,
+      "learning_rate": 6.271930770534093e-07,
+      "loss": 1.6119,
+      "step": 3169
+    },
+    {
+      "epoch": 0.22088283454691146,
+      "grad_norm": 0.70856599701562,
+      "learning_rate": 6.271489795446038e-07,
+      "loss": 1.6155,
+      "step": 3170
+    },
+    {
+      "epoch": 0.22095251367452878,
+      "grad_norm": 0.755885503203361,
+      "learning_rate": 6.27104870431522e-07,
+      "loss": 1.6431,
+      "step": 3171
+    },
+    {
+      "epoch": 0.2210221928021461,
+      "grad_norm": 0.7865730300687626,
+      "learning_rate": 6.27060749716278e-07,
+      "loss": 1.4305,
+      "step": 3172
+    },
+    {
+      "epoch": 0.22109187192976343,
+      "grad_norm": 0.7013495707737089,
+      "learning_rate": 6.270166174009857e-07,
+      "loss": 1.4782,
+      "step": 3173
+    },
+    {
+      "epoch": 0.22116155105738075,
+      "grad_norm": 0.6739764223939262,
+      "learning_rate": 6.269724734877604e-07,
+      "loss": 1.5228,
+      "step": 3174
+    },
+    {
+      "epoch": 0.22123123018499807,
+      "grad_norm": 0.7092554794529752,
+      "learning_rate": 6.269283179787171e-07,
+      "loss": 1.5998,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2213009093126154,
+      "grad_norm": 0.68071132595703,
+      "learning_rate": 6.268841508759725e-07,
+      "loss": 1.5796,
+      "step": 3176
+    },
+    {
+      "epoch": 0.22137058844023272,
+      "grad_norm": 0.6705866455284171,
+      "learning_rate": 6.268399721816427e-07,
+      "loss": 1.5333,
+      "step": 3177
+    },
+    {
+      "epoch": 0.22144026756785004,
+      "grad_norm": 0.7316307681953858,
+      "learning_rate": 6.267957818978449e-07,
+      "loss": 1.5732,
+      "step": 3178
+    },
+    {
+      "epoch": 0.22150994669546736,
+      "grad_norm": 0.7036388021018205,
+      "learning_rate": 6.267515800266969e-07,
+      "loss": 1.4423,
+      "step": 3179
+    },
+    {
+      "epoch": 0.22157962582308469,
+      "grad_norm": 0.701486644574639,
+      "learning_rate": 6.267073665703168e-07,
+      "loss": 1.5843,
+      "step": 3180
+    },
+    {
+      "epoch": 0.221649304950702,
+      "grad_norm": 0.7315723398194484,
+      "learning_rate": 6.266631415308236e-07,
+      "loss": 1.5913,
+      "step": 3181
+    },
+    {
+      "epoch": 0.22171898407831933,
+      "grad_norm": 0.7610730029702969,
+      "learning_rate": 6.266189049103364e-07,
+      "loss": 1.5173,
+      "step": 3182
+    },
+    {
+      "epoch": 0.22178866320593665,
+      "grad_norm": 0.7267234976534973,
+      "learning_rate": 6.265746567109752e-07,
+      "loss": 1.5464,
+      "step": 3183
+    },
+    {
+      "epoch": 0.22185834233355398,
+      "grad_norm": 0.7581939561966579,
+      "learning_rate": 6.265303969348606e-07,
+      "loss": 1.5121,
+      "step": 3184
+    },
+    {
+      "epoch": 0.2219280214611713,
+      "grad_norm": 0.7818869908905624,
+      "learning_rate": 6.264861255841136e-07,
+      "loss": 1.6076,
+      "step": 3185
+    },
+    {
+      "epoch": 0.22199770058878862,
+      "grad_norm": 0.7334237426975201,
+      "learning_rate": 6.264418426608556e-07,
+      "loss": 1.5393,
+      "step": 3186
+    },
+    {
+      "epoch": 0.22206737971640594,
+      "grad_norm": 0.666609199162287,
+      "learning_rate": 6.263975481672088e-07,
+      "loss": 1.4628,
+      "step": 3187
+    },
+    {
+      "epoch": 0.22213705884402327,
+      "grad_norm": 0.7218983139497597,
+      "learning_rate": 6.26353242105296e-07,
+      "loss": 1.5396,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2222067379716406,
+      "grad_norm": 0.7243553909759005,
+      "learning_rate": 6.263089244772403e-07,
+      "loss": 1.4566,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2222764170992579,
+      "grad_norm": 0.7754149969874606,
+      "learning_rate": 6.262645952851656e-07,
+      "loss": 1.6057,
+      "step": 3190
+    },
+    {
+      "epoch": 0.22234609622687523,
+      "grad_norm": 0.7007605347212531,
+      "learning_rate": 6.262202545311961e-07,
+      "loss": 1.5546,
+      "step": 3191
+    },
+    {
+      "epoch": 0.22241577535449256,
+      "grad_norm": 0.7193548432527994,
+      "learning_rate": 6.261759022174569e-07,
+      "loss": 1.5214,
+      "step": 3192
+    },
+    {
+      "epoch": 0.22248545448210988,
+      "grad_norm": 0.7430888845944795,
+      "learning_rate": 6.261315383460732e-07,
+      "loss": 1.5679,
+      "step": 3193
+    },
+    {
+      "epoch": 0.2225551336097272,
+      "grad_norm": 0.7067278834368642,
+      "learning_rate": 6.260871629191712e-07,
+      "loss": 1.6559,
+      "step": 3194
+    },
+    {
+      "epoch": 0.22262481273734452,
+      "grad_norm": 0.679596178691687,
+      "learning_rate": 6.260427759388774e-07,
+      "loss": 1.4959,
+      "step": 3195
+    },
+    {
+      "epoch": 0.22269449186496185,
+      "grad_norm": 0.736290631526461,
+      "learning_rate": 6.25998377407319e-07,
+      "loss": 1.4611,
+      "step": 3196
+    },
+    {
+      "epoch": 0.22276417099257917,
+      "grad_norm": 0.7230193809098943,
+      "learning_rate": 6.259539673266236e-07,
+      "loss": 1.5939,
+      "step": 3197
+    },
+    {
+      "epoch": 0.2228338501201965,
+      "grad_norm": 0.7508187483513579,
+      "learning_rate": 6.259095456989193e-07,
+      "loss": 1.5279,
+      "step": 3198
+    },
+    {
+      "epoch": 0.22290352924781381,
+      "grad_norm": 0.7190325739384879,
+      "learning_rate": 6.258651125263351e-07,
+      "loss": 1.5578,
+      "step": 3199
+    },
+    {
+      "epoch": 0.22297320837543114,
+      "grad_norm": 0.6868033846137466,
+      "learning_rate": 6.258206678110001e-07,
+      "loss": 1.5303,
+      "step": 3200
+    },
+    {
+      "epoch": 0.22304288750304846,
+      "grad_norm": 0.7103286386533224,
+      "learning_rate": 6.257762115550445e-07,
+      "loss": 1.5063,
+      "step": 3201
+    },
+    {
+      "epoch": 0.22311256663066578,
+      "grad_norm": 0.7159508152038254,
+      "learning_rate": 6.257317437605984e-07,
+      "loss": 1.5417,
+      "step": 3202
+    },
+    {
+      "epoch": 0.2231822457582831,
+      "grad_norm": 0.754030296647133,
+      "learning_rate": 6.256872644297928e-07,
+      "loss": 1.593,
+      "step": 3203
+    },
+    {
+      "epoch": 0.22325192488590043,
+      "grad_norm": 0.7129339711853596,
+      "learning_rate": 6.256427735647596e-07,
+      "loss": 1.5354,
+      "step": 3204
+    },
+    {
+      "epoch": 0.22332160401351775,
+      "grad_norm": 0.7398839096789644,
+      "learning_rate": 6.255982711676306e-07,
+      "loss": 1.6543,
+      "step": 3205
+    },
+    {
+      "epoch": 0.22339128314113507,
+      "grad_norm": 0.7076361329437126,
+      "learning_rate": 6.255537572405385e-07,
+      "loss": 1.5077,
+      "step": 3206
+    },
+    {
+      "epoch": 0.2234609622687524,
+      "grad_norm": 0.7558888533563315,
+      "learning_rate": 6.255092317856164e-07,
+      "loss": 1.5365,
+      "step": 3207
+    },
+    {
+      "epoch": 0.22353064139636972,
+      "grad_norm": 0.6913289274165797,
+      "learning_rate": 6.254646948049982e-07,
+      "loss": 1.4091,
+      "step": 3208
+    },
+    {
+      "epoch": 0.22360032052398704,
+      "grad_norm": 0.6862729311578976,
+      "learning_rate": 6.254201463008183e-07,
+      "loss": 1.5289,
+      "step": 3209
+    },
+    {
+      "epoch": 0.22366999965160436,
+      "grad_norm": 0.7032564177413655,
+      "learning_rate": 6.253755862752113e-07,
+      "loss": 1.6496,
+      "step": 3210
+    },
+    {
+      "epoch": 0.22373967877922168,
+      "grad_norm": 0.672748516405407,
+      "learning_rate": 6.253310147303128e-07,
+      "loss": 1.3317,
+      "step": 3211
+    },
+    {
+      "epoch": 0.223809357906839,
+      "grad_norm": 0.6511809527835895,
+      "learning_rate": 6.252864316682587e-07,
+      "loss": 1.4124,
+      "step": 3212
+    },
+    {
+      "epoch": 0.22387903703445633,
+      "grad_norm": 0.785166870644245,
+      "learning_rate": 6.252418370911853e-07,
+      "loss": 1.4795,
+      "step": 3213
+    },
+    {
+      "epoch": 0.22394871616207365,
+      "grad_norm": 0.7156469515997128,
+      "learning_rate": 6.2519723100123e-07,
+      "loss": 1.61,
+      "step": 3214
+    },
+    {
+      "epoch": 0.22401839528969097,
+      "grad_norm": 0.7322960624451204,
+      "learning_rate": 6.251526134005302e-07,
+      "loss": 1.6137,
+      "step": 3215
+    },
+    {
+      "epoch": 0.2240880744173083,
+      "grad_norm": 0.7298373702823864,
+      "learning_rate": 6.251079842912242e-07,
+      "loss": 1.5798,
+      "step": 3216
+    },
+    {
+      "epoch": 0.22415775354492562,
+      "grad_norm": 0.7420559329181559,
+      "learning_rate": 6.250633436754507e-07,
+      "loss": 1.569,
+      "step": 3217
+    },
+    {
+      "epoch": 0.22422743267254294,
+      "grad_norm": 0.6912194583934528,
+      "learning_rate": 6.25018691555349e-07,
+      "loss": 1.5474,
+      "step": 3218
+    },
+    {
+      "epoch": 0.22429711180016026,
+      "grad_norm": 0.8192448526866579,
+      "learning_rate": 6.249740279330586e-07,
+      "loss": 1.691,
+      "step": 3219
+    },
+    {
+      "epoch": 0.2243667909277776,
+      "grad_norm": 0.712953611035292,
+      "learning_rate": 6.249293528107203e-07,
+      "loss": 1.5323,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2244364700553949,
+      "grad_norm": 0.7296944378767534,
+      "learning_rate": 6.248846661904748e-07,
+      "loss": 1.5539,
+      "step": 3221
+    },
+    {
+      "epoch": 0.22450614918301223,
+      "grad_norm": 0.7445076520950352,
+      "learning_rate": 6.248399680744637e-07,
+      "loss": 1.5898,
+      "step": 3222
+    },
+    {
+      "epoch": 0.22457582831062955,
+      "grad_norm": 0.7085715501497795,
+      "learning_rate": 6.247952584648289e-07,
+      "loss": 1.6631,
+      "step": 3223
+    },
+    {
+      "epoch": 0.22464550743824688,
+      "grad_norm": 0.73313301704967,
+      "learning_rate": 6.247505373637131e-07,
+      "loss": 1.6425,
+      "step": 3224
+    },
+    {
+      "epoch": 0.2247151865658642,
+      "grad_norm": 0.7468409271178961,
+      "learning_rate": 6.247058047732591e-07,
+      "loss": 1.4843,
+      "step": 3225
+    },
+    {
+      "epoch": 0.22478486569348152,
+      "grad_norm": 0.7406310185623337,
+      "learning_rate": 6.24661060695611e-07,
+      "loss": 1.5649,
+      "step": 3226
+    },
+    {
+      "epoch": 0.22485454482109885,
+      "grad_norm": 0.7286940068337966,
+      "learning_rate": 6.246163051329129e-07,
+      "loss": 1.5882,
+      "step": 3227
+    },
+    {
+      "epoch": 0.22492422394871617,
+      "grad_norm": 0.7416579984863053,
+      "learning_rate": 6.245715380873094e-07,
+      "loss": 1.5131,
+      "step": 3228
+    },
+    {
+      "epoch": 0.2249939030763335,
+      "grad_norm": 0.678153652186507,
+      "learning_rate": 6.245267595609461e-07,
+      "loss": 1.4576,
+      "step": 3229
+    },
+    {
+      "epoch": 0.2250635822039508,
+      "grad_norm": 0.7760279305504684,
+      "learning_rate": 6.244819695559686e-07,
+      "loss": 1.4768,
+      "step": 3230
+    },
+    {
+      "epoch": 0.22513326133156814,
+      "grad_norm": 0.6735251740482139,
+      "learning_rate": 6.244371680745236e-07,
+      "loss": 1.4129,
+      "step": 3231
+    },
+    {
+      "epoch": 0.22520294045918546,
+      "grad_norm": 0.7608296780045913,
+      "learning_rate": 6.243923551187581e-07,
+      "loss": 1.5914,
+      "step": 3232
+    },
+    {
+      "epoch": 0.22527261958680278,
+      "grad_norm": 0.7262753496045481,
+      "learning_rate": 6.243475306908191e-07,
+      "loss": 1.5632,
+      "step": 3233
+    },
+    {
+      "epoch": 0.2253422987144201,
+      "grad_norm": 0.7168146417665728,
+      "learning_rate": 6.243026947928552e-07,
+      "loss": 1.5634,
+      "step": 3234
+    },
+    {
+      "epoch": 0.22541197784203743,
+      "grad_norm": 0.7666596628392616,
+      "learning_rate": 6.24257847427015e-07,
+      "loss": 1.5986,
+      "step": 3235
+    },
+    {
+      "epoch": 0.22548165696965475,
+      "grad_norm": 0.7096138543338777,
+      "learning_rate": 6.242129885954475e-07,
+      "loss": 1.5359,
+      "step": 3236
+    },
+    {
+      "epoch": 0.22555133609727207,
+      "grad_norm": 0.7387515003643389,
+      "learning_rate": 6.241681183003026e-07,
+      "loss": 1.533,
+      "step": 3237
+    },
+    {
+      "epoch": 0.2256210152248894,
+      "grad_norm": 0.6686624505246506,
+      "learning_rate": 6.241232365437303e-07,
+      "loss": 1.4926,
+      "step": 3238
+    },
+    {
+      "epoch": 0.22569069435250672,
+      "grad_norm": 0.7244864005371732,
+      "learning_rate": 6.240783433278816e-07,
+      "loss": 1.5109,
+      "step": 3239
+    },
+    {
+      "epoch": 0.22576037348012404,
+      "grad_norm": 0.7300898381688603,
+      "learning_rate": 6.240334386549079e-07,
+      "loss": 1.4947,
+      "step": 3240
+    },
+    {
+      "epoch": 0.22583005260774136,
+      "grad_norm": 0.694543083655283,
+      "learning_rate": 6.239885225269611e-07,
+      "loss": 1.5542,
+      "step": 3241
+    },
+    {
+      "epoch": 0.22589973173535868,
+      "grad_norm": 0.6938706067751353,
+      "learning_rate": 6.239435949461937e-07,
+      "loss": 1.5457,
+      "step": 3242
+    },
+    {
+      "epoch": 0.225969410862976,
+      "grad_norm": 0.8120680247426197,
+      "learning_rate": 6.238986559147587e-07,
+      "loss": 1.4822,
+      "step": 3243
+    },
+    {
+      "epoch": 0.22603908999059333,
+      "grad_norm": 1.0393662763357927,
+      "learning_rate": 6.238537054348097e-07,
+      "loss": 1.5993,
+      "step": 3244
+    },
+    {
+      "epoch": 0.22610876911821065,
+      "grad_norm": 0.6939123246891056,
+      "learning_rate": 6.238087435085006e-07,
+      "loss": 1.5299,
+      "step": 3245
+    },
+    {
+      "epoch": 0.22617844824582797,
+      "grad_norm": 0.7103542178040833,
+      "learning_rate": 6.237637701379864e-07,
+      "loss": 1.5642,
+      "step": 3246
+    },
+    {
+      "epoch": 0.2262481273734453,
+      "grad_norm": 0.6790085012080815,
+      "learning_rate": 6.237187853254221e-07,
+      "loss": 1.4859,
+      "step": 3247
+    },
+    {
+      "epoch": 0.22631780650106262,
+      "grad_norm": 0.7613369388575224,
+      "learning_rate": 6.236737890729635e-07,
+      "loss": 1.5173,
+      "step": 3248
+    },
+    {
+      "epoch": 0.22638748562867994,
+      "grad_norm": 0.7266168413309799,
+      "learning_rate": 6.23628781382767e-07,
+      "loss": 1.5771,
+      "step": 3249
+    },
+    {
+      "epoch": 0.22645716475629726,
+      "grad_norm": 0.7838417336138501,
+      "learning_rate": 6.235837622569894e-07,
+      "loss": 1.6464,
+      "step": 3250
+    },
+    {
+      "epoch": 0.22652684388391459,
+      "grad_norm": 0.6878127982873993,
+      "learning_rate": 6.235387316977881e-07,
+      "loss": 1.4028,
+      "step": 3251
+    },
+    {
+      "epoch": 0.2265965230115319,
+      "grad_norm": 0.6759827777027689,
+      "learning_rate": 6.23493689707321e-07,
+      "loss": 1.4111,
+      "step": 3252
+    },
+    {
+      "epoch": 0.22666620213914923,
+      "grad_norm": 0.7563518707403992,
+      "learning_rate": 6.234486362877468e-07,
+      "loss": 1.6306,
+      "step": 3253
+    },
+    {
+      "epoch": 0.22673588126676655,
+      "grad_norm": 0.705109375720128,
+      "learning_rate": 6.234035714412243e-07,
+      "loss": 1.4413,
+      "step": 3254
+    },
+    {
+      "epoch": 0.22680556039438388,
+      "grad_norm": 0.6972099729276918,
+      "learning_rate": 6.233584951699133e-07,
+      "loss": 1.5258,
+      "step": 3255
+    },
+    {
+      "epoch": 0.22687523952200117,
+      "grad_norm": 0.7448910365246851,
+      "learning_rate": 6.233134074759739e-07,
+      "loss": 1.5063,
+      "step": 3256
+    },
+    {
+      "epoch": 0.2269449186496185,
+      "grad_norm": 0.7259674297447937,
+      "learning_rate": 6.232683083615668e-07,
+      "loss": 1.6128,
+      "step": 3257
+    },
+    {
+      "epoch": 0.22701459777723582,
+      "grad_norm": 0.6804071352169611,
+      "learning_rate": 6.23223197828853e-07,
+      "loss": 1.4106,
+      "step": 3258
+    },
+    {
+      "epoch": 0.22708427690485314,
+      "grad_norm": 0.7116889143448107,
+      "learning_rate": 6.231780758799946e-07,
+      "loss": 1.628,
+      "step": 3259
+    },
+    {
+      "epoch": 0.22715395603247046,
+      "grad_norm": 0.7161916261000766,
+      "learning_rate": 6.231329425171538e-07,
+      "loss": 1.6124,
+      "step": 3260
+    },
+    {
+      "epoch": 0.22722363516008778,
+      "grad_norm": 0.7276125930319626,
+      "learning_rate": 6.230877977424936e-07,
+      "loss": 1.5841,
+      "step": 3261
+    },
+    {
+      "epoch": 0.2272933142877051,
+      "grad_norm": 0.7186008176701532,
+      "learning_rate": 6.230426415581773e-07,
+      "loss": 1.5515,
+      "step": 3262
+    },
+    {
+      "epoch": 0.22736299341532243,
+      "grad_norm": 0.7167681974231933,
+      "learning_rate": 6.229974739663689e-07,
+      "loss": 1.6088,
+      "step": 3263
+    },
+    {
+      "epoch": 0.22743267254293975,
+      "grad_norm": 0.706065353428564,
+      "learning_rate": 6.229522949692331e-07,
+      "loss": 1.556,
+      "step": 3264
+    },
+    {
+      "epoch": 0.22750235167055707,
+      "grad_norm": 0.7269340461885619,
+      "learning_rate": 6.229071045689346e-07,
+      "loss": 1.5334,
+      "step": 3265
+    },
+    {
+      "epoch": 0.2275720307981744,
+      "grad_norm": 0.7456034153946854,
+      "learning_rate": 6.228619027676394e-07,
+      "loss": 1.4739,
+      "step": 3266
+    },
+    {
+      "epoch": 0.22764170992579172,
+      "grad_norm": 0.7358562726000603,
+      "learning_rate": 6.228166895675134e-07,
+      "loss": 1.6216,
+      "step": 3267
+    },
+    {
+      "epoch": 0.22771138905340904,
+      "grad_norm": 0.7259410956511649,
+      "learning_rate": 6.227714649707234e-07,
+      "loss": 1.5702,
+      "step": 3268
+    },
+    {
+      "epoch": 0.22778106818102636,
+      "grad_norm": 0.6678683131356811,
+      "learning_rate": 6.227262289794368e-07,
+      "loss": 1.451,
+      "step": 3269
+    },
+    {
+      "epoch": 0.2278507473086437,
+      "grad_norm": 0.8432593479988705,
+      "learning_rate": 6.226809815958212e-07,
+      "loss": 1.749,
+      "step": 3270
+    },
+    {
+      "epoch": 0.227920426436261,
+      "grad_norm": 0.6699718251478374,
+      "learning_rate": 6.22635722822045e-07,
+      "loss": 1.3672,
+      "step": 3271
+    },
+    {
+      "epoch": 0.22799010556387833,
+      "grad_norm": 0.7134528447425973,
+      "learning_rate": 6.22590452660277e-07,
+      "loss": 1.4964,
+      "step": 3272
+    },
+    {
+      "epoch": 0.22805978469149565,
+      "grad_norm": 0.8470823062532795,
+      "learning_rate": 6.22545171112687e-07,
+      "loss": 1.7031,
+      "step": 3273
+    },
+    {
+      "epoch": 0.22812946381911298,
+      "grad_norm": 0.7057268115009272,
+      "learning_rate": 6.224998781814445e-07,
+      "loss": 1.5361,
+      "step": 3274
+    },
+    {
+      "epoch": 0.2281991429467303,
+      "grad_norm": 0.7369497880685594,
+      "learning_rate": 6.224545738687203e-07,
+      "loss": 1.5211,
+      "step": 3275
+    },
+    {
+      "epoch": 0.22826882207434762,
+      "grad_norm": 0.7145201468391852,
+      "learning_rate": 6.224092581766854e-07,
+      "loss": 1.5296,
+      "step": 3276
+    },
+    {
+      "epoch": 0.22833850120196494,
+      "grad_norm": 0.7103105050401702,
+      "learning_rate": 6.223639311075114e-07,
+      "loss": 1.4471,
+      "step": 3277
+    },
+    {
+      "epoch": 0.22840818032958227,
+      "grad_norm": 0.844762788612772,
+      "learning_rate": 6.223185926633709e-07,
+      "loss": 1.5651,
+      "step": 3278
+    },
+    {
+      "epoch": 0.2284778594571996,
+      "grad_norm": 0.7034913972328873,
+      "learning_rate": 6.22273242846436e-07,
+      "loss": 1.5359,
+      "step": 3279
+    },
+    {
+      "epoch": 0.2285475385848169,
+      "grad_norm": 0.7461126329176588,
+      "learning_rate": 6.2222788165888e-07,
+      "loss": 1.6537,
+      "step": 3280
+    },
+    {
+      "epoch": 0.22861721771243423,
+      "grad_norm": 0.7269639564103746,
+      "learning_rate": 6.221825091028772e-07,
+      "loss": 1.4432,
+      "step": 3281
+    },
+    {
+      "epoch": 0.22868689684005156,
+      "grad_norm": 0.6712173022990764,
+      "learning_rate": 6.221371251806014e-07,
+      "loss": 1.5733,
+      "step": 3282
+    },
+    {
+      "epoch": 0.22875657596766888,
+      "grad_norm": 0.7475145129410613,
+      "learning_rate": 6.220917298942278e-07,
+      "loss": 1.5138,
+      "step": 3283
+    },
+    {
+      "epoch": 0.2288262550952862,
+      "grad_norm": 0.7008954768552739,
+      "learning_rate": 6.220463232459318e-07,
+      "loss": 1.4898,
+      "step": 3284
+    },
+    {
+      "epoch": 0.22889593422290352,
+      "grad_norm": 0.7248710827182362,
+      "learning_rate": 6.220009052378892e-07,
+      "loss": 1.5659,
+      "step": 3285
+    },
+    {
+      "epoch": 0.22896561335052085,
+      "grad_norm": 0.7064860142782986,
+      "learning_rate": 6.21955475872277e-07,
+      "loss": 1.5297,
+      "step": 3286
+    },
+    {
+      "epoch": 0.22903529247813817,
+      "grad_norm": 0.7221219290864086,
+      "learning_rate": 6.219100351512717e-07,
+      "loss": 1.5716,
+      "step": 3287
+    },
+    {
+      "epoch": 0.2291049716057555,
+      "grad_norm": 0.7565337053436231,
+      "learning_rate": 6.218645830770511e-07,
+      "loss": 1.6439,
+      "step": 3288
+    },
+    {
+      "epoch": 0.22917465073337281,
+      "grad_norm": 0.7723173370820656,
+      "learning_rate": 6.218191196517935e-07,
+      "loss": 1.5419,
+      "step": 3289
+    },
+    {
+      "epoch": 0.22924432986099014,
+      "grad_norm": 0.6800261948404732,
+      "learning_rate": 6.217736448776775e-07,
+      "loss": 1.5262,
+      "step": 3290
+    },
+    {
+      "epoch": 0.22931400898860746,
+      "grad_norm": 0.6742666990820785,
+      "learning_rate": 6.217281587568823e-07,
+      "loss": 1.4581,
+      "step": 3291
+    },
+    {
+      "epoch": 0.22938368811622478,
+      "grad_norm": 0.6924075637187566,
+      "learning_rate": 6.216826612915877e-07,
+      "loss": 1.5268,
+      "step": 3292
+    },
+    {
+      "epoch": 0.2294533672438421,
+      "grad_norm": 0.6770195243966041,
+      "learning_rate": 6.216371524839743e-07,
+      "loss": 1.5681,
+      "step": 3293
+    },
+    {
+      "epoch": 0.22952304637145943,
+      "grad_norm": 0.6831251809175937,
+      "learning_rate": 6.215916323362225e-07,
+      "loss": 1.4456,
+      "step": 3294
+    },
+    {
+      "epoch": 0.22959272549907675,
+      "grad_norm": 0.6829693545829544,
+      "learning_rate": 6.215461008505141e-07,
+      "loss": 1.616,
+      "step": 3295
+    },
+    {
+      "epoch": 0.22966240462669407,
+      "grad_norm": 0.6949054818073175,
+      "learning_rate": 6.215005580290309e-07,
+      "loss": 1.5678,
+      "step": 3296
+    },
+    {
+      "epoch": 0.2297320837543114,
+      "grad_norm": 0.6812757453937095,
+      "learning_rate": 6.214550038739554e-07,
+      "loss": 1.5277,
+      "step": 3297
+    },
+    {
+      "epoch": 0.22980176288192872,
+      "grad_norm": 0.6736556299964148,
+      "learning_rate": 6.214094383874707e-07,
+      "loss": 1.5254,
+      "step": 3298
+    },
+    {
+      "epoch": 0.22987144200954604,
+      "grad_norm": 0.6692168534392988,
+      "learning_rate": 6.213638615717605e-07,
+      "loss": 1.5795,
+      "step": 3299
+    },
+    {
+      "epoch": 0.22994112113716336,
+      "grad_norm": 0.7090645204789643,
+      "learning_rate": 6.213182734290085e-07,
+      "loss": 1.5528,
+      "step": 3300
+    },
+    {
+      "epoch": 0.23001080026478068,
+      "grad_norm": 0.7001362981176051,
+      "learning_rate": 6.212726739613998e-07,
+      "loss": 1.4763,
+      "step": 3301
+    },
+    {
+      "epoch": 0.230080479392398,
+      "grad_norm": 0.7586834311900935,
+      "learning_rate": 6.212270631711197e-07,
+      "loss": 1.6276,
+      "step": 3302
+    },
+    {
+      "epoch": 0.23015015852001533,
+      "grad_norm": 0.7934441272813372,
+      "learning_rate": 6.211814410603536e-07,
+      "loss": 1.5447,
+      "step": 3303
+    },
+    {
+      "epoch": 0.23021983764763265,
+      "grad_norm": 0.7482907591214504,
+      "learning_rate": 6.211358076312881e-07,
+      "loss": 1.4875,
+      "step": 3304
+    },
+    {
+      "epoch": 0.23028951677524997,
+      "grad_norm": 0.6763665819146414,
+      "learning_rate": 6.210901628861098e-07,
+      "loss": 1.4554,
+      "step": 3305
+    },
+    {
+      "epoch": 0.2303591959028673,
+      "grad_norm": 0.7312207211708315,
+      "learning_rate": 6.210445068270063e-07,
+      "loss": 1.5047,
+      "step": 3306
+    },
+    {
+      "epoch": 0.23042887503048462,
+      "grad_norm": 0.7269062813292612,
+      "learning_rate": 6.209988394561652e-07,
+      "loss": 1.6715,
+      "step": 3307
+    },
+    {
+      "epoch": 0.23049855415810194,
+      "grad_norm": 0.7149134256932522,
+      "learning_rate": 6.209531607757755e-07,
+      "loss": 1.4913,
+      "step": 3308
+    },
+    {
+      "epoch": 0.23056823328571927,
+      "grad_norm": 0.6878175606991938,
+      "learning_rate": 6.209074707880259e-07,
+      "loss": 1.4401,
+      "step": 3309
+    },
+    {
+      "epoch": 0.2306379124133366,
+      "grad_norm": 0.7116693886252153,
+      "learning_rate": 6.208617694951059e-07,
+      "loss": 1.4861,
+      "step": 3310
+    },
+    {
+      "epoch": 0.2307075915409539,
+      "grad_norm": 0.6677721593746183,
+      "learning_rate": 6.208160568992057e-07,
+      "loss": 1.575,
+      "step": 3311
+    },
+    {
+      "epoch": 0.23077727066857123,
+      "grad_norm": 0.7350179713720592,
+      "learning_rate": 6.20770333002516e-07,
+      "loss": 1.5758,
+      "step": 3312
+    },
+    {
+      "epoch": 0.23084694979618856,
+      "grad_norm": 0.6928707351994486,
+      "learning_rate": 6.207245978072279e-07,
+      "loss": 1.4518,
+      "step": 3313
+    },
+    {
+      "epoch": 0.23091662892380588,
+      "grad_norm": 0.6738940033331204,
+      "learning_rate": 6.206788513155331e-07,
+      "loss": 1.4948,
+      "step": 3314
+    },
+    {
+      "epoch": 0.2309863080514232,
+      "grad_norm": 0.7204099383196804,
+      "learning_rate": 6.206330935296239e-07,
+      "loss": 1.5013,
+      "step": 3315
+    },
+    {
+      "epoch": 0.23105598717904052,
+      "grad_norm": 0.7320799572268409,
+      "learning_rate": 6.205873244516931e-07,
+      "loss": 1.6149,
+      "step": 3316
+    },
+    {
+      "epoch": 0.23112566630665785,
+      "grad_norm": 0.679257456943743,
+      "learning_rate": 6.20541544083934e-07,
+      "loss": 1.5815,
+      "step": 3317
+    },
+    {
+      "epoch": 0.23119534543427517,
+      "grad_norm": 0.759769092880823,
+      "learning_rate": 6.204957524285407e-07,
+      "loss": 1.659,
+      "step": 3318
+    },
+    {
+      "epoch": 0.2312650245618925,
+      "grad_norm": 0.7434624189053274,
+      "learning_rate": 6.204499494877074e-07,
+      "loss": 1.5217,
+      "step": 3319
+    },
+    {
+      "epoch": 0.2313347036895098,
+      "grad_norm": 0.7239122991163198,
+      "learning_rate": 6.204041352636293e-07,
+      "loss": 1.494,
+      "step": 3320
+    },
+    {
+      "epoch": 0.23140438281712714,
+      "grad_norm": 0.730206357832636,
+      "learning_rate": 6.203583097585015e-07,
+      "loss": 1.598,
+      "step": 3321
+    },
+    {
+      "epoch": 0.23147406194474446,
+      "grad_norm": 0.7236108764408529,
+      "learning_rate": 6.203124729745206e-07,
+      "loss": 1.5986,
+      "step": 3322
+    },
+    {
+      "epoch": 0.23154374107236178,
+      "grad_norm": 0.7670495414586959,
+      "learning_rate": 6.202666249138827e-07,
+      "loss": 1.6228,
+      "step": 3323
+    },
+    {
+      "epoch": 0.2316134201999791,
+      "grad_norm": 0.7049449713350656,
+      "learning_rate": 6.202207655787851e-07,
+      "loss": 1.4829,
+      "step": 3324
+    },
+    {
+      "epoch": 0.23168309932759643,
+      "grad_norm": 0.7305992229489915,
+      "learning_rate": 6.201748949714257e-07,
+      "loss": 1.6744,
+      "step": 3325
+    },
+    {
+      "epoch": 0.23175277845521375,
+      "grad_norm": 0.7119919576125097,
+      "learning_rate": 6.201290130940024e-07,
+      "loss": 1.4452,
+      "step": 3326
+    },
+    {
+      "epoch": 0.23182245758283107,
+      "grad_norm": 0.6822509435659513,
+      "learning_rate": 6.200831199487141e-07,
+      "loss": 1.4767,
+      "step": 3327
+    },
+    {
+      "epoch": 0.2318921367104484,
+      "grad_norm": 0.7289185593864849,
+      "learning_rate": 6.200372155377601e-07,
+      "loss": 1.5892,
+      "step": 3328
+    },
+    {
+      "epoch": 0.23196181583806572,
+      "grad_norm": 0.6963543315365277,
+      "learning_rate": 6.199912998633401e-07,
+      "loss": 1.5631,
+      "step": 3329
+    },
+    {
+      "epoch": 0.23203149496568304,
+      "grad_norm": 0.6930151776226096,
+      "learning_rate": 6.199453729276547e-07,
+      "loss": 1.6462,
+      "step": 3330
+    },
+    {
+      "epoch": 0.23210117409330036,
+      "grad_norm": 0.69238854833818,
+      "learning_rate": 6.198994347329047e-07,
+      "loss": 1.5036,
+      "step": 3331
+    },
+    {
+      "epoch": 0.23217085322091768,
+      "grad_norm": 0.7474791474026496,
+      "learning_rate": 6.198534852812916e-07,
+      "loss": 1.4649,
+      "step": 3332
+    },
+    {
+      "epoch": 0.232240532348535,
+      "grad_norm": 0.725018393418341,
+      "learning_rate": 6.198075245750173e-07,
+      "loss": 1.5417,
+      "step": 3333
+    },
+    {
+      "epoch": 0.23231021147615233,
+      "grad_norm": 0.7522669532923917,
+      "learning_rate": 6.197615526162843e-07,
+      "loss": 1.5881,
+      "step": 3334
+    },
+    {
+      "epoch": 0.23237989060376965,
+      "grad_norm": 0.7084064750535294,
+      "learning_rate": 6.197155694072958e-07,
+      "loss": 1.5404,
+      "step": 3335
+    },
+    {
+      "epoch": 0.23244956973138697,
+      "grad_norm": 0.7518904247342453,
+      "learning_rate": 6.196695749502553e-07,
+      "loss": 1.6081,
+      "step": 3336
+    },
+    {
+      "epoch": 0.2325192488590043,
+      "grad_norm": 0.7462448084328533,
+      "learning_rate": 6.19623569247367e-07,
+      "loss": 1.6945,
+      "step": 3337
+    },
+    {
+      "epoch": 0.23258892798662162,
+      "grad_norm": 0.7352790334391689,
+      "learning_rate": 6.195775523008357e-07,
+      "loss": 1.6254,
+      "step": 3338
+    },
+    {
+      "epoch": 0.23265860711423894,
+      "grad_norm": 0.692602365608469,
+      "learning_rate": 6.195315241128664e-07,
+      "loss": 1.5174,
+      "step": 3339
+    },
+    {
+      "epoch": 0.23272828624185626,
+      "grad_norm": 0.7557983065691252,
+      "learning_rate": 6.19485484685665e-07,
+      "loss": 1.7174,
+      "step": 3340
+    },
+    {
+      "epoch": 0.2327979653694736,
+      "grad_norm": 0.7820184701909594,
+      "learning_rate": 6.194394340214378e-07,
+      "loss": 1.629,
+      "step": 3341
+    },
+    {
+      "epoch": 0.2328676444970909,
+      "grad_norm": 0.741342894454099,
+      "learning_rate": 6.193933721223916e-07,
+      "loss": 1.5313,
+      "step": 3342
+    },
+    {
+      "epoch": 0.23293732362470823,
+      "grad_norm": 0.78097018285349,
+      "learning_rate": 6.193472989907339e-07,
+      "loss": 1.6066,
+      "step": 3343
+    },
+    {
+      "epoch": 0.23300700275232555,
+      "grad_norm": 0.7650208326074384,
+      "learning_rate": 6.193012146286725e-07,
+      "loss": 1.5455,
+      "step": 3344
+    },
+    {
+      "epoch": 0.23307668187994288,
+      "grad_norm": 0.8015620503451505,
+      "learning_rate": 6.192551190384158e-07,
+      "loss": 1.468,
+      "step": 3345
+    },
+    {
+      "epoch": 0.2331463610075602,
+      "grad_norm": 0.7433641051339869,
+      "learning_rate": 6.192090122221729e-07,
+      "loss": 1.5386,
+      "step": 3346
+    },
+    {
+      "epoch": 0.23321604013517752,
+      "grad_norm": 0.7174021653327852,
+      "learning_rate": 6.191628941821534e-07,
+      "loss": 1.7446,
+      "step": 3347
+    },
+    {
+      "epoch": 0.23328571926279482,
+      "grad_norm": 0.6836406889381682,
+      "learning_rate": 6.191167649205672e-07,
+      "loss": 1.5269,
+      "step": 3348
+    },
+    {
+      "epoch": 0.23335539839041214,
+      "grad_norm": 0.7099080109048698,
+      "learning_rate": 6.190706244396251e-07,
+      "loss": 1.5104,
+      "step": 3349
+    },
+    {
+      "epoch": 0.23342507751802946,
+      "grad_norm": 0.663174403385495,
+      "learning_rate": 6.19024472741538e-07,
+      "loss": 1.5063,
+      "step": 3350
+    },
+    {
+      "epoch": 0.23349475664564678,
+      "grad_norm": 0.7500897550374107,
+      "learning_rate": 6.189783098285178e-07,
+      "loss": 1.5334,
+      "step": 3351
+    },
+    {
+      "epoch": 0.2335644357732641,
+      "grad_norm": 0.7258609882745846,
+      "learning_rate": 6.189321357027766e-07,
+      "loss": 1.4554,
+      "step": 3352
+    },
+    {
+      "epoch": 0.23363411490088143,
+      "grad_norm": 0.7432031040341716,
+      "learning_rate": 6.188859503665272e-07,
+      "loss": 1.6481,
+      "step": 3353
+    },
+    {
+      "epoch": 0.23370379402849875,
+      "grad_norm": 0.7166121656267893,
+      "learning_rate": 6.188397538219829e-07,
+      "loss": 1.595,
+      "step": 3354
+    },
+    {
+      "epoch": 0.23377347315611607,
+      "grad_norm": 0.6927380416464944,
+      "learning_rate": 6.187935460713575e-07,
+      "loss": 1.4588,
+      "step": 3355
+    },
+    {
+      "epoch": 0.2338431522837334,
+      "grad_norm": 0.7540129503422559,
+      "learning_rate": 6.187473271168655e-07,
+      "loss": 1.5675,
+      "step": 3356
+    },
+    {
+      "epoch": 0.23391283141135072,
+      "grad_norm": 0.7308119077873947,
+      "learning_rate": 6.187010969607217e-07,
+      "loss": 1.6251,
+      "step": 3357
+    },
+    {
+      "epoch": 0.23398251053896804,
+      "grad_norm": 0.6693944602398552,
+      "learning_rate": 6.186548556051415e-07,
+      "loss": 1.5107,
+      "step": 3358
+    },
+    {
+      "epoch": 0.23405218966658536,
+      "grad_norm": 0.7086992977341247,
+      "learning_rate": 6.18608603052341e-07,
+      "loss": 1.5787,
+      "step": 3359
+    },
+    {
+      "epoch": 0.2341218687942027,
+      "grad_norm": 0.6669225432306505,
+      "learning_rate": 6.185623393045367e-07,
+      "loss": 1.5375,
+      "step": 3360
+    },
+    {
+      "epoch": 0.23419154792182,
+      "grad_norm": 0.6538619423769466,
+      "learning_rate": 6.185160643639454e-07,
+      "loss": 1.4486,
+      "step": 3361
+    },
+    {
+      "epoch": 0.23426122704943733,
+      "grad_norm": 0.7317000886747869,
+      "learning_rate": 6.184697782327851e-07,
+      "loss": 1.5035,
+      "step": 3362
+    },
+    {
+      "epoch": 0.23433090617705465,
+      "grad_norm": 0.680179715005664,
+      "learning_rate": 6.184234809132737e-07,
+      "loss": 1.4639,
+      "step": 3363
+    },
+    {
+      "epoch": 0.23440058530467198,
+      "grad_norm": 0.8368622763484619,
+      "learning_rate": 6.183771724076298e-07,
+      "loss": 1.6463,
+      "step": 3364
+    },
+    {
+      "epoch": 0.2344702644322893,
+      "grad_norm": 0.8092067224444279,
+      "learning_rate": 6.183308527180727e-07,
+      "loss": 1.6825,
+      "step": 3365
+    },
+    {
+      "epoch": 0.23453994355990662,
+      "grad_norm": 0.727191601732615,
+      "learning_rate": 6.182845218468222e-07,
+      "loss": 1.5721,
+      "step": 3366
+    },
+    {
+      "epoch": 0.23460962268752394,
+      "grad_norm": 0.7190081206311841,
+      "learning_rate": 6.182381797960983e-07,
+      "loss": 1.4984,
+      "step": 3367
+    },
+    {
+      "epoch": 0.23467930181514127,
+      "grad_norm": 0.721597520938331,
+      "learning_rate": 6.181918265681221e-07,
+      "loss": 1.3929,
+      "step": 3368
+    },
+    {
+      "epoch": 0.2347489809427586,
+      "grad_norm": 0.7270928500070124,
+      "learning_rate": 6.181454621651149e-07,
+      "loss": 1.4872,
+      "step": 3369
+    },
+    {
+      "epoch": 0.2348186600703759,
+      "grad_norm": 0.7859246873980587,
+      "learning_rate": 6.180990865892984e-07,
+      "loss": 1.6426,
+      "step": 3370
+    },
+    {
+      "epoch": 0.23488833919799323,
+      "grad_norm": 0.7658917026133951,
+      "learning_rate": 6.180526998428953e-07,
+      "loss": 1.4258,
+      "step": 3371
+    },
+    {
+      "epoch": 0.23495801832561056,
+      "grad_norm": 0.7703297932331463,
+      "learning_rate": 6.180063019281282e-07,
+      "loss": 1.7079,
+      "step": 3372
+    },
+    {
+      "epoch": 0.23502769745322788,
+      "grad_norm": 0.7604510847538044,
+      "learning_rate": 6.179598928472208e-07,
+      "loss": 1.6182,
+      "step": 3373
+    },
+    {
+      "epoch": 0.2350973765808452,
+      "grad_norm": 0.7085186689342032,
+      "learning_rate": 6.179134726023971e-07,
+      "loss": 1.5748,
+      "step": 3374
+    },
+    {
+      "epoch": 0.23516705570846252,
+      "grad_norm": 0.7427333380583586,
+      "learning_rate": 6.178670411958817e-07,
+      "loss": 1.4754,
+      "step": 3375
+    },
+    {
+      "epoch": 0.23523673483607985,
+      "grad_norm": 0.7173447702200312,
+      "learning_rate": 6.178205986298996e-07,
+      "loss": 1.5343,
+      "step": 3376
+    },
+    {
+      "epoch": 0.23530641396369717,
+      "grad_norm": 0.7043449607322789,
+      "learning_rate": 6.177741449066763e-07,
+      "loss": 1.4961,
+      "step": 3377
+    },
+    {
+      "epoch": 0.2353760930913145,
+      "grad_norm": 0.7456152236319041,
+      "learning_rate": 6.177276800284382e-07,
+      "loss": 1.4932,
+      "step": 3378
+    },
+    {
+      "epoch": 0.23544577221893181,
+      "grad_norm": 0.7424025461195325,
+      "learning_rate": 6.176812039974119e-07,
+      "loss": 1.4947,
+      "step": 3379
+    },
+    {
+      "epoch": 0.23551545134654914,
+      "grad_norm": 0.7232428004855358,
+      "learning_rate": 6.176347168158246e-07,
+      "loss": 1.552,
+      "step": 3380
+    },
+    {
+      "epoch": 0.23558513047416646,
+      "grad_norm": 0.6779956331594822,
+      "learning_rate": 6.175882184859041e-07,
+      "loss": 1.4757,
+      "step": 3381
+    },
+    {
+      "epoch": 0.23565480960178378,
+      "grad_norm": 0.7171144453890865,
+      "learning_rate": 6.175417090098787e-07,
+      "loss": 1.5146,
+      "step": 3382
+    },
+    {
+      "epoch": 0.2357244887294011,
+      "grad_norm": 0.7121967513830976,
+      "learning_rate": 6.174951883899771e-07,
+      "loss": 1.5433,
+      "step": 3383
+    },
+    {
+      "epoch": 0.23579416785701843,
+      "grad_norm": 0.6891281985086806,
+      "learning_rate": 6.17448656628429e-07,
+      "loss": 1.6011,
+      "step": 3384
+    },
+    {
+      "epoch": 0.23586384698463575,
+      "grad_norm": 0.6951321103009805,
+      "learning_rate": 6.174021137274638e-07,
+      "loss": 1.486,
+      "step": 3385
+    },
+    {
+      "epoch": 0.23593352611225307,
+      "grad_norm": 0.7260711549908915,
+      "learning_rate": 6.173555596893123e-07,
+      "loss": 1.4865,
+      "step": 3386
+    },
+    {
+      "epoch": 0.2360032052398704,
+      "grad_norm": 0.6964511042381395,
+      "learning_rate": 6.173089945162053e-07,
+      "loss": 1.5472,
+      "step": 3387
+    },
+    {
+      "epoch": 0.23607288436748772,
+      "grad_norm": 0.7433377677254371,
+      "learning_rate": 6.172624182103744e-07,
+      "loss": 1.5743,
+      "step": 3388
+    },
+    {
+      "epoch": 0.23614256349510504,
+      "grad_norm": 0.7242210025791754,
+      "learning_rate": 6.172158307740517e-07,
+      "loss": 1.5983,
+      "step": 3389
+    },
+    {
+      "epoch": 0.23621224262272236,
+      "grad_norm": 0.7789543820644682,
+      "learning_rate": 6.171692322094696e-07,
+      "loss": 1.4545,
+      "step": 3390
+    },
+    {
+      "epoch": 0.23628192175033969,
+      "grad_norm": 0.751532733815692,
+      "learning_rate": 6.171226225188612e-07,
+      "loss": 1.5104,
+      "step": 3391
+    },
+    {
+      "epoch": 0.236351600877957,
+      "grad_norm": 0.7256233715460338,
+      "learning_rate": 6.170760017044602e-07,
+      "loss": 1.5217,
+      "step": 3392
+    },
+    {
+      "epoch": 0.23642128000557433,
+      "grad_norm": 0.695169404445898,
+      "learning_rate": 6.170293697685008e-07,
+      "loss": 1.4637,
+      "step": 3393
+    },
+    {
+      "epoch": 0.23649095913319165,
+      "grad_norm": 0.7268883932193495,
+      "learning_rate": 6.169827267132177e-07,
+      "loss": 1.5521,
+      "step": 3394
+    },
+    {
+      "epoch": 0.23656063826080898,
+      "grad_norm": 0.8191933964958839,
+      "learning_rate": 6.169360725408461e-07,
+      "loss": 1.5085,
+      "step": 3395
+    },
+    {
+      "epoch": 0.2366303173884263,
+      "grad_norm": 0.7463896155748065,
+      "learning_rate": 6.168894072536215e-07,
+      "loss": 1.631,
+      "step": 3396
+    },
+    {
+      "epoch": 0.23669999651604362,
+      "grad_norm": 0.7538657089938234,
+      "learning_rate": 6.168427308537807e-07,
+      "loss": 1.6949,
+      "step": 3397
+    },
+    {
+      "epoch": 0.23676967564366094,
+      "grad_norm": 0.6922451358544848,
+      "learning_rate": 6.167960433435602e-07,
+      "loss": 1.5991,
+      "step": 3398
+    },
+    {
+      "epoch": 0.23683935477127827,
+      "grad_norm": 0.7173957121519011,
+      "learning_rate": 6.167493447251974e-07,
+      "loss": 1.5608,
+      "step": 3399
+    },
+    {
+      "epoch": 0.2369090338988956,
+      "grad_norm": 0.6702020251445877,
+      "learning_rate": 6.167026350009302e-07,
+      "loss": 1.5437,
+      "step": 3400
+    },
+    {
+      "epoch": 0.2369787130265129,
+      "grad_norm": 0.8055652122242971,
+      "learning_rate": 6.166559141729971e-07,
+      "loss": 1.506,
+      "step": 3401
+    },
+    {
+      "epoch": 0.23704839215413023,
+      "grad_norm": 0.7832627703406165,
+      "learning_rate": 6.166091822436371e-07,
+      "loss": 1.6436,
+      "step": 3402
+    },
+    {
+      "epoch": 0.23711807128174756,
+      "grad_norm": 0.6791136166363546,
+      "learning_rate": 6.165624392150895e-07,
+      "loss": 1.4786,
+      "step": 3403
+    },
+    {
+      "epoch": 0.23718775040936488,
+      "grad_norm": 0.772012569377695,
+      "learning_rate": 6.165156850895944e-07,
+      "loss": 1.56,
+      "step": 3404
+    },
+    {
+      "epoch": 0.2372574295369822,
+      "grad_norm": 0.7148878873649895,
+      "learning_rate": 6.164689198693925e-07,
+      "loss": 1.5209,
+      "step": 3405
+    },
+    {
+      "epoch": 0.23732710866459952,
+      "grad_norm": 0.7661887521009225,
+      "learning_rate": 6.164221435567247e-07,
+      "loss": 1.4896,
+      "step": 3406
+    },
+    {
+      "epoch": 0.23739678779221685,
+      "grad_norm": 0.8144245815031177,
+      "learning_rate": 6.163753561538325e-07,
+      "loss": 1.3957,
+      "step": 3407
+    },
+    {
+      "epoch": 0.23746646691983417,
+      "grad_norm": 0.7057628127124091,
+      "learning_rate": 6.163285576629585e-07,
+      "loss": 1.5469,
+      "step": 3408
+    },
+    {
+      "epoch": 0.2375361460474515,
+      "grad_norm": 0.7065635665133951,
+      "learning_rate": 6.16281748086345e-07,
+      "loss": 1.6138,
+      "step": 3409
+    },
+    {
+      "epoch": 0.2376058251750688,
+      "grad_norm": 0.7702791357963328,
+      "learning_rate": 6.162349274262353e-07,
+      "loss": 1.6062,
+      "step": 3410
+    },
+    {
+      "epoch": 0.23767550430268614,
+      "grad_norm": 0.7129163308798244,
+      "learning_rate": 6.161880956848732e-07,
+      "loss": 1.5728,
+      "step": 3411
+    },
+    {
+      "epoch": 0.23774518343030346,
+      "grad_norm": 0.7367937664828406,
+      "learning_rate": 6.16141252864503e-07,
+      "loss": 1.5456,
+      "step": 3412
+    },
+    {
+      "epoch": 0.23781486255792078,
+      "grad_norm": 0.7393353115134411,
+      "learning_rate": 6.160943989673692e-07,
+      "loss": 1.5298,
+      "step": 3413
+    },
+    {
+      "epoch": 0.2378845416855381,
+      "grad_norm": 0.7583212519478039,
+      "learning_rate": 6.160475339957176e-07,
+      "loss": 1.543,
+      "step": 3414
+    },
+    {
+      "epoch": 0.23795422081315543,
+      "grad_norm": 0.7226687449053614,
+      "learning_rate": 6.160006579517937e-07,
+      "loss": 1.5804,
+      "step": 3415
+    },
+    {
+      "epoch": 0.23802389994077275,
+      "grad_norm": 0.7184778025421876,
+      "learning_rate": 6.159537708378441e-07,
+      "loss": 1.5452,
+      "step": 3416
+    },
+    {
+      "epoch": 0.23809357906839007,
+      "grad_norm": 0.7181375940406853,
+      "learning_rate": 6.159068726561158e-07,
+      "loss": 1.418,
+      "step": 3417
+    },
+    {
+      "epoch": 0.2381632581960074,
+      "grad_norm": 0.7104117242390996,
+      "learning_rate": 6.158599634088559e-07,
+      "loss": 1.538,
+      "step": 3418
+    },
+    {
+      "epoch": 0.23823293732362472,
+      "grad_norm": 0.7020110087325926,
+      "learning_rate": 6.158130430983127e-07,
+      "loss": 1.516,
+      "step": 3419
+    },
+    {
+      "epoch": 0.23830261645124204,
+      "grad_norm": 0.705561198667519,
+      "learning_rate": 6.157661117267347e-07,
+      "loss": 1.6158,
+      "step": 3420
+    },
+    {
+      "epoch": 0.23837229557885936,
+      "grad_norm": 0.6897473127244094,
+      "learning_rate": 6.157191692963706e-07,
+      "loss": 1.5299,
+      "step": 3421
+    },
+    {
+      "epoch": 0.23844197470647668,
+      "grad_norm": 0.7142173166774628,
+      "learning_rate": 6.156722158094705e-07,
+      "loss": 1.6315,
+      "step": 3422
+    },
+    {
+      "epoch": 0.238511653834094,
+      "grad_norm": 0.7311353735621938,
+      "learning_rate": 6.156252512682842e-07,
+      "loss": 1.5065,
+      "step": 3423
+    },
+    {
+      "epoch": 0.23858133296171133,
+      "grad_norm": 0.7386698453917857,
+      "learning_rate": 6.155782756750624e-07,
+      "loss": 1.5852,
+      "step": 3424
+    },
+    {
+      "epoch": 0.23865101208932865,
+      "grad_norm": 0.7174450568898096,
+      "learning_rate": 6.155312890320563e-07,
+      "loss": 1.6653,
+      "step": 3425
+    },
+    {
+      "epoch": 0.23872069121694597,
+      "grad_norm": 0.6757530045972093,
+      "learning_rate": 6.154842913415175e-07,
+      "loss": 1.519,
+      "step": 3426
+    },
+    {
+      "epoch": 0.2387903703445633,
+      "grad_norm": 0.7123687389973483,
+      "learning_rate": 6.154372826056983e-07,
+      "loss": 1.5688,
+      "step": 3427
+    },
+    {
+      "epoch": 0.23886004947218062,
+      "grad_norm": 0.731046582437802,
+      "learning_rate": 6.153902628268514e-07,
+      "loss": 1.6112,
+      "step": 3428
+    },
+    {
+      "epoch": 0.23892972859979794,
+      "grad_norm": 0.6917579901963694,
+      "learning_rate": 6.153432320072301e-07,
+      "loss": 1.5759,
+      "step": 3429
+    },
+    {
+      "epoch": 0.23899940772741526,
+      "grad_norm": 0.6981507803157918,
+      "learning_rate": 6.152961901490884e-07,
+      "loss": 1.6106,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2390690868550326,
+      "grad_norm": 0.7125774274588189,
+      "learning_rate": 6.152491372546804e-07,
+      "loss": 1.5118,
+      "step": 3431
+    },
+    {
+      "epoch": 0.2391387659826499,
+      "grad_norm": 0.7006860564252603,
+      "learning_rate": 6.15202073326261e-07,
+      "loss": 1.4689,
+      "step": 3432
+    },
+    {
+      "epoch": 0.23920844511026723,
+      "grad_norm": 0.7095641180879917,
+      "learning_rate": 6.151549983660856e-07,
+      "loss": 1.4697,
+      "step": 3433
+    },
+    {
+      "epoch": 0.23927812423788455,
+      "grad_norm": 0.7259126769570512,
+      "learning_rate": 6.151079123764104e-07,
+      "loss": 1.5463,
+      "step": 3434
+    },
+    {
+      "epoch": 0.23934780336550188,
+      "grad_norm": 0.8038698655154628,
+      "learning_rate": 6.150608153594915e-07,
+      "loss": 1.5724,
+      "step": 3435
+    },
+    {
+      "epoch": 0.2394174824931192,
+      "grad_norm": 0.7405163393205098,
+      "learning_rate": 6.150137073175859e-07,
+      "loss": 1.5711,
+      "step": 3436
+    },
+    {
+      "epoch": 0.23948716162073652,
+      "grad_norm": 0.697010624031155,
+      "learning_rate": 6.149665882529513e-07,
+      "loss": 1.4704,
+      "step": 3437
+    },
+    {
+      "epoch": 0.23955684074835384,
+      "grad_norm": 0.7037022988926063,
+      "learning_rate": 6.149194581678457e-07,
+      "loss": 1.5133,
+      "step": 3438
+    },
+    {
+      "epoch": 0.23962651987597114,
+      "grad_norm": 0.7656938726176664,
+      "learning_rate": 6.148723170645277e-07,
+      "loss": 1.5413,
+      "step": 3439
+    },
+    {
+      "epoch": 0.23969619900358846,
+      "grad_norm": 0.7258567001381848,
+      "learning_rate": 6.148251649452564e-07,
+      "loss": 1.4647,
+      "step": 3440
+    },
+    {
+      "epoch": 0.23976587813120578,
+      "grad_norm": 0.721650592199456,
+      "learning_rate": 6.147780018122912e-07,
+      "loss": 1.5456,
+      "step": 3441
+    },
+    {
+      "epoch": 0.2398355572588231,
+      "grad_norm": 0.7469447720018643,
+      "learning_rate": 6.147308276678926e-07,
+      "loss": 1.6895,
+      "step": 3442
+    },
+    {
+      "epoch": 0.23990523638644043,
+      "grad_norm": 0.6885474505548068,
+      "learning_rate": 6.14683642514321e-07,
+      "loss": 1.5356,
+      "step": 3443
+    },
+    {
+      "epoch": 0.23997491551405775,
+      "grad_norm": 0.7302283015243304,
+      "learning_rate": 6.146364463538377e-07,
+      "loss": 1.6192,
+      "step": 3444
+    },
+    {
+      "epoch": 0.24004459464167507,
+      "grad_norm": 0.7402686792467148,
+      "learning_rate": 6.145892391887046e-07,
+      "loss": 1.6829,
+      "step": 3445
+    },
+    {
+      "epoch": 0.2401142737692924,
+      "grad_norm": 0.752528153723735,
+      "learning_rate": 6.145420210211837e-07,
+      "loss": 1.5519,
+      "step": 3446
+    },
+    {
+      "epoch": 0.24018395289690972,
+      "grad_norm": 0.6629770876275467,
+      "learning_rate": 6.144947918535379e-07,
+      "loss": 1.4805,
+      "step": 3447
+    },
+    {
+      "epoch": 0.24025363202452704,
+      "grad_norm": 0.699902908697308,
+      "learning_rate": 6.144475516880307e-07,
+      "loss": 1.4926,
+      "step": 3448
+    },
+    {
+      "epoch": 0.24032331115214436,
+      "grad_norm": 0.7410783032423637,
+      "learning_rate": 6.144003005269256e-07,
+      "loss": 1.4714,
+      "step": 3449
+    },
+    {
+      "epoch": 0.2403929902797617,
+      "grad_norm": 0.7367406097734193,
+      "learning_rate": 6.143530383724872e-07,
+      "loss": 1.5405,
+      "step": 3450
+    },
+    {
+      "epoch": 0.240462669407379,
+      "grad_norm": 0.771099676513623,
+      "learning_rate": 6.143057652269803e-07,
+      "loss": 1.5602,
+      "step": 3451
+    },
+    {
+      "epoch": 0.24053234853499633,
+      "grad_norm": 0.7071012761640494,
+      "learning_rate": 6.142584810926704e-07,
+      "loss": 1.5461,
+      "step": 3452
+    },
+    {
+      "epoch": 0.24060202766261365,
+      "grad_norm": 0.7170107410005159,
+      "learning_rate": 6.142111859718235e-07,
+      "loss": 1.4895,
+      "step": 3453
+    },
+    {
+      "epoch": 0.24067170679023098,
+      "grad_norm": 0.7104958509461615,
+      "learning_rate": 6.141638798667058e-07,
+      "loss": 1.4503,
+      "step": 3454
+    },
+    {
+      "epoch": 0.2407413859178483,
+      "grad_norm": 0.8801778611348274,
+      "learning_rate": 6.141165627795848e-07,
+      "loss": 1.6851,
+      "step": 3455
+    },
+    {
+      "epoch": 0.24081106504546562,
+      "grad_norm": 0.7737450201788408,
+      "learning_rate": 6.140692347127276e-07,
+      "loss": 1.4547,
+      "step": 3456
+    },
+    {
+      "epoch": 0.24088074417308294,
+      "grad_norm": 0.7241405832766921,
+      "learning_rate": 6.140218956684024e-07,
+      "loss": 1.6129,
+      "step": 3457
+    },
+    {
+      "epoch": 0.24095042330070027,
+      "grad_norm": 0.7633311415281308,
+      "learning_rate": 6.139745456488778e-07,
+      "loss": 1.6202,
+      "step": 3458
+    },
+    {
+      "epoch": 0.2410201024283176,
+      "grad_norm": 0.7854016040714885,
+      "learning_rate": 6.139271846564229e-07,
+      "loss": 1.7454,
+      "step": 3459
+    },
+    {
+      "epoch": 0.2410897815559349,
+      "grad_norm": 0.7161975046816076,
+      "learning_rate": 6.138798126933074e-07,
+      "loss": 1.6831,
+      "step": 3460
+    },
+    {
+      "epoch": 0.24115946068355223,
+      "grad_norm": 0.6605931548102567,
+      "learning_rate": 6.138324297618012e-07,
+      "loss": 1.5774,
+      "step": 3461
+    },
+    {
+      "epoch": 0.24122913981116956,
+      "grad_norm": 0.6311640853920627,
+      "learning_rate": 6.137850358641754e-07,
+      "loss": 1.4281,
+      "step": 3462
+    },
+    {
+      "epoch": 0.24129881893878688,
+      "grad_norm": 0.7140109842514096,
+      "learning_rate": 6.137376310027008e-07,
+      "loss": 1.5915,
+      "step": 3463
+    },
+    {
+      "epoch": 0.2413684980664042,
+      "grad_norm": 0.7039807459381068,
+      "learning_rate": 6.136902151796495e-07,
+      "loss": 1.5709,
+      "step": 3464
+    },
+    {
+      "epoch": 0.24143817719402152,
+      "grad_norm": 0.6721860425881251,
+      "learning_rate": 6.136427883972935e-07,
+      "loss": 1.5871,
+      "step": 3465
+    },
+    {
+      "epoch": 0.24150785632163885,
+      "grad_norm": 0.7615669920967632,
+      "learning_rate": 6.135953506579057e-07,
+      "loss": 1.5575,
+      "step": 3466
+    },
+    {
+      "epoch": 0.24157753544925617,
+      "grad_norm": 0.6938093239176405,
+      "learning_rate": 6.135479019637593e-07,
+      "loss": 1.5081,
+      "step": 3467
+    },
+    {
+      "epoch": 0.2416472145768735,
+      "grad_norm": 0.6998087722630385,
+      "learning_rate": 6.135004423171284e-07,
+      "loss": 1.475,
+      "step": 3468
+    },
+    {
+      "epoch": 0.24171689370449082,
+      "grad_norm": 0.7284799755105115,
+      "learning_rate": 6.134529717202873e-07,
+      "loss": 1.5256,
+      "step": 3469
+    },
+    {
+      "epoch": 0.24178657283210814,
+      "grad_norm": 0.7339876042097843,
+      "learning_rate": 6.134054901755106e-07,
+      "loss": 1.4791,
+      "step": 3470
+    },
+    {
+      "epoch": 0.24185625195972546,
+      "grad_norm": 0.7344813016736725,
+      "learning_rate": 6.133579976850738e-07,
+      "loss": 1.6821,
+      "step": 3471
+    },
+    {
+      "epoch": 0.24192593108734278,
+      "grad_norm": 0.8034131484006056,
+      "learning_rate": 6.133104942512532e-07,
+      "loss": 1.5536,
+      "step": 3472
+    },
+    {
+      "epoch": 0.2419956102149601,
+      "grad_norm": 0.7091929112333546,
+      "learning_rate": 6.132629798763249e-07,
+      "loss": 1.5343,
+      "step": 3473
+    },
+    {
+      "epoch": 0.24206528934257743,
+      "grad_norm": 0.7372457302786409,
+      "learning_rate": 6.13215454562566e-07,
+      "loss": 1.5681,
+      "step": 3474
+    },
+    {
+      "epoch": 0.24213496847019475,
+      "grad_norm": 0.7163122501788207,
+      "learning_rate": 6.131679183122539e-07,
+      "loss": 1.5721,
+      "step": 3475
+    },
+    {
+      "epoch": 0.24220464759781207,
+      "grad_norm": 0.7644696332870136,
+      "learning_rate": 6.131203711276669e-07,
+      "loss": 1.5467,
+      "step": 3476
+    },
+    {
+      "epoch": 0.2422743267254294,
+      "grad_norm": 0.659718083288887,
+      "learning_rate": 6.130728130110833e-07,
+      "loss": 1.5017,
+      "step": 3477
+    },
+    {
+      "epoch": 0.24234400585304672,
+      "grad_norm": 0.6883044205305839,
+      "learning_rate": 6.130252439647823e-07,
+      "loss": 1.5113,
+      "step": 3478
+    },
+    {
+      "epoch": 0.24241368498066404,
+      "grad_norm": 0.7201350430252552,
+      "learning_rate": 6.129776639910434e-07,
+      "loss": 1.5899,
+      "step": 3479
+    },
+    {
+      "epoch": 0.24248336410828136,
+      "grad_norm": 0.7132112424660706,
+      "learning_rate": 6.129300730921468e-07,
+      "loss": 1.4527,
+      "step": 3480
+    },
+    {
+      "epoch": 0.24255304323589869,
+      "grad_norm": 0.7118297627685076,
+      "learning_rate": 6.128824712703734e-07,
+      "loss": 1.6241,
+      "step": 3481
+    },
+    {
+      "epoch": 0.242622722363516,
+      "grad_norm": 0.7700354394842043,
+      "learning_rate": 6.128348585280039e-07,
+      "loss": 1.5645,
+      "step": 3482
+    },
+    {
+      "epoch": 0.24269240149113333,
+      "grad_norm": 0.7532057834209641,
+      "learning_rate": 6.127872348673204e-07,
+      "loss": 1.5646,
+      "step": 3483
+    },
+    {
+      "epoch": 0.24276208061875065,
+      "grad_norm": 0.7249802886382719,
+      "learning_rate": 6.127396002906049e-07,
+      "loss": 1.4486,
+      "step": 3484
+    },
+    {
+      "epoch": 0.24283175974636798,
+      "grad_norm": 0.7016431329438683,
+      "learning_rate": 6.126919548001403e-07,
+      "loss": 1.6213,
+      "step": 3485
+    },
+    {
+      "epoch": 0.2429014388739853,
+      "grad_norm": 0.7030641444243805,
+      "learning_rate": 6.126442983982096e-07,
+      "loss": 1.5588,
+      "step": 3486
+    },
+    {
+      "epoch": 0.24297111800160262,
+      "grad_norm": 0.7413791208700851,
+      "learning_rate": 6.125966310870968e-07,
+      "loss": 1.6794,
+      "step": 3487
+    },
+    {
+      "epoch": 0.24304079712921994,
+      "grad_norm": 0.6779401090861621,
+      "learning_rate": 6.125489528690863e-07,
+      "loss": 1.6256,
+      "step": 3488
+    },
+    {
+      "epoch": 0.24311047625683727,
+      "grad_norm": 0.6852961562158302,
+      "learning_rate": 6.125012637464628e-07,
+      "loss": 1.578,
+      "step": 3489
+    },
+    {
+      "epoch": 0.2431801553844546,
+      "grad_norm": 0.7876391800882989,
+      "learning_rate": 6.124535637215116e-07,
+      "loss": 1.6865,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2432498345120719,
+      "grad_norm": 0.7089052160927828,
+      "learning_rate": 6.124058527965189e-07,
+      "loss": 1.4728,
+      "step": 3491
+    },
+    {
+      "epoch": 0.24331951363968923,
+      "grad_norm": 0.6710252479864719,
+      "learning_rate": 6.123581309737707e-07,
+      "loss": 1.5328,
+      "step": 3492
+    },
+    {
+      "epoch": 0.24338919276730656,
+      "grad_norm": 7.417390349939993,
+      "learning_rate": 6.12310398255554e-07,
+      "loss": 1.468,
+      "step": 3493
+    },
+    {
+      "epoch": 0.24345887189492388,
+      "grad_norm": 0.6818307687602854,
+      "learning_rate": 6.122626546441567e-07,
+      "loss": 1.6087,
+      "step": 3494
+    },
+    {
+      "epoch": 0.2435285510225412,
+      "grad_norm": 0.7113519070432572,
+      "learning_rate": 6.122149001418661e-07,
+      "loss": 1.5702,
+      "step": 3495
+    },
+    {
+      "epoch": 0.24359823015015852,
+      "grad_norm": 0.7331933417391607,
+      "learning_rate": 6.121671347509712e-07,
+      "loss": 1.6366,
+      "step": 3496
+    },
+    {
+      "epoch": 0.24366790927777585,
+      "grad_norm": 0.7353912377560792,
+      "learning_rate": 6.121193584737607e-07,
+      "loss": 1.4549,
+      "step": 3497
+    },
+    {
+      "epoch": 0.24373758840539317,
+      "grad_norm": 0.747218662762241,
+      "learning_rate": 6.120715713125245e-07,
+      "loss": 1.6776,
+      "step": 3498
+    },
+    {
+      "epoch": 0.2438072675330105,
+      "grad_norm": 0.7783135396187291,
+      "learning_rate": 6.120237732695521e-07,
+      "loss": 1.5549,
+      "step": 3499
+    },
+    {
+      "epoch": 0.2438769466606278,
+      "grad_norm": 0.7866557794969244,
+      "learning_rate": 6.119759643471347e-07,
+      "loss": 1.3877,
+      "step": 3500
+    },
+    {
+      "epoch": 0.24394662578824514,
+      "grad_norm": 0.7104409997433944,
+      "learning_rate": 6.11928144547563e-07,
+      "loss": 1.5761,
+      "step": 3501
+    },
+    {
+      "epoch": 0.24401630491586246,
+      "grad_norm": 0.7434088447922035,
+      "learning_rate": 6.118803138731287e-07,
+      "loss": 1.513,
+      "step": 3502
+    },
+    {
+      "epoch": 0.24408598404347978,
+      "grad_norm": 0.719581279474016,
+      "learning_rate": 6.118324723261241e-07,
+      "loss": 1.4972,
+      "step": 3503
+    },
+    {
+      "epoch": 0.2441556631710971,
+      "grad_norm": 0.643472580546137,
+      "learning_rate": 6.117846199088417e-07,
+      "loss": 1.4641,
+      "step": 3504
+    },
+    {
+      "epoch": 0.24422534229871443,
+      "grad_norm": 0.7445412858204775,
+      "learning_rate": 6.117367566235748e-07,
+      "loss": 1.5576,
+      "step": 3505
+    },
+    {
+      "epoch": 0.24429502142633175,
+      "grad_norm": 0.7239385537409243,
+      "learning_rate": 6.11688882472617e-07,
+      "loss": 1.5952,
+      "step": 3506
+    },
+    {
+      "epoch": 0.24436470055394907,
+      "grad_norm": 0.7764917647022519,
+      "learning_rate": 6.116409974582625e-07,
+      "loss": 1.6206,
+      "step": 3507
+    },
+    {
+      "epoch": 0.2444343796815664,
+      "grad_norm": 0.74202044266512,
+      "learning_rate": 6.115931015828062e-07,
+      "loss": 1.5645,
+      "step": 3508
+    },
+    {
+      "epoch": 0.24450405880918372,
+      "grad_norm": 0.6935767282043237,
+      "learning_rate": 6.115451948485431e-07,
+      "loss": 1.5034,
+      "step": 3509
+    },
+    {
+      "epoch": 0.24457373793680104,
+      "grad_norm": 0.7039628910097253,
+      "learning_rate": 6.114972772577693e-07,
+      "loss": 1.5239,
+      "step": 3510
+    },
+    {
+      "epoch": 0.24464341706441836,
+      "grad_norm": 0.6874285293399682,
+      "learning_rate": 6.11449348812781e-07,
+      "loss": 1.6456,
+      "step": 3511
+    },
+    {
+      "epoch": 0.24471309619203568,
+      "grad_norm": 0.6522379381976608,
+      "learning_rate": 6.11401409515875e-07,
+      "loss": 1.5646,
+      "step": 3512
+    },
+    {
+      "epoch": 0.244782775319653,
+      "grad_norm": 0.8171749552425288,
+      "learning_rate": 6.113534593693486e-07,
+      "loss": 1.5602,
+      "step": 3513
+    },
+    {
+      "epoch": 0.24485245444727033,
+      "grad_norm": 0.7396199021021814,
+      "learning_rate": 6.113054983754999e-07,
+      "loss": 1.5303,
+      "step": 3514
+    },
+    {
+      "epoch": 0.24492213357488765,
+      "grad_norm": 0.7662564325697268,
+      "learning_rate": 6.11257526536627e-07,
+      "loss": 1.5687,
+      "step": 3515
+    },
+    {
+      "epoch": 0.24499181270250497,
+      "grad_norm": 0.6908179397521864,
+      "learning_rate": 6.11209543855029e-07,
+      "loss": 1.5637,
+      "step": 3516
+    },
+    {
+      "epoch": 0.2450614918301223,
+      "grad_norm": 0.6836038980227871,
+      "learning_rate": 6.111615503330051e-07,
+      "loss": 1.5156,
+      "step": 3517
+    },
+    {
+      "epoch": 0.24513117095773962,
+      "grad_norm": 0.73581436271446,
+      "learning_rate": 6.111135459728556e-07,
+      "loss": 1.6565,
+      "step": 3518
+    },
+    {
+      "epoch": 0.24520085008535694,
+      "grad_norm": 0.7405195856434742,
+      "learning_rate": 6.110655307768808e-07,
+      "loss": 1.4523,
+      "step": 3519
+    },
+    {
+      "epoch": 0.24527052921297426,
+      "grad_norm": 0.7585536397564181,
+      "learning_rate": 6.110175047473816e-07,
+      "loss": 1.3932,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2453402083405916,
+      "grad_norm": 0.709276744464445,
+      "learning_rate": 6.109694678866594e-07,
+      "loss": 1.4675,
+      "step": 3521
+    },
+    {
+      "epoch": 0.2454098874682089,
+      "grad_norm": 0.750704891947952,
+      "learning_rate": 6.109214201970165e-07,
+      "loss": 1.4458,
+      "step": 3522
+    },
+    {
+      "epoch": 0.24547956659582623,
+      "grad_norm": 0.772678393930129,
+      "learning_rate": 6.108733616807554e-07,
+      "loss": 1.6841,
+      "step": 3523
+    },
+    {
+      "epoch": 0.24554924572344355,
+      "grad_norm": 0.6817202500029117,
+      "learning_rate": 6.10825292340179e-07,
+      "loss": 1.5587,
+      "step": 3524
+    },
+    {
+      "epoch": 0.24561892485106088,
+      "grad_norm": 0.8040107414482522,
+      "learning_rate": 6.10777212177591e-07,
+      "loss": 1.5774,
+      "step": 3525
+    },
+    {
+      "epoch": 0.2456886039786782,
+      "grad_norm": 1.796804413019553,
+      "learning_rate": 6.107291211952956e-07,
+      "loss": 1.4213,
+      "step": 3526
+    },
+    {
+      "epoch": 0.24575828310629552,
+      "grad_norm": 0.7693183737224776,
+      "learning_rate": 6.106810193955972e-07,
+      "loss": 1.6151,
+      "step": 3527
+    },
+    {
+      "epoch": 0.24582796223391284,
+      "grad_norm": 0.7251457431661789,
+      "learning_rate": 6.10632906780801e-07,
+      "loss": 1.5434,
+      "step": 3528
+    },
+    {
+      "epoch": 0.24589764136153017,
+      "grad_norm": 0.7261229116118169,
+      "learning_rate": 6.105847833532127e-07,
+      "loss": 1.4882,
+      "step": 3529
+    },
+    {
+      "epoch": 0.2459673204891475,
+      "grad_norm": 0.7457010223359175,
+      "learning_rate": 6.105366491151387e-07,
+      "loss": 1.7156,
+      "step": 3530
+    },
+    {
+      "epoch": 0.24603699961676478,
+      "grad_norm": 0.7071322583151507,
+      "learning_rate": 6.104885040688851e-07,
+      "loss": 1.5663,
+      "step": 3531
+    },
+    {
+      "epoch": 0.2461066787443821,
+      "grad_norm": 0.6879911743671988,
+      "learning_rate": 6.104403482167596e-07,
+      "loss": 1.5783,
+      "step": 3532
+    },
+    {
+      "epoch": 0.24617635787199943,
+      "grad_norm": 0.744748622425573,
+      "learning_rate": 6.103921815610699e-07,
+      "loss": 1.6289,
+      "step": 3533
+    },
+    {
+      "epoch": 0.24624603699961675,
+      "grad_norm": 0.691139277317366,
+      "learning_rate": 6.10344004104124e-07,
+      "loss": 1.4919,
+      "step": 3534
+    },
+    {
+      "epoch": 0.24631571612723407,
+      "grad_norm": 0.6674381484094699,
+      "learning_rate": 6.102958158482309e-07,
+      "loss": 1.344,
+      "step": 3535
+    },
+    {
+      "epoch": 0.2463853952548514,
+      "grad_norm": 0.7468365533757317,
+      "learning_rate": 6.102476167956997e-07,
+      "loss": 1.5975,
+      "step": 3536
+    },
+    {
+      "epoch": 0.24645507438246872,
+      "grad_norm": 0.7617737698264097,
+      "learning_rate": 6.101994069488403e-07,
+      "loss": 1.6178,
+      "step": 3537
+    },
+    {
+      "epoch": 0.24652475351008604,
+      "grad_norm": 0.6924214940752028,
+      "learning_rate": 6.10151186309963e-07,
+      "loss": 1.5002,
+      "step": 3538
+    },
+    {
+      "epoch": 0.24659443263770336,
+      "grad_norm": 0.7349604377014846,
+      "learning_rate": 6.101029548813787e-07,
+      "loss": 1.6046,
+      "step": 3539
+    },
+    {
+      "epoch": 0.2466641117653207,
+      "grad_norm": 0.7471521985422922,
+      "learning_rate": 6.100547126653986e-07,
+      "loss": 1.5667,
+      "step": 3540
+    },
+    {
+      "epoch": 0.246733790892938,
+      "grad_norm": 0.7367598222463619,
+      "learning_rate": 6.100064596643346e-07,
+      "loss": 1.6448,
+      "step": 3541
+    },
+    {
+      "epoch": 0.24680347002055533,
+      "grad_norm": 0.7776711532694474,
+      "learning_rate": 6.099581958804993e-07,
+      "loss": 1.4532,
+      "step": 3542
+    },
+    {
+      "epoch": 0.24687314914817265,
+      "grad_norm": 0.7806167018788672,
+      "learning_rate": 6.099099213162053e-07,
+      "loss": 1.638,
+      "step": 3543
+    },
+    {
+      "epoch": 0.24694282827578998,
+      "grad_norm": 0.7661398733958934,
+      "learning_rate": 6.098616359737661e-07,
+      "loss": 1.6671,
+      "step": 3544
+    },
+    {
+      "epoch": 0.2470125074034073,
+      "grad_norm": 0.7234253141818819,
+      "learning_rate": 6.098133398554956e-07,
+      "loss": 1.524,
+      "step": 3545
+    },
+    {
+      "epoch": 0.24708218653102462,
+      "grad_norm": 0.7359219209086516,
+      "learning_rate": 6.097650329637085e-07,
+      "loss": 1.6014,
+      "step": 3546
+    },
+    {
+      "epoch": 0.24715186565864194,
+      "grad_norm": 0.681303466428281,
+      "learning_rate": 6.097167153007195e-07,
+      "loss": 1.6173,
+      "step": 3547
+    },
+    {
+      "epoch": 0.24722154478625927,
+      "grad_norm": 0.6865820922181248,
+      "learning_rate": 6.096683868688443e-07,
+      "loss": 1.5139,
+      "step": 3548
+    },
+    {
+      "epoch": 0.2472912239138766,
+      "grad_norm": 0.6920372454919099,
+      "learning_rate": 6.096200476703986e-07,
+      "loss": 1.3954,
+      "step": 3549
+    },
+    {
+      "epoch": 0.2473609030414939,
+      "grad_norm": 0.7476041916829527,
+      "learning_rate": 6.095716977076992e-07,
+      "loss": 1.3092,
+      "step": 3550
+    },
+    {
+      "epoch": 0.24743058216911124,
+      "grad_norm": 0.8072680023419683,
+      "learning_rate": 6.095233369830628e-07,
+      "loss": 1.5462,
+      "step": 3551
+    },
+    {
+      "epoch": 0.24750026129672856,
+      "grad_norm": 0.8213723342959321,
+      "learning_rate": 6.094749654988073e-07,
+      "loss": 1.7721,
+      "step": 3552
+    },
+    {
+      "epoch": 0.24756994042434588,
+      "grad_norm": 0.6873792435636162,
+      "learning_rate": 6.094265832572506e-07,
+      "loss": 1.5889,
+      "step": 3553
+    },
+    {
+      "epoch": 0.2476396195519632,
+      "grad_norm": 0.7316708923206956,
+      "learning_rate": 6.093781902607114e-07,
+      "loss": 1.523,
+      "step": 3554
+    },
+    {
+      "epoch": 0.24770929867958053,
+      "grad_norm": 0.7469485247831148,
+      "learning_rate": 6.093297865115086e-07,
+      "loss": 1.6226,
+      "step": 3555
+    },
+    {
+      "epoch": 0.24777897780719785,
+      "grad_norm": 0.6725940287268622,
+      "learning_rate": 6.092813720119618e-07,
+      "loss": 1.435,
+      "step": 3556
+    },
+    {
+      "epoch": 0.24784865693481517,
+      "grad_norm": 0.6954422705334324,
+      "learning_rate": 6.092329467643914e-07,
+      "loss": 1.4995,
+      "step": 3557
+    },
+    {
+      "epoch": 0.2479183360624325,
+      "grad_norm": 0.7117543686613589,
+      "learning_rate": 6.091845107711177e-07,
+      "loss": 1.384,
+      "step": 3558
+    },
+    {
+      "epoch": 0.24798801519004982,
+      "grad_norm": 0.7432403113648027,
+      "learning_rate": 6.091360640344619e-07,
+      "loss": 1.6325,
+      "step": 3559
+    },
+    {
+      "epoch": 0.24805769431766714,
+      "grad_norm": 0.7451833768308808,
+      "learning_rate": 6.09087606556746e-07,
+      "loss": 1.6012,
+      "step": 3560
+    },
+    {
+      "epoch": 0.24812737344528446,
+      "grad_norm": 1.4392707418778636,
+      "learning_rate": 6.090391383402919e-07,
+      "loss": 1.5517,
+      "step": 3561
+    },
+    {
+      "epoch": 0.24819705257290178,
+      "grad_norm": 0.726831617640144,
+      "learning_rate": 6.089906593874222e-07,
+      "loss": 1.492,
+      "step": 3562
+    },
+    {
+      "epoch": 0.2482667317005191,
+      "grad_norm": 0.768420711031291,
+      "learning_rate": 6.089421697004604e-07,
+      "loss": 1.6141,
+      "step": 3563
+    },
+    {
+      "epoch": 0.24833641082813643,
+      "grad_norm": 0.7361593118586319,
+      "learning_rate": 6.088936692817301e-07,
+      "loss": 1.7136,
+      "step": 3564
+    },
+    {
+      "epoch": 0.24840608995575375,
+      "grad_norm": 0.7033722393985278,
+      "learning_rate": 6.088451581335555e-07,
+      "loss": 1.4913,
+      "step": 3565
+    },
+    {
+      "epoch": 0.24847576908337107,
+      "grad_norm": 0.7232376051382627,
+      "learning_rate": 6.087966362582614e-07,
+      "loss": 1.5684,
+      "step": 3566
+    },
+    {
+      "epoch": 0.2485454482109884,
+      "grad_norm": 0.7350353882547026,
+      "learning_rate": 6.087481036581729e-07,
+      "loss": 1.54,
+      "step": 3567
+    },
+    {
+      "epoch": 0.24861512733860572,
+      "grad_norm": 0.7768085786334259,
+      "learning_rate": 6.08699560335616e-07,
+      "loss": 1.6108,
+      "step": 3568
+    },
+    {
+      "epoch": 0.24868480646622304,
+      "grad_norm": 0.7023493785204465,
+      "learning_rate": 6.08651006292917e-07,
+      "loss": 1.4761,
+      "step": 3569
+    },
+    {
+      "epoch": 0.24875448559384036,
+      "grad_norm": 0.699437447824012,
+      "learning_rate": 6.086024415324025e-07,
+      "loss": 1.5039,
+      "step": 3570
+    },
+    {
+      "epoch": 0.24882416472145769,
+      "grad_norm": 0.7063756052529659,
+      "learning_rate": 6.085538660564001e-07,
+      "loss": 1.4976,
+      "step": 3571
+    },
+    {
+      "epoch": 0.248893843849075,
+      "grad_norm": 0.7142139878054212,
+      "learning_rate": 6.085052798672374e-07,
+      "loss": 1.3809,
+      "step": 3572
+    },
+    {
+      "epoch": 0.24896352297669233,
+      "grad_norm": 0.6959967770546999,
+      "learning_rate": 6.084566829672429e-07,
+      "loss": 1.5112,
+      "step": 3573
+    },
+    {
+      "epoch": 0.24903320210430965,
+      "grad_norm": 0.6939186177106823,
+      "learning_rate": 6.084080753587453e-07,
+      "loss": 1.5446,
+      "step": 3574
+    },
+    {
+      "epoch": 0.24910288123192698,
+      "grad_norm": 0.7029487486994859,
+      "learning_rate": 6.083594570440742e-07,
+      "loss": 1.4871,
+      "step": 3575
+    },
+    {
+      "epoch": 0.2491725603595443,
+      "grad_norm": 0.742816701423201,
+      "learning_rate": 6.083108280255593e-07,
+      "loss": 1.6708,
+      "step": 3576
+    },
+    {
+      "epoch": 0.24924223948716162,
+      "grad_norm": 0.7137477992334339,
+      "learning_rate": 6.08262188305531e-07,
+      "loss": 1.431,
+      "step": 3577
+    },
+    {
+      "epoch": 0.24931191861477894,
+      "grad_norm": 0.7036883536953713,
+      "learning_rate": 6.082135378863204e-07,
+      "loss": 1.5361,
+      "step": 3578
+    },
+    {
+      "epoch": 0.24938159774239627,
+      "grad_norm": 0.807427431662475,
+      "learning_rate": 6.081648767702589e-07,
+      "loss": 1.6449,
+      "step": 3579
+    },
+    {
+      "epoch": 0.2494512768700136,
+      "grad_norm": 0.6729778495040805,
+      "learning_rate": 6.081162049596781e-07,
+      "loss": 1.5132,
+      "step": 3580
+    },
+    {
+      "epoch": 0.2495209559976309,
+      "grad_norm": 0.6724953431528643,
+      "learning_rate": 6.080675224569108e-07,
+      "loss": 1.4788,
+      "step": 3581
+    },
+    {
+      "epoch": 0.24959063512524823,
+      "grad_norm": 0.7154576876166014,
+      "learning_rate": 6.080188292642901e-07,
+      "loss": 1.5311,
+      "step": 3582
+    },
+    {
+      "epoch": 0.24966031425286556,
+      "grad_norm": 0.7493444527994712,
+      "learning_rate": 6.07970125384149e-07,
+      "loss": 1.5591,
+      "step": 3583
+    },
+    {
+      "epoch": 0.24972999338048288,
+      "grad_norm": 0.7266179891707369,
+      "learning_rate": 6.079214108188219e-07,
+      "loss": 1.5179,
+      "step": 3584
+    },
+    {
+      "epoch": 0.2497996725081002,
+      "grad_norm": 0.7217565810210871,
+      "learning_rate": 6.07872685570643e-07,
+      "loss": 1.5389,
+      "step": 3585
+    },
+    {
+      "epoch": 0.24986935163571752,
+      "grad_norm": 0.6896161569382101,
+      "learning_rate": 6.078239496419476e-07,
+      "loss": 1.5998,
+      "step": 3586
+    },
+    {
+      "epoch": 0.24993903076333485,
+      "grad_norm": 0.7371361263420952,
+      "learning_rate": 6.07775203035071e-07,
+      "loss": 1.5602,
+      "step": 3587
+    },
+    {
+      "epoch": 0.25000870989095214,
+      "grad_norm": 0.7334616155478288,
+      "learning_rate": 6.077264457523493e-07,
+      "loss": 1.5357,
+      "step": 3588
+    },
+    {
+      "epoch": 0.2500783890185695,
+      "grad_norm": 0.6977451408653291,
+      "learning_rate": 6.076776777961192e-07,
+      "loss": 1.7069,
+      "step": 3589
+    },
+    {
+      "epoch": 0.2501480681461868,
+      "grad_norm": 0.6836085763879641,
+      "learning_rate": 6.076288991687174e-07,
+      "loss": 1.5012,
+      "step": 3590
+    },
+    {
+      "epoch": 0.25021774727380414,
+      "grad_norm": 0.724668401614909,
+      "learning_rate": 6.075801098724819e-07,
+      "loss": 1.4189,
+      "step": 3591
+    },
+    {
+      "epoch": 0.25028742640142143,
+      "grad_norm": 0.7218857976482577,
+      "learning_rate": 6.075313099097505e-07,
+      "loss": 1.533,
+      "step": 3592
+    },
+    {
+      "epoch": 0.2503571055290388,
+      "grad_norm": 0.7454390129326977,
+      "learning_rate": 6.07482499282862e-07,
+      "loss": 1.6404,
+      "step": 3593
+    },
+    {
+      "epoch": 0.2504267846566561,
+      "grad_norm": 0.7276507976619264,
+      "learning_rate": 6.074336779941551e-07,
+      "loss": 1.5575,
+      "step": 3594
+    },
+    {
+      "epoch": 0.2504964637842734,
+      "grad_norm": 0.8181538349091576,
+      "learning_rate": 6.0738484604597e-07,
+      "loss": 1.5305,
+      "step": 3595
+    },
+    {
+      "epoch": 0.2505661429118907,
+      "grad_norm": 0.7022901688141522,
+      "learning_rate": 6.073360034406465e-07,
+      "loss": 1.4621,
+      "step": 3596
+    },
+    {
+      "epoch": 0.25063582203950807,
+      "grad_norm": 0.7053344078146021,
+      "learning_rate": 6.072871501805251e-07,
+      "loss": 1.6526,
+      "step": 3597
+    },
+    {
+      "epoch": 0.25070550116712537,
+      "grad_norm": 0.6812712012110201,
+      "learning_rate": 6.072382862679472e-07,
+      "loss": 1.4954,
+      "step": 3598
+    },
+    {
+      "epoch": 0.2507751802947427,
+      "grad_norm": 0.6865562783180164,
+      "learning_rate": 6.071894117052545e-07,
+      "loss": 1.5444,
+      "step": 3599
+    },
+    {
+      "epoch": 0.25084485942236,
+      "grad_norm": 0.7130584509089323,
+      "learning_rate": 6.071405264947889e-07,
+      "loss": 1.5434,
+      "step": 3600
+    },
+    {
+      "epoch": 0.25091453854997736,
+      "grad_norm": 0.7343554727384859,
+      "learning_rate": 6.070916306388933e-07,
+      "loss": 1.4086,
+      "step": 3601
+    },
+    {
+      "epoch": 0.25098421767759466,
+      "grad_norm": 0.7308227185668241,
+      "learning_rate": 6.070427241399108e-07,
+      "loss": 1.533,
+      "step": 3602
+    },
+    {
+      "epoch": 0.251053896805212,
+      "grad_norm": 0.7054271106104274,
+      "learning_rate": 6.06993807000185e-07,
+      "loss": 1.5204,
+      "step": 3603
+    },
+    {
+      "epoch": 0.2511235759328293,
+      "grad_norm": 0.7238952984007729,
+      "learning_rate": 6.069448792220603e-07,
+      "loss": 1.5857,
+      "step": 3604
+    },
+    {
+      "epoch": 0.25119325506044665,
+      "grad_norm": 0.7619163189051473,
+      "learning_rate": 6.068959408078813e-07,
+      "loss": 1.6456,
+      "step": 3605
+    },
+    {
+      "epoch": 0.25126293418806395,
+      "grad_norm": 0.7691821142914758,
+      "learning_rate": 6.068469917599934e-07,
+      "loss": 1.6207,
+      "step": 3606
+    },
+    {
+      "epoch": 0.2513326133156813,
+      "grad_norm": 0.7844297210575869,
+      "learning_rate": 6.067980320807421e-07,
+      "loss": 1.6005,
+      "step": 3607
+    },
+    {
+      "epoch": 0.2514022924432986,
+      "grad_norm": 0.767312581828994,
+      "learning_rate": 6.067490617724739e-07,
+      "loss": 1.5474,
+      "step": 3608
+    },
+    {
+      "epoch": 0.25147197157091594,
+      "grad_norm": 0.7466302509493756,
+      "learning_rate": 6.067000808375353e-07,
+      "loss": 1.5245,
+      "step": 3609
+    },
+    {
+      "epoch": 0.25154165069853324,
+      "grad_norm": 0.7194713716465715,
+      "learning_rate": 6.066510892782737e-07,
+      "loss": 1.5255,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2516113298261506,
+      "grad_norm": 0.6903691577711621,
+      "learning_rate": 6.066020870970368e-07,
+      "loss": 1.5452,
+      "step": 3611
+    },
+    {
+      "epoch": 0.2516810089537679,
+      "grad_norm": 0.8469633001258798,
+      "learning_rate": 6.065530742961731e-07,
+      "loss": 1.5305,
+      "step": 3612
+    },
+    {
+      "epoch": 0.25175068808138523,
+      "grad_norm": 0.6957640057146053,
+      "learning_rate": 6.065040508780312e-07,
+      "loss": 1.4241,
+      "step": 3613
+    },
+    {
+      "epoch": 0.2518203672090025,
+      "grad_norm": 0.7070417869184299,
+      "learning_rate": 6.064550168449603e-07,
+      "loss": 1.5566,
+      "step": 3614
+    },
+    {
+      "epoch": 0.2518900463366199,
+      "grad_norm": 0.7573782379837182,
+      "learning_rate": 6.064059721993104e-07,
+      "loss": 1.5595,
+      "step": 3615
+    },
+    {
+      "epoch": 0.25195972546423717,
+      "grad_norm": 0.7678449791648668,
+      "learning_rate": 6.063569169434319e-07,
+      "loss": 1.5895,
+      "step": 3616
+    },
+    {
+      "epoch": 0.2520294045918545,
+      "grad_norm": 0.7780040963314879,
+      "learning_rate": 6.063078510796754e-07,
+      "loss": 1.5685,
+      "step": 3617
+    },
+    {
+      "epoch": 0.2520990837194718,
+      "grad_norm": 0.697366601001487,
+      "learning_rate": 6.062587746103924e-07,
+      "loss": 1.6299,
+      "step": 3618
+    },
+    {
+      "epoch": 0.25216876284708917,
+      "grad_norm": 0.6918393842439775,
+      "learning_rate": 6.062096875379346e-07,
+      "loss": 1.6065,
+      "step": 3619
+    },
+    {
+      "epoch": 0.25223844197470646,
+      "grad_norm": 0.7765055919439487,
+      "learning_rate": 6.061605898646545e-07,
+      "loss": 1.6526,
+      "step": 3620
+    },
+    {
+      "epoch": 0.2523081211023238,
+      "grad_norm": 0.7472761617400331,
+      "learning_rate": 6.06111481592905e-07,
+      "loss": 1.542,
+      "step": 3621
+    },
+    {
+      "epoch": 0.2523778002299411,
+      "grad_norm": 0.7486342805260572,
+      "learning_rate": 6.060623627250391e-07,
+      "loss": 1.6035,
+      "step": 3622
+    },
+    {
+      "epoch": 0.25244747935755846,
+      "grad_norm": 0.7500589843074995,
+      "learning_rate": 6.060132332634111e-07,
+      "loss": 1.5478,
+      "step": 3623
+    },
+    {
+      "epoch": 0.25251715848517575,
+      "grad_norm": 0.7331824331258279,
+      "learning_rate": 6.059640932103753e-07,
+      "loss": 1.4816,
+      "step": 3624
+    },
+    {
+      "epoch": 0.2525868376127931,
+      "grad_norm": 0.6785392683089051,
+      "learning_rate": 6.059149425682865e-07,
+      "loss": 1.4475,
+      "step": 3625
+    },
+    {
+      "epoch": 0.2526565167404104,
+      "grad_norm": 0.7152394572979834,
+      "learning_rate": 6.058657813395e-07,
+      "loss": 1.6863,
+      "step": 3626
+    },
+    {
+      "epoch": 0.25272619586802775,
+      "grad_norm": 0.7231210956234625,
+      "learning_rate": 6.058166095263719e-07,
+      "loss": 1.583,
+      "step": 3627
+    },
+    {
+      "epoch": 0.25279587499564504,
+      "grad_norm": 0.7245911280542969,
+      "learning_rate": 6.057674271312585e-07,
+      "loss": 1.64,
+      "step": 3628
+    },
+    {
+      "epoch": 0.2528655541232624,
+      "grad_norm": 0.7097081774448422,
+      "learning_rate": 6.057182341565169e-07,
+      "loss": 1.5683,
+      "step": 3629
+    },
+    {
+      "epoch": 0.2529352332508797,
+      "grad_norm": 0.7191293124979857,
+      "learning_rate": 6.056690306045043e-07,
+      "loss": 1.5606,
+      "step": 3630
+    },
+    {
+      "epoch": 0.25300491237849704,
+      "grad_norm": 0.7614871453223097,
+      "learning_rate": 6.056198164775786e-07,
+      "loss": 1.5594,
+      "step": 3631
+    },
+    {
+      "epoch": 0.25307459150611433,
+      "grad_norm": 0.8002734672962935,
+      "learning_rate": 6.055705917780987e-07,
+      "loss": 1.5608,
+      "step": 3632
+    },
+    {
+      "epoch": 0.2531442706337317,
+      "grad_norm": 0.7754927308854186,
+      "learning_rate": 6.055213565084229e-07,
+      "loss": 1.678,
+      "step": 3633
+    },
+    {
+      "epoch": 0.253213949761349,
+      "grad_norm": 0.832975102500679,
+      "learning_rate": 6.054721106709111e-07,
+      "loss": 1.5249,
+      "step": 3634
+    },
+    {
+      "epoch": 0.25328362888896633,
+      "grad_norm": 0.697784227911405,
+      "learning_rate": 6.054228542679231e-07,
+      "loss": 1.5183,
+      "step": 3635
+    },
+    {
+      "epoch": 0.2533533080165836,
+      "grad_norm": 0.7001080333414058,
+      "learning_rate": 6.053735873018195e-07,
+      "loss": 1.5209,
+      "step": 3636
+    },
+    {
+      "epoch": 0.253422987144201,
+      "grad_norm": 0.6923807006499639,
+      "learning_rate": 6.053243097749611e-07,
+      "loss": 1.5297,
+      "step": 3637
+    },
+    {
+      "epoch": 0.25349266627181827,
+      "grad_norm": 0.8379293298094277,
+      "learning_rate": 6.052750216897095e-07,
+      "loss": 1.5739,
+      "step": 3638
+    },
+    {
+      "epoch": 0.2535623453994356,
+      "grad_norm": 0.6882683163202826,
+      "learning_rate": 6.052257230484266e-07,
+      "loss": 1.4844,
+      "step": 3639
+    },
+    {
+      "epoch": 0.2536320245270529,
+      "grad_norm": 0.7033906157479533,
+      "learning_rate": 6.05176413853475e-07,
+      "loss": 1.435,
+      "step": 3640
+    },
+    {
+      "epoch": 0.25370170365467026,
+      "grad_norm": 0.6733432624515764,
+      "learning_rate": 6.051270941072176e-07,
+      "loss": 1.3151,
+      "step": 3641
+    },
+    {
+      "epoch": 0.25377138278228756,
+      "grad_norm": 0.6875760437295031,
+      "learning_rate": 6.050777638120179e-07,
+      "loss": 1.5586,
+      "step": 3642
+    },
+    {
+      "epoch": 0.2538410619099049,
+      "grad_norm": 0.7383449477336569,
+      "learning_rate": 6.050284229702399e-07,
+      "loss": 1.468,
+      "step": 3643
+    },
+    {
+      "epoch": 0.2539107410375222,
+      "grad_norm": 0.7189865672271544,
+      "learning_rate": 6.049790715842483e-07,
+      "loss": 1.5838,
+      "step": 3644
+    },
+    {
+      "epoch": 0.25398042016513955,
+      "grad_norm": 0.6813274745603186,
+      "learning_rate": 6.04929709656408e-07,
+      "loss": 1.4486,
+      "step": 3645
+    },
+    {
+      "epoch": 0.25405009929275685,
+      "grad_norm": 0.7344695622304153,
+      "learning_rate": 6.048803371890844e-07,
+      "loss": 1.499,
+      "step": 3646
+    },
+    {
+      "epoch": 0.2541197784203742,
+      "grad_norm": 0.7489876891373174,
+      "learning_rate": 6.048309541846436e-07,
+      "loss": 1.573,
+      "step": 3647
+    },
+    {
+      "epoch": 0.2541894575479915,
+      "grad_norm": 0.6704671719396704,
+      "learning_rate": 6.047815606454523e-07,
+      "loss": 1.4899,
+      "step": 3648
+    },
+    {
+      "epoch": 0.25425913667560884,
+      "grad_norm": 0.7009858120136003,
+      "learning_rate": 6.047321565738773e-07,
+      "loss": 1.5259,
+      "step": 3649
+    },
+    {
+      "epoch": 0.25432881580322614,
+      "grad_norm": 0.6893590317823148,
+      "learning_rate": 6.046827419722863e-07,
+      "loss": 1.5622,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2543984949308435,
+      "grad_norm": 0.7623085036945566,
+      "learning_rate": 6.046333168430474e-07,
+      "loss": 1.7747,
+      "step": 3651
+    },
+    {
+      "epoch": 0.2544681740584608,
+      "grad_norm": 0.6939366234602085,
+      "learning_rate": 6.045838811885289e-07,
+      "loss": 1.6092,
+      "step": 3652
+    },
+    {
+      "epoch": 0.25453785318607813,
+      "grad_norm": 0.7396889519355708,
+      "learning_rate": 6.045344350111001e-07,
+      "loss": 1.5682,
+      "step": 3653
+    },
+    {
+      "epoch": 0.25460753231369543,
+      "grad_norm": 0.7759652144680148,
+      "learning_rate": 6.044849783131304e-07,
+      "loss": 1.6867,
+      "step": 3654
+    },
+    {
+      "epoch": 0.2546772114413128,
+      "grad_norm": 0.6993831801702053,
+      "learning_rate": 6.044355110969901e-07,
+      "loss": 1.5669,
+      "step": 3655
+    },
+    {
+      "epoch": 0.2547468905689301,
+      "grad_norm": 0.738306542563742,
+      "learning_rate": 6.043860333650495e-07,
+      "loss": 1.4476,
+      "step": 3656
+    },
+    {
+      "epoch": 0.2548165696965474,
+      "grad_norm": 0.6945902157434075,
+      "learning_rate": 6.043365451196799e-07,
+      "loss": 1.5581,
+      "step": 3657
+    },
+    {
+      "epoch": 0.2548862488241647,
+      "grad_norm": 0.7798903180473131,
+      "learning_rate": 6.042870463632525e-07,
+      "loss": 1.5781,
+      "step": 3658
+    },
+    {
+      "epoch": 0.25495592795178207,
+      "grad_norm": 0.6992002259989656,
+      "learning_rate": 6.042375370981399e-07,
+      "loss": 1.5645,
+      "step": 3659
+    },
+    {
+      "epoch": 0.25502560707939936,
+      "grad_norm": 0.6678133223554386,
+      "learning_rate": 6.041880173267144e-07,
+      "loss": 1.4591,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2550952862070167,
+      "grad_norm": 0.6941086204495733,
+      "learning_rate": 6.041384870513491e-07,
+      "loss": 1.5182,
+      "step": 3661
+    },
+    {
+      "epoch": 0.255164965334634,
+      "grad_norm": 0.7269003400483975,
+      "learning_rate": 6.040889462744175e-07,
+      "loss": 1.4942,
+      "step": 3662
+    },
+    {
+      "epoch": 0.25523464446225136,
+      "grad_norm": 0.6982954121554564,
+      "learning_rate": 6.04039394998294e-07,
+      "loss": 1.5595,
+      "step": 3663
+    },
+    {
+      "epoch": 0.25530432358986865,
+      "grad_norm": 0.7186993415671693,
+      "learning_rate": 6.039898332253529e-07,
+      "loss": 1.4928,
+      "step": 3664
+    },
+    {
+      "epoch": 0.255374002717486,
+      "grad_norm": 0.7386453946777918,
+      "learning_rate": 6.039402609579695e-07,
+      "loss": 1.514,
+      "step": 3665
+    },
+    {
+      "epoch": 0.2554436818451033,
+      "grad_norm": 0.7891239561661717,
+      "learning_rate": 6.038906781985192e-07,
+      "loss": 1.5981,
+      "step": 3666
+    },
+    {
+      "epoch": 0.25551336097272065,
+      "grad_norm": 0.7333305917018511,
+      "learning_rate": 6.038410849493784e-07,
+      "loss": 1.6627,
+      "step": 3667
+    },
+    {
+      "epoch": 0.25558304010033794,
+      "grad_norm": 0.7238548352809335,
+      "learning_rate": 6.037914812129233e-07,
+      "loss": 1.5441,
+      "step": 3668
+    },
+    {
+      "epoch": 0.25565271922795524,
+      "grad_norm": 0.7443351369453468,
+      "learning_rate": 6.037418669915314e-07,
+      "loss": 1.5688,
+      "step": 3669
+    },
+    {
+      "epoch": 0.2557223983555726,
+      "grad_norm": 0.766995135505678,
+      "learning_rate": 6.036922422875802e-07,
+      "loss": 1.6102,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2557920774831899,
+      "grad_norm": 0.7051706694725376,
+      "learning_rate": 6.036426071034477e-07,
+      "loss": 1.6172,
+      "step": 3671
+    },
+    {
+      "epoch": 0.25586175661080723,
+      "grad_norm": 0.7193831318040109,
+      "learning_rate": 6.035929614415127e-07,
+      "loss": 1.4728,
+      "step": 3672
+    },
+    {
+      "epoch": 0.25593143573842453,
+      "grad_norm": 0.6910373279699918,
+      "learning_rate": 6.035433053041542e-07,
+      "loss": 1.4988,
+      "step": 3673
+    },
+    {
+      "epoch": 0.2560011148660419,
+      "grad_norm": 0.6740834430678302,
+      "learning_rate": 6.034936386937517e-07,
+      "loss": 1.4256,
+      "step": 3674
+    },
+    {
+      "epoch": 0.2560707939936592,
+      "grad_norm": 0.7120350209786269,
+      "learning_rate": 6.034439616126855e-07,
+      "loss": 1.4566,
+      "step": 3675
+    },
+    {
+      "epoch": 0.2561404731212765,
+      "grad_norm": 0.7113209360044545,
+      "learning_rate": 6.033942740633364e-07,
+      "loss": 1.5773,
+      "step": 3676
+    },
+    {
+      "epoch": 0.2562101522488938,
+      "grad_norm": 0.7035221439888028,
+      "learning_rate": 6.033445760480852e-07,
+      "loss": 1.5885,
+      "step": 3677
+    },
+    {
+      "epoch": 0.25627983137651117,
+      "grad_norm": 0.7274637005286824,
+      "learning_rate": 6.032948675693137e-07,
+      "loss": 1.6685,
+      "step": 3678
+    },
+    {
+      "epoch": 0.25634951050412846,
+      "grad_norm": 0.7882564844836323,
+      "learning_rate": 6.03245148629404e-07,
+      "loss": 1.5647,
+      "step": 3679
+    },
+    {
+      "epoch": 0.2564191896317458,
+      "grad_norm": 0.7194757461164577,
+      "learning_rate": 6.031954192307387e-07,
+      "loss": 1.6198,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2564888687593631,
+      "grad_norm": 0.7119046320308019,
+      "learning_rate": 6.031456793757009e-07,
+      "loss": 1.5605,
+      "step": 3681
+    },
+    {
+      "epoch": 0.25655854788698046,
+      "grad_norm": 0.7037393192757397,
+      "learning_rate": 6.030959290666744e-07,
+      "loss": 1.5774,
+      "step": 3682
+    },
+    {
+      "epoch": 0.25662822701459775,
+      "grad_norm": 0.7657401184144824,
+      "learning_rate": 6.030461683060431e-07,
+      "loss": 1.5446,
+      "step": 3683
+    },
+    {
+      "epoch": 0.2566979061422151,
+      "grad_norm": 0.7131206536769255,
+      "learning_rate": 6.02996397096192e-07,
+      "loss": 1.5325,
+      "step": 3684
+    },
+    {
+      "epoch": 0.2567675852698324,
+      "grad_norm": 0.9433307003881891,
+      "learning_rate": 6.029466154395059e-07,
+      "loss": 1.5525,
+      "step": 3685
+    },
+    {
+      "epoch": 0.25683726439744975,
+      "grad_norm": 0.7386253299710522,
+      "learning_rate": 6.028968233383705e-07,
+      "loss": 1.4944,
+      "step": 3686
+    },
+    {
+      "epoch": 0.25690694352506704,
+      "grad_norm": 0.7490420571721698,
+      "learning_rate": 6.028470207951719e-07,
+      "loss": 1.5285,
+      "step": 3687
+    },
+    {
+      "epoch": 0.2569766226526844,
+      "grad_norm": 0.7557014064188572,
+      "learning_rate": 6.027972078122972e-07,
+      "loss": 1.6712,
+      "step": 3688
+    },
+    {
+      "epoch": 0.2570463017803017,
+      "grad_norm": 0.7693733995532998,
+      "learning_rate": 6.027473843921329e-07,
+      "loss": 1.6772,
+      "step": 3689
+    },
+    {
+      "epoch": 0.25711598090791904,
+      "grad_norm": 0.7186790659646887,
+      "learning_rate": 6.026975505370669e-07,
+      "loss": 1.5178,
+      "step": 3690
+    },
+    {
+      "epoch": 0.25718566003553633,
+      "grad_norm": 0.7160997653437787,
+      "learning_rate": 6.026477062494874e-07,
+      "loss": 1.4966,
+      "step": 3691
+    },
+    {
+      "epoch": 0.2572553391631537,
+      "grad_norm": 0.758441640620066,
+      "learning_rate": 6.025978515317831e-07,
+      "loss": 1.6479,
+      "step": 3692
+    },
+    {
+      "epoch": 0.257325018290771,
+      "grad_norm": 0.7598611067484194,
+      "learning_rate": 6.025479863863429e-07,
+      "loss": 1.6224,
+      "step": 3693
+    },
+    {
+      "epoch": 0.25739469741838833,
+      "grad_norm": 0.7295528775548727,
+      "learning_rate": 6.024981108155564e-07,
+      "loss": 1.6253,
+      "step": 3694
+    },
+    {
+      "epoch": 0.2574643765460056,
+      "grad_norm": 0.8061244876510675,
+      "learning_rate": 6.024482248218143e-07,
+      "loss": 1.5635,
+      "step": 3695
+    },
+    {
+      "epoch": 0.257534055673623,
+      "grad_norm": 0.7410224621117784,
+      "learning_rate": 6.023983284075067e-07,
+      "loss": 1.5478,
+      "step": 3696
+    },
+    {
+      "epoch": 0.25760373480124027,
+      "grad_norm": 0.784543350091465,
+      "learning_rate": 6.023484215750249e-07,
+      "loss": 1.498,
+      "step": 3697
+    },
+    {
+      "epoch": 0.2576734139288576,
+      "grad_norm": 0.7045651266569742,
+      "learning_rate": 6.022985043267604e-07,
+      "loss": 1.4791,
+      "step": 3698
+    },
+    {
+      "epoch": 0.2577430930564749,
+      "grad_norm": 0.6606350336266129,
+      "learning_rate": 6.022485766651056e-07,
+      "loss": 1.54,
+      "step": 3699
+    },
+    {
+      "epoch": 0.25781277218409226,
+      "grad_norm": 0.707716525899006,
+      "learning_rate": 6.02198638592453e-07,
+      "loss": 1.4594,
+      "step": 3700
+    },
+    {
+      "epoch": 0.25788245131170956,
+      "grad_norm": 0.7500573443667033,
+      "learning_rate": 6.021486901111957e-07,
+      "loss": 1.6371,
+      "step": 3701
+    },
+    {
+      "epoch": 0.2579521304393269,
+      "grad_norm": 0.7185889496885267,
+      "learning_rate": 6.020987312237273e-07,
+      "loss": 1.5289,
+      "step": 3702
+    },
+    {
+      "epoch": 0.2580218095669442,
+      "grad_norm": 0.7717476468655174,
+      "learning_rate": 6.020487619324421e-07,
+      "loss": 1.5658,
+      "step": 3703
+    },
+    {
+      "epoch": 0.25809148869456155,
+      "grad_norm": 0.7144107425409536,
+      "learning_rate": 6.019987822397346e-07,
+      "loss": 1.4534,
+      "step": 3704
+    },
+    {
+      "epoch": 0.25816116782217885,
+      "grad_norm": 0.6853109997445458,
+      "learning_rate": 6.019487921480001e-07,
+      "loss": 1.5396,
+      "step": 3705
+    },
+    {
+      "epoch": 0.2582308469497962,
+      "grad_norm": 0.6840861044462495,
+      "learning_rate": 6.018987916596339e-07,
+      "loss": 1.5586,
+      "step": 3706
+    },
+    {
+      "epoch": 0.2583005260774135,
+      "grad_norm": 0.7329357026479503,
+      "learning_rate": 6.018487807770325e-07,
+      "loss": 1.5688,
+      "step": 3707
+    },
+    {
+      "epoch": 0.25837020520503085,
+      "grad_norm": 0.7701244503623086,
+      "learning_rate": 6.017987595025921e-07,
+      "loss": 1.6769,
+      "step": 3708
+    },
+    {
+      "epoch": 0.25843988433264814,
+      "grad_norm": 0.7615865858349096,
+      "learning_rate": 6.017487278387103e-07,
+      "loss": 1.6313,
+      "step": 3709
+    },
+    {
+      "epoch": 0.2585095634602655,
+      "grad_norm": 0.7481709300334196,
+      "learning_rate": 6.016986857877844e-07,
+      "loss": 1.6588,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2585792425878828,
+      "grad_norm": 0.7067990272665308,
+      "learning_rate": 6.016486333522125e-07,
+      "loss": 1.6063,
+      "step": 3711
+    },
+    {
+      "epoch": 0.25864892171550014,
+      "grad_norm": 0.7049759736878695,
+      "learning_rate": 6.015985705343935e-07,
+      "loss": 1.6585,
+      "step": 3712
+    },
+    {
+      "epoch": 0.25871860084311743,
+      "grad_norm": 0.6755596694999774,
+      "learning_rate": 6.015484973367262e-07,
+      "loss": 1.4303,
+      "step": 3713
+    },
+    {
+      "epoch": 0.2587882799707348,
+      "grad_norm": 0.7357990096575269,
+      "learning_rate": 6.014984137616104e-07,
+      "loss": 1.5035,
+      "step": 3714
+    },
+    {
+      "epoch": 0.2588579590983521,
+      "grad_norm": 0.7120439456701165,
+      "learning_rate": 6.014483198114461e-07,
+      "loss": 1.5557,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2589276382259694,
+      "grad_norm": 0.7012771032934041,
+      "learning_rate": 6.01398215488634e-07,
+      "loss": 1.5811,
+      "step": 3716
+    },
+    {
+      "epoch": 0.2589973173535867,
+      "grad_norm": 0.7305842165014204,
+      "learning_rate": 6.013481007955752e-07,
+      "loss": 1.5853,
+      "step": 3717
+    },
+    {
+      "epoch": 0.25906699648120407,
+      "grad_norm": 0.7919793887127279,
+      "learning_rate": 6.012979757346712e-07,
+      "loss": 1.5073,
+      "step": 3718
+    },
+    {
+      "epoch": 0.25913667560882137,
+      "grad_norm": 0.7068150998297997,
+      "learning_rate": 6.012478403083242e-07,
+      "loss": 1.5391,
+      "step": 3719
+    },
+    {
+      "epoch": 0.2592063547364387,
+      "grad_norm": 0.6993516094342813,
+      "learning_rate": 6.011976945189366e-07,
+      "loss": 1.6703,
+      "step": 3720
+    },
+    {
+      "epoch": 0.259276033864056,
+      "grad_norm": 0.7608863004195396,
+      "learning_rate": 6.011475383689118e-07,
+      "loss": 1.5703,
+      "step": 3721
+    },
+    {
+      "epoch": 0.25934571299167336,
+      "grad_norm": 0.6914592273320355,
+      "learning_rate": 6.010973718606531e-07,
+      "loss": 1.4446,
+      "step": 3722
+    },
+    {
+      "epoch": 0.25941539211929066,
+      "grad_norm": 0.6983881594270946,
+      "learning_rate": 6.010471949965648e-07,
+      "loss": 1.4893,
+      "step": 3723
+    },
+    {
+      "epoch": 0.259485071246908,
+      "grad_norm": 0.7046492246974446,
+      "learning_rate": 6.009970077790513e-07,
+      "loss": 1.5758,
+      "step": 3724
+    },
+    {
+      "epoch": 0.2595547503745253,
+      "grad_norm": 0.7441364234564669,
+      "learning_rate": 6.009468102105178e-07,
+      "loss": 1.4596,
+      "step": 3725
+    },
+    {
+      "epoch": 0.25962442950214265,
+      "grad_norm": 0.7656422173680353,
+      "learning_rate": 6.008966022933698e-07,
+      "loss": 1.6614,
+      "step": 3726
+    },
+    {
+      "epoch": 0.25969410862975995,
+      "grad_norm": 0.7559299822870974,
+      "learning_rate": 6.008463840300134e-07,
+      "loss": 1.5062,
+      "step": 3727
+    },
+    {
+      "epoch": 0.2597637877573773,
+      "grad_norm": 0.7669212531621459,
+      "learning_rate": 6.007961554228552e-07,
+      "loss": 1.7258,
+      "step": 3728
+    },
+    {
+      "epoch": 0.2598334668849946,
+      "grad_norm": 0.6772176348456944,
+      "learning_rate": 6.007459164743022e-07,
+      "loss": 1.4701,
+      "step": 3729
+    },
+    {
+      "epoch": 0.25990314601261194,
+      "grad_norm": 0.7499056635774635,
+      "learning_rate": 6.006956671867618e-07,
+      "loss": 1.6034,
+      "step": 3730
+    },
+    {
+      "epoch": 0.25997282514022924,
+      "grad_norm": 0.7225704922652993,
+      "learning_rate": 6.006454075626425e-07,
+      "loss": 1.5875,
+      "step": 3731
+    },
+    {
+      "epoch": 0.2600425042678466,
+      "grad_norm": 0.6980878984816783,
+      "learning_rate": 6.005951376043523e-07,
+      "loss": 1.5183,
+      "step": 3732
+    },
+    {
+      "epoch": 0.2601121833954639,
+      "grad_norm": 0.7491400607437037,
+      "learning_rate": 6.005448573143007e-07,
+      "loss": 1.5805,
+      "step": 3733
+    },
+    {
+      "epoch": 0.26018186252308123,
+      "grad_norm": 0.8043793521506405,
+      "learning_rate": 6.004945666948968e-07,
+      "loss": 1.5908,
+      "step": 3734
+    },
+    {
+      "epoch": 0.2602515416506985,
+      "grad_norm": 0.691452771123545,
+      "learning_rate": 6.004442657485511e-07,
+      "loss": 1.4322,
+      "step": 3735
+    },
+    {
+      "epoch": 0.2603212207783159,
+      "grad_norm": 0.7392515048446673,
+      "learning_rate": 6.003939544776738e-07,
+      "loss": 1.4845,
+      "step": 3736
+    },
+    {
+      "epoch": 0.26039089990593317,
+      "grad_norm": 0.7362991468882549,
+      "learning_rate": 6.003436328846759e-07,
+      "loss": 1.6551,
+      "step": 3737
+    },
+    {
+      "epoch": 0.2604605790335505,
+      "grad_norm": 0.7294508844401375,
+      "learning_rate": 6.002933009719691e-07,
+      "loss": 1.5713,
+      "step": 3738
+    },
+    {
+      "epoch": 0.2605302581611678,
+      "grad_norm": 0.7223413768091034,
+      "learning_rate": 6.002429587419654e-07,
+      "loss": 1.6509,
+      "step": 3739
+    },
+    {
+      "epoch": 0.26059993728878517,
+      "grad_norm": 0.7098332952131358,
+      "learning_rate": 6.001926061970771e-07,
+      "loss": 1.5559,
+      "step": 3740
+    },
+    {
+      "epoch": 0.26066961641640246,
+      "grad_norm": 0.683662666270985,
+      "learning_rate": 6.001422433397174e-07,
+      "loss": 1.5261,
+      "step": 3741
+    },
+    {
+      "epoch": 0.2607392955440198,
+      "grad_norm": 0.7534065865215693,
+      "learning_rate": 6.000918701722998e-07,
+      "loss": 1.498,
+      "step": 3742
+    },
+    {
+      "epoch": 0.2608089746716371,
+      "grad_norm": 0.6924815806601603,
+      "learning_rate": 6.00041486697238e-07,
+      "loss": 1.5978,
+      "step": 3743
+    },
+    {
+      "epoch": 0.26087865379925446,
+      "grad_norm": 0.739100535066475,
+      "learning_rate": 5.99991092916947e-07,
+      "loss": 1.5646,
+      "step": 3744
+    },
+    {
+      "epoch": 0.26094833292687175,
+      "grad_norm": 0.7083409366002702,
+      "learning_rate": 5.999406888338412e-07,
+      "loss": 1.5164,
+      "step": 3745
+    },
+    {
+      "epoch": 0.2610180120544891,
+      "grad_norm": 0.6906008402718623,
+      "learning_rate": 5.998902744503363e-07,
+      "loss": 1.5732,
+      "step": 3746
+    },
+    {
+      "epoch": 0.2610876911821064,
+      "grad_norm": 0.8148269080140913,
+      "learning_rate": 5.998398497688484e-07,
+      "loss": 1.5443,
+      "step": 3747
+    },
+    {
+      "epoch": 0.26115737030972375,
+      "grad_norm": 0.7406162957615721,
+      "learning_rate": 5.99789414791794e-07,
+      "loss": 1.5242,
+      "step": 3748
+    },
+    {
+      "epoch": 0.26122704943734104,
+      "grad_norm": 0.7937614876731822,
+      "learning_rate": 5.997389695215896e-07,
+      "loss": 1.5805,
+      "step": 3749
+    },
+    {
+      "epoch": 0.2612967285649584,
+      "grad_norm": 0.7573554802703404,
+      "learning_rate": 5.99688513960653e-07,
+      "loss": 1.5713,
+      "step": 3750
+    },
+    {
+      "epoch": 0.2613664076925757,
+      "grad_norm": 0.7646561069072213,
+      "learning_rate": 5.996380481114021e-07,
+      "loss": 1.5567,
+      "step": 3751
+    },
+    {
+      "epoch": 0.26143608682019304,
+      "grad_norm": 0.6612253926929647,
+      "learning_rate": 5.995875719762554e-07,
+      "loss": 1.5712,
+      "step": 3752
+    },
+    {
+      "epoch": 0.26150576594781033,
+      "grad_norm": 0.692424855543187,
+      "learning_rate": 5.995370855576315e-07,
+      "loss": 1.579,
+      "step": 3753
+    },
+    {
+      "epoch": 0.2615754450754277,
+      "grad_norm": 0.7442513843985062,
+      "learning_rate": 5.994865888579501e-07,
+      "loss": 1.4847,
+      "step": 3754
+    },
+    {
+      "epoch": 0.261645124203045,
+      "grad_norm": 0.7325976152475112,
+      "learning_rate": 5.994360818796312e-07,
+      "loss": 1.2739,
+      "step": 3755
+    },
+    {
+      "epoch": 0.2617148033306623,
+      "grad_norm": 0.6912059227547824,
+      "learning_rate": 5.993855646250948e-07,
+      "loss": 1.4752,
+      "step": 3756
+    },
+    {
+      "epoch": 0.2617844824582796,
+      "grad_norm": 0.7352347757664525,
+      "learning_rate": 5.993350370967621e-07,
+      "loss": 1.5596,
+      "step": 3757
+    },
+    {
+      "epoch": 0.26185416158589697,
+      "grad_norm": 0.6756060946917455,
+      "learning_rate": 5.992844992970544e-07,
+      "loss": 1.6212,
+      "step": 3758
+    },
+    {
+      "epoch": 0.26192384071351427,
+      "grad_norm": 0.6798226503928068,
+      "learning_rate": 5.992339512283936e-07,
+      "loss": 1.4912,
+      "step": 3759
+    },
+    {
+      "epoch": 0.26199351984113156,
+      "grad_norm": 0.7513251410438295,
+      "learning_rate": 5.991833928932022e-07,
+      "loss": 1.6829,
+      "step": 3760
+    },
+    {
+      "epoch": 0.2620631989687489,
+      "grad_norm": 0.6865988084084071,
+      "learning_rate": 5.991328242939027e-07,
+      "loss": 1.4194,
+      "step": 3761
+    },
+    {
+      "epoch": 0.2621328780963662,
+      "grad_norm": 0.6832901524123712,
+      "learning_rate": 5.990822454329185e-07,
+      "loss": 1.4876,
+      "step": 3762
+    },
+    {
+      "epoch": 0.26220255722398356,
+      "grad_norm": 0.7563513546614019,
+      "learning_rate": 5.990316563126739e-07,
+      "loss": 1.5627,
+      "step": 3763
+    },
+    {
+      "epoch": 0.26227223635160085,
+      "grad_norm": 0.7891804562004142,
+      "learning_rate": 5.989810569355928e-07,
+      "loss": 1.5799,
+      "step": 3764
+    },
+    {
+      "epoch": 0.2623419154792182,
+      "grad_norm": 0.7740143811084153,
+      "learning_rate": 5.989304473041002e-07,
+      "loss": 1.5068,
+      "step": 3765
+    },
+    {
+      "epoch": 0.2624115946068355,
+      "grad_norm": 0.6755385725152862,
+      "learning_rate": 5.988798274206213e-07,
+      "loss": 1.5628,
+      "step": 3766
+    },
+    {
+      "epoch": 0.26248127373445285,
+      "grad_norm": 0.67862127255677,
+      "learning_rate": 5.98829197287582e-07,
+      "loss": 1.4632,
+      "step": 3767
+    },
+    {
+      "epoch": 0.26255095286207014,
+      "grad_norm": 0.7277909881984042,
+      "learning_rate": 5.987785569074086e-07,
+      "loss": 1.4731,
+      "step": 3768
+    },
+    {
+      "epoch": 0.2626206319896875,
+      "grad_norm": 0.7746379657241323,
+      "learning_rate": 5.987279062825278e-07,
+      "loss": 1.5755,
+      "step": 3769
+    },
+    {
+      "epoch": 0.2626903111173048,
+      "grad_norm": 0.7188812385609231,
+      "learning_rate": 5.986772454153671e-07,
+      "loss": 1.5729,
+      "step": 3770
+    },
+    {
+      "epoch": 0.26275999024492214,
+      "grad_norm": 0.7012551759451631,
+      "learning_rate": 5.98626574308354e-07,
+      "loss": 1.579,
+      "step": 3771
+    },
+    {
+      "epoch": 0.26282966937253943,
+      "grad_norm": 0.6879359396906378,
+      "learning_rate": 5.985758929639171e-07,
+      "loss": 1.5858,
+      "step": 3772
+    },
+    {
+      "epoch": 0.2628993485001568,
+      "grad_norm": 0.7367205083171565,
+      "learning_rate": 5.985252013844848e-07,
+      "loss": 1.6155,
+      "step": 3773
+    },
+    {
+      "epoch": 0.2629690276277741,
+      "grad_norm": 0.6755010882852025,
+      "learning_rate": 5.984744995724865e-07,
+      "loss": 1.544,
+      "step": 3774
+    },
+    {
+      "epoch": 0.2630387067553914,
+      "grad_norm": 0.7220361856984341,
+      "learning_rate": 5.984237875303518e-07,
+      "loss": 1.5533,
+      "step": 3775
+    },
+    {
+      "epoch": 0.2631083858830087,
+      "grad_norm": 0.6939224696872534,
+      "learning_rate": 5.983730652605112e-07,
+      "loss": 1.5252,
+      "step": 3776
+    },
+    {
+      "epoch": 0.26317806501062607,
+      "grad_norm": 0.7790602795147022,
+      "learning_rate": 5.983223327653953e-07,
+      "loss": 1.5188,
+      "step": 3777
+    },
+    {
+      "epoch": 0.26324774413824337,
+      "grad_norm": 0.6997604006917725,
+      "learning_rate": 5.982715900474351e-07,
+      "loss": 1.5381,
+      "step": 3778
+    },
+    {
+      "epoch": 0.2633174232658607,
+      "grad_norm": 0.7261413109184857,
+      "learning_rate": 5.982208371090626e-07,
+      "loss": 1.5845,
+      "step": 3779
+    },
+    {
+      "epoch": 0.263387102393478,
+      "grad_norm": 0.7014290816886654,
+      "learning_rate": 5.981700739527099e-07,
+      "loss": 1.4236,
+      "step": 3780
+    },
+    {
+      "epoch": 0.26345678152109536,
+      "grad_norm": 0.7262834750193699,
+      "learning_rate": 5.981193005808095e-07,
+      "loss": 1.5169,
+      "step": 3781
+    },
+    {
+      "epoch": 0.26352646064871266,
+      "grad_norm": 0.7254743030652147,
+      "learning_rate": 5.980685169957948e-07,
+      "loss": 1.5322,
+      "step": 3782
+    },
+    {
+      "epoch": 0.26359613977633,
+      "grad_norm": 0.7169497660477526,
+      "learning_rate": 5.980177232000992e-07,
+      "loss": 1.6798,
+      "step": 3783
+    },
+    {
+      "epoch": 0.2636658189039473,
+      "grad_norm": 0.7078883148268486,
+      "learning_rate": 5.979669191961572e-07,
+      "loss": 1.5295,
+      "step": 3784
+    },
+    {
+      "epoch": 0.26373549803156465,
+      "grad_norm": 0.6876144944709754,
+      "learning_rate": 5.979161049864031e-07,
+      "loss": 1.4183,
+      "step": 3785
+    },
+    {
+      "epoch": 0.26380517715918195,
+      "grad_norm": 0.7007535327284083,
+      "learning_rate": 5.978652805732721e-07,
+      "loss": 1.5592,
+      "step": 3786
+    },
+    {
+      "epoch": 0.2638748562867993,
+      "grad_norm": 0.7329176497548947,
+      "learning_rate": 5.978144459591999e-07,
+      "loss": 1.6272,
+      "step": 3787
+    },
+    {
+      "epoch": 0.2639445354144166,
+      "grad_norm": 0.7552545836313325,
+      "learning_rate": 5.977636011466225e-07,
+      "loss": 1.6171,
+      "step": 3788
+    },
+    {
+      "epoch": 0.26401421454203394,
+      "grad_norm": 0.7909895832044823,
+      "learning_rate": 5.977127461379767e-07,
+      "loss": 1.586,
+      "step": 3789
+    },
+    {
+      "epoch": 0.26408389366965124,
+      "grad_norm": 0.7172742828183447,
+      "learning_rate": 5.976618809356991e-07,
+      "loss": 1.5546,
+      "step": 3790
+    },
+    {
+      "epoch": 0.2641535727972686,
+      "grad_norm": 0.7377383764434615,
+      "learning_rate": 5.976110055422278e-07,
+      "loss": 1.592,
+      "step": 3791
+    },
+    {
+      "epoch": 0.2642232519248859,
+      "grad_norm": 0.7335537641666481,
+      "learning_rate": 5.975601199600006e-07,
+      "loss": 1.7052,
+      "step": 3792
+    },
+    {
+      "epoch": 0.26429293105250323,
+      "grad_norm": 0.7697742371966286,
+      "learning_rate": 5.975092241914562e-07,
+      "loss": 1.5435,
+      "step": 3793
+    },
+    {
+      "epoch": 0.2643626101801205,
+      "grad_norm": 0.7501305244135283,
+      "learning_rate": 5.974583182390333e-07,
+      "loss": 1.4065,
+      "step": 3794
+    },
+    {
+      "epoch": 0.2644322893077379,
+      "grad_norm": 0.721222982117427,
+      "learning_rate": 5.974074021051717e-07,
+      "loss": 1.6465,
+      "step": 3795
+    },
+    {
+      "epoch": 0.2645019684353552,
+      "grad_norm": 0.7255111363489318,
+      "learning_rate": 5.973564757923113e-07,
+      "loss": 1.5097,
+      "step": 3796
+    },
+    {
+      "epoch": 0.2645716475629725,
+      "grad_norm": 0.7725503160769095,
+      "learning_rate": 5.973055393028927e-07,
+      "loss": 1.5885,
+      "step": 3797
+    },
+    {
+      "epoch": 0.2646413266905898,
+      "grad_norm": 0.722794112667975,
+      "learning_rate": 5.972545926393567e-07,
+      "loss": 1.4518,
+      "step": 3798
+    },
+    {
+      "epoch": 0.26471100581820717,
+      "grad_norm": 0.750610046876159,
+      "learning_rate": 5.97203635804145e-07,
+      "loss": 1.4518,
+      "step": 3799
+    },
+    {
+      "epoch": 0.26478068494582446,
+      "grad_norm": 0.7373683509483345,
+      "learning_rate": 5.971526687996992e-07,
+      "loss": 1.6069,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2648503640734418,
+      "grad_norm": 0.7197788095293496,
+      "learning_rate": 5.971016916284623e-07,
+      "loss": 1.4767,
+      "step": 3801
+    },
+    {
+      "epoch": 0.2649200432010591,
+      "grad_norm": 0.7489669567408815,
+      "learning_rate": 5.970507042928765e-07,
+      "loss": 1.5974,
+      "step": 3802
+    },
+    {
+      "epoch": 0.26498972232867646,
+      "grad_norm": 0.7200758831805548,
+      "learning_rate": 5.969997067953859e-07,
+      "loss": 1.4463,
+      "step": 3803
+    },
+    {
+      "epoch": 0.26505940145629375,
+      "grad_norm": 0.7146080622081296,
+      "learning_rate": 5.969486991384342e-07,
+      "loss": 1.4771,
+      "step": 3804
+    },
+    {
+      "epoch": 0.2651290805839111,
+      "grad_norm": 0.712843526792142,
+      "learning_rate": 5.968976813244654e-07,
+      "loss": 1.5082,
+      "step": 3805
+    },
+    {
+      "epoch": 0.2651987597115284,
+      "grad_norm": 0.7322880790478435,
+      "learning_rate": 5.968466533559249e-07,
+      "loss": 1.517,
+      "step": 3806
+    },
+    {
+      "epoch": 0.26526843883914575,
+      "grad_norm": 0.7157357524960127,
+      "learning_rate": 5.967956152352578e-07,
+      "loss": 1.6056,
+      "step": 3807
+    },
+    {
+      "epoch": 0.26533811796676304,
+      "grad_norm": 0.7047036158964357,
+      "learning_rate": 5.967445669649101e-07,
+      "loss": 1.4682,
+      "step": 3808
+    },
+    {
+      "epoch": 0.2654077970943804,
+      "grad_norm": 0.7234814253316819,
+      "learning_rate": 5.96693508547328e-07,
+      "loss": 1.4715,
+      "step": 3809
+    },
+    {
+      "epoch": 0.2654774762219977,
+      "grad_norm": 0.7030356586208922,
+      "learning_rate": 5.966424399849583e-07,
+      "loss": 1.5992,
+      "step": 3810
+    },
+    {
+      "epoch": 0.26554715534961504,
+      "grad_norm": 0.7439713742620705,
+      "learning_rate": 5.965913612802485e-07,
+      "loss": 1.5361,
+      "step": 3811
+    },
+    {
+      "epoch": 0.26561683447723233,
+      "grad_norm": 0.739432309777115,
+      "learning_rate": 5.965402724356462e-07,
+      "loss": 1.6648,
+      "step": 3812
+    },
+    {
+      "epoch": 0.2656865136048497,
+      "grad_norm": 0.7282252553032253,
+      "learning_rate": 5.964891734535997e-07,
+      "loss": 1.5186,
+      "step": 3813
+    },
+    {
+      "epoch": 0.265756192732467,
+      "grad_norm": 0.7404702107542114,
+      "learning_rate": 5.964380643365579e-07,
+      "loss": 1.6763,
+      "step": 3814
+    },
+    {
+      "epoch": 0.26582587186008433,
+      "grad_norm": 0.7245406478505013,
+      "learning_rate": 5.9638694508697e-07,
+      "loss": 1.5752,
+      "step": 3815
+    },
+    {
+      "epoch": 0.2658955509877016,
+      "grad_norm": 0.7015576237292017,
+      "learning_rate": 5.963358157072858e-07,
+      "loss": 1.5034,
+      "step": 3816
+    },
+    {
+      "epoch": 0.265965230115319,
+      "grad_norm": 0.7258510311401303,
+      "learning_rate": 5.962846761999553e-07,
+      "loss": 1.499,
+      "step": 3817
+    },
+    {
+      "epoch": 0.26603490924293627,
+      "grad_norm": 0.7092976051769906,
+      "learning_rate": 5.962335265674295e-07,
+      "loss": 1.5071,
+      "step": 3818
+    },
+    {
+      "epoch": 0.2661045883705536,
+      "grad_norm": 0.7026549998783386,
+      "learning_rate": 5.961823668121593e-07,
+      "loss": 1.444,
+      "step": 3819
+    },
+    {
+      "epoch": 0.2661742674981709,
+      "grad_norm": 0.7836415542286314,
+      "learning_rate": 5.961311969365966e-07,
+      "loss": 1.6607,
+      "step": 3820
+    },
+    {
+      "epoch": 0.26624394662578826,
+      "grad_norm": 0.7517676503653571,
+      "learning_rate": 5.960800169431935e-07,
+      "loss": 1.6165,
+      "step": 3821
+    },
+    {
+      "epoch": 0.26631362575340556,
+      "grad_norm": 0.7071949248379115,
+      "learning_rate": 5.960288268344027e-07,
+      "loss": 1.6095,
+      "step": 3822
+    },
+    {
+      "epoch": 0.2663833048810229,
+      "grad_norm": 0.7524462940338275,
+      "learning_rate": 5.959776266126772e-07,
+      "loss": 1.5655,
+      "step": 3823
+    },
+    {
+      "epoch": 0.2664529840086402,
+      "grad_norm": 0.7905471950668109,
+      "learning_rate": 5.959264162804707e-07,
+      "loss": 1.5676,
+      "step": 3824
+    },
+    {
+      "epoch": 0.26652266313625755,
+      "grad_norm": 0.7576424220733925,
+      "learning_rate": 5.958751958402374e-07,
+      "loss": 1.5781,
+      "step": 3825
+    },
+    {
+      "epoch": 0.26659234226387485,
+      "grad_norm": 0.7559598623005291,
+      "learning_rate": 5.958239652944317e-07,
+      "loss": 1.3891,
+      "step": 3826
+    },
+    {
+      "epoch": 0.2666620213914922,
+      "grad_norm": 0.7388353146062195,
+      "learning_rate": 5.957727246455088e-07,
+      "loss": 1.399,
+      "step": 3827
+    },
+    {
+      "epoch": 0.2667317005191095,
+      "grad_norm": 0.7097430840949527,
+      "learning_rate": 5.957214738959243e-07,
+      "loss": 1.4865,
+      "step": 3828
+    },
+    {
+      "epoch": 0.26680137964672684,
+      "grad_norm": 0.7271421482613403,
+      "learning_rate": 5.95670213048134e-07,
+      "loss": 1.4648,
+      "step": 3829
+    },
+    {
+      "epoch": 0.26687105877434414,
+      "grad_norm": 0.7505684797204005,
+      "learning_rate": 5.956189421045947e-07,
+      "loss": 1.5283,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2669407379019615,
+      "grad_norm": 0.7260488787468974,
+      "learning_rate": 5.955676610677633e-07,
+      "loss": 1.4497,
+      "step": 3831
+    },
+    {
+      "epoch": 0.2670104170295788,
+      "grad_norm": 0.7848335820145739,
+      "learning_rate": 5.955163699400973e-07,
+      "loss": 1.8227,
+      "step": 3832
+    },
+    {
+      "epoch": 0.26708009615719613,
+      "grad_norm": 0.6623961844248426,
+      "learning_rate": 5.954650687240547e-07,
+      "loss": 1.5249,
+      "step": 3833
+    },
+    {
+      "epoch": 0.26714977528481343,
+      "grad_norm": 0.6783606484283297,
+      "learning_rate": 5.954137574220939e-07,
+      "loss": 1.5641,
+      "step": 3834
+    },
+    {
+      "epoch": 0.2672194544124308,
+      "grad_norm": 0.6901063180908451,
+      "learning_rate": 5.953624360366739e-07,
+      "loss": 1.4457,
+      "step": 3835
+    },
+    {
+      "epoch": 0.2672891335400481,
+      "grad_norm": 0.6583457935445118,
+      "learning_rate": 5.95311104570254e-07,
+      "loss": 1.481,
+      "step": 3836
+    },
+    {
+      "epoch": 0.2673588126676654,
+      "grad_norm": 0.7459974223450113,
+      "learning_rate": 5.952597630252943e-07,
+      "loss": 1.6124,
+      "step": 3837
+    },
+    {
+      "epoch": 0.2674284917952827,
+      "grad_norm": 0.7200644745261434,
+      "learning_rate": 5.952084114042551e-07,
+      "loss": 1.5149,
+      "step": 3838
+    },
+    {
+      "epoch": 0.26749817092290007,
+      "grad_norm": 0.8150337193713256,
+      "learning_rate": 5.951570497095973e-07,
+      "loss": 1.5111,
+      "step": 3839
+    },
+    {
+      "epoch": 0.26756785005051736,
+      "grad_norm": 0.6647113024228307,
+      "learning_rate": 5.951056779437821e-07,
+      "loss": 1.5736,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2676375291781347,
+      "grad_norm": 0.6977037781613992,
+      "learning_rate": 5.950542961092715e-07,
+      "loss": 1.4829,
+      "step": 3841
+    },
+    {
+      "epoch": 0.267707208305752,
+      "grad_norm": 0.7263040973123264,
+      "learning_rate": 5.950029042085279e-07,
+      "loss": 1.4914,
+      "step": 3842
+    },
+    {
+      "epoch": 0.26777688743336936,
+      "grad_norm": 0.7059006018144998,
+      "learning_rate": 5.949515022440141e-07,
+      "loss": 1.4649,
+      "step": 3843
+    },
+    {
+      "epoch": 0.26784656656098665,
+      "grad_norm": 0.7146294619304026,
+      "learning_rate": 5.949000902181932e-07,
+      "loss": 1.5275,
+      "step": 3844
+    },
+    {
+      "epoch": 0.267916245688604,
+      "grad_norm": 0.7528834166368202,
+      "learning_rate": 5.948486681335289e-07,
+      "loss": 1.4976,
+      "step": 3845
+    },
+    {
+      "epoch": 0.2679859248162213,
+      "grad_norm": 0.6506852625680217,
+      "learning_rate": 5.947972359924857e-07,
+      "loss": 1.4088,
+      "step": 3846
+    },
+    {
+      "epoch": 0.26805560394383865,
+      "grad_norm": 0.7112999509070153,
+      "learning_rate": 5.947457937975282e-07,
+      "loss": 1.5822,
+      "step": 3847
+    },
+    {
+      "epoch": 0.26812528307145594,
+      "grad_norm": 0.6937904601326073,
+      "learning_rate": 5.946943415511218e-07,
+      "loss": 1.5623,
+      "step": 3848
+    },
+    {
+      "epoch": 0.2681949621990733,
+      "grad_norm": 0.7137558963987013,
+      "learning_rate": 5.946428792557321e-07,
+      "loss": 1.535,
+      "step": 3849
+    },
+    {
+      "epoch": 0.2682646413266906,
+      "grad_norm": 0.6923843161735229,
+      "learning_rate": 5.94591406913825e-07,
+      "loss": 1.4935,
+      "step": 3850
+    },
+    {
+      "epoch": 0.26833432045430794,
+      "grad_norm": 0.6818469415482451,
+      "learning_rate": 5.945399245278675e-07,
+      "loss": 1.5496,
+      "step": 3851
+    },
+    {
+      "epoch": 0.26840399958192523,
+      "grad_norm": 0.6965188523204926,
+      "learning_rate": 5.944884321003267e-07,
+      "loss": 1.4331,
+      "step": 3852
+    },
+    {
+      "epoch": 0.26847367870954253,
+      "grad_norm": 0.7738072123350715,
+      "learning_rate": 5.944369296336701e-07,
+      "loss": 1.782,
+      "step": 3853
+    },
+    {
+      "epoch": 0.2685433578371599,
+      "grad_norm": 0.7172882849130879,
+      "learning_rate": 5.943854171303659e-07,
+      "loss": 1.5333,
+      "step": 3854
+    },
+    {
+      "epoch": 0.2686130369647772,
+      "grad_norm": 0.7399299275792292,
+      "learning_rate": 5.943338945928827e-07,
+      "loss": 1.4818,
+      "step": 3855
+    },
+    {
+      "epoch": 0.2686827160923945,
+      "grad_norm": 0.7494982746550088,
+      "learning_rate": 5.942823620236894e-07,
+      "loss": 1.5849,
+      "step": 3856
+    },
+    {
+      "epoch": 0.2687523952200118,
+      "grad_norm": 1.3592300358320584,
+      "learning_rate": 5.942308194252557e-07,
+      "loss": 1.6021,
+      "step": 3857
+    },
+    {
+      "epoch": 0.26882207434762917,
+      "grad_norm": 0.6870898200709619,
+      "learning_rate": 5.941792668000517e-07,
+      "loss": 1.5074,
+      "step": 3858
+    },
+    {
+      "epoch": 0.26889175347524646,
+      "grad_norm": 0.7256808114459625,
+      "learning_rate": 5.941277041505477e-07,
+      "loss": 1.5594,
+      "step": 3859
+    },
+    {
+      "epoch": 0.2689614326028638,
+      "grad_norm": 0.6883037203550751,
+      "learning_rate": 5.940761314792148e-07,
+      "loss": 1.5388,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2690311117304811,
+      "grad_norm": 0.7187263352455807,
+      "learning_rate": 5.940245487885244e-07,
+      "loss": 1.4854,
+      "step": 3861
+    },
+    {
+      "epoch": 0.26910079085809846,
+      "grad_norm": 0.7166388609160896,
+      "learning_rate": 5.939729560809486e-07,
+      "loss": 1.4961,
+      "step": 3862
+    },
+    {
+      "epoch": 0.26917046998571575,
+      "grad_norm": 0.7070290799262712,
+      "learning_rate": 5.939213533589596e-07,
+      "loss": 1.5669,
+      "step": 3863
+    },
+    {
+      "epoch": 0.2692401491133331,
+      "grad_norm": 0.6967168983725861,
+      "learning_rate": 5.938697406250307e-07,
+      "loss": 1.3717,
+      "step": 3864
+    },
+    {
+      "epoch": 0.2693098282409504,
+      "grad_norm": 0.681060189861049,
+      "learning_rate": 5.938181178816349e-07,
+      "loss": 1.5998,
+      "step": 3865
+    },
+    {
+      "epoch": 0.26937950736856775,
+      "grad_norm": 0.8336457379589953,
+      "learning_rate": 5.93766485131246e-07,
+      "loss": 1.5647,
+      "step": 3866
+    },
+    {
+      "epoch": 0.26944918649618504,
+      "grad_norm": 0.6879300359719382,
+      "learning_rate": 5.937148423763387e-07,
+      "loss": 1.4075,
+      "step": 3867
+    },
+    {
+      "epoch": 0.2695188656238024,
+      "grad_norm": 0.7410474852761608,
+      "learning_rate": 5.936631896193877e-07,
+      "loss": 1.5457,
+      "step": 3868
+    },
+    {
+      "epoch": 0.2695885447514197,
+      "grad_norm": 0.7548080374598897,
+      "learning_rate": 5.936115268628682e-07,
+      "loss": 1.5415,
+      "step": 3869
+    },
+    {
+      "epoch": 0.26965822387903704,
+      "grad_norm": 0.7264228100243976,
+      "learning_rate": 5.935598541092561e-07,
+      "loss": 1.5861,
+      "step": 3870
+    },
+    {
+      "epoch": 0.26972790300665433,
+      "grad_norm": 0.7722431128355013,
+      "learning_rate": 5.935081713610277e-07,
+      "loss": 1.6332,
+      "step": 3871
+    },
+    {
+      "epoch": 0.2697975821342717,
+      "grad_norm": 0.754648281167634,
+      "learning_rate": 5.934564786206595e-07,
+      "loss": 1.5744,
+      "step": 3872
+    },
+    {
+      "epoch": 0.269867261261889,
+      "grad_norm": 0.6926936779814764,
+      "learning_rate": 5.934047758906291e-07,
+      "loss": 1.4116,
+      "step": 3873
+    },
+    {
+      "epoch": 0.26993694038950633,
+      "grad_norm": 0.7420917039064506,
+      "learning_rate": 5.933530631734138e-07,
+      "loss": 1.5542,
+      "step": 3874
+    },
+    {
+      "epoch": 0.2700066195171236,
+      "grad_norm": 0.7718671864174498,
+      "learning_rate": 5.93301340471492e-07,
+      "loss": 1.5018,
+      "step": 3875
+    },
+    {
+      "epoch": 0.270076298644741,
+      "grad_norm": 0.7436769872040874,
+      "learning_rate": 5.932496077873425e-07,
+      "loss": 1.5629,
+      "step": 3876
+    },
+    {
+      "epoch": 0.27014597777235827,
+      "grad_norm": 0.7024002295532155,
+      "learning_rate": 5.93197865123444e-07,
+      "loss": 1.5561,
+      "step": 3877
+    },
+    {
+      "epoch": 0.2702156568999756,
+      "grad_norm": 0.7036180595380055,
+      "learning_rate": 5.931461124822766e-07,
+      "loss": 1.4487,
+      "step": 3878
+    },
+    {
+      "epoch": 0.2702853360275929,
+      "grad_norm": 0.7291033209991897,
+      "learning_rate": 5.9309434986632e-07,
+      "loss": 1.6397,
+      "step": 3879
+    },
+    {
+      "epoch": 0.27035501515521027,
+      "grad_norm": 0.7270278064368998,
+      "learning_rate": 5.930425772780551e-07,
+      "loss": 1.6212,
+      "step": 3880
+    },
+    {
+      "epoch": 0.27042469428282756,
+      "grad_norm": 0.7196922674297855,
+      "learning_rate": 5.929907947199628e-07,
+      "loss": 1.5628,
+      "step": 3881
+    },
+    {
+      "epoch": 0.2704943734104449,
+      "grad_norm": 0.7405096746089026,
+      "learning_rate": 5.929390021945246e-07,
+      "loss": 1.3544,
+      "step": 3882
+    },
+    {
+      "epoch": 0.2705640525380622,
+      "grad_norm": 0.7450568923816304,
+      "learning_rate": 5.928871997042224e-07,
+      "loss": 1.4874,
+      "step": 3883
+    },
+    {
+      "epoch": 0.27063373166567956,
+      "grad_norm": 0.7056369007831684,
+      "learning_rate": 5.928353872515389e-07,
+      "loss": 1.4927,
+      "step": 3884
+    },
+    {
+      "epoch": 0.27070341079329685,
+      "grad_norm": 0.761581874927409,
+      "learning_rate": 5.92783564838957e-07,
+      "loss": 1.5912,
+      "step": 3885
+    },
+    {
+      "epoch": 0.2707730899209142,
+      "grad_norm": 0.6921846983017318,
+      "learning_rate": 5.9273173246896e-07,
+      "loss": 1.5517,
+      "step": 3886
+    },
+    {
+      "epoch": 0.2708427690485315,
+      "grad_norm": 0.7002922072429691,
+      "learning_rate": 5.926798901440321e-07,
+      "loss": 1.5203,
+      "step": 3887
+    },
+    {
+      "epoch": 0.27091244817614885,
+      "grad_norm": 0.683729525033674,
+      "learning_rate": 5.926280378666573e-07,
+      "loss": 1.4792,
+      "step": 3888
+    },
+    {
+      "epoch": 0.27098212730376614,
+      "grad_norm": 0.6414599752660434,
+      "learning_rate": 5.925761756393207e-07,
+      "loss": 1.5579,
+      "step": 3889
+    },
+    {
+      "epoch": 0.2710518064313835,
+      "grad_norm": 0.7010029702153505,
+      "learning_rate": 5.925243034645077e-07,
+      "loss": 1.3604,
+      "step": 3890
+    },
+    {
+      "epoch": 0.2711214855590008,
+      "grad_norm": 0.6772803213294848,
+      "learning_rate": 5.92472421344704e-07,
+      "loss": 1.4497,
+      "step": 3891
+    },
+    {
+      "epoch": 0.27119116468661814,
+      "grad_norm": 0.7423859825107686,
+      "learning_rate": 5.92420529282396e-07,
+      "loss": 1.6493,
+      "step": 3892
+    },
+    {
+      "epoch": 0.27126084381423543,
+      "grad_norm": 0.7069407783189319,
+      "learning_rate": 5.923686272800703e-07,
+      "loss": 1.4724,
+      "step": 3893
+    },
+    {
+      "epoch": 0.2713305229418528,
+      "grad_norm": 0.7288513991064071,
+      "learning_rate": 5.923167153402144e-07,
+      "loss": 1.6363,
+      "step": 3894
+    },
+    {
+      "epoch": 0.2714002020694701,
+      "grad_norm": 0.6983283286669905,
+      "learning_rate": 5.922647934653158e-07,
+      "loss": 1.5045,
+      "step": 3895
+    },
+    {
+      "epoch": 0.2714698811970874,
+      "grad_norm": 0.729073744684144,
+      "learning_rate": 5.922128616578627e-07,
+      "loss": 1.497,
+      "step": 3896
+    },
+    {
+      "epoch": 0.2715395603247047,
+      "grad_norm": 0.6998014105535623,
+      "learning_rate": 5.92160919920344e-07,
+      "loss": 1.6703,
+      "step": 3897
+    },
+    {
+      "epoch": 0.27160923945232207,
+      "grad_norm": 0.7285550394088254,
+      "learning_rate": 5.921089682552487e-07,
+      "loss": 1.5291,
+      "step": 3898
+    },
+    {
+      "epoch": 0.27167891857993937,
+      "grad_norm": 0.7584854068310711,
+      "learning_rate": 5.920570066650665e-07,
+      "loss": 1.4431,
+      "step": 3899
+    },
+    {
+      "epoch": 0.2717485977075567,
+      "grad_norm": 0.7782366181505483,
+      "learning_rate": 5.920050351522874e-07,
+      "loss": 1.5376,
+      "step": 3900
+    },
+    {
+      "epoch": 0.271818276835174,
+      "grad_norm": 0.7645158556386534,
+      "learning_rate": 5.91953053719402e-07,
+      "loss": 1.6062,
+      "step": 3901
+    },
+    {
+      "epoch": 0.27188795596279136,
+      "grad_norm": 0.6900052460278817,
+      "learning_rate": 5.919010623689015e-07,
+      "loss": 1.5655,
+      "step": 3902
+    },
+    {
+      "epoch": 0.27195763509040866,
+      "grad_norm": 0.7540368159165663,
+      "learning_rate": 5.918490611032772e-07,
+      "loss": 1.6374,
+      "step": 3903
+    },
+    {
+      "epoch": 0.272027314218026,
+      "grad_norm": 0.7339110598734926,
+      "learning_rate": 5.917970499250214e-07,
+      "loss": 1.6146,
+      "step": 3904
+    },
+    {
+      "epoch": 0.2720969933456433,
+      "grad_norm": 0.7138874824396794,
+      "learning_rate": 5.917450288366263e-07,
+      "loss": 1.5524,
+      "step": 3905
+    },
+    {
+      "epoch": 0.27216667247326065,
+      "grad_norm": 0.6772495253223015,
+      "learning_rate": 5.916929978405849e-07,
+      "loss": 1.6166,
+      "step": 3906
+    },
+    {
+      "epoch": 0.27223635160087795,
+      "grad_norm": 0.7887647513498364,
+      "learning_rate": 5.916409569393909e-07,
+      "loss": 1.5605,
+      "step": 3907
+    },
+    {
+      "epoch": 0.2723060307284953,
+      "grad_norm": 0.6988054387986751,
+      "learning_rate": 5.915889061355379e-07,
+      "loss": 1.5141,
+      "step": 3908
+    },
+    {
+      "epoch": 0.2723757098561126,
+      "grad_norm": 0.7296767353519555,
+      "learning_rate": 5.915368454315205e-07,
+      "loss": 1.5563,
+      "step": 3909
+    },
+    {
+      "epoch": 0.27244538898372994,
+      "grad_norm": 0.7497446074632108,
+      "learning_rate": 5.914847748298333e-07,
+      "loss": 1.6255,
+      "step": 3910
+    },
+    {
+      "epoch": 0.27251506811134724,
+      "grad_norm": 0.6639192068050025,
+      "learning_rate": 5.914326943329719e-07,
+      "loss": 1.5285,
+      "step": 3911
+    },
+    {
+      "epoch": 0.2725847472389646,
+      "grad_norm": 0.7428035774613251,
+      "learning_rate": 5.913806039434321e-07,
+      "loss": 1.4783,
+      "step": 3912
+    },
+    {
+      "epoch": 0.2726544263665819,
+      "grad_norm": 0.7281922760552719,
+      "learning_rate": 5.913285036637098e-07,
+      "loss": 1.5469,
+      "step": 3913
+    },
+    {
+      "epoch": 0.27272410549419923,
+      "grad_norm": 0.7333014735087869,
+      "learning_rate": 5.912763934963022e-07,
+      "loss": 1.592,
+      "step": 3914
+    },
+    {
+      "epoch": 0.2727937846218165,
+      "grad_norm": 0.7798965806035877,
+      "learning_rate": 5.912242734437064e-07,
+      "loss": 1.6835,
+      "step": 3915
+    },
+    {
+      "epoch": 0.2728634637494339,
+      "grad_norm": 0.7536671752599499,
+      "learning_rate": 5.911721435084199e-07,
+      "loss": 1.7395,
+      "step": 3916
+    },
+    {
+      "epoch": 0.27293314287705117,
+      "grad_norm": 0.6902780463980486,
+      "learning_rate": 5.911200036929411e-07,
+      "loss": 1.6335,
+      "step": 3917
+    },
+    {
+      "epoch": 0.2730028220046685,
+      "grad_norm": 0.6967623275559346,
+      "learning_rate": 5.910678539997686e-07,
+      "loss": 1.5714,
+      "step": 3918
+    },
+    {
+      "epoch": 0.2730725011322858,
+      "grad_norm": 0.7397568053887077,
+      "learning_rate": 5.910156944314013e-07,
+      "loss": 1.6657,
+      "step": 3919
+    },
+    {
+      "epoch": 0.27314218025990317,
+      "grad_norm": 0.701659478786721,
+      "learning_rate": 5.909635249903391e-07,
+      "loss": 1.5462,
+      "step": 3920
+    },
+    {
+      "epoch": 0.27321185938752046,
+      "grad_norm": 0.7725419571633488,
+      "learning_rate": 5.90911345679082e-07,
+      "loss": 1.6394,
+      "step": 3921
+    },
+    {
+      "epoch": 0.2732815385151378,
+      "grad_norm": 0.6980051136129517,
+      "learning_rate": 5.908591565001304e-07,
+      "loss": 1.4364,
+      "step": 3922
+    },
+    {
+      "epoch": 0.2733512176427551,
+      "grad_norm": 0.8135918312405112,
+      "learning_rate": 5.908069574559854e-07,
+      "loss": 1.5069,
+      "step": 3923
+    },
+    {
+      "epoch": 0.27342089677037246,
+      "grad_norm": 0.7053317409211138,
+      "learning_rate": 5.907547485491487e-07,
+      "loss": 1.5591,
+      "step": 3924
+    },
+    {
+      "epoch": 0.27349057589798975,
+      "grad_norm": 0.7018480395090504,
+      "learning_rate": 5.907025297821218e-07,
+      "loss": 1.5574,
+      "step": 3925
+    },
+    {
+      "epoch": 0.2735602550256071,
+      "grad_norm": 0.7720773415373251,
+      "learning_rate": 5.906503011574075e-07,
+      "loss": 1.4853,
+      "step": 3926
+    },
+    {
+      "epoch": 0.2736299341532244,
+      "grad_norm": 0.6976717600818462,
+      "learning_rate": 5.905980626775085e-07,
+      "loss": 1.542,
+      "step": 3927
+    },
+    {
+      "epoch": 0.27369961328084175,
+      "grad_norm": 0.7211622466811443,
+      "learning_rate": 5.905458143449282e-07,
+      "loss": 1.6,
+      "step": 3928
+    },
+    {
+      "epoch": 0.27376929240845904,
+      "grad_norm": 0.695676948180807,
+      "learning_rate": 5.904935561621705e-07,
+      "loss": 1.61,
+      "step": 3929
+    },
+    {
+      "epoch": 0.2738389715360764,
+      "grad_norm": 0.7392444394382608,
+      "learning_rate": 5.904412881317398e-07,
+      "loss": 1.4697,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2739086506636937,
+      "grad_norm": 0.7379040097850592,
+      "learning_rate": 5.903890102561409e-07,
+      "loss": 1.6316,
+      "step": 3931
+    },
+    {
+      "epoch": 0.27397832979131104,
+      "grad_norm": 0.6838537018000045,
+      "learning_rate": 5.903367225378788e-07,
+      "loss": 1.508,
+      "step": 3932
+    },
+    {
+      "epoch": 0.27404800891892833,
+      "grad_norm": 0.723306590283538,
+      "learning_rate": 5.902844249794595e-07,
+      "loss": 1.5367,
+      "step": 3933
+    },
+    {
+      "epoch": 0.2741176880465457,
+      "grad_norm": 0.7380752803903438,
+      "learning_rate": 5.902321175833891e-07,
+      "loss": 1.5289,
+      "step": 3934
+    },
+    {
+      "epoch": 0.274187367174163,
+      "grad_norm": 0.7063438845989196,
+      "learning_rate": 5.901798003521742e-07,
+      "loss": 1.4684,
+      "step": 3935
+    },
+    {
+      "epoch": 0.2742570463017803,
+      "grad_norm": 0.7569904066641995,
+      "learning_rate": 5.901274732883223e-07,
+      "loss": 1.5033,
+      "step": 3936
+    },
+    {
+      "epoch": 0.2743267254293976,
+      "grad_norm": 0.7296337737834608,
+      "learning_rate": 5.900751363943405e-07,
+      "loss": 1.5454,
+      "step": 3937
+    },
+    {
+      "epoch": 0.274396404557015,
+      "grad_norm": 0.7029267716615997,
+      "learning_rate": 5.900227896727372e-07,
+      "loss": 1.5316,
+      "step": 3938
+    },
+    {
+      "epoch": 0.27446608368463227,
+      "grad_norm": 0.7178752281258443,
+      "learning_rate": 5.89970433126021e-07,
+      "loss": 1.5698,
+      "step": 3939
+    },
+    {
+      "epoch": 0.2745357628122496,
+      "grad_norm": 0.7567086642839586,
+      "learning_rate": 5.899180667567008e-07,
+      "loss": 1.516,
+      "step": 3940
+    },
+    {
+      "epoch": 0.2746054419398669,
+      "grad_norm": 0.6923104599232259,
+      "learning_rate": 5.898656905672861e-07,
+      "loss": 1.5541,
+      "step": 3941
+    },
+    {
+      "epoch": 0.27467512106748426,
+      "grad_norm": 0.710308756305562,
+      "learning_rate": 5.89813304560287e-07,
+      "loss": 1.5168,
+      "step": 3942
+    },
+    {
+      "epoch": 0.27474480019510156,
+      "grad_norm": 0.7618485460806338,
+      "learning_rate": 5.897609087382139e-07,
+      "loss": 1.5854,
+      "step": 3943
+    },
+    {
+      "epoch": 0.27481447932271885,
+      "grad_norm": 0.7897506539988088,
+      "learning_rate": 5.897085031035776e-07,
+      "loss": 1.5347,
+      "step": 3944
+    },
+    {
+      "epoch": 0.2748841584503362,
+      "grad_norm": 0.7576815455048657,
+      "learning_rate": 5.896560876588897e-07,
+      "loss": 1.5786,
+      "step": 3945
+    },
+    {
+      "epoch": 0.2749538375779535,
+      "grad_norm": 0.6614648949215904,
+      "learning_rate": 5.896036624066618e-07,
+      "loss": 1.429,
+      "step": 3946
+    },
+    {
+      "epoch": 0.27502351670557085,
+      "grad_norm": 0.7016203280395538,
+      "learning_rate": 5.895512273494064e-07,
+      "loss": 1.5163,
+      "step": 3947
+    },
+    {
+      "epoch": 0.27509319583318814,
+      "grad_norm": 0.6931782798530423,
+      "learning_rate": 5.894987824896362e-07,
+      "loss": 1.5627,
+      "step": 3948
+    },
+    {
+      "epoch": 0.2751628749608055,
+      "grad_norm": 0.7451596533117754,
+      "learning_rate": 5.894463278298647e-07,
+      "loss": 1.559,
+      "step": 3949
+    },
+    {
+      "epoch": 0.2752325540884228,
+      "grad_norm": 0.8392928248647462,
+      "learning_rate": 5.893938633726052e-07,
+      "loss": 1.5077,
+      "step": 3950
+    },
+    {
+      "epoch": 0.27530223321604014,
+      "grad_norm": 0.7496311483780644,
+      "learning_rate": 5.893413891203723e-07,
+      "loss": 1.6871,
+      "step": 3951
+    },
+    {
+      "epoch": 0.27537191234365743,
+      "grad_norm": 0.6866597786792935,
+      "learning_rate": 5.892889050756805e-07,
+      "loss": 1.5015,
+      "step": 3952
+    },
+    {
+      "epoch": 0.2754415914712748,
+      "grad_norm": 0.7539980443871873,
+      "learning_rate": 5.89236411241045e-07,
+      "loss": 1.5783,
+      "step": 3953
+    },
+    {
+      "epoch": 0.2755112705988921,
+      "grad_norm": 0.7120384308434919,
+      "learning_rate": 5.891839076189814e-07,
+      "loss": 1.4073,
+      "step": 3954
+    },
+    {
+      "epoch": 0.2755809497265094,
+      "grad_norm": 0.7563875392467487,
+      "learning_rate": 5.891313942120056e-07,
+      "loss": 1.4677,
+      "step": 3955
+    },
+    {
+      "epoch": 0.2756506288541267,
+      "grad_norm": 0.6614944369011492,
+      "learning_rate": 5.890788710226344e-07,
+      "loss": 1.5537,
+      "step": 3956
+    },
+    {
+      "epoch": 0.2757203079817441,
+      "grad_norm": 0.6610667999394185,
+      "learning_rate": 5.890263380533848e-07,
+      "loss": 1.4159,
+      "step": 3957
+    },
+    {
+      "epoch": 0.27578998710936137,
+      "grad_norm": 0.7281030442813066,
+      "learning_rate": 5.889737953067742e-07,
+      "loss": 1.4733,
+      "step": 3958
+    },
+    {
+      "epoch": 0.2758596662369787,
+      "grad_norm": 0.7187232399212806,
+      "learning_rate": 5.889212427853205e-07,
+      "loss": 1.5881,
+      "step": 3959
+    },
+    {
+      "epoch": 0.275929345364596,
+      "grad_norm": 0.7783928113723274,
+      "learning_rate": 5.888686804915423e-07,
+      "loss": 1.7394,
+      "step": 3960
+    },
+    {
+      "epoch": 0.27599902449221336,
+      "grad_norm": 0.7800510926427071,
+      "learning_rate": 5.888161084279584e-07,
+      "loss": 1.5735,
+      "step": 3961
+    },
+    {
+      "epoch": 0.27606870361983066,
+      "grad_norm": 0.6975159879900961,
+      "learning_rate": 5.887635265970882e-07,
+      "loss": 1.5884,
+      "step": 3962
+    },
+    {
+      "epoch": 0.276138382747448,
+      "grad_norm": 0.8555360238849415,
+      "learning_rate": 5.887109350014513e-07,
+      "loss": 1.5826,
+      "step": 3963
+    },
+    {
+      "epoch": 0.2762080618750653,
+      "grad_norm": 0.7028176190290327,
+      "learning_rate": 5.886583336435683e-07,
+      "loss": 1.7248,
+      "step": 3964
+    },
+    {
+      "epoch": 0.27627774100268265,
+      "grad_norm": 0.7432092852493237,
+      "learning_rate": 5.886057225259598e-07,
+      "loss": 1.6639,
+      "step": 3965
+    },
+    {
+      "epoch": 0.27634742013029995,
+      "grad_norm": 0.6896597838445886,
+      "learning_rate": 5.885531016511472e-07,
+      "loss": 1.447,
+      "step": 3966
+    },
+    {
+      "epoch": 0.2764170992579173,
+      "grad_norm": 0.720433857116974,
+      "learning_rate": 5.885004710216519e-07,
+      "loss": 1.6086,
+      "step": 3967
+    },
+    {
+      "epoch": 0.2764867783855346,
+      "grad_norm": 0.7668175443909877,
+      "learning_rate": 5.884478306399965e-07,
+      "loss": 1.5574,
+      "step": 3968
+    },
+    {
+      "epoch": 0.27655645751315194,
+      "grad_norm": 0.7209226325400172,
+      "learning_rate": 5.883951805087033e-07,
+      "loss": 1.6687,
+      "step": 3969
+    },
+    {
+      "epoch": 0.27662613664076924,
+      "grad_norm": 0.714311583829575,
+      "learning_rate": 5.883425206302952e-07,
+      "loss": 1.5147,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2766958157683866,
+      "grad_norm": 0.6742840641439894,
+      "learning_rate": 5.882898510072964e-07,
+      "loss": 1.4254,
+      "step": 3971
+    },
+    {
+      "epoch": 0.2767654948960039,
+      "grad_norm": 0.7348778247791798,
+      "learning_rate": 5.882371716422306e-07,
+      "loss": 1.4907,
+      "step": 3972
+    },
+    {
+      "epoch": 0.27683517402362123,
+      "grad_norm": 0.6579709981742292,
+      "learning_rate": 5.88184482537622e-07,
+      "loss": 1.4232,
+      "step": 3973
+    },
+    {
+      "epoch": 0.27690485315123853,
+      "grad_norm": 0.7227626916827766,
+      "learning_rate": 5.881317836959961e-07,
+      "loss": 1.4126,
+      "step": 3974
+    },
+    {
+      "epoch": 0.2769745322788559,
+      "grad_norm": 0.699676254107354,
+      "learning_rate": 5.880790751198782e-07,
+      "loss": 1.564,
+      "step": 3975
+    },
+    {
+      "epoch": 0.2770442114064732,
+      "grad_norm": 0.6913603310218001,
+      "learning_rate": 5.880263568117939e-07,
+      "loss": 1.6078,
+      "step": 3976
+    },
+    {
+      "epoch": 0.2771138905340905,
+      "grad_norm": 0.6826484309669981,
+      "learning_rate": 5.879736287742698e-07,
+      "loss": 1.6187,
+      "step": 3977
+    },
+    {
+      "epoch": 0.2771835696617078,
+      "grad_norm": 0.7649106951889587,
+      "learning_rate": 5.879208910098327e-07,
+      "loss": 1.6896,
+      "step": 3978
+    },
+    {
+      "epoch": 0.27725324878932517,
+      "grad_norm": 0.686641906084249,
+      "learning_rate": 5.878681435210099e-07,
+      "loss": 1.4748,
+      "step": 3979
+    },
+    {
+      "epoch": 0.27732292791694246,
+      "grad_norm": 0.7413788736872214,
+      "learning_rate": 5.878153863103294e-07,
+      "loss": 1.6367,
+      "step": 3980
+    },
+    {
+      "epoch": 0.2773926070445598,
+      "grad_norm": 0.7146264681803929,
+      "learning_rate": 5.87762619380319e-07,
+      "loss": 1.423,
+      "step": 3981
+    },
+    {
+      "epoch": 0.2774622861721771,
+      "grad_norm": 0.7193547975548729,
+      "learning_rate": 5.877098427335077e-07,
+      "loss": 1.5612,
+      "step": 3982
+    },
+    {
+      "epoch": 0.27753196529979446,
+      "grad_norm": 0.6911172013518755,
+      "learning_rate": 5.876570563724246e-07,
+      "loss": 1.6294,
+      "step": 3983
+    },
+    {
+      "epoch": 0.27760164442741175,
+      "grad_norm": 0.7974987700336611,
+      "learning_rate": 5.876042602995991e-07,
+      "loss": 1.5826,
+      "step": 3984
+    },
+    {
+      "epoch": 0.2776713235550291,
+      "grad_norm": 0.7523075392775425,
+      "learning_rate": 5.875514545175619e-07,
+      "loss": 1.4486,
+      "step": 3985
+    },
+    {
+      "epoch": 0.2777410026826464,
+      "grad_norm": 0.7105791555841411,
+      "learning_rate": 5.874986390288428e-07,
+      "loss": 1.526,
+      "step": 3986
+    },
+    {
+      "epoch": 0.27781068181026375,
+      "grad_norm": 0.6911026667088176,
+      "learning_rate": 5.874458138359734e-07,
+      "loss": 1.4773,
+      "step": 3987
+    },
+    {
+      "epoch": 0.27788036093788104,
+      "grad_norm": 0.769290264633232,
+      "learning_rate": 5.873929789414849e-07,
+      "loss": 1.5241,
+      "step": 3988
+    },
+    {
+      "epoch": 0.2779500400654984,
+      "grad_norm": 0.7393921952918916,
+      "learning_rate": 5.873401343479093e-07,
+      "loss": 1.5214,
+      "step": 3989
+    },
+    {
+      "epoch": 0.2780197191931157,
+      "grad_norm": 0.6990692068762214,
+      "learning_rate": 5.872872800577792e-07,
+      "loss": 1.4944,
+      "step": 3990
+    },
+    {
+      "epoch": 0.27808939832073304,
+      "grad_norm": 0.7482070834918554,
+      "learning_rate": 5.872344160736273e-07,
+      "loss": 1.4291,
+      "step": 3991
+    },
+    {
+      "epoch": 0.27815907744835033,
+      "grad_norm": 0.7684926653711424,
+      "learning_rate": 5.87181542397987e-07,
+      "loss": 1.449,
+      "step": 3992
+    },
+    {
+      "epoch": 0.2782287565759677,
+      "grad_norm": 0.7400846004615286,
+      "learning_rate": 5.871286590333921e-07,
+      "loss": 1.571,
+      "step": 3993
+    },
+    {
+      "epoch": 0.278298435703585,
+      "grad_norm": 0.7117553166829831,
+      "learning_rate": 5.870757659823769e-07,
+      "loss": 1.5885,
+      "step": 3994
+    },
+    {
+      "epoch": 0.27836811483120233,
+      "grad_norm": 0.7295296260321572,
+      "learning_rate": 5.870228632474761e-07,
+      "loss": 1.573,
+      "step": 3995
+    },
+    {
+      "epoch": 0.2784377939588196,
+      "grad_norm": 0.7096933784710855,
+      "learning_rate": 5.869699508312251e-07,
+      "loss": 1.6313,
+      "step": 3996
+    },
+    {
+      "epoch": 0.278507473086437,
+      "grad_norm": 0.74818849404883,
+      "learning_rate": 5.869170287361592e-07,
+      "loss": 1.6776,
+      "step": 3997
+    },
+    {
+      "epoch": 0.27857715221405427,
+      "grad_norm": 0.6633573767886243,
+      "learning_rate": 5.868640969648149e-07,
+      "loss": 1.5791,
+      "step": 3998
+    },
+    {
+      "epoch": 0.2786468313416716,
+      "grad_norm": 0.7037151501018837,
+      "learning_rate": 5.868111555197287e-07,
+      "loss": 1.5421,
+      "step": 3999
+    },
+    {
+      "epoch": 0.2787165104692889,
+      "grad_norm": 0.7677023462476129,
+      "learning_rate": 5.867582044034374e-07,
+      "loss": 1.4824,
+      "step": 4000
+    },
+    {
+      "epoch": 0.27878618959690626,
+      "grad_norm": 0.6969241390359108,
+      "learning_rate": 5.86705243618479e-07,
+      "loss": 1.6216,
+      "step": 4001
+    },
+    {
+      "epoch": 0.27885586872452356,
+      "grad_norm": 0.6722182109071275,
+      "learning_rate": 5.86652273167391e-07,
+      "loss": 1.5251,
+      "step": 4002
+    },
+    {
+      "epoch": 0.2789255478521409,
+      "grad_norm": 0.7489623752803286,
+      "learning_rate": 5.865992930527123e-07,
+      "loss": 1.4817,
+      "step": 4003
+    },
+    {
+      "epoch": 0.2789952269797582,
+      "grad_norm": 0.7007812459698798,
+      "learning_rate": 5.865463032769814e-07,
+      "loss": 1.5225,
+      "step": 4004
+    },
+    {
+      "epoch": 0.27906490610737555,
+      "grad_norm": 0.7121219333031956,
+      "learning_rate": 5.86493303842738e-07,
+      "loss": 1.5117,
+      "step": 4005
+    },
+    {
+      "epoch": 0.27913458523499285,
+      "grad_norm": 0.6799027790844323,
+      "learning_rate": 5.864402947525218e-07,
+      "loss": 1.5251,
+      "step": 4006
+    },
+    {
+      "epoch": 0.2792042643626102,
+      "grad_norm": 0.7083461002334056,
+      "learning_rate": 5.863872760088732e-07,
+      "loss": 1.526,
+      "step": 4007
+    },
+    {
+      "epoch": 0.2792739434902275,
+      "grad_norm": 0.6869380534638632,
+      "learning_rate": 5.863342476143329e-07,
+      "loss": 1.548,
+      "step": 4008
+    },
+    {
+      "epoch": 0.27934362261784484,
+      "grad_norm": 0.7917821268802485,
+      "learning_rate": 5.86281209571442e-07,
+      "loss": 1.4851,
+      "step": 4009
+    },
+    {
+      "epoch": 0.27941330174546214,
+      "grad_norm": 0.7288384722578055,
+      "learning_rate": 5.862281618827423e-07,
+      "loss": 1.5221,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2794829808730795,
+      "grad_norm": 0.737623419529946,
+      "learning_rate": 5.861751045507761e-07,
+      "loss": 1.5496,
+      "step": 4011
+    },
+    {
+      "epoch": 0.2795526600006968,
+      "grad_norm": 0.7023565178847206,
+      "learning_rate": 5.861220375780858e-07,
+      "loss": 1.6295,
+      "step": 4012
+    },
+    {
+      "epoch": 0.27962233912831413,
+      "grad_norm": 0.8024061445348584,
+      "learning_rate": 5.860689609672146e-07,
+      "loss": 1.6185,
+      "step": 4013
+    },
+    {
+      "epoch": 0.27969201825593143,
+      "grad_norm": 0.7407329543983926,
+      "learning_rate": 5.86015874720706e-07,
+      "loss": 1.6296,
+      "step": 4014
+    },
+    {
+      "epoch": 0.2797616973835488,
+      "grad_norm": 0.656469116523215,
+      "learning_rate": 5.85962778841104e-07,
+      "loss": 1.4228,
+      "step": 4015
+    },
+    {
+      "epoch": 0.2798313765111661,
+      "grad_norm": 0.6946187772623037,
+      "learning_rate": 5.85909673330953e-07,
+      "loss": 1.4918,
+      "step": 4016
+    },
+    {
+      "epoch": 0.2799010556387834,
+      "grad_norm": 0.7373641595783593,
+      "learning_rate": 5.858565581927981e-07,
+      "loss": 1.6121,
+      "step": 4017
+    },
+    {
+      "epoch": 0.2799707347664007,
+      "grad_norm": 0.7532752424720006,
+      "learning_rate": 5.858034334291845e-07,
+      "loss": 1.5603,
+      "step": 4018
+    },
+    {
+      "epoch": 0.28004041389401807,
+      "grad_norm": 0.7264475887960263,
+      "learning_rate": 5.857502990426582e-07,
+      "loss": 1.4764,
+      "step": 4019
+    },
+    {
+      "epoch": 0.28011009302163536,
+      "grad_norm": 0.798784286364674,
+      "learning_rate": 5.856971550357653e-07,
+      "loss": 1.5344,
+      "step": 4020
+    },
+    {
+      "epoch": 0.2801797721492527,
+      "grad_norm": 0.7176879232704745,
+      "learning_rate": 5.856440014110529e-07,
+      "loss": 1.6604,
+      "step": 4021
+    },
+    {
+      "epoch": 0.28024945127687,
+      "grad_norm": 0.6773894896943677,
+      "learning_rate": 5.855908381710679e-07,
+      "loss": 1.4547,
+      "step": 4022
+    },
+    {
+      "epoch": 0.28031913040448736,
+      "grad_norm": 0.7127957867651743,
+      "learning_rate": 5.855376653183582e-07,
+      "loss": 1.5388,
+      "step": 4023
+    },
+    {
+      "epoch": 0.28038880953210465,
+      "grad_norm": 0.7166579306263381,
+      "learning_rate": 5.854844828554719e-07,
+      "loss": 1.5743,
+      "step": 4024
+    },
+    {
+      "epoch": 0.280458488659722,
+      "grad_norm": 0.7410980916162924,
+      "learning_rate": 5.854312907849575e-07,
+      "loss": 1.4147,
+      "step": 4025
+    },
+    {
+      "epoch": 0.2805281677873393,
+      "grad_norm": 0.7268041723692737,
+      "learning_rate": 5.853780891093643e-07,
+      "loss": 1.5175,
+      "step": 4026
+    },
+    {
+      "epoch": 0.28059784691495665,
+      "grad_norm": 0.7321423766832356,
+      "learning_rate": 5.853248778312416e-07,
+      "loss": 1.5852,
+      "step": 4027
+    },
+    {
+      "epoch": 0.28066752604257394,
+      "grad_norm": 0.6934596891068451,
+      "learning_rate": 5.852716569531395e-07,
+      "loss": 1.5311,
+      "step": 4028
+    },
+    {
+      "epoch": 0.2807372051701913,
+      "grad_norm": 0.7081751885804427,
+      "learning_rate": 5.852184264776085e-07,
+      "loss": 1.5793,
+      "step": 4029
+    },
+    {
+      "epoch": 0.2808068842978086,
+      "grad_norm": 0.715593124627546,
+      "learning_rate": 5.851651864071994e-07,
+      "loss": 1.5773,
+      "step": 4030
+    },
+    {
+      "epoch": 0.28087656342542594,
+      "grad_norm": 0.8029234811798415,
+      "learning_rate": 5.851119367444636e-07,
+      "loss": 1.5563,
+      "step": 4031
+    },
+    {
+      "epoch": 0.28094624255304324,
+      "grad_norm": 0.7836241719894106,
+      "learning_rate": 5.850586774919531e-07,
+      "loss": 1.5504,
+      "step": 4032
+    },
+    {
+      "epoch": 0.2810159216806606,
+      "grad_norm": 0.7444625174947146,
+      "learning_rate": 5.8500540865222e-07,
+      "loss": 1.5774,
+      "step": 4033
+    },
+    {
+      "epoch": 0.2810856008082779,
+      "grad_norm": 0.7682038540781259,
+      "learning_rate": 5.849521302278171e-07,
+      "loss": 1.6068,
+      "step": 4034
+    },
+    {
+      "epoch": 0.2811552799358952,
+      "grad_norm": 0.6977564285986588,
+      "learning_rate": 5.848988422212977e-07,
+      "loss": 1.539,
+      "step": 4035
+    },
+    {
+      "epoch": 0.2812249590635125,
+      "grad_norm": 0.6861978787885887,
+      "learning_rate": 5.848455446352152e-07,
+      "loss": 1.5767,
+      "step": 4036
+    },
+    {
+      "epoch": 0.2812946381911298,
+      "grad_norm": 0.7061641280191499,
+      "learning_rate": 5.847922374721241e-07,
+      "loss": 1.5446,
+      "step": 4037
+    },
+    {
+      "epoch": 0.28136431731874717,
+      "grad_norm": 0.7162478499272291,
+      "learning_rate": 5.847389207345788e-07,
+      "loss": 1.5821,
+      "step": 4038
+    },
+    {
+      "epoch": 0.28143399644636446,
+      "grad_norm": 0.704977072425101,
+      "learning_rate": 5.846855944251343e-07,
+      "loss": 1.6258,
+      "step": 4039
+    },
+    {
+      "epoch": 0.2815036755739818,
+      "grad_norm": 0.7989239527820603,
+      "learning_rate": 5.846322585463462e-07,
+      "loss": 1.5252,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2815733547015991,
+      "grad_norm": 0.7620500570883405,
+      "learning_rate": 5.845789131007705e-07,
+      "loss": 1.4945,
+      "step": 4041
+    },
+    {
+      "epoch": 0.28164303382921646,
+      "grad_norm": 0.7065137235650111,
+      "learning_rate": 5.845255580909634e-07,
+      "loss": 1.5273,
+      "step": 4042
+    },
+    {
+      "epoch": 0.28171271295683376,
+      "grad_norm": 0.72292442941867,
+      "learning_rate": 5.844721935194821e-07,
+      "loss": 1.6439,
+      "step": 4043
+    },
+    {
+      "epoch": 0.2817823920844511,
+      "grad_norm": 0.7135494906603791,
+      "learning_rate": 5.844188193888838e-07,
+      "loss": 1.6209,
+      "step": 4044
+    },
+    {
+      "epoch": 0.2818520712120684,
+      "grad_norm": 0.7238594633780343,
+      "learning_rate": 5.843654357017261e-07,
+      "loss": 1.6056,
+      "step": 4045
+    },
+    {
+      "epoch": 0.28192175033968575,
+      "grad_norm": 0.7195992753862425,
+      "learning_rate": 5.843120424605675e-07,
+      "loss": 1.6093,
+      "step": 4046
+    },
+    {
+      "epoch": 0.28199142946730305,
+      "grad_norm": 0.7102552233590842,
+      "learning_rate": 5.842586396679666e-07,
+      "loss": 1.5801,
+      "step": 4047
+    },
+    {
+      "epoch": 0.2820611085949204,
+      "grad_norm": 0.7237963977095361,
+      "learning_rate": 5.842052273264828e-07,
+      "loss": 1.599,
+      "step": 4048
+    },
+    {
+      "epoch": 0.2821307877225377,
+      "grad_norm": 0.7388017933797084,
+      "learning_rate": 5.841518054386754e-07,
+      "loss": 1.4611,
+      "step": 4049
+    },
+    {
+      "epoch": 0.28220046685015504,
+      "grad_norm": 0.7169800257321843,
+      "learning_rate": 5.840983740071046e-07,
+      "loss": 1.5368,
+      "step": 4050
+    },
+    {
+      "epoch": 0.28227014597777234,
+      "grad_norm": 0.7591467618718327,
+      "learning_rate": 5.840449330343311e-07,
+      "loss": 1.4682,
+      "step": 4051
+    },
+    {
+      "epoch": 0.2823398251053897,
+      "grad_norm": 0.7816992052749729,
+      "learning_rate": 5.839914825229157e-07,
+      "loss": 1.5462,
+      "step": 4052
+    },
+    {
+      "epoch": 0.282409504233007,
+      "grad_norm": 0.7678402184087277,
+      "learning_rate": 5.839380224754199e-07,
+      "loss": 1.411,
+      "step": 4053
+    },
+    {
+      "epoch": 0.28247918336062433,
+      "grad_norm": 0.7008928225544877,
+      "learning_rate": 5.838845528944057e-07,
+      "loss": 1.5995,
+      "step": 4054
+    },
+    {
+      "epoch": 0.2825488624882416,
+      "grad_norm": 0.7036680440880047,
+      "learning_rate": 5.838310737824353e-07,
+      "loss": 1.5138,
+      "step": 4055
+    },
+    {
+      "epoch": 0.282618541615859,
+      "grad_norm": 0.6918801095209847,
+      "learning_rate": 5.837775851420719e-07,
+      "loss": 1.5122,
+      "step": 4056
+    },
+    {
+      "epoch": 0.28268822074347627,
+      "grad_norm": 0.710151105832471,
+      "learning_rate": 5.837240869758785e-07,
+      "loss": 1.6328,
+      "step": 4057
+    },
+    {
+      "epoch": 0.2827578998710936,
+      "grad_norm": 6.128195129119697,
+      "learning_rate": 5.836705792864187e-07,
+      "loss": 1.4178,
+      "step": 4058
+    },
+    {
+      "epoch": 0.2828275789987109,
+      "grad_norm": 0.6829741022309734,
+      "learning_rate": 5.83617062076257e-07,
+      "loss": 1.5073,
+      "step": 4059
+    },
+    {
+      "epoch": 0.28289725812632827,
+      "grad_norm": 0.715180603423623,
+      "learning_rate": 5.835635353479579e-07,
+      "loss": 1.7047,
+      "step": 4060
+    },
+    {
+      "epoch": 0.28296693725394556,
+      "grad_norm": 0.7240989089434184,
+      "learning_rate": 5.835099991040865e-07,
+      "loss": 1.6419,
+      "step": 4061
+    },
+    {
+      "epoch": 0.2830366163815629,
+      "grad_norm": 1.4884192988469906,
+      "learning_rate": 5.834564533472084e-07,
+      "loss": 1.5715,
+      "step": 4062
+    },
+    {
+      "epoch": 0.2831062955091802,
+      "grad_norm": 0.6869430028687137,
+      "learning_rate": 5.834028980798897e-07,
+      "loss": 1.6057,
+      "step": 4063
+    },
+    {
+      "epoch": 0.28317597463679756,
+      "grad_norm": 0.7492810286800211,
+      "learning_rate": 5.833493333046969e-07,
+      "loss": 1.5047,
+      "step": 4064
+    },
+    {
+      "epoch": 0.28324565376441485,
+      "grad_norm": 0.6528949631329686,
+      "learning_rate": 5.832957590241967e-07,
+      "loss": 1.4532,
+      "step": 4065
+    },
+    {
+      "epoch": 0.2833153328920322,
+      "grad_norm": 0.69702222687072,
+      "learning_rate": 5.832421752409567e-07,
+      "loss": 1.5506,
+      "step": 4066
+    },
+    {
+      "epoch": 0.2833850120196495,
+      "grad_norm": 0.7244086971828632,
+      "learning_rate": 5.831885819575447e-07,
+      "loss": 1.5833,
+      "step": 4067
+    },
+    {
+      "epoch": 0.28345469114726685,
+      "grad_norm": 0.7019005426267827,
+      "learning_rate": 5.831349791765289e-07,
+      "loss": 1.4391,
+      "step": 4068
+    },
+    {
+      "epoch": 0.28352437027488414,
+      "grad_norm": 0.7200978367651489,
+      "learning_rate": 5.830813669004781e-07,
+      "loss": 1.477,
+      "step": 4069
+    },
+    {
+      "epoch": 0.2835940494025015,
+      "grad_norm": 0.6700016328199216,
+      "learning_rate": 5.830277451319616e-07,
+      "loss": 1.4821,
+      "step": 4070
+    },
+    {
+      "epoch": 0.2836637285301188,
+      "grad_norm": 0.7013166629088055,
+      "learning_rate": 5.829741138735491e-07,
+      "loss": 1.5592,
+      "step": 4071
+    },
+    {
+      "epoch": 0.28373340765773614,
+      "grad_norm": 0.7464737988751529,
+      "learning_rate": 5.829204731278105e-07,
+      "loss": 1.5095,
+      "step": 4072
+    },
+    {
+      "epoch": 0.28380308678535343,
+      "grad_norm": 0.7175291663947816,
+      "learning_rate": 5.828668228973166e-07,
+      "loss": 1.6289,
+      "step": 4073
+    },
+    {
+      "epoch": 0.2838727659129708,
+      "grad_norm": 0.772606784486943,
+      "learning_rate": 5.828131631846383e-07,
+      "loss": 1.4315,
+      "step": 4074
+    },
+    {
+      "epoch": 0.2839424450405881,
+      "grad_norm": 0.6821940080682076,
+      "learning_rate": 5.82759493992347e-07,
+      "loss": 1.603,
+      "step": 4075
+    },
+    {
+      "epoch": 0.2840121241682054,
+      "grad_norm": 0.6781925194254677,
+      "learning_rate": 5.827058153230149e-07,
+      "loss": 1.4863,
+      "step": 4076
+    },
+    {
+      "epoch": 0.2840818032958227,
+      "grad_norm": 0.7357770305547745,
+      "learning_rate": 5.826521271792142e-07,
+      "loss": 1.6178,
+      "step": 4077
+    },
+    {
+      "epoch": 0.28415148242344007,
+      "grad_norm": 0.6701447160402433,
+      "learning_rate": 5.825984295635178e-07,
+      "loss": 1.4617,
+      "step": 4078
+    },
+    {
+      "epoch": 0.28422116155105737,
+      "grad_norm": 0.7456256029388817,
+      "learning_rate": 5.82544722478499e-07,
+      "loss": 1.5877,
+      "step": 4079
+    },
+    {
+      "epoch": 0.2842908406786747,
+      "grad_norm": 0.7150322734011164,
+      "learning_rate": 5.824910059267316e-07,
+      "loss": 1.4595,
+      "step": 4080
+    },
+    {
+      "epoch": 0.284360519806292,
+      "grad_norm": 0.7762158243973858,
+      "learning_rate": 5.824372799107898e-07,
+      "loss": 1.4091,
+      "step": 4081
+    },
+    {
+      "epoch": 0.28443019893390936,
+      "grad_norm": 0.6999482983696204,
+      "learning_rate": 5.823835444332481e-07,
+      "loss": 1.5803,
+      "step": 4082
+    },
+    {
+      "epoch": 0.28449987806152666,
+      "grad_norm": 0.7179465348485711,
+      "learning_rate": 5.823297994966817e-07,
+      "loss": 1.5976,
+      "step": 4083
+    },
+    {
+      "epoch": 0.284569557189144,
+      "grad_norm": 0.6886196672918611,
+      "learning_rate": 5.822760451036663e-07,
+      "loss": 1.4657,
+      "step": 4084
+    },
+    {
+      "epoch": 0.2846392363167613,
+      "grad_norm": 0.7343911706726971,
+      "learning_rate": 5.822222812567777e-07,
+      "loss": 1.5281,
+      "step": 4085
+    },
+    {
+      "epoch": 0.28470891544437865,
+      "grad_norm": 0.7747706894286983,
+      "learning_rate": 5.821685079585925e-07,
+      "loss": 1.6097,
+      "step": 4086
+    },
+    {
+      "epoch": 0.28477859457199595,
+      "grad_norm": 0.7459684826116418,
+      "learning_rate": 5.821147252116877e-07,
+      "loss": 1.6015,
+      "step": 4087
+    },
+    {
+      "epoch": 0.2848482736996133,
+      "grad_norm": 0.7068888527292315,
+      "learning_rate": 5.820609330186406e-07,
+      "loss": 1.552,
+      "step": 4088
+    },
+    {
+      "epoch": 0.2849179528272306,
+      "grad_norm": 0.6849166782257857,
+      "learning_rate": 5.82007131382029e-07,
+      "loss": 1.6113,
+      "step": 4089
+    },
+    {
+      "epoch": 0.28498763195484794,
+      "grad_norm": 0.7440776028099806,
+      "learning_rate": 5.819533203044312e-07,
+      "loss": 1.8013,
+      "step": 4090
+    },
+    {
+      "epoch": 0.28505731108246524,
+      "grad_norm": 0.6927799473369971,
+      "learning_rate": 5.81899499788426e-07,
+      "loss": 1.5277,
+      "step": 4091
+    },
+    {
+      "epoch": 0.2851269902100826,
+      "grad_norm": 0.7629208548117576,
+      "learning_rate": 5.818456698365925e-07,
+      "loss": 1.5684,
+      "step": 4092
+    },
+    {
+      "epoch": 0.2851966693376999,
+      "grad_norm": 0.7884306874549732,
+      "learning_rate": 5.817918304515104e-07,
+      "loss": 1.5751,
+      "step": 4093
+    },
+    {
+      "epoch": 0.28526634846531723,
+      "grad_norm": 0.7391133979352249,
+      "learning_rate": 5.817379816357597e-07,
+      "loss": 1.5001,
+      "step": 4094
+    },
+    {
+      "epoch": 0.2853360275929345,
+      "grad_norm": 0.6698145354407297,
+      "learning_rate": 5.816841233919212e-07,
+      "loss": 1.4165,
+      "step": 4095
+    },
+    {
+      "epoch": 0.2854057067205519,
+      "grad_norm": 0.792643478722233,
+      "learning_rate": 5.816302557225756e-07,
+      "loss": 1.5447,
+      "step": 4096
+    },
+    {
+      "epoch": 0.28547538584816917,
+      "grad_norm": 0.7136178558225388,
+      "learning_rate": 5.815763786303045e-07,
+      "loss": 1.5373,
+      "step": 4097
+    },
+    {
+      "epoch": 0.2855450649757865,
+      "grad_norm": 0.7575039417167665,
+      "learning_rate": 5.815224921176897e-07,
+      "loss": 1.6437,
+      "step": 4098
+    },
+    {
+      "epoch": 0.2856147441034038,
+      "grad_norm": 0.7584843652406438,
+      "learning_rate": 5.814685961873138e-07,
+      "loss": 1.583,
+      "step": 4099
+    },
+    {
+      "epoch": 0.28568442323102117,
+      "grad_norm": 0.7576203179121972,
+      "learning_rate": 5.814146908417594e-07,
+      "loss": 1.5119,
+      "step": 4100
+    },
+    {
+      "epoch": 0.28575410235863846,
+      "grad_norm": 0.7598044895034622,
+      "learning_rate": 5.813607760836097e-07,
+      "loss": 1.574,
+      "step": 4101
+    },
+    {
+      "epoch": 0.2858237814862558,
+      "grad_norm": 0.6735244316784529,
+      "learning_rate": 5.813068519154485e-07,
+      "loss": 1.5738,
+      "step": 4102
+    },
+    {
+      "epoch": 0.2858934606138731,
+      "grad_norm": 0.6868829077636657,
+      "learning_rate": 5.812529183398598e-07,
+      "loss": 1.5322,
+      "step": 4103
+    },
+    {
+      "epoch": 0.28596313974149046,
+      "grad_norm": 0.7419971118801489,
+      "learning_rate": 5.811989753594286e-07,
+      "loss": 1.5351,
+      "step": 4104
+    },
+    {
+      "epoch": 0.28603281886910775,
+      "grad_norm": 0.7270043780129994,
+      "learning_rate": 5.811450229767396e-07,
+      "loss": 1.3956,
+      "step": 4105
+    },
+    {
+      "epoch": 0.2861024979967251,
+      "grad_norm": 0.7054461678186529,
+      "learning_rate": 5.810910611943784e-07,
+      "loss": 1.5752,
+      "step": 4106
+    },
+    {
+      "epoch": 0.2861721771243424,
+      "grad_norm": 0.8449176564524208,
+      "learning_rate": 5.810370900149311e-07,
+      "loss": 1.5074,
+      "step": 4107
+    },
+    {
+      "epoch": 0.28624185625195975,
+      "grad_norm": 0.7140677921475737,
+      "learning_rate": 5.809831094409838e-07,
+      "loss": 1.5529,
+      "step": 4108
+    },
+    {
+      "epoch": 0.28631153537957704,
+      "grad_norm": 0.7074402998095285,
+      "learning_rate": 5.809291194751236e-07,
+      "loss": 1.4844,
+      "step": 4109
+    },
+    {
+      "epoch": 0.2863812145071944,
+      "grad_norm": 0.7303614882913259,
+      "learning_rate": 5.808751201199379e-07,
+      "loss": 1.5742,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2864508936348117,
+      "grad_norm": 0.7183991860832905,
+      "learning_rate": 5.808211113780142e-07,
+      "loss": 1.4775,
+      "step": 4111
+    },
+    {
+      "epoch": 0.28652057276242904,
+      "grad_norm": 0.8003586154955813,
+      "learning_rate": 5.807670932519409e-07,
+      "loss": 1.5617,
+      "step": 4112
+    },
+    {
+      "epoch": 0.28659025189004633,
+      "grad_norm": 0.7474000305808857,
+      "learning_rate": 5.807130657443066e-07,
+      "loss": 1.4609,
+      "step": 4113
+    },
+    {
+      "epoch": 0.2866599310176637,
+      "grad_norm": 0.6674786568948675,
+      "learning_rate": 5.806590288577002e-07,
+      "loss": 1.473,
+      "step": 4114
+    },
+    {
+      "epoch": 0.286729610145281,
+      "grad_norm": 0.7505481394579682,
+      "learning_rate": 5.806049825947117e-07,
+      "loss": 1.5761,
+      "step": 4115
+    },
+    {
+      "epoch": 0.28679928927289833,
+      "grad_norm": 0.7296521930923773,
+      "learning_rate": 5.805509269579308e-07,
+      "loss": 1.4665,
+      "step": 4116
+    },
+    {
+      "epoch": 0.2868689684005156,
+      "grad_norm": 0.687910460173145,
+      "learning_rate": 5.804968619499479e-07,
+      "loss": 1.5137,
+      "step": 4117
+    },
+    {
+      "epoch": 0.286938647528133,
+      "grad_norm": 0.7760861013120258,
+      "learning_rate": 5.804427875733541e-07,
+      "loss": 1.5681,
+      "step": 4118
+    },
+    {
+      "epoch": 0.28700832665575027,
+      "grad_norm": 0.6496291087556836,
+      "learning_rate": 5.803887038307407e-07,
+      "loss": 1.4863,
+      "step": 4119
+    },
+    {
+      "epoch": 0.2870780057833676,
+      "grad_norm": 0.6963744044109444,
+      "learning_rate": 5.803346107246995e-07,
+      "loss": 1.4003,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2871476849109849,
+      "grad_norm": 0.7441529427156599,
+      "learning_rate": 5.802805082578228e-07,
+      "loss": 1.4934,
+      "step": 4121
+    },
+    {
+      "epoch": 0.28721736403860226,
+      "grad_norm": 0.6879418056717342,
+      "learning_rate": 5.802263964327031e-07,
+      "loss": 1.4888,
+      "step": 4122
+    },
+    {
+      "epoch": 0.28728704316621956,
+      "grad_norm": 0.7485835173848666,
+      "learning_rate": 5.801722752519338e-07,
+      "loss": 1.5704,
+      "step": 4123
+    },
+    {
+      "epoch": 0.2873567222938369,
+      "grad_norm": 0.702037064253478,
+      "learning_rate": 5.801181447181083e-07,
+      "loss": 1.5496,
+      "step": 4124
+    },
+    {
+      "epoch": 0.2874264014214542,
+      "grad_norm": 0.7026946239503448,
+      "learning_rate": 5.800640048338209e-07,
+      "loss": 1.5774,
+      "step": 4125
+    },
+    {
+      "epoch": 0.28749608054907155,
+      "grad_norm": 0.7491119945124249,
+      "learning_rate": 5.800098556016658e-07,
+      "loss": 1.4153,
+      "step": 4126
+    },
+    {
+      "epoch": 0.28756575967668885,
+      "grad_norm": 0.6902498898799309,
+      "learning_rate": 5.79955697024238e-07,
+      "loss": 1.5892,
+      "step": 4127
+    },
+    {
+      "epoch": 0.28763543880430614,
+      "grad_norm": 0.7063696011417466,
+      "learning_rate": 5.79901529104133e-07,
+      "loss": 1.5036,
+      "step": 4128
+    },
+    {
+      "epoch": 0.2877051179319235,
+      "grad_norm": 0.6764332043828444,
+      "learning_rate": 5.798473518439467e-07,
+      "loss": 1.4542,
+      "step": 4129
+    },
+    {
+      "epoch": 0.2877747970595408,
+      "grad_norm": 0.7574873462504658,
+      "learning_rate": 5.797931652462752e-07,
+      "loss": 1.5444,
+      "step": 4130
+    },
+    {
+      "epoch": 0.28784447618715814,
+      "grad_norm": 0.6919679634888707,
+      "learning_rate": 5.797389693137154e-07,
+      "loss": 1.5891,
+      "step": 4131
+    },
+    {
+      "epoch": 0.28791415531477543,
+      "grad_norm": 0.7380176334601344,
+      "learning_rate": 5.796847640488644e-07,
+      "loss": 1.6671,
+      "step": 4132
+    },
+    {
+      "epoch": 0.2879838344423928,
+      "grad_norm": 0.8050340076019923,
+      "learning_rate": 5.796305494543197e-07,
+      "loss": 1.5894,
+      "step": 4133
+    },
+    {
+      "epoch": 0.2880535135700101,
+      "grad_norm": 0.7487002362938737,
+      "learning_rate": 5.795763255326796e-07,
+      "loss": 1.4812,
+      "step": 4134
+    },
+    {
+      "epoch": 0.28812319269762743,
+      "grad_norm": 0.7369552008147401,
+      "learning_rate": 5.795220922865426e-07,
+      "loss": 1.4874,
+      "step": 4135
+    },
+    {
+      "epoch": 0.2881928718252447,
+      "grad_norm": 5.4765373368585175,
+      "learning_rate": 5.794678497185075e-07,
+      "loss": 1.5605,
+      "step": 4136
+    },
+    {
+      "epoch": 0.2882625509528621,
+      "grad_norm": 0.7284276944673779,
+      "learning_rate": 5.794135978311737e-07,
+      "loss": 1.5034,
+      "step": 4137
+    },
+    {
+      "epoch": 0.28833223008047937,
+      "grad_norm": 0.7995026203041848,
+      "learning_rate": 5.793593366271413e-07,
+      "loss": 1.5816,
+      "step": 4138
+    },
+    {
+      "epoch": 0.2884019092080967,
+      "grad_norm": 0.7179892897013607,
+      "learning_rate": 5.793050661090105e-07,
+      "loss": 1.5214,
+      "step": 4139
+    },
+    {
+      "epoch": 0.288471588335714,
+      "grad_norm": 0.7446689591066641,
+      "learning_rate": 5.79250786279382e-07,
+      "loss": 1.6026,
+      "step": 4140
+    },
+    {
+      "epoch": 0.28854126746333136,
+      "grad_norm": 0.7414451978063824,
+      "learning_rate": 5.791964971408569e-07,
+      "loss": 1.507,
+      "step": 4141
+    },
+    {
+      "epoch": 0.28861094659094866,
+      "grad_norm": 0.7118717774250654,
+      "learning_rate": 5.791421986960371e-07,
+      "loss": 1.4607,
+      "step": 4142
+    },
+    {
+      "epoch": 0.288680625718566,
+      "grad_norm": 0.7274986192626761,
+      "learning_rate": 5.790878909475246e-07,
+      "loss": 1.6512,
+      "step": 4143
+    },
+    {
+      "epoch": 0.2887503048461833,
+      "grad_norm": 0.7470589216705154,
+      "learning_rate": 5.790335738979218e-07,
+      "loss": 1.5379,
+      "step": 4144
+    },
+    {
+      "epoch": 0.28881998397380065,
+      "grad_norm": 0.7192847739232004,
+      "learning_rate": 5.789792475498319e-07,
+      "loss": 1.5424,
+      "step": 4145
+    },
+    {
+      "epoch": 0.28888966310141795,
+      "grad_norm": 0.7808717081028412,
+      "learning_rate": 5.789249119058582e-07,
+      "loss": 1.5689,
+      "step": 4146
+    },
+    {
+      "epoch": 0.2889593422290353,
+      "grad_norm": 0.7165423217935094,
+      "learning_rate": 5.788705669686047e-07,
+      "loss": 1.5309,
+      "step": 4147
+    },
+    {
+      "epoch": 0.2890290213566526,
+      "grad_norm": 0.7413403373071283,
+      "learning_rate": 5.788162127406755e-07,
+      "loss": 1.6779,
+      "step": 4148
+    },
+    {
+      "epoch": 0.28909870048426994,
+      "grad_norm": 0.7088883743890833,
+      "learning_rate": 5.787618492246754e-07,
+      "loss": 1.638,
+      "step": 4149
+    },
+    {
+      "epoch": 0.28916837961188724,
+      "grad_norm": 0.6863263458241512,
+      "learning_rate": 5.787074764232098e-07,
+      "loss": 1.5491,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2892380587395046,
+      "grad_norm": 0.7306183676069988,
+      "learning_rate": 5.786530943388843e-07,
+      "loss": 1.5322,
+      "step": 4151
+    },
+    {
+      "epoch": 0.2893077378671219,
+      "grad_norm": 0.718996793284283,
+      "learning_rate": 5.78598702974305e-07,
+      "loss": 1.567,
+      "step": 4152
+    },
+    {
+      "epoch": 0.28937741699473923,
+      "grad_norm": 0.7430070802419222,
+      "learning_rate": 5.785443023320782e-07,
+      "loss": 1.5341,
+      "step": 4153
+    },
+    {
+      "epoch": 0.28944709612235653,
+      "grad_norm": 0.7173099226439454,
+      "learning_rate": 5.784898924148112e-07,
+      "loss": 1.5728,
+      "step": 4154
+    },
+    {
+      "epoch": 0.2895167752499739,
+      "grad_norm": 0.7298294434551115,
+      "learning_rate": 5.784354732251114e-07,
+      "loss": 1.5361,
+      "step": 4155
+    },
+    {
+      "epoch": 0.2895864543775912,
+      "grad_norm": 0.7063577470300629,
+      "learning_rate": 5.783810447655865e-07,
+      "loss": 1.5852,
+      "step": 4156
+    },
+    {
+      "epoch": 0.2896561335052085,
+      "grad_norm": 0.746458774354094,
+      "learning_rate": 5.78326607038845e-07,
+      "loss": 1.5576,
+      "step": 4157
+    },
+    {
+      "epoch": 0.2897258126328258,
+      "grad_norm": 0.7408594387198453,
+      "learning_rate": 5.782721600474956e-07,
+      "loss": 1.5814,
+      "step": 4158
+    },
+    {
+      "epoch": 0.28979549176044317,
+      "grad_norm": 0.7430836064324008,
+      "learning_rate": 5.782177037941475e-07,
+      "loss": 1.467,
+      "step": 4159
+    },
+    {
+      "epoch": 0.28986517088806046,
+      "grad_norm": 0.7198430677931947,
+      "learning_rate": 5.781632382814104e-07,
+      "loss": 1.4687,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2899348500156778,
+      "grad_norm": 0.7252227953435738,
+      "learning_rate": 5.781087635118942e-07,
+      "loss": 1.6479,
+      "step": 4161
+    },
+    {
+      "epoch": 0.2900045291432951,
+      "grad_norm": 0.686562113644517,
+      "learning_rate": 5.780542794882098e-07,
+      "loss": 1.482,
+      "step": 4162
+    },
+    {
+      "epoch": 0.29007420827091246,
+      "grad_norm": 0.7181583474767024,
+      "learning_rate": 5.779997862129679e-07,
+      "loss": 1.5554,
+      "step": 4163
+    },
+    {
+      "epoch": 0.29014388739852975,
+      "grad_norm": 0.7422299136780489,
+      "learning_rate": 5.779452836887801e-07,
+      "loss": 1.6733,
+      "step": 4164
+    },
+    {
+      "epoch": 0.2902135665261471,
+      "grad_norm": 0.6927016820321847,
+      "learning_rate": 5.77890771918258e-07,
+      "loss": 1.5271,
+      "step": 4165
+    },
+    {
+      "epoch": 0.2902832456537644,
+      "grad_norm": 0.6829369165723617,
+      "learning_rate": 5.778362509040143e-07,
+      "loss": 1.5014,
+      "step": 4166
+    },
+    {
+      "epoch": 0.29035292478138175,
+      "grad_norm": 0.6620607702711369,
+      "learning_rate": 5.777817206486616e-07,
+      "loss": 1.6308,
+      "step": 4167
+    },
+    {
+      "epoch": 0.29042260390899904,
+      "grad_norm": 0.7073017154807985,
+      "learning_rate": 5.77727181154813e-07,
+      "loss": 1.5599,
+      "step": 4168
+    },
+    {
+      "epoch": 0.2904922830366164,
+      "grad_norm": 0.8033268829212622,
+      "learning_rate": 5.776726324250822e-07,
+      "loss": 1.5079,
+      "step": 4169
+    },
+    {
+      "epoch": 0.2905619621642337,
+      "grad_norm": 0.7209486887046835,
+      "learning_rate": 5.776180744620833e-07,
+      "loss": 1.5049,
+      "step": 4170
+    },
+    {
+      "epoch": 0.29063164129185104,
+      "grad_norm": 0.8022012464666788,
+      "learning_rate": 5.775635072684308e-07,
+      "loss": 1.4163,
+      "step": 4171
+    },
+    {
+      "epoch": 0.29070132041946833,
+      "grad_norm": 0.7277687745750111,
+      "learning_rate": 5.775089308467398e-07,
+      "loss": 1.5629,
+      "step": 4172
+    },
+    {
+      "epoch": 0.2907709995470857,
+      "grad_norm": 0.783412324097584,
+      "learning_rate": 5.774543451996256e-07,
+      "loss": 1.6353,
+      "step": 4173
+    },
+    {
+      "epoch": 0.290840678674703,
+      "grad_norm": 0.7209177364651561,
+      "learning_rate": 5.773997503297041e-07,
+      "loss": 1.5835,
+      "step": 4174
+    },
+    {
+      "epoch": 0.29091035780232033,
+      "grad_norm": 0.7199756416577356,
+      "learning_rate": 5.773451462395915e-07,
+      "loss": 1.5281,
+      "step": 4175
+    },
+    {
+      "epoch": 0.2909800369299376,
+      "grad_norm": 0.7322879678085863,
+      "learning_rate": 5.772905329319047e-07,
+      "loss": 1.3603,
+      "step": 4176
+    },
+    {
+      "epoch": 0.291049716057555,
+      "grad_norm": 0.6590412334963223,
+      "learning_rate": 5.772359104092607e-07,
+      "loss": 1.4824,
+      "step": 4177
+    },
+    {
+      "epoch": 0.29111939518517227,
+      "grad_norm": 0.7445913105232103,
+      "learning_rate": 5.771812786742773e-07,
+      "loss": 1.5666,
+      "step": 4178
+    },
+    {
+      "epoch": 0.2911890743127896,
+      "grad_norm": 0.7165402829789416,
+      "learning_rate": 5.771266377295725e-07,
+      "loss": 1.5071,
+      "step": 4179
+    },
+    {
+      "epoch": 0.2912587534404069,
+      "grad_norm": 0.6794848939177912,
+      "learning_rate": 5.770719875777647e-07,
+      "loss": 1.4203,
+      "step": 4180
+    },
+    {
+      "epoch": 0.29132843256802426,
+      "grad_norm": 0.7532430788398345,
+      "learning_rate": 5.77017328221473e-07,
+      "loss": 1.6119,
+      "step": 4181
+    },
+    {
+      "epoch": 0.29139811169564156,
+      "grad_norm": 0.7048663985537814,
+      "learning_rate": 5.769626596633167e-07,
+      "loss": 1.6076,
+      "step": 4182
+    },
+    {
+      "epoch": 0.2914677908232589,
+      "grad_norm": 0.7260710522617362,
+      "learning_rate": 5.769079819059156e-07,
+      "loss": 1.5931,
+      "step": 4183
+    },
+    {
+      "epoch": 0.2915374699508762,
+      "grad_norm": 0.7085546777892404,
+      "learning_rate": 5.768532949518901e-07,
+      "loss": 1.5788,
+      "step": 4184
+    },
+    {
+      "epoch": 0.29160714907849355,
+      "grad_norm": 0.7351359290443842,
+      "learning_rate": 5.767985988038609e-07,
+      "loss": 1.4791,
+      "step": 4185
+    },
+    {
+      "epoch": 0.29167682820611085,
+      "grad_norm": 0.7345803290205484,
+      "learning_rate": 5.767438934644489e-07,
+      "loss": 1.498,
+      "step": 4186
+    },
+    {
+      "epoch": 0.2917465073337282,
+      "grad_norm": 0.7275728966569317,
+      "learning_rate": 5.766891789362761e-07,
+      "loss": 1.5793,
+      "step": 4187
+    },
+    {
+      "epoch": 0.2918161864613455,
+      "grad_norm": 0.6385572204281225,
+      "learning_rate": 5.766344552219643e-07,
+      "loss": 1.5234,
+      "step": 4188
+    },
+    {
+      "epoch": 0.29188586558896285,
+      "grad_norm": 0.6878039246866837,
+      "learning_rate": 5.765797223241358e-07,
+      "loss": 1.529,
+      "step": 4189
+    },
+    {
+      "epoch": 0.29195554471658014,
+      "grad_norm": 0.7407162221510748,
+      "learning_rate": 5.765249802454138e-07,
+      "loss": 1.4297,
+      "step": 4190
+    },
+    {
+      "epoch": 0.2920252238441975,
+      "grad_norm": 0.6954475217623283,
+      "learning_rate": 5.764702289884216e-07,
+      "loss": 1.5887,
+      "step": 4191
+    },
+    {
+      "epoch": 0.2920949029718148,
+      "grad_norm": 0.7438942748646914,
+      "learning_rate": 5.764154685557829e-07,
+      "loss": 1.5754,
+      "step": 4192
+    },
+    {
+      "epoch": 0.29216458209943214,
+      "grad_norm": 0.7878280970340501,
+      "learning_rate": 5.763606989501221e-07,
+      "loss": 1.5281,
+      "step": 4193
+    },
+    {
+      "epoch": 0.29223426122704943,
+      "grad_norm": 0.6856592601877685,
+      "learning_rate": 5.763059201740636e-07,
+      "loss": 1.5176,
+      "step": 4194
+    },
+    {
+      "epoch": 0.2923039403546668,
+      "grad_norm": 0.8063588131594583,
+      "learning_rate": 5.762511322302326e-07,
+      "loss": 1.6245,
+      "step": 4195
+    },
+    {
+      "epoch": 0.2923736194822841,
+      "grad_norm": 0.7078649971602377,
+      "learning_rate": 5.761963351212548e-07,
+      "loss": 1.5293,
+      "step": 4196
+    },
+    {
+      "epoch": 0.2924432986099014,
+      "grad_norm": 0.6867274360359023,
+      "learning_rate": 5.761415288497562e-07,
+      "loss": 1.4691,
+      "step": 4197
+    },
+    {
+      "epoch": 0.2925129777375187,
+      "grad_norm": 0.6938725129579134,
+      "learning_rate": 5.76086713418363e-07,
+      "loss": 1.5709,
+      "step": 4198
+    },
+    {
+      "epoch": 0.29258265686513607,
+      "grad_norm": 0.6964301335240399,
+      "learning_rate": 5.760318888297023e-07,
+      "loss": 1.5175,
+      "step": 4199
+    },
+    {
+      "epoch": 0.29265233599275337,
+      "grad_norm": 0.7176137893004337,
+      "learning_rate": 5.759770550864012e-07,
+      "loss": 1.6088,
+      "step": 4200
+    },
+    {
+      "epoch": 0.2927220151203707,
+      "grad_norm": 0.6600968939310983,
+      "learning_rate": 5.759222121910876e-07,
+      "loss": 1.5252,
+      "step": 4201
+    },
+    {
+      "epoch": 0.292791694247988,
+      "grad_norm": 0.7713602795893594,
+      "learning_rate": 5.758673601463897e-07,
+      "loss": 1.7007,
+      "step": 4202
+    },
+    {
+      "epoch": 0.29286137337560536,
+      "grad_norm": 0.8521496590271287,
+      "learning_rate": 5.758124989549359e-07,
+      "loss": 1.3938,
+      "step": 4203
+    },
+    {
+      "epoch": 0.29293105250322266,
+      "grad_norm": 0.7064840849897108,
+      "learning_rate": 5.757576286193557e-07,
+      "loss": 1.6102,
+      "step": 4204
+    },
+    {
+      "epoch": 0.29300073163084,
+      "grad_norm": 0.7073473475058616,
+      "learning_rate": 5.75702749142278e-07,
+      "loss": 1.4099,
+      "step": 4205
+    },
+    {
+      "epoch": 0.2930704107584573,
+      "grad_norm": 0.7497971375461359,
+      "learning_rate": 5.756478605263332e-07,
+      "loss": 1.59,
+      "step": 4206
+    },
+    {
+      "epoch": 0.29314008988607465,
+      "grad_norm": 0.7483325900051251,
+      "learning_rate": 5.755929627741515e-07,
+      "loss": 1.5683,
+      "step": 4207
+    },
+    {
+      "epoch": 0.29320976901369195,
+      "grad_norm": 0.677131842423103,
+      "learning_rate": 5.755380558883638e-07,
+      "loss": 1.5654,
+      "step": 4208
+    },
+    {
+      "epoch": 0.2932794481413093,
+      "grad_norm": 0.7090599829679397,
+      "learning_rate": 5.754831398716012e-07,
+      "loss": 1.5853,
+      "step": 4209
+    },
+    {
+      "epoch": 0.2933491272689266,
+      "grad_norm": 0.6531789620922398,
+      "learning_rate": 5.754282147264955e-07,
+      "loss": 1.5282,
+      "step": 4210
+    },
+    {
+      "epoch": 0.29341880639654394,
+      "grad_norm": 0.7290449241279219,
+      "learning_rate": 5.753732804556789e-07,
+      "loss": 1.5285,
+      "step": 4211
+    },
+    {
+      "epoch": 0.29348848552416124,
+      "grad_norm": 0.696260443545441,
+      "learning_rate": 5.753183370617839e-07,
+      "loss": 1.5345,
+      "step": 4212
+    },
+    {
+      "epoch": 0.2935581646517786,
+      "grad_norm": 0.7252444210476912,
+      "learning_rate": 5.752633845474433e-07,
+      "loss": 1.5398,
+      "step": 4213
+    },
+    {
+      "epoch": 0.2936278437793959,
+      "grad_norm": 0.7312914360002516,
+      "learning_rate": 5.752084229152909e-07,
+      "loss": 1.5642,
+      "step": 4214
+    },
+    {
+      "epoch": 0.29369752290701323,
+      "grad_norm": 0.724005882627236,
+      "learning_rate": 5.751534521679603e-07,
+      "loss": 1.4495,
+      "step": 4215
+    },
+    {
+      "epoch": 0.2937672020346305,
+      "grad_norm": 0.7720337660286007,
+      "learning_rate": 5.750984723080859e-07,
+      "loss": 1.5852,
+      "step": 4216
+    },
+    {
+      "epoch": 0.2938368811622479,
+      "grad_norm": 0.6758714480813982,
+      "learning_rate": 5.750434833383024e-07,
+      "loss": 1.4688,
+      "step": 4217
+    },
+    {
+      "epoch": 0.29390656028986517,
+      "grad_norm": 0.6951183763692467,
+      "learning_rate": 5.74988485261245e-07,
+      "loss": 1.5259,
+      "step": 4218
+    },
+    {
+      "epoch": 0.29397623941748247,
+      "grad_norm": 0.6941762057762565,
+      "learning_rate": 5.749334780795495e-07,
+      "loss": 1.4688,
+      "step": 4219
+    },
+    {
+      "epoch": 0.2940459185450998,
+      "grad_norm": 0.7445950618080178,
+      "learning_rate": 5.748784617958516e-07,
+      "loss": 1.5763,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2941155976727171,
+      "grad_norm": 0.7090111805366819,
+      "learning_rate": 5.748234364127881e-07,
+      "loss": 1.6148,
+      "step": 4221
+    },
+    {
+      "epoch": 0.29418527680033446,
+      "grad_norm": 0.6931440631837924,
+      "learning_rate": 5.747684019329958e-07,
+      "loss": 1.4726,
+      "step": 4222
+    },
+    {
+      "epoch": 0.29425495592795176,
+      "grad_norm": 0.6616793591516503,
+      "learning_rate": 5.747133583591122e-07,
+      "loss": 1.495,
+      "step": 4223
+    },
+    {
+      "epoch": 0.2943246350555691,
+      "grad_norm": 0.7047857600329261,
+      "learning_rate": 5.746583056937749e-07,
+      "loss": 1.4379,
+      "step": 4224
+    },
+    {
+      "epoch": 0.2943943141831864,
+      "grad_norm": 0.7691171573820788,
+      "learning_rate": 5.746032439396223e-07,
+      "loss": 1.7682,
+      "step": 4225
+    },
+    {
+      "epoch": 0.29446399331080375,
+      "grad_norm": 0.6925654078498741,
+      "learning_rate": 5.745481730992929e-07,
+      "loss": 1.5043,
+      "step": 4226
+    },
+    {
+      "epoch": 0.29453367243842105,
+      "grad_norm": 0.7204041277305288,
+      "learning_rate": 5.74493093175426e-07,
+      "loss": 1.62,
+      "step": 4227
+    },
+    {
+      "epoch": 0.2946033515660384,
+      "grad_norm": 0.784007626003321,
+      "learning_rate": 5.74438004170661e-07,
+      "loss": 1.5427,
+      "step": 4228
+    },
+    {
+      "epoch": 0.2946730306936557,
+      "grad_norm": 0.7054496982926721,
+      "learning_rate": 5.743829060876379e-07,
+      "loss": 1.6203,
+      "step": 4229
+    },
+    {
+      "epoch": 0.29474270982127304,
+      "grad_norm": 0.7036820710436857,
+      "learning_rate": 5.743277989289972e-07,
+      "loss": 1.4374,
+      "step": 4230
+    },
+    {
+      "epoch": 0.29481238894889034,
+      "grad_norm": 0.7043837784856916,
+      "learning_rate": 5.742726826973797e-07,
+      "loss": 1.5413,
+      "step": 4231
+    },
+    {
+      "epoch": 0.2948820680765077,
+      "grad_norm": 0.7269167952414226,
+      "learning_rate": 5.742175573954266e-07,
+      "loss": 1.4719,
+      "step": 4232
+    },
+    {
+      "epoch": 0.294951747204125,
+      "grad_norm": 0.706476929301725,
+      "learning_rate": 5.741624230257798e-07,
+      "loss": 1.533,
+      "step": 4233
+    },
+    {
+      "epoch": 0.29502142633174233,
+      "grad_norm": 0.7356825908807404,
+      "learning_rate": 5.741072795910813e-07,
+      "loss": 1.437,
+      "step": 4234
+    },
+    {
+      "epoch": 0.2950911054593596,
+      "grad_norm": 0.7040136084896028,
+      "learning_rate": 5.740521270939737e-07,
+      "loss": 1.4609,
+      "step": 4235
+    },
+    {
+      "epoch": 0.295160784586977,
+      "grad_norm": 0.7340794981592375,
+      "learning_rate": 5.739969655371e-07,
+      "loss": 1.5591,
+      "step": 4236
+    },
+    {
+      "epoch": 0.29523046371459427,
+      "grad_norm": 0.6603536126599652,
+      "learning_rate": 5.739417949231038e-07,
+      "loss": 1.4947,
+      "step": 4237
+    },
+    {
+      "epoch": 0.2953001428422116,
+      "grad_norm": 0.7751036977717093,
+      "learning_rate": 5.73886615254629e-07,
+      "loss": 1.5717,
+      "step": 4238
+    },
+    {
+      "epoch": 0.2953698219698289,
+      "grad_norm": 0.791476669810281,
+      "learning_rate": 5.738314265343196e-07,
+      "loss": 1.5748,
+      "step": 4239
+    },
+    {
+      "epoch": 0.29543950109744627,
+      "grad_norm": 0.7362906055325384,
+      "learning_rate": 5.737762287648207e-07,
+      "loss": 1.605,
+      "step": 4240
+    },
+    {
+      "epoch": 0.29550918022506356,
+      "grad_norm": 0.723257144222618,
+      "learning_rate": 5.737210219487774e-07,
+      "loss": 1.4971,
+      "step": 4241
+    },
+    {
+      "epoch": 0.2955788593526809,
+      "grad_norm": 0.6947517250277025,
+      "learning_rate": 5.736658060888352e-07,
+      "loss": 1.5147,
+      "step": 4242
+    },
+    {
+      "epoch": 0.2956485384802982,
+      "grad_norm": 0.7270369921297178,
+      "learning_rate": 5.736105811876403e-07,
+      "loss": 1.492,
+      "step": 4243
+    },
+    {
+      "epoch": 0.29571821760791556,
+      "grad_norm": 0.708560313274988,
+      "learning_rate": 5.735553472478391e-07,
+      "loss": 1.3891,
+      "step": 4244
+    },
+    {
+      "epoch": 0.29578789673553285,
+      "grad_norm": 0.7285901567444775,
+      "learning_rate": 5.735001042720786e-07,
+      "loss": 1.6297,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2958575758631502,
+      "grad_norm": 0.7960332124526845,
+      "learning_rate": 5.734448522630062e-07,
+      "loss": 1.5527,
+      "step": 4246
+    },
+    {
+      "epoch": 0.2959272549907675,
+      "grad_norm": 0.7020497113259067,
+      "learning_rate": 5.733895912232694e-07,
+      "loss": 1.6506,
+      "step": 4247
+    },
+    {
+      "epoch": 0.29599693411838485,
+      "grad_norm": 0.715234323713793,
+      "learning_rate": 5.733343211555169e-07,
+      "loss": 1.5195,
+      "step": 4248
+    },
+    {
+      "epoch": 0.29606661324600214,
+      "grad_norm": 0.7032592778948775,
+      "learning_rate": 5.732790420623969e-07,
+      "loss": 1.5182,
+      "step": 4249
+    },
+    {
+      "epoch": 0.2961362923736195,
+      "grad_norm": 0.6762321778401789,
+      "learning_rate": 5.732237539465586e-07,
+      "loss": 1.5153,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2962059715012368,
+      "grad_norm": 0.6585325758077796,
+      "learning_rate": 5.731684568106518e-07,
+      "loss": 1.4654,
+      "step": 4251
+    },
+    {
+      "epoch": 0.29627565062885414,
+      "grad_norm": 0.7659023292111404,
+      "learning_rate": 5.731131506573262e-07,
+      "loss": 1.5343,
+      "step": 4252
+    },
+    {
+      "epoch": 0.29634532975647143,
+      "grad_norm": 0.7176805803072737,
+      "learning_rate": 5.730578354892322e-07,
+      "loss": 1.5905,
+      "step": 4253
+    },
+    {
+      "epoch": 0.2964150088840888,
+      "grad_norm": 0.698554222722932,
+      "learning_rate": 5.730025113090206e-07,
+      "loss": 1.486,
+      "step": 4254
+    },
+    {
+      "epoch": 0.2964846880117061,
+      "grad_norm": 0.7916029679362232,
+      "learning_rate": 5.729471781193427e-07,
+      "loss": 1.5445,
+      "step": 4255
+    },
+    {
+      "epoch": 0.2965543671393234,
+      "grad_norm": 0.6874181377702069,
+      "learning_rate": 5.728918359228502e-07,
+      "loss": 1.4352,
+      "step": 4256
+    },
+    {
+      "epoch": 0.2966240462669407,
+      "grad_norm": 0.7579198602498406,
+      "learning_rate": 5.728364847221953e-07,
+      "loss": 1.4514,
+      "step": 4257
+    },
+    {
+      "epoch": 0.29669372539455807,
+      "grad_norm": 0.7097472989631317,
+      "learning_rate": 5.727811245200302e-07,
+      "loss": 1.5043,
+      "step": 4258
+    },
+    {
+      "epoch": 0.29676340452217537,
+      "grad_norm": 0.6564182632603695,
+      "learning_rate": 5.727257553190083e-07,
+      "loss": 1.4526,
+      "step": 4259
+    },
+    {
+      "epoch": 0.2968330836497927,
+      "grad_norm": 0.7058523448481508,
+      "learning_rate": 5.726703771217827e-07,
+      "loss": 1.6799,
+      "step": 4260
+    },
+    {
+      "epoch": 0.29690276277741,
+      "grad_norm": 0.7245857438315936,
+      "learning_rate": 5.726149899310075e-07,
+      "loss": 1.5592,
+      "step": 4261
+    },
+    {
+      "epoch": 0.29697244190502736,
+      "grad_norm": 0.7315326028212079,
+      "learning_rate": 5.725595937493366e-07,
+      "loss": 1.5439,
+      "step": 4262
+    },
+    {
+      "epoch": 0.29704212103264466,
+      "grad_norm": 0.7195140259415446,
+      "learning_rate": 5.72504188579425e-07,
+      "loss": 1.4122,
+      "step": 4263
+    },
+    {
+      "epoch": 0.297111800160262,
+      "grad_norm": 0.7164306723650375,
+      "learning_rate": 5.724487744239278e-07,
+      "loss": 1.5216,
+      "step": 4264
+    },
+    {
+      "epoch": 0.2971814792878793,
+      "grad_norm": 0.7744066165079407,
+      "learning_rate": 5.723933512855005e-07,
+      "loss": 1.5427,
+      "step": 4265
+    },
+    {
+      "epoch": 0.29725115841549665,
+      "grad_norm": 0.6820385126931663,
+      "learning_rate": 5.72337919166799e-07,
+      "loss": 1.5005,
+      "step": 4266
+    },
+    {
+      "epoch": 0.29732083754311395,
+      "grad_norm": 0.7602095640104902,
+      "learning_rate": 5.7228247807048e-07,
+      "loss": 1.5856,
+      "step": 4267
+    },
+    {
+      "epoch": 0.2973905166707313,
+      "grad_norm": 0.7376863216947709,
+      "learning_rate": 5.722270279992e-07,
+      "loss": 1.5361,
+      "step": 4268
+    },
+    {
+      "epoch": 0.2974601957983486,
+      "grad_norm": 0.6976848335979555,
+      "learning_rate": 5.721715689556165e-07,
+      "loss": 1.6066,
+      "step": 4269
+    },
+    {
+      "epoch": 0.29752987492596594,
+      "grad_norm": 0.7387189895719177,
+      "learning_rate": 5.721161009423872e-07,
+      "loss": 1.5509,
+      "step": 4270
+    },
+    {
+      "epoch": 0.29759955405358324,
+      "grad_norm": 0.7470502890559497,
+      "learning_rate": 5.720606239621701e-07,
+      "loss": 1.5248,
+      "step": 4271
+    },
+    {
+      "epoch": 0.2976692331812006,
+      "grad_norm": 0.7507968058964529,
+      "learning_rate": 5.72005138017624e-07,
+      "loss": 1.5589,
+      "step": 4272
+    },
+    {
+      "epoch": 0.2977389123088179,
+      "grad_norm": 0.7697138737872022,
+      "learning_rate": 5.719496431114077e-07,
+      "loss": 1.4206,
+      "step": 4273
+    },
+    {
+      "epoch": 0.29780859143643523,
+      "grad_norm": 0.8122208452859123,
+      "learning_rate": 5.718941392461806e-07,
+      "loss": 1.7525,
+      "step": 4274
+    },
+    {
+      "epoch": 0.2978782705640525,
+      "grad_norm": 1.025275901310073,
+      "learning_rate": 5.718386264246029e-07,
+      "loss": 1.6115,
+      "step": 4275
+    },
+    {
+      "epoch": 0.2979479496916699,
+      "grad_norm": 0.7000608371774311,
+      "learning_rate": 5.717831046493345e-07,
+      "loss": 1.4787,
+      "step": 4276
+    },
+    {
+      "epoch": 0.2980176288192872,
+      "grad_norm": 0.7136322606534233,
+      "learning_rate": 5.717275739230363e-07,
+      "loss": 1.5461,
+      "step": 4277
+    },
+    {
+      "epoch": 0.2980873079469045,
+      "grad_norm": 0.747023635229759,
+      "learning_rate": 5.716720342483693e-07,
+      "loss": 1.6262,
+      "step": 4278
+    },
+    {
+      "epoch": 0.2981569870745218,
+      "grad_norm": 0.6962780662742107,
+      "learning_rate": 5.716164856279952e-07,
+      "loss": 1.5011,
+      "step": 4279
+    },
+    {
+      "epoch": 0.29822666620213917,
+      "grad_norm": 0.7586413862543497,
+      "learning_rate": 5.715609280645762e-07,
+      "loss": 1.5255,
+      "step": 4280
+    },
+    {
+      "epoch": 0.29829634532975646,
+      "grad_norm": 0.7048885106327465,
+      "learning_rate": 5.715053615607744e-07,
+      "loss": 1.5426,
+      "step": 4281
+    },
+    {
+      "epoch": 0.2983660244573738,
+      "grad_norm": 0.6822702776681451,
+      "learning_rate": 5.714497861192527e-07,
+      "loss": 1.564,
+      "step": 4282
+    },
+    {
+      "epoch": 0.2984357035849911,
+      "grad_norm": 0.738493339810172,
+      "learning_rate": 5.713942017426747e-07,
+      "loss": 1.4922,
+      "step": 4283
+    },
+    {
+      "epoch": 0.29850538271260846,
+      "grad_norm": 0.7083738712536051,
+      "learning_rate": 5.713386084337038e-07,
+      "loss": 1.3278,
+      "step": 4284
+    },
+    {
+      "epoch": 0.29857506184022575,
+      "grad_norm": 0.819880393354702,
+      "learning_rate": 5.712830061950042e-07,
+      "loss": 1.6089,
+      "step": 4285
+    },
+    {
+      "epoch": 0.2986447409678431,
+      "grad_norm": 0.7375639649661877,
+      "learning_rate": 5.712273950292404e-07,
+      "loss": 1.5276,
+      "step": 4286
+    },
+    {
+      "epoch": 0.2987144200954604,
+      "grad_norm": 0.7537522147243475,
+      "learning_rate": 5.711717749390776e-07,
+      "loss": 1.4919,
+      "step": 4287
+    },
+    {
+      "epoch": 0.29878409922307775,
+      "grad_norm": 0.6781966488725063,
+      "learning_rate": 5.711161459271812e-07,
+      "loss": 1.5011,
+      "step": 4288
+    },
+    {
+      "epoch": 0.29885377835069504,
+      "grad_norm": 0.8326959141972974,
+      "learning_rate": 5.710605079962171e-07,
+      "loss": 1.5864,
+      "step": 4289
+    },
+    {
+      "epoch": 0.2989234574783124,
+      "grad_norm": 0.7750678881838795,
+      "learning_rate": 5.710048611488512e-07,
+      "loss": 1.6775,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2989931366059297,
+      "grad_norm": 0.7339000247889513,
+      "learning_rate": 5.709492053877506e-07,
+      "loss": 1.627,
+      "step": 4291
+    },
+    {
+      "epoch": 0.29906281573354704,
+      "grad_norm": 0.7341569990588388,
+      "learning_rate": 5.708935407155824e-07,
+      "loss": 1.5164,
+      "step": 4292
+    },
+    {
+      "epoch": 0.29913249486116433,
+      "grad_norm": 0.8037081802813852,
+      "learning_rate": 5.708378671350141e-07,
+      "loss": 1.4966,
+      "step": 4293
+    },
+    {
+      "epoch": 0.2992021739887817,
+      "grad_norm": 0.6627371432304198,
+      "learning_rate": 5.707821846487136e-07,
+      "loss": 1.3968,
+      "step": 4294
+    },
+    {
+      "epoch": 0.299271853116399,
+      "grad_norm": 0.767241188025273,
+      "learning_rate": 5.707264932593494e-07,
+      "loss": 1.6399,
+      "step": 4295
+    },
+    {
+      "epoch": 0.29934153224401633,
+      "grad_norm": 0.6914039403317371,
+      "learning_rate": 5.706707929695905e-07,
+      "loss": 1.5115,
+      "step": 4296
+    },
+    {
+      "epoch": 0.2994112113716336,
+      "grad_norm": 0.6980323825789414,
+      "learning_rate": 5.706150837821059e-07,
+      "loss": 1.4639,
+      "step": 4297
+    },
+    {
+      "epoch": 0.299480890499251,
+      "grad_norm": 0.7171078956726246,
+      "learning_rate": 5.705593656995654e-07,
+      "loss": 1.611,
+      "step": 4298
+    },
+    {
+      "epoch": 0.29955056962686827,
+      "grad_norm": 0.7332984331849871,
+      "learning_rate": 5.705036387246393e-07,
+      "loss": 1.6235,
+      "step": 4299
+    },
+    {
+      "epoch": 0.2996202487544856,
+      "grad_norm": 0.7008562486112845,
+      "learning_rate": 5.704479028599979e-07,
+      "loss": 1.486,
+      "step": 4300
+    },
+    {
+      "epoch": 0.2996899278821029,
+      "grad_norm": 0.764475428172602,
+      "learning_rate": 5.703921581083123e-07,
+      "loss": 1.5944,
+      "step": 4301
+    },
+    {
+      "epoch": 0.29975960700972026,
+      "grad_norm": 0.9630472761992942,
+      "learning_rate": 5.703364044722539e-07,
+      "loss": 1.6101,
+      "step": 4302
+    },
+    {
+      "epoch": 0.29982928613733756,
+      "grad_norm": 0.7357985164170848,
+      "learning_rate": 5.702806419544945e-07,
+      "loss": 1.5459,
+      "step": 4303
+    },
+    {
+      "epoch": 0.2998989652649549,
+      "grad_norm": 0.6860971242574258,
+      "learning_rate": 5.702248705577064e-07,
+      "loss": 1.3803,
+      "step": 4304
+    },
+    {
+      "epoch": 0.2999686443925722,
+      "grad_norm": 0.7307420289792195,
+      "learning_rate": 5.701690902845622e-07,
+      "loss": 1.5831,
+      "step": 4305
+    },
+    {
+      "epoch": 0.30003832352018955,
+      "grad_norm": 0.7160864235298279,
+      "learning_rate": 5.701133011377349e-07,
+      "loss": 1.4446,
+      "step": 4306
+    },
+    {
+      "epoch": 0.30010800264780685,
+      "grad_norm": 0.7127607913585317,
+      "learning_rate": 5.700575031198983e-07,
+      "loss": 1.6148,
+      "step": 4307
+    },
+    {
+      "epoch": 0.3001776817754242,
+      "grad_norm": 0.7371661893242732,
+      "learning_rate": 5.700016962337264e-07,
+      "loss": 1.584,
+      "step": 4308
+    },
+    {
+      "epoch": 0.3002473609030415,
+      "grad_norm": 0.6862056831220482,
+      "learning_rate": 5.699458804818933e-07,
+      "loss": 1.5426,
+      "step": 4309
+    },
+    {
+      "epoch": 0.30031704003065884,
+      "grad_norm": 0.9409326647764209,
+      "learning_rate": 5.698900558670737e-07,
+      "loss": 1.5583,
+      "step": 4310
+    },
+    {
+      "epoch": 0.30038671915827614,
+      "grad_norm": 0.707354464843671,
+      "learning_rate": 5.698342223919433e-07,
+      "loss": 1.3867,
+      "step": 4311
+    },
+    {
+      "epoch": 0.30045639828589343,
+      "grad_norm": 0.6662497690540862,
+      "learning_rate": 5.697783800591775e-07,
+      "loss": 1.5286,
+      "step": 4312
+    },
+    {
+      "epoch": 0.3005260774135108,
+      "grad_norm": 0.6995012600576833,
+      "learning_rate": 5.697225288714523e-07,
+      "loss": 1.4394,
+      "step": 4313
+    },
+    {
+      "epoch": 0.3005957565411281,
+      "grad_norm": 0.6587298537731207,
+      "learning_rate": 5.696666688314442e-07,
+      "loss": 1.5089,
+      "step": 4314
+    },
+    {
+      "epoch": 0.30066543566874543,
+      "grad_norm": 0.6730697833442947,
+      "learning_rate": 5.696107999418305e-07,
+      "loss": 1.5451,
+      "step": 4315
+    },
+    {
+      "epoch": 0.3007351147963627,
+      "grad_norm": 0.7094606803913679,
+      "learning_rate": 5.69554922205288e-07,
+      "loss": 1.4908,
+      "step": 4316
+    },
+    {
+      "epoch": 0.3008047939239801,
+      "grad_norm": 0.6509023335460346,
+      "learning_rate": 5.69499035624495e-07,
+      "loss": 1.5202,
+      "step": 4317
+    },
+    {
+      "epoch": 0.30087447305159737,
+      "grad_norm": 0.7491001156354492,
+      "learning_rate": 5.694431402021292e-07,
+      "loss": 1.5196,
+      "step": 4318
+    },
+    {
+      "epoch": 0.3009441521792147,
+      "grad_norm": 0.6870728473339202,
+      "learning_rate": 5.693872359408696e-07,
+      "loss": 1.4587,
+      "step": 4319
+    },
+    {
+      "epoch": 0.301013831306832,
+      "grad_norm": 0.6913259421135415,
+      "learning_rate": 5.69331322843395e-07,
+      "loss": 1.6039,
+      "step": 4320
+    },
+    {
+      "epoch": 0.30108351043444936,
+      "grad_norm": 0.7570791386505102,
+      "learning_rate": 5.69275400912385e-07,
+      "loss": 1.6491,
+      "step": 4321
+    },
+    {
+      "epoch": 0.30115318956206666,
+      "grad_norm": 0.7419386183508703,
+      "learning_rate": 5.692194701505195e-07,
+      "loss": 1.5979,
+      "step": 4322
+    },
+    {
+      "epoch": 0.301222868689684,
+      "grad_norm": 0.7168472397929991,
+      "learning_rate": 5.691635305604789e-07,
+      "loss": 1.6062,
+      "step": 4323
+    },
+    {
+      "epoch": 0.3012925478173013,
+      "grad_norm": 0.7273318013404496,
+      "learning_rate": 5.691075821449437e-07,
+      "loss": 1.5357,
+      "step": 4324
+    },
+    {
+      "epoch": 0.30136222694491865,
+      "grad_norm": 0.703456514068365,
+      "learning_rate": 5.690516249065953e-07,
+      "loss": 1.7195,
+      "step": 4325
+    },
+    {
+      "epoch": 0.30143190607253595,
+      "grad_norm": 0.7146216193702607,
+      "learning_rate": 5.689956588481151e-07,
+      "loss": 1.5897,
+      "step": 4326
+    },
+    {
+      "epoch": 0.3015015852001533,
+      "grad_norm": 0.6979947721924662,
+      "learning_rate": 5.689396839721853e-07,
+      "loss": 1.5119,
+      "step": 4327
+    },
+    {
+      "epoch": 0.3015712643277706,
+      "grad_norm": 0.7286369517144264,
+      "learning_rate": 5.688837002814881e-07,
+      "loss": 1.5625,
+      "step": 4328
+    },
+    {
+      "epoch": 0.30164094345538794,
+      "grad_norm": 0.7002598809033803,
+      "learning_rate": 5.688277077787065e-07,
+      "loss": 1.4626,
+      "step": 4329
+    },
+    {
+      "epoch": 0.30171062258300524,
+      "grad_norm": 0.7152829336567289,
+      "learning_rate": 5.687717064665239e-07,
+      "loss": 1.6106,
+      "step": 4330
+    },
+    {
+      "epoch": 0.3017803017106226,
+      "grad_norm": 0.7341268521670578,
+      "learning_rate": 5.687156963476236e-07,
+      "loss": 1.5397,
+      "step": 4331
+    },
+    {
+      "epoch": 0.3018499808382399,
+      "grad_norm": 0.6993141305033542,
+      "learning_rate": 5.686596774246903e-07,
+      "loss": 1.4845,
+      "step": 4332
+    },
+    {
+      "epoch": 0.30191965996585723,
+      "grad_norm": 0.7425939500950345,
+      "learning_rate": 5.686036497004079e-07,
+      "loss": 1.5275,
+      "step": 4333
+    },
+    {
+      "epoch": 0.30198933909347453,
+      "grad_norm": 0.8728770492993477,
+      "learning_rate": 5.685476131774617e-07,
+      "loss": 1.617,
+      "step": 4334
+    },
+    {
+      "epoch": 0.3020590182210919,
+      "grad_norm": 0.7165920132387991,
+      "learning_rate": 5.684915678585372e-07,
+      "loss": 1.4761,
+      "step": 4335
+    },
+    {
+      "epoch": 0.3021286973487092,
+      "grad_norm": 0.7206231125446654,
+      "learning_rate": 5.684355137463201e-07,
+      "loss": 1.6038,
+      "step": 4336
+    },
+    {
+      "epoch": 0.3021983764763265,
+      "grad_norm": 0.7461449598093063,
+      "learning_rate": 5.683794508434965e-07,
+      "loss": 1.4806,
+      "step": 4337
+    },
+    {
+      "epoch": 0.3022680556039438,
+      "grad_norm": 0.7326703682099187,
+      "learning_rate": 5.683233791527532e-07,
+      "loss": 1.5679,
+      "step": 4338
+    },
+    {
+      "epoch": 0.30233773473156117,
+      "grad_norm": 0.7098570289778796,
+      "learning_rate": 5.682672986767771e-07,
+      "loss": 1.6465,
+      "step": 4339
+    },
+    {
+      "epoch": 0.30240741385917846,
+      "grad_norm": 0.7338137182161039,
+      "learning_rate": 5.682112094182559e-07,
+      "loss": 1.5681,
+      "step": 4340
+    },
+    {
+      "epoch": 0.3024770929867958,
+      "grad_norm": 0.7598766115415722,
+      "learning_rate": 5.681551113798774e-07,
+      "loss": 1.7009,
+      "step": 4341
+    },
+    {
+      "epoch": 0.3025467721144131,
+      "grad_norm": 0.6912780506537121,
+      "learning_rate": 5.680990045643299e-07,
+      "loss": 1.4817,
+      "step": 4342
+    },
+    {
+      "epoch": 0.30261645124203046,
+      "grad_norm": 0.7008636273971501,
+      "learning_rate": 5.680428889743023e-07,
+      "loss": 1.5617,
+      "step": 4343
+    },
+    {
+      "epoch": 0.30268613036964775,
+      "grad_norm": 0.7017393439038174,
+      "learning_rate": 5.679867646124837e-07,
+      "loss": 1.5042,
+      "step": 4344
+    },
+    {
+      "epoch": 0.3027558094972651,
+      "grad_norm": 0.7272041861547763,
+      "learning_rate": 5.679306314815636e-07,
+      "loss": 1.4487,
+      "step": 4345
+    },
+    {
+      "epoch": 0.3028254886248824,
+      "grad_norm": 0.7117474417769055,
+      "learning_rate": 5.678744895842321e-07,
+      "loss": 1.4574,
+      "step": 4346
+    },
+    {
+      "epoch": 0.30289516775249975,
+      "grad_norm": 0.7026478170110361,
+      "learning_rate": 5.678183389231796e-07,
+      "loss": 1.4453,
+      "step": 4347
+    },
+    {
+      "epoch": 0.30296484688011704,
+      "grad_norm": 0.713541128071128,
+      "learning_rate": 5.67762179501097e-07,
+      "loss": 1.5474,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3030345260077344,
+      "grad_norm": 0.7447825123847247,
+      "learning_rate": 5.677060113206756e-07,
+      "loss": 1.5962,
+      "step": 4349
+    },
+    {
+      "epoch": 0.3031042051353517,
+      "grad_norm": 0.6790629074737846,
+      "learning_rate": 5.67649834384607e-07,
+      "loss": 1.629,
+      "step": 4350
+    },
+    {
+      "epoch": 0.30317388426296904,
+      "grad_norm": 0.7346346969447306,
+      "learning_rate": 5.675936486955834e-07,
+      "loss": 1.6648,
+      "step": 4351
+    },
+    {
+      "epoch": 0.30324356339058633,
+      "grad_norm": 0.7110715667981556,
+      "learning_rate": 5.675374542562973e-07,
+      "loss": 1.5896,
+      "step": 4352
+    },
+    {
+      "epoch": 0.3033132425182037,
+      "grad_norm": 0.7142436991764997,
+      "learning_rate": 5.674812510694416e-07,
+      "loss": 1.5098,
+      "step": 4353
+    },
+    {
+      "epoch": 0.303382921645821,
+      "grad_norm": 0.7105548896306177,
+      "learning_rate": 5.674250391377097e-07,
+      "loss": 1.5081,
+      "step": 4354
+    },
+    {
+      "epoch": 0.30345260077343833,
+      "grad_norm": 0.7322611956071579,
+      "learning_rate": 5.673688184637956e-07,
+      "loss": 1.5588,
+      "step": 4355
+    },
+    {
+      "epoch": 0.3035222799010556,
+      "grad_norm": 0.6954733835570996,
+      "learning_rate": 5.673125890503932e-07,
+      "loss": 1.4469,
+      "step": 4356
+    },
+    {
+      "epoch": 0.303591959028673,
+      "grad_norm": 0.7341910322709105,
+      "learning_rate": 5.672563509001972e-07,
+      "loss": 1.4872,
+      "step": 4357
+    },
+    {
+      "epoch": 0.30366163815629027,
+      "grad_norm": 0.688288685718414,
+      "learning_rate": 5.672001040159026e-07,
+      "loss": 1.4781,
+      "step": 4358
+    },
+    {
+      "epoch": 0.3037313172839076,
+      "grad_norm": 0.7411868407592087,
+      "learning_rate": 5.67143848400205e-07,
+      "loss": 1.6026,
+      "step": 4359
+    },
+    {
+      "epoch": 0.3038009964115249,
+      "grad_norm": 0.6503515017270078,
+      "learning_rate": 5.670875840558005e-07,
+      "loss": 1.4431,
+      "step": 4360
+    },
+    {
+      "epoch": 0.30387067553914227,
+      "grad_norm": 0.7775967475885618,
+      "learning_rate": 5.67031310985385e-07,
+      "loss": 1.5814,
+      "step": 4361
+    },
+    {
+      "epoch": 0.30394035466675956,
+      "grad_norm": 0.6685209900171873,
+      "learning_rate": 5.669750291916554e-07,
+      "loss": 1.5307,
+      "step": 4362
+    },
+    {
+      "epoch": 0.3040100337943769,
+      "grad_norm": 0.7203373515957084,
+      "learning_rate": 5.669187386773087e-07,
+      "loss": 1.5483,
+      "step": 4363
+    },
+    {
+      "epoch": 0.3040797129219942,
+      "grad_norm": 0.7274571102331661,
+      "learning_rate": 5.668624394450428e-07,
+      "loss": 1.4758,
+      "step": 4364
+    },
+    {
+      "epoch": 0.30414939204961156,
+      "grad_norm": 0.7206220254266728,
+      "learning_rate": 5.668061314975553e-07,
+      "loss": 1.4899,
+      "step": 4365
+    },
+    {
+      "epoch": 0.30421907117722885,
+      "grad_norm": 0.710236911651575,
+      "learning_rate": 5.667498148375447e-07,
+      "loss": 1.642,
+      "step": 4366
+    },
+    {
+      "epoch": 0.3042887503048462,
+      "grad_norm": 0.7139976815719326,
+      "learning_rate": 5.666934894677099e-07,
+      "loss": 1.5409,
+      "step": 4367
+    },
+    {
+      "epoch": 0.3043584294324635,
+      "grad_norm": 0.6954576104030126,
+      "learning_rate": 5.666371553907501e-07,
+      "loss": 1.434,
+      "step": 4368
+    },
+    {
+      "epoch": 0.30442810856008085,
+      "grad_norm": 0.7030492892506666,
+      "learning_rate": 5.665808126093649e-07,
+      "loss": 1.3851,
+      "step": 4369
+    },
+    {
+      "epoch": 0.30449778768769814,
+      "grad_norm": 0.6610603452433989,
+      "learning_rate": 5.665244611262543e-07,
+      "loss": 1.4934,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3045674668153155,
+      "grad_norm": 0.7353844159537593,
+      "learning_rate": 5.66468100944119e-07,
+      "loss": 1.5959,
+      "step": 4371
+    },
+    {
+      "epoch": 0.3046371459429328,
+      "grad_norm": 0.7180170447023938,
+      "learning_rate": 5.664117320656596e-07,
+      "loss": 1.5525,
+      "step": 4372
+    },
+    {
+      "epoch": 0.30470682507055014,
+      "grad_norm": 0.7076752320951577,
+      "learning_rate": 5.663553544935777e-07,
+      "loss": 1.4768,
+      "step": 4373
+    },
+    {
+      "epoch": 0.30477650419816743,
+      "grad_norm": 0.7305586044258503,
+      "learning_rate": 5.662989682305748e-07,
+      "loss": 1.6246,
+      "step": 4374
+    },
+    {
+      "epoch": 0.3048461833257848,
+      "grad_norm": 0.7558819635443185,
+      "learning_rate": 5.662425732793532e-07,
+      "loss": 1.5732,
+      "step": 4375
+    },
+    {
+      "epoch": 0.3049158624534021,
+      "grad_norm": 0.7153477087807026,
+      "learning_rate": 5.661861696426154e-07,
+      "loss": 1.4463,
+      "step": 4376
+    },
+    {
+      "epoch": 0.3049855415810194,
+      "grad_norm": 0.7263139382865508,
+      "learning_rate": 5.661297573230644e-07,
+      "loss": 1.5023,
+      "step": 4377
+    },
+    {
+      "epoch": 0.3050552207086367,
+      "grad_norm": 0.7001941576629662,
+      "learning_rate": 5.660733363234035e-07,
+      "loss": 1.4831,
+      "step": 4378
+    },
+    {
+      "epoch": 0.30512489983625407,
+      "grad_norm": 0.7056967216013887,
+      "learning_rate": 5.660169066463367e-07,
+      "loss": 1.4512,
+      "step": 4379
+    },
+    {
+      "epoch": 0.30519457896387137,
+      "grad_norm": 0.722321962015521,
+      "learning_rate": 5.659604682945681e-07,
+      "loss": 1.6328,
+      "step": 4380
+    },
+    {
+      "epoch": 0.3052642580914887,
+      "grad_norm": 0.7459925387500808,
+      "learning_rate": 5.659040212708024e-07,
+      "loss": 1.5597,
+      "step": 4381
+    },
+    {
+      "epoch": 0.305333937219106,
+      "grad_norm": 0.6985403031047768,
+      "learning_rate": 5.658475655777445e-07,
+      "loss": 1.5844,
+      "step": 4382
+    },
+    {
+      "epoch": 0.30540361634672336,
+      "grad_norm": 0.6924190337217615,
+      "learning_rate": 5.657911012180999e-07,
+      "loss": 1.4897,
+      "step": 4383
+    },
+    {
+      "epoch": 0.30547329547434066,
+      "grad_norm": 0.6857257208268763,
+      "learning_rate": 5.657346281945748e-07,
+      "loss": 1.5364,
+      "step": 4384
+    },
+    {
+      "epoch": 0.305542974601958,
+      "grad_norm": 0.6960778169774027,
+      "learning_rate": 5.65678146509875e-07,
+      "loss": 1.6121,
+      "step": 4385
+    },
+    {
+      "epoch": 0.3056126537295753,
+      "grad_norm": 0.7042377335398031,
+      "learning_rate": 5.656216561667078e-07,
+      "loss": 1.6275,
+      "step": 4386
+    },
+    {
+      "epoch": 0.30568233285719265,
+      "grad_norm": 0.6764060757496547,
+      "learning_rate": 5.655651571677797e-07,
+      "loss": 1.5309,
+      "step": 4387
+    },
+    {
+      "epoch": 0.30575201198480995,
+      "grad_norm": 0.7420151820599352,
+      "learning_rate": 5.655086495157989e-07,
+      "loss": 1.515,
+      "step": 4388
+    },
+    {
+      "epoch": 0.3058216911124273,
+      "grad_norm": 0.7454176035431609,
+      "learning_rate": 5.654521332134729e-07,
+      "loss": 1.5449,
+      "step": 4389
+    },
+    {
+      "epoch": 0.3058913702400446,
+      "grad_norm": 0.7166026979384447,
+      "learning_rate": 5.653956082635102e-07,
+      "loss": 1.4952,
+      "step": 4390
+    },
+    {
+      "epoch": 0.30596104936766194,
+      "grad_norm": 0.696252526086783,
+      "learning_rate": 5.653390746686195e-07,
+      "loss": 1.4843,
+      "step": 4391
+    },
+    {
+      "epoch": 0.30603072849527924,
+      "grad_norm": 0.7198107998452128,
+      "learning_rate": 5.652825324315103e-07,
+      "loss": 1.5638,
+      "step": 4392
+    },
+    {
+      "epoch": 0.3061004076228966,
+      "grad_norm": 0.7111816471169524,
+      "learning_rate": 5.652259815548919e-07,
+      "loss": 1.6544,
+      "step": 4393
+    },
+    {
+      "epoch": 0.3061700867505139,
+      "grad_norm": 0.7687614335199426,
+      "learning_rate": 5.651694220414745e-07,
+      "loss": 1.5855,
+      "step": 4394
+    },
+    {
+      "epoch": 0.30623976587813123,
+      "grad_norm": 0.7454634904231932,
+      "learning_rate": 5.651128538939687e-07,
+      "loss": 1.6859,
+      "step": 4395
+    },
+    {
+      "epoch": 0.3063094450057485,
+      "grad_norm": 0.6789328892800132,
+      "learning_rate": 5.65056277115085e-07,
+      "loss": 1.4511,
+      "step": 4396
+    },
+    {
+      "epoch": 0.3063791241333659,
+      "grad_norm": 0.6967181617814054,
+      "learning_rate": 5.649996917075348e-07,
+      "loss": 1.5627,
+      "step": 4397
+    },
+    {
+      "epoch": 0.30644880326098317,
+      "grad_norm": 0.7221567523229675,
+      "learning_rate": 5.649430976740299e-07,
+      "loss": 1.4447,
+      "step": 4398
+    },
+    {
+      "epoch": 0.3065184823886005,
+      "grad_norm": 0.7419218988374826,
+      "learning_rate": 5.648864950172825e-07,
+      "loss": 1.6038,
+      "step": 4399
+    },
+    {
+      "epoch": 0.3065881615162178,
+      "grad_norm": 0.7543528666991685,
+      "learning_rate": 5.648298837400047e-07,
+      "loss": 1.5557,
+      "step": 4400
+    },
+    {
+      "epoch": 0.30665784064383517,
+      "grad_norm": 0.672566336166264,
+      "learning_rate": 5.647732638449098e-07,
+      "loss": 1.5224,
+      "step": 4401
+    },
+    {
+      "epoch": 0.30672751977145246,
+      "grad_norm": 0.6647018335496818,
+      "learning_rate": 5.64716635334711e-07,
+      "loss": 1.5032,
+      "step": 4402
+    },
+    {
+      "epoch": 0.30679719889906976,
+      "grad_norm": 0.7035658712733628,
+      "learning_rate": 5.646599982121222e-07,
+      "loss": 1.499,
+      "step": 4403
+    },
+    {
+      "epoch": 0.3068668780266871,
+      "grad_norm": 0.7324652829978336,
+      "learning_rate": 5.646033524798572e-07,
+      "loss": 1.6015,
+      "step": 4404
+    },
+    {
+      "epoch": 0.3069365571543044,
+      "grad_norm": 0.6581572710664613,
+      "learning_rate": 5.645466981406311e-07,
+      "loss": 1.4681,
+      "step": 4405
+    },
+    {
+      "epoch": 0.30700623628192175,
+      "grad_norm": 0.7379880634218996,
+      "learning_rate": 5.644900351971586e-07,
+      "loss": 1.5125,
+      "step": 4406
+    },
+    {
+      "epoch": 0.30707591540953905,
+      "grad_norm": 0.6937990883394584,
+      "learning_rate": 5.644333636521549e-07,
+      "loss": 1.534,
+      "step": 4407
+    },
+    {
+      "epoch": 0.3071455945371564,
+      "grad_norm": 0.709535109038696,
+      "learning_rate": 5.643766835083363e-07,
+      "loss": 1.5428,
+      "step": 4408
+    },
+    {
+      "epoch": 0.3072152736647737,
+      "grad_norm": 0.7080500790253447,
+      "learning_rate": 5.643199947684187e-07,
+      "loss": 1.6857,
+      "step": 4409
+    },
+    {
+      "epoch": 0.30728495279239104,
+      "grad_norm": 0.7306898899751795,
+      "learning_rate": 5.642632974351187e-07,
+      "loss": 1.5052,
+      "step": 4410
+    },
+    {
+      "epoch": 0.30735463192000834,
+      "grad_norm": 0.6916557398320974,
+      "learning_rate": 5.642065915111535e-07,
+      "loss": 1.5313,
+      "step": 4411
+    },
+    {
+      "epoch": 0.3074243110476257,
+      "grad_norm": 0.7456137117488936,
+      "learning_rate": 5.641498769992406e-07,
+      "loss": 1.6002,
+      "step": 4412
+    },
+    {
+      "epoch": 0.307493990175243,
+      "grad_norm": 0.7286947464768576,
+      "learning_rate": 5.640931539020978e-07,
+      "loss": 1.5603,
+      "step": 4413
+    },
+    {
+      "epoch": 0.30756366930286033,
+      "grad_norm": 0.7296045551328778,
+      "learning_rate": 5.640364222224435e-07,
+      "loss": 1.472,
+      "step": 4414
+    },
+    {
+      "epoch": 0.3076333484304776,
+      "grad_norm": 0.7188974984129131,
+      "learning_rate": 5.63979681962996e-07,
+      "loss": 1.5777,
+      "step": 4415
+    },
+    {
+      "epoch": 0.307703027558095,
+      "grad_norm": 0.6987921504744042,
+      "learning_rate": 5.639229331264748e-07,
+      "loss": 1.5582,
+      "step": 4416
+    },
+    {
+      "epoch": 0.30777270668571227,
+      "grad_norm": 0.6733705792089942,
+      "learning_rate": 5.638661757155995e-07,
+      "loss": 1.4362,
+      "step": 4417
+    },
+    {
+      "epoch": 0.3078423858133296,
+      "grad_norm": 0.7532019974955833,
+      "learning_rate": 5.638094097330898e-07,
+      "loss": 1.5592,
+      "step": 4418
+    },
+    {
+      "epoch": 0.3079120649409469,
+      "grad_norm": 0.7362829024275824,
+      "learning_rate": 5.63752635181666e-07,
+      "loss": 1.5044,
+      "step": 4419
+    },
+    {
+      "epoch": 0.30798174406856427,
+      "grad_norm": 0.6879131082588306,
+      "learning_rate": 5.63695852064049e-07,
+      "loss": 1.5981,
+      "step": 4420
+    },
+    {
+      "epoch": 0.30805142319618156,
+      "grad_norm": 0.7261822695523283,
+      "learning_rate": 5.636390603829599e-07,
+      "loss": 1.7112,
+      "step": 4421
+    },
+    {
+      "epoch": 0.3081211023237989,
+      "grad_norm": 0.8396134526299301,
+      "learning_rate": 5.635822601411203e-07,
+      "loss": 1.5046,
+      "step": 4422
+    },
+    {
+      "epoch": 0.3081907814514162,
+      "grad_norm": 0.6924912458603729,
+      "learning_rate": 5.635254513412522e-07,
+      "loss": 1.5348,
+      "step": 4423
+    },
+    {
+      "epoch": 0.30826046057903356,
+      "grad_norm": 0.7804285234866137,
+      "learning_rate": 5.634686339860779e-07,
+      "loss": 1.5138,
+      "step": 4424
+    },
+    {
+      "epoch": 0.30833013970665085,
+      "grad_norm": 0.7762150856399619,
+      "learning_rate": 5.634118080783203e-07,
+      "loss": 1.5775,
+      "step": 4425
+    },
+    {
+      "epoch": 0.3083998188342682,
+      "grad_norm": 0.7025050135748799,
+      "learning_rate": 5.633549736207026e-07,
+      "loss": 1.4535,
+      "step": 4426
+    },
+    {
+      "epoch": 0.3084694979618855,
+      "grad_norm": 0.7092815664252987,
+      "learning_rate": 5.632981306159483e-07,
+      "loss": 1.5258,
+      "step": 4427
+    },
+    {
+      "epoch": 0.30853917708950285,
+      "grad_norm": 0.7130146811467492,
+      "learning_rate": 5.632412790667818e-07,
+      "loss": 1.6053,
+      "step": 4428
+    },
+    {
+      "epoch": 0.30860885621712014,
+      "grad_norm": 0.7943626302984487,
+      "learning_rate": 5.631844189759271e-07,
+      "loss": 1.5869,
+      "step": 4429
+    },
+    {
+      "epoch": 0.3086785353447375,
+      "grad_norm": 0.7409423035412404,
+      "learning_rate": 5.631275503461091e-07,
+      "loss": 1.4726,
+      "step": 4430
+    },
+    {
+      "epoch": 0.3087482144723548,
+      "grad_norm": 0.7429732104868719,
+      "learning_rate": 5.630706731800535e-07,
+      "loss": 1.5541,
+      "step": 4431
+    },
+    {
+      "epoch": 0.30881789359997214,
+      "grad_norm": 0.7551340123253788,
+      "learning_rate": 5.630137874804855e-07,
+      "loss": 1.711,
+      "step": 4432
+    },
+    {
+      "epoch": 0.30888757272758943,
+      "grad_norm": 0.6698510994317796,
+      "learning_rate": 5.629568932501314e-07,
+      "loss": 1.4954,
+      "step": 4433
+    },
+    {
+      "epoch": 0.3089572518552068,
+      "grad_norm": 0.7099811918718965,
+      "learning_rate": 5.628999904917175e-07,
+      "loss": 1.4815,
+      "step": 4434
+    },
+    {
+      "epoch": 0.3090269309828241,
+      "grad_norm": 0.7550362727314823,
+      "learning_rate": 5.62843079207971e-07,
+      "loss": 1.5581,
+      "step": 4435
+    },
+    {
+      "epoch": 0.3090966101104414,
+      "grad_norm": 0.7185400605901908,
+      "learning_rate": 5.62786159401619e-07,
+      "loss": 1.5663,
+      "step": 4436
+    },
+    {
+      "epoch": 0.3091662892380587,
+      "grad_norm": 0.6874944643037497,
+      "learning_rate": 5.627292310753892e-07,
+      "loss": 1.6189,
+      "step": 4437
+    },
+    {
+      "epoch": 0.3092359683656761,
+      "grad_norm": 0.6862872862618756,
+      "learning_rate": 5.626722942320098e-07,
+      "loss": 1.526,
+      "step": 4438
+    },
+    {
+      "epoch": 0.30930564749329337,
+      "grad_norm": 0.777117696116382,
+      "learning_rate": 5.626153488742094e-07,
+      "loss": 1.609,
+      "step": 4439
+    },
+    {
+      "epoch": 0.3093753266209107,
+      "grad_norm": 0.7148010353520651,
+      "learning_rate": 5.625583950047168e-07,
+      "loss": 1.5535,
+      "step": 4440
+    },
+    {
+      "epoch": 0.309445005748528,
+      "grad_norm": 0.7045498396728237,
+      "learning_rate": 5.625014326262612e-07,
+      "loss": 1.5452,
+      "step": 4441
+    },
+    {
+      "epoch": 0.30951468487614536,
+      "grad_norm": 0.6919278984568086,
+      "learning_rate": 5.624444617415727e-07,
+      "loss": 1.4555,
+      "step": 4442
+    },
+    {
+      "epoch": 0.30958436400376266,
+      "grad_norm": 0.6905139175639234,
+      "learning_rate": 5.623874823533813e-07,
+      "loss": 1.5853,
+      "step": 4443
+    },
+    {
+      "epoch": 0.30965404313138,
+      "grad_norm": 0.7165338368059605,
+      "learning_rate": 5.623304944644177e-07,
+      "loss": 1.5571,
+      "step": 4444
+    },
+    {
+      "epoch": 0.3097237222589973,
+      "grad_norm": 0.7097791827485946,
+      "learning_rate": 5.622734980774126e-07,
+      "loss": 1.4597,
+      "step": 4445
+    },
+    {
+      "epoch": 0.30979340138661465,
+      "grad_norm": 0.6880495509954674,
+      "learning_rate": 5.622164931950975e-07,
+      "loss": 1.4994,
+      "step": 4446
+    },
+    {
+      "epoch": 0.30986308051423195,
+      "grad_norm": 0.7265350063500531,
+      "learning_rate": 5.621594798202044e-07,
+      "loss": 1.5729,
+      "step": 4447
+    },
+    {
+      "epoch": 0.3099327596418493,
+      "grad_norm": 0.7271006631212064,
+      "learning_rate": 5.621024579554652e-07,
+      "loss": 1.6254,
+      "step": 4448
+    },
+    {
+      "epoch": 0.3100024387694666,
+      "grad_norm": 0.7172550132449438,
+      "learning_rate": 5.620454276036126e-07,
+      "loss": 1.5689,
+      "step": 4449
+    },
+    {
+      "epoch": 0.31007211789708394,
+      "grad_norm": 0.6859933153200322,
+      "learning_rate": 5.619883887673798e-07,
+      "loss": 1.5004,
+      "step": 4450
+    },
+    {
+      "epoch": 0.31014179702470124,
+      "grad_norm": 0.747765073619636,
+      "learning_rate": 5.619313414494999e-07,
+      "loss": 1.5306,
+      "step": 4451
+    },
+    {
+      "epoch": 0.3102114761523186,
+      "grad_norm": 0.7128484452214556,
+      "learning_rate": 5.618742856527069e-07,
+      "loss": 1.5639,
+      "step": 4452
+    },
+    {
+      "epoch": 0.3102811552799359,
+      "grad_norm": 0.7681613031122471,
+      "learning_rate": 5.618172213797351e-07,
+      "loss": 1.6,
+      "step": 4453
+    },
+    {
+      "epoch": 0.31035083440755323,
+      "grad_norm": 0.7524879920443474,
+      "learning_rate": 5.617601486333189e-07,
+      "loss": 1.5381,
+      "step": 4454
+    },
+    {
+      "epoch": 0.31042051353517053,
+      "grad_norm": 0.7252015588543342,
+      "learning_rate": 5.617030674161936e-07,
+      "loss": 1.6141,
+      "step": 4455
+    },
+    {
+      "epoch": 0.3104901926627879,
+      "grad_norm": 0.7135283316475586,
+      "learning_rate": 5.616459777310946e-07,
+      "loss": 1.6824,
+      "step": 4456
+    },
+    {
+      "epoch": 0.3105598717904052,
+      "grad_norm": 0.7234761560418117,
+      "learning_rate": 5.615888795807577e-07,
+      "loss": 1.5512,
+      "step": 4457
+    },
+    {
+      "epoch": 0.3106295509180225,
+      "grad_norm": 0.6888581693005156,
+      "learning_rate": 5.61531772967919e-07,
+      "loss": 1.4997,
+      "step": 4458
+    },
+    {
+      "epoch": 0.3106992300456398,
+      "grad_norm": 0.7150965852736714,
+      "learning_rate": 5.614746578953155e-07,
+      "loss": 1.6749,
+      "step": 4459
+    },
+    {
+      "epoch": 0.31076890917325717,
+      "grad_norm": 0.685877969315356,
+      "learning_rate": 5.61417534365684e-07,
+      "loss": 1.5597,
+      "step": 4460
+    },
+    {
+      "epoch": 0.31083858830087446,
+      "grad_norm": 0.6710895076279765,
+      "learning_rate": 5.613604023817622e-07,
+      "loss": 1.5663,
+      "step": 4461
+    },
+    {
+      "epoch": 0.3109082674284918,
+      "grad_norm": 0.7201758805730143,
+      "learning_rate": 5.613032619462877e-07,
+      "loss": 1.534,
+      "step": 4462
+    },
+    {
+      "epoch": 0.3109779465561091,
+      "grad_norm": 0.7009125282431902,
+      "learning_rate": 5.612461130619991e-07,
+      "loss": 1.5826,
+      "step": 4463
+    },
+    {
+      "epoch": 0.31104762568372646,
+      "grad_norm": 0.7343800397441643,
+      "learning_rate": 5.611889557316349e-07,
+      "loss": 1.6119,
+      "step": 4464
+    },
+    {
+      "epoch": 0.31111730481134375,
+      "grad_norm": 0.6762556387092465,
+      "learning_rate": 5.611317899579342e-07,
+      "loss": 1.5175,
+      "step": 4465
+    },
+    {
+      "epoch": 0.3111869839389611,
+      "grad_norm": 0.7320624779852928,
+      "learning_rate": 5.610746157436364e-07,
+      "loss": 1.5275,
+      "step": 4466
+    },
+    {
+      "epoch": 0.3112566630665784,
+      "grad_norm": 0.6697280425825357,
+      "learning_rate": 5.610174330914817e-07,
+      "loss": 1.4777,
+      "step": 4467
+    },
+    {
+      "epoch": 0.31132634219419575,
+      "grad_norm": 0.7468461994111188,
+      "learning_rate": 5.609602420042102e-07,
+      "loss": 1.6182,
+      "step": 4468
+    },
+    {
+      "epoch": 0.31139602132181304,
+      "grad_norm": 0.7355989195989832,
+      "learning_rate": 5.609030424845627e-07,
+      "loss": 1.5292,
+      "step": 4469
+    },
+    {
+      "epoch": 0.3114657004494304,
+      "grad_norm": 0.7209632348788007,
+      "learning_rate": 5.608458345352802e-07,
+      "loss": 1.5888,
+      "step": 4470
+    },
+    {
+      "epoch": 0.3115353795770477,
+      "grad_norm": 0.6846198121785194,
+      "learning_rate": 5.607886181591043e-07,
+      "loss": 1.487,
+      "step": 4471
+    },
+    {
+      "epoch": 0.31160505870466504,
+      "grad_norm": 0.7210029872091488,
+      "learning_rate": 5.60731393358777e-07,
+      "loss": 1.5018,
+      "step": 4472
+    },
+    {
+      "epoch": 0.31167473783228233,
+      "grad_norm": 0.7261186538094028,
+      "learning_rate": 5.606741601370406e-07,
+      "loss": 1.4957,
+      "step": 4473
+    },
+    {
+      "epoch": 0.3117444169598997,
+      "grad_norm": 0.7004930303444802,
+      "learning_rate": 5.606169184966377e-07,
+      "loss": 1.4821,
+      "step": 4474
+    },
+    {
+      "epoch": 0.311814096087517,
+      "grad_norm": 0.6922215262031016,
+      "learning_rate": 5.605596684403115e-07,
+      "loss": 1.4644,
+      "step": 4475
+    },
+    {
+      "epoch": 0.31188377521513433,
+      "grad_norm": 0.7123564721272889,
+      "learning_rate": 5.605024099708058e-07,
+      "loss": 1.4204,
+      "step": 4476
+    },
+    {
+      "epoch": 0.3119534543427516,
+      "grad_norm": 0.6901484934307104,
+      "learning_rate": 5.60445143090864e-07,
+      "loss": 1.4364,
+      "step": 4477
+    },
+    {
+      "epoch": 0.312023133470369,
+      "grad_norm": 0.7434500197807802,
+      "learning_rate": 5.60387867803231e-07,
+      "loss": 1.5128,
+      "step": 4478
+    },
+    {
+      "epoch": 0.31209281259798627,
+      "grad_norm": 0.6970836184981862,
+      "learning_rate": 5.603305841106511e-07,
+      "loss": 1.5619,
+      "step": 4479
+    },
+    {
+      "epoch": 0.3121624917256036,
+      "grad_norm": 0.7172415376307196,
+      "learning_rate": 5.6027329201587e-07,
+      "loss": 1.6662,
+      "step": 4480
+    },
+    {
+      "epoch": 0.3122321708532209,
+      "grad_norm": 0.724545533459188,
+      "learning_rate": 5.602159915216326e-07,
+      "loss": 1.4708,
+      "step": 4481
+    },
+    {
+      "epoch": 0.31230184998083826,
+      "grad_norm": 0.6768537107167042,
+      "learning_rate": 5.601586826306853e-07,
+      "loss": 1.5061,
+      "step": 4482
+    },
+    {
+      "epoch": 0.31237152910845556,
+      "grad_norm": 0.7007067821148911,
+      "learning_rate": 5.601013653457743e-07,
+      "loss": 1.503,
+      "step": 4483
+    },
+    {
+      "epoch": 0.3124412082360729,
+      "grad_norm": 0.7139890398072082,
+      "learning_rate": 5.600440396696465e-07,
+      "loss": 1.5266,
+      "step": 4484
+    },
+    {
+      "epoch": 0.3125108873636902,
+      "grad_norm": 0.7267770350004812,
+      "learning_rate": 5.599867056050489e-07,
+      "loss": 1.4927,
+      "step": 4485
+    },
+    {
+      "epoch": 0.31258056649130755,
+      "grad_norm": 0.7346092705739481,
+      "learning_rate": 5.599293631547289e-07,
+      "loss": 1.5618,
+      "step": 4486
+    },
+    {
+      "epoch": 0.31265024561892485,
+      "grad_norm": 0.6883420305557534,
+      "learning_rate": 5.59872012321435e-07,
+      "loss": 1.4841,
+      "step": 4487
+    },
+    {
+      "epoch": 0.3127199247465422,
+      "grad_norm": 0.8690928931163032,
+      "learning_rate": 5.598146531079151e-07,
+      "loss": 1.6252,
+      "step": 4488
+    },
+    {
+      "epoch": 0.3127896038741595,
+      "grad_norm": 0.7386897051184834,
+      "learning_rate": 5.597572855169182e-07,
+      "loss": 1.5784,
+      "step": 4489
+    },
+    {
+      "epoch": 0.31285928300177684,
+      "grad_norm": 0.7397508406451234,
+      "learning_rate": 5.596999095511935e-07,
+      "loss": 1.5716,
+      "step": 4490
+    },
+    {
+      "epoch": 0.31292896212939414,
+      "grad_norm": 0.7126692429917938,
+      "learning_rate": 5.596425252134903e-07,
+      "loss": 1.5979,
+      "step": 4491
+    },
+    {
+      "epoch": 0.3129986412570115,
+      "grad_norm": 0.6840682521617562,
+      "learning_rate": 5.595851325065588e-07,
+      "loss": 1.6575,
+      "step": 4492
+    },
+    {
+      "epoch": 0.3130683203846288,
+      "grad_norm": 0.7597470315412264,
+      "learning_rate": 5.595277314331495e-07,
+      "loss": 1.5514,
+      "step": 4493
+    },
+    {
+      "epoch": 0.3131379995122461,
+      "grad_norm": 0.6596496330675966,
+      "learning_rate": 5.594703219960127e-07,
+      "loss": 1.4768,
+      "step": 4494
+    },
+    {
+      "epoch": 0.31320767863986343,
+      "grad_norm": 0.7101073477506992,
+      "learning_rate": 5.594129041979001e-07,
+      "loss": 1.4752,
+      "step": 4495
+    },
+    {
+      "epoch": 0.3132773577674807,
+      "grad_norm": 0.8087865878948145,
+      "learning_rate": 5.593554780415632e-07,
+      "loss": 1.5729,
+      "step": 4496
+    },
+    {
+      "epoch": 0.3133470368950981,
+      "grad_norm": 0.7625992856773187,
+      "learning_rate": 5.592980435297535e-07,
+      "loss": 1.5422,
+      "step": 4497
+    },
+    {
+      "epoch": 0.31341671602271537,
+      "grad_norm": 0.6867096727586379,
+      "learning_rate": 5.592406006652241e-07,
+      "loss": 1.4298,
+      "step": 4498
+    },
+    {
+      "epoch": 0.3134863951503327,
+      "grad_norm": 0.7457698816149559,
+      "learning_rate": 5.591831494507271e-07,
+      "loss": 1.5903,
+      "step": 4499
+    },
+    {
+      "epoch": 0.31355607427795,
+      "grad_norm": 0.7503797775935128,
+      "learning_rate": 5.591256898890162e-07,
+      "loss": 1.5962,
+      "step": 4500
+    },
+    {
+      "epoch": 0.31362575340556736,
+      "grad_norm": 0.7085594299784876,
+      "learning_rate": 5.590682219828446e-07,
+      "loss": 1.5023,
+      "step": 4501
+    },
+    {
+      "epoch": 0.31369543253318466,
+      "grad_norm": 0.7243760005682286,
+      "learning_rate": 5.590107457349667e-07,
+      "loss": 1.6518,
+      "step": 4502
+    },
+    {
+      "epoch": 0.313765111660802,
+      "grad_norm": 0.7448276470045246,
+      "learning_rate": 5.589532611481363e-07,
+      "loss": 1.6848,
+      "step": 4503
+    },
+    {
+      "epoch": 0.3138347907884193,
+      "grad_norm": 0.7494794861319963,
+      "learning_rate": 5.588957682251087e-07,
+      "loss": 1.5629,
+      "step": 4504
+    },
+    {
+      "epoch": 0.31390446991603665,
+      "grad_norm": 0.7213716227937492,
+      "learning_rate": 5.588382669686389e-07,
+      "loss": 1.5777,
+      "step": 4505
+    },
+    {
+      "epoch": 0.31397414904365395,
+      "grad_norm": 0.7646428526365497,
+      "learning_rate": 5.587807573814824e-07,
+      "loss": 1.5049,
+      "step": 4506
+    },
+    {
+      "epoch": 0.3140438281712713,
+      "grad_norm": 0.767120792824794,
+      "learning_rate": 5.587232394663951e-07,
+      "loss": 1.4491,
+      "step": 4507
+    },
+    {
+      "epoch": 0.3141135072988886,
+      "grad_norm": 0.7105591775635199,
+      "learning_rate": 5.586657132261337e-07,
+      "loss": 1.4434,
+      "step": 4508
+    },
+    {
+      "epoch": 0.31418318642650594,
+      "grad_norm": 0.758203714015006,
+      "learning_rate": 5.586081786634549e-07,
+      "loss": 1.7566,
+      "step": 4509
+    },
+    {
+      "epoch": 0.31425286555412324,
+      "grad_norm": 0.7786093682246001,
+      "learning_rate": 5.585506357811156e-07,
+      "loss": 1.5759,
+      "step": 4510
+    },
+    {
+      "epoch": 0.3143225446817406,
+      "grad_norm": 0.6442824769449441,
+      "learning_rate": 5.584930845818736e-07,
+      "loss": 1.5383,
+      "step": 4511
+    },
+    {
+      "epoch": 0.3143922238093579,
+      "grad_norm": 0.6987033992122678,
+      "learning_rate": 5.584355250684867e-07,
+      "loss": 1.4782,
+      "step": 4512
+    },
+    {
+      "epoch": 0.31446190293697523,
+      "grad_norm": 0.7123821044662978,
+      "learning_rate": 5.583779572437135e-07,
+      "loss": 1.4858,
+      "step": 4513
+    },
+    {
+      "epoch": 0.31453158206459253,
+      "grad_norm": 0.7612746697091406,
+      "learning_rate": 5.583203811103125e-07,
+      "loss": 1.5794,
+      "step": 4514
+    },
+    {
+      "epoch": 0.3146012611922099,
+      "grad_norm": 0.6833845167460846,
+      "learning_rate": 5.582627966710432e-07,
+      "loss": 1.4549,
+      "step": 4515
+    },
+    {
+      "epoch": 0.3146709403198272,
+      "grad_norm": 0.7935778077228008,
+      "learning_rate": 5.582052039286649e-07,
+      "loss": 1.6172,
+      "step": 4516
+    },
+    {
+      "epoch": 0.3147406194474445,
+      "grad_norm": 0.7301222093535693,
+      "learning_rate": 5.581476028859377e-07,
+      "loss": 1.5544,
+      "step": 4517
+    },
+    {
+      "epoch": 0.3148102985750618,
+      "grad_norm": 0.7456619945461759,
+      "learning_rate": 5.580899935456218e-07,
+      "loss": 1.588,
+      "step": 4518
+    },
+    {
+      "epoch": 0.31487997770267917,
+      "grad_norm": 0.7179618775554296,
+      "learning_rate": 5.580323759104781e-07,
+      "loss": 1.5222,
+      "step": 4519
+    },
+    {
+      "epoch": 0.31494965683029646,
+      "grad_norm": 0.7132286895714686,
+      "learning_rate": 5.579747499832679e-07,
+      "loss": 1.434,
+      "step": 4520
+    },
+    {
+      "epoch": 0.3150193359579138,
+      "grad_norm": 0.7465885391719073,
+      "learning_rate": 5.579171157667522e-07,
+      "loss": 1.5557,
+      "step": 4521
+    },
+    {
+      "epoch": 0.3150890150855311,
+      "grad_norm": 0.749662806631823,
+      "learning_rate": 5.578594732636936e-07,
+      "loss": 1.6624,
+      "step": 4522
+    },
+    {
+      "epoch": 0.31515869421314846,
+      "grad_norm": 0.949124921903201,
+      "learning_rate": 5.578018224768542e-07,
+      "loss": 1.4182,
+      "step": 4523
+    },
+    {
+      "epoch": 0.31522837334076576,
+      "grad_norm": 0.744353130865854,
+      "learning_rate": 5.577441634089965e-07,
+      "loss": 1.6209,
+      "step": 4524
+    },
+    {
+      "epoch": 0.3152980524683831,
+      "grad_norm": 0.6873792099313955,
+      "learning_rate": 5.576864960628839e-07,
+      "loss": 1.4797,
+      "step": 4525
+    },
+    {
+      "epoch": 0.3153677315960004,
+      "grad_norm": 0.6994386653153584,
+      "learning_rate": 5.5762882044128e-07,
+      "loss": 1.4613,
+      "step": 4526
+    },
+    {
+      "epoch": 0.31543741072361775,
+      "grad_norm": 0.705522154107607,
+      "learning_rate": 5.575711365469486e-07,
+      "loss": 1.6182,
+      "step": 4527
+    },
+    {
+      "epoch": 0.31550708985123505,
+      "grad_norm": 0.7725865357784353,
+      "learning_rate": 5.57513444382654e-07,
+      "loss": 1.569,
+      "step": 4528
+    },
+    {
+      "epoch": 0.3155767689788524,
+      "grad_norm": 0.6845243748298593,
+      "learning_rate": 5.574557439511612e-07,
+      "loss": 1.5217,
+      "step": 4529
+    },
+    {
+      "epoch": 0.3156464481064697,
+      "grad_norm": 0.7261032424467299,
+      "learning_rate": 5.573980352552348e-07,
+      "loss": 1.6041,
+      "step": 4530
+    },
+    {
+      "epoch": 0.31571612723408704,
+      "grad_norm": 0.773279494758788,
+      "learning_rate": 5.573403182976408e-07,
+      "loss": 1.6184,
+      "step": 4531
+    },
+    {
+      "epoch": 0.31578580636170434,
+      "grad_norm": 0.9622361724689067,
+      "learning_rate": 5.572825930811449e-07,
+      "loss": 1.6083,
+      "step": 4532
+    },
+    {
+      "epoch": 0.3158554854893217,
+      "grad_norm": 0.7432584934801373,
+      "learning_rate": 5.572248596085133e-07,
+      "loss": 1.436,
+      "step": 4533
+    },
+    {
+      "epoch": 0.315925164616939,
+      "grad_norm": 0.7427295237224485,
+      "learning_rate": 5.571671178825131e-07,
+      "loss": 1.5585,
+      "step": 4534
+    },
+    {
+      "epoch": 0.31599484374455633,
+      "grad_norm": 0.6928269582168132,
+      "learning_rate": 5.57109367905911e-07,
+      "loss": 1.4105,
+      "step": 4535
+    },
+    {
+      "epoch": 0.3160645228721736,
+      "grad_norm": 0.6728242767255252,
+      "learning_rate": 5.570516096814747e-07,
+      "loss": 1.5522,
+      "step": 4536
+    },
+    {
+      "epoch": 0.316134201999791,
+      "grad_norm": 0.7263367317524648,
+      "learning_rate": 5.569938432119721e-07,
+      "loss": 1.6393,
+      "step": 4537
+    },
+    {
+      "epoch": 0.31620388112740827,
+      "grad_norm": 0.715112801174288,
+      "learning_rate": 5.569360685001715e-07,
+      "loss": 1.4807,
+      "step": 4538
+    },
+    {
+      "epoch": 0.3162735602550256,
+      "grad_norm": 0.7492932996364073,
+      "learning_rate": 5.568782855488413e-07,
+      "loss": 1.6866,
+      "step": 4539
+    },
+    {
+      "epoch": 0.3163432393826429,
+      "grad_norm": 0.6986507984234215,
+      "learning_rate": 5.568204943607508e-07,
+      "loss": 1.5787,
+      "step": 4540
+    },
+    {
+      "epoch": 0.31641291851026027,
+      "grad_norm": 0.6888654643145271,
+      "learning_rate": 5.567626949386696e-07,
+      "loss": 1.6045,
+      "step": 4541
+    },
+    {
+      "epoch": 0.31648259763787756,
+      "grad_norm": 0.6805162691732184,
+      "learning_rate": 5.567048872853675e-07,
+      "loss": 1.5848,
+      "step": 4542
+    },
+    {
+      "epoch": 0.3165522767654949,
+      "grad_norm": 0.6884368305487023,
+      "learning_rate": 5.566470714036145e-07,
+      "loss": 1.55,
+      "step": 4543
+    },
+    {
+      "epoch": 0.3166219558931122,
+      "grad_norm": 0.7570329409020771,
+      "learning_rate": 5.565892472961816e-07,
+      "loss": 1.6468,
+      "step": 4544
+    },
+    {
+      "epoch": 0.31669163502072956,
+      "grad_norm": 0.7059467717662286,
+      "learning_rate": 5.565314149658398e-07,
+      "loss": 1.4445,
+      "step": 4545
+    },
+    {
+      "epoch": 0.31676131414834685,
+      "grad_norm": 0.6657695365455025,
+      "learning_rate": 5.564735744153601e-07,
+      "loss": 1.4957,
+      "step": 4546
+    },
+    {
+      "epoch": 0.3168309932759642,
+      "grad_norm": 0.7407102775583433,
+      "learning_rate": 5.56415725647515e-07,
+      "loss": 1.5459,
+      "step": 4547
+    },
+    {
+      "epoch": 0.3169006724035815,
+      "grad_norm": 0.7106313167914474,
+      "learning_rate": 5.563578686650763e-07,
+      "loss": 1.5298,
+      "step": 4548
+    },
+    {
+      "epoch": 0.31697035153119885,
+      "grad_norm": 0.682333055169887,
+      "learning_rate": 5.563000034708168e-07,
+      "loss": 1.5829,
+      "step": 4549
+    },
+    {
+      "epoch": 0.31704003065881614,
+      "grad_norm": 0.7192818302564468,
+      "learning_rate": 5.562421300675094e-07,
+      "loss": 1.419,
+      "step": 4550
+    },
+    {
+      "epoch": 0.3171097097864335,
+      "grad_norm": 0.7430656876330435,
+      "learning_rate": 5.561842484579276e-07,
+      "loss": 1.5114,
+      "step": 4551
+    },
+    {
+      "epoch": 0.3171793889140508,
+      "grad_norm": 0.7111699564860365,
+      "learning_rate": 5.561263586448452e-07,
+      "loss": 1.6344,
+      "step": 4552
+    },
+    {
+      "epoch": 0.31724906804166814,
+      "grad_norm": 0.7041991520578657,
+      "learning_rate": 5.560684606310363e-07,
+      "loss": 1.4087,
+      "step": 4553
+    },
+    {
+      "epoch": 0.31731874716928543,
+      "grad_norm": 0.6955203999717197,
+      "learning_rate": 5.560105544192756e-07,
+      "loss": 1.5001,
+      "step": 4554
+    },
+    {
+      "epoch": 0.3173884262969028,
+      "grad_norm": 0.642220065364471,
+      "learning_rate": 5.559526400123382e-07,
+      "loss": 1.3332,
+      "step": 4555
+    },
+    {
+      "epoch": 0.3174581054245201,
+      "grad_norm": 0.7146233847081206,
+      "learning_rate": 5.558947174129991e-07,
+      "loss": 1.5475,
+      "step": 4556
+    },
+    {
+      "epoch": 0.3175277845521374,
+      "grad_norm": 0.6547314881825679,
+      "learning_rate": 5.558367866240346e-07,
+      "loss": 1.451,
+      "step": 4557
+    },
+    {
+      "epoch": 0.3175974636797547,
+      "grad_norm": 0.7557513478775116,
+      "learning_rate": 5.557788476482202e-07,
+      "loss": 1.512,
+      "step": 4558
+    },
+    {
+      "epoch": 0.31766714280737207,
+      "grad_norm": 0.7504684331261248,
+      "learning_rate": 5.557209004883331e-07,
+      "loss": 1.4838,
+      "step": 4559
+    },
+    {
+      "epoch": 0.31773682193498937,
+      "grad_norm": 0.6641025593399148,
+      "learning_rate": 5.556629451471498e-07,
+      "loss": 1.4582,
+      "step": 4560
+    },
+    {
+      "epoch": 0.3178065010626067,
+      "grad_norm": 0.6858127362399964,
+      "learning_rate": 5.556049816274479e-07,
+      "loss": 1.4477,
+      "step": 4561
+    },
+    {
+      "epoch": 0.317876180190224,
+      "grad_norm": 0.7816268642559431,
+      "learning_rate": 5.555470099320049e-07,
+      "loss": 1.4744,
+      "step": 4562
+    },
+    {
+      "epoch": 0.31794585931784136,
+      "grad_norm": 0.7280784746335369,
+      "learning_rate": 5.554890300635992e-07,
+      "loss": 1.602,
+      "step": 4563
+    },
+    {
+      "epoch": 0.31801553844545866,
+      "grad_norm": 0.7235496817025429,
+      "learning_rate": 5.554310420250091e-07,
+      "loss": 1.5873,
+      "step": 4564
+    },
+    {
+      "epoch": 0.318085217573076,
+      "grad_norm": 0.7145943660212166,
+      "learning_rate": 5.553730458190136e-07,
+      "loss": 1.5606,
+      "step": 4565
+    },
+    {
+      "epoch": 0.3181548967006933,
+      "grad_norm": 0.7405716180465298,
+      "learning_rate": 5.55315041448392e-07,
+      "loss": 1.5548,
+      "step": 4566
+    },
+    {
+      "epoch": 0.31822457582831065,
+      "grad_norm": 0.699821597252651,
+      "learning_rate": 5.55257028915924e-07,
+      "loss": 1.5007,
+      "step": 4567
+    },
+    {
+      "epoch": 0.31829425495592795,
+      "grad_norm": 0.7264404039646922,
+      "learning_rate": 5.551990082243896e-07,
+      "loss": 1.5778,
+      "step": 4568
+    },
+    {
+      "epoch": 0.3183639340835453,
+      "grad_norm": 0.7018136293753561,
+      "learning_rate": 5.551409793765692e-07,
+      "loss": 1.5822,
+      "step": 4569
+    },
+    {
+      "epoch": 0.3184336132111626,
+      "grad_norm": 0.7085191412090573,
+      "learning_rate": 5.55082942375244e-07,
+      "loss": 1.566,
+      "step": 4570
+    },
+    {
+      "epoch": 0.31850329233877994,
+      "grad_norm": 0.6959965920230744,
+      "learning_rate": 5.550248972231949e-07,
+      "loss": 1.424,
+      "step": 4571
+    },
+    {
+      "epoch": 0.31857297146639724,
+      "grad_norm": 0.7573753306961281,
+      "learning_rate": 5.549668439232036e-07,
+      "loss": 1.7044,
+      "step": 4572
+    },
+    {
+      "epoch": 0.3186426505940146,
+      "grad_norm": 0.7796455158742204,
+      "learning_rate": 5.549087824780523e-07,
+      "loss": 1.4707,
+      "step": 4573
+    },
+    {
+      "epoch": 0.3187123297216319,
+      "grad_norm": 0.8073763532584914,
+      "learning_rate": 5.548507128905233e-07,
+      "loss": 1.6353,
+      "step": 4574
+    },
+    {
+      "epoch": 0.31878200884924923,
+      "grad_norm": 0.6759903199641083,
+      "learning_rate": 5.547926351633995e-07,
+      "loss": 1.4381,
+      "step": 4575
+    },
+    {
+      "epoch": 0.3188516879768665,
+      "grad_norm": 0.8112029098972438,
+      "learning_rate": 5.54734549299464e-07,
+      "loss": 1.4541,
+      "step": 4576
+    },
+    {
+      "epoch": 0.3189213671044839,
+      "grad_norm": 0.668927646315414,
+      "learning_rate": 5.546764553015004e-07,
+      "loss": 1.5504,
+      "step": 4577
+    },
+    {
+      "epoch": 0.31899104623210117,
+      "grad_norm": 0.728448386639013,
+      "learning_rate": 5.546183531722927e-07,
+      "loss": 1.6189,
+      "step": 4578
+    },
+    {
+      "epoch": 0.3190607253597185,
+      "grad_norm": 0.7022708932585651,
+      "learning_rate": 5.545602429146254e-07,
+      "loss": 1.4315,
+      "step": 4579
+    },
+    {
+      "epoch": 0.3191304044873358,
+      "grad_norm": 0.7663878863520591,
+      "learning_rate": 5.54502124531283e-07,
+      "loss": 1.5433,
+      "step": 4580
+    },
+    {
+      "epoch": 0.31920008361495317,
+      "grad_norm": 0.7464042904379575,
+      "learning_rate": 5.544439980250511e-07,
+      "loss": 1.7276,
+      "step": 4581
+    },
+    {
+      "epoch": 0.31926976274257046,
+      "grad_norm": 0.761958715868217,
+      "learning_rate": 5.543858633987147e-07,
+      "loss": 1.4849,
+      "step": 4582
+    },
+    {
+      "epoch": 0.3193394418701878,
+      "grad_norm": 0.7402266957550955,
+      "learning_rate": 5.5432772065506e-07,
+      "loss": 1.619,
+      "step": 4583
+    },
+    {
+      "epoch": 0.3194091209978051,
+      "grad_norm": 0.676648377109348,
+      "learning_rate": 5.542695697968735e-07,
+      "loss": 1.5384,
+      "step": 4584
+    },
+    {
+      "epoch": 0.31947880012542246,
+      "grad_norm": 0.7194002923819981,
+      "learning_rate": 5.542114108269416e-07,
+      "loss": 1.499,
+      "step": 4585
+    },
+    {
+      "epoch": 0.31954847925303975,
+      "grad_norm": 0.7094758253955392,
+      "learning_rate": 5.541532437480514e-07,
+      "loss": 1.5803,
+      "step": 4586
+    },
+    {
+      "epoch": 0.31961815838065705,
+      "grad_norm": 0.6828922434293256,
+      "learning_rate": 5.540950685629905e-07,
+      "loss": 1.4883,
+      "step": 4587
+    },
+    {
+      "epoch": 0.3196878375082744,
+      "grad_norm": 0.7289414195466307,
+      "learning_rate": 5.540368852745469e-07,
+      "loss": 1.6278,
+      "step": 4588
+    },
+    {
+      "epoch": 0.3197575166358917,
+      "grad_norm": 0.7408149230732806,
+      "learning_rate": 5.539786938855087e-07,
+      "loss": 1.5954,
+      "step": 4589
+    },
+    {
+      "epoch": 0.31982719576350904,
+      "grad_norm": 0.6675256658590901,
+      "learning_rate": 5.539204943986645e-07,
+      "loss": 1.517,
+      "step": 4590
+    },
+    {
+      "epoch": 0.31989687489112634,
+      "grad_norm": 0.6800644937580433,
+      "learning_rate": 5.538622868168034e-07,
+      "loss": 1.5081,
+      "step": 4591
+    },
+    {
+      "epoch": 0.3199665540187437,
+      "grad_norm": 0.729421915525262,
+      "learning_rate": 5.53804071142715e-07,
+      "loss": 1.609,
+      "step": 4592
+    },
+    {
+      "epoch": 0.320036233146361,
+      "grad_norm": 0.6810241293051824,
+      "learning_rate": 5.537458473791889e-07,
+      "loss": 1.4824,
+      "step": 4593
+    },
+    {
+      "epoch": 0.32010591227397833,
+      "grad_norm": 0.6577963086195658,
+      "learning_rate": 5.536876155290153e-07,
+      "loss": 1.4177,
+      "step": 4594
+    },
+    {
+      "epoch": 0.3201755914015956,
+      "grad_norm": 0.7461821891392836,
+      "learning_rate": 5.53629375594985e-07,
+      "loss": 1.5158,
+      "step": 4595
+    },
+    {
+      "epoch": 0.320245270529213,
+      "grad_norm": 0.7391443519888712,
+      "learning_rate": 5.535711275798887e-07,
+      "loss": 1.4028,
+      "step": 4596
+    },
+    {
+      "epoch": 0.3203149496568303,
+      "grad_norm": 0.7217676239781258,
+      "learning_rate": 5.53512871486518e-07,
+      "loss": 1.6136,
+      "step": 4597
+    },
+    {
+      "epoch": 0.3203846287844476,
+      "grad_norm": 0.7728777338722576,
+      "learning_rate": 5.534546073176645e-07,
+      "loss": 1.4466,
+      "step": 4598
+    },
+    {
+      "epoch": 0.3204543079120649,
+      "grad_norm": 0.6624023550793688,
+      "learning_rate": 5.533963350761203e-07,
+      "loss": 1.5125,
+      "step": 4599
+    },
+    {
+      "epoch": 0.32052398703968227,
+      "grad_norm": 0.7744005505294767,
+      "learning_rate": 5.533380547646781e-07,
+      "loss": 1.4948,
+      "step": 4600
+    },
+    {
+      "epoch": 0.32059366616729956,
+      "grad_norm": 0.6691477498617161,
+      "learning_rate": 5.532797663861307e-07,
+      "loss": 1.5916,
+      "step": 4601
+    },
+    {
+      "epoch": 0.3206633452949169,
+      "grad_norm": 0.6918371725583984,
+      "learning_rate": 5.532214699432715e-07,
+      "loss": 1.5005,
+      "step": 4602
+    },
+    {
+      "epoch": 0.3207330244225342,
+      "grad_norm": 0.7497768594055559,
+      "learning_rate": 5.53163165438894e-07,
+      "loss": 1.4548,
+      "step": 4603
+    },
+    {
+      "epoch": 0.32080270355015156,
+      "grad_norm": 0.7488556365884924,
+      "learning_rate": 5.531048528757924e-07,
+      "loss": 1.5472,
+      "step": 4604
+    },
+    {
+      "epoch": 0.32087238267776885,
+      "grad_norm": 0.7579239463318133,
+      "learning_rate": 5.530465322567612e-07,
+      "loss": 1.5825,
+      "step": 4605
+    },
+    {
+      "epoch": 0.3209420618053862,
+      "grad_norm": 0.7919011629759329,
+      "learning_rate": 5.529882035845952e-07,
+      "loss": 1.5038,
+      "step": 4606
+    },
+    {
+      "epoch": 0.3210117409330035,
+      "grad_norm": 0.7674975520767843,
+      "learning_rate": 5.529298668620894e-07,
+      "loss": 1.6024,
+      "step": 4607
+    },
+    {
+      "epoch": 0.32108142006062085,
+      "grad_norm": 0.7096403207271622,
+      "learning_rate": 5.528715220920397e-07,
+      "loss": 1.5253,
+      "step": 4608
+    },
+    {
+      "epoch": 0.32115109918823814,
+      "grad_norm": 0.7134317507024582,
+      "learning_rate": 5.528131692772423e-07,
+      "loss": 1.6089,
+      "step": 4609
+    },
+    {
+      "epoch": 0.3212207783158555,
+      "grad_norm": 0.7671723571840533,
+      "learning_rate": 5.52754808420493e-07,
+      "loss": 1.6016,
+      "step": 4610
+    },
+    {
+      "epoch": 0.3212904574434728,
+      "grad_norm": 0.688495820899462,
+      "learning_rate": 5.52696439524589e-07,
+      "loss": 1.5726,
+      "step": 4611
+    },
+    {
+      "epoch": 0.32136013657109014,
+      "grad_norm": 0.6678380593790649,
+      "learning_rate": 5.526380625923274e-07,
+      "loss": 1.5112,
+      "step": 4612
+    },
+    {
+      "epoch": 0.32142981569870743,
+      "grad_norm": 0.745762956582063,
+      "learning_rate": 5.525796776265057e-07,
+      "loss": 1.5177,
+      "step": 4613
+    },
+    {
+      "epoch": 0.3214994948263248,
+      "grad_norm": 0.7112486718285385,
+      "learning_rate": 5.525212846299217e-07,
+      "loss": 1.5147,
+      "step": 4614
+    },
+    {
+      "epoch": 0.3215691739539421,
+      "grad_norm": 0.724836975919717,
+      "learning_rate": 5.524628836053739e-07,
+      "loss": 1.6959,
+      "step": 4615
+    },
+    {
+      "epoch": 0.32163885308155943,
+      "grad_norm": 0.6913389692644148,
+      "learning_rate": 5.524044745556608e-07,
+      "loss": 1.5525,
+      "step": 4616
+    },
+    {
+      "epoch": 0.3217085322091767,
+      "grad_norm": 0.7223242124116028,
+      "learning_rate": 5.523460574835818e-07,
+      "loss": 1.6182,
+      "step": 4617
+    },
+    {
+      "epoch": 0.3217782113367941,
+      "grad_norm": 0.6972269063425547,
+      "learning_rate": 5.52287632391936e-07,
+      "loss": 1.614,
+      "step": 4618
+    },
+    {
+      "epoch": 0.32184789046441137,
+      "grad_norm": 0.6855602585457429,
+      "learning_rate": 5.522291992835234e-07,
+      "loss": 1.5322,
+      "step": 4619
+    },
+    {
+      "epoch": 0.3219175695920287,
+      "grad_norm": 0.7295488072661531,
+      "learning_rate": 5.521707581611445e-07,
+      "loss": 1.4599,
+      "step": 4620
+    },
+    {
+      "epoch": 0.321987248719646,
+      "grad_norm": 0.7602169389635018,
+      "learning_rate": 5.521123090275996e-07,
+      "loss": 1.4392,
+      "step": 4621
+    },
+    {
+      "epoch": 0.32205692784726336,
+      "grad_norm": 0.7611538203907603,
+      "learning_rate": 5.520538518856896e-07,
+      "loss": 1.6624,
+      "step": 4622
+    },
+    {
+      "epoch": 0.32212660697488066,
+      "grad_norm": 0.7382468747722052,
+      "learning_rate": 5.519953867382163e-07,
+      "loss": 1.5633,
+      "step": 4623
+    },
+    {
+      "epoch": 0.322196286102498,
+      "grad_norm": 0.6877266006916888,
+      "learning_rate": 5.51936913587981e-07,
+      "loss": 1.4824,
+      "step": 4624
+    },
+    {
+      "epoch": 0.3222659652301153,
+      "grad_norm": 0.7238249354166105,
+      "learning_rate": 5.518784324377861e-07,
+      "loss": 1.6241,
+      "step": 4625
+    },
+    {
+      "epoch": 0.32233564435773265,
+      "grad_norm": 0.7876389933251687,
+      "learning_rate": 5.518199432904342e-07,
+      "loss": 1.6126,
+      "step": 4626
+    },
+    {
+      "epoch": 0.32240532348534995,
+      "grad_norm": 0.7244708081210876,
+      "learning_rate": 5.517614461487283e-07,
+      "loss": 1.57,
+      "step": 4627
+    },
+    {
+      "epoch": 0.3224750026129673,
+      "grad_norm": 0.6877563899579893,
+      "learning_rate": 5.517029410154713e-07,
+      "loss": 1.4658,
+      "step": 4628
+    },
+    {
+      "epoch": 0.3225446817405846,
+      "grad_norm": 0.7381068738510921,
+      "learning_rate": 5.516444278934672e-07,
+      "loss": 1.472,
+      "step": 4629
+    },
+    {
+      "epoch": 0.32261436086820194,
+      "grad_norm": 0.7661324152373148,
+      "learning_rate": 5.5158590678552e-07,
+      "loss": 1.4738,
+      "step": 4630
+    },
+    {
+      "epoch": 0.32268403999581924,
+      "grad_norm": 0.7180663750359986,
+      "learning_rate": 5.515273776944343e-07,
+      "loss": 1.4249,
+      "step": 4631
+    },
+    {
+      "epoch": 0.3227537191234366,
+      "grad_norm": 0.7314590378305785,
+      "learning_rate": 5.514688406230145e-07,
+      "loss": 1.5152,
+      "step": 4632
+    },
+    {
+      "epoch": 0.3228233982510539,
+      "grad_norm": 0.6801266282448725,
+      "learning_rate": 5.514102955740663e-07,
+      "loss": 1.5268,
+      "step": 4633
+    },
+    {
+      "epoch": 0.32289307737867123,
+      "grad_norm": 0.7305814534042752,
+      "learning_rate": 5.51351742550395e-07,
+      "loss": 1.495,
+      "step": 4634
+    },
+    {
+      "epoch": 0.32296275650628853,
+      "grad_norm": 0.6815117848190505,
+      "learning_rate": 5.512931815548069e-07,
+      "loss": 1.4861,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3230324356339059,
+      "grad_norm": 0.7455133929413991,
+      "learning_rate": 5.512346125901079e-07,
+      "loss": 1.4462,
+      "step": 4636
+    },
+    {
+      "epoch": 0.3231021147615232,
+      "grad_norm": 0.7508120732279528,
+      "learning_rate": 5.511760356591052e-07,
+      "loss": 1.5483,
+      "step": 4637
+    },
+    {
+      "epoch": 0.3231717938891405,
+      "grad_norm": 0.7014812959074661,
+      "learning_rate": 5.511174507646055e-07,
+      "loss": 1.609,
+      "step": 4638
+    },
+    {
+      "epoch": 0.3232414730167578,
+      "grad_norm": 0.7421777110315984,
+      "learning_rate": 5.510588579094168e-07,
+      "loss": 1.4491,
+      "step": 4639
+    },
+    {
+      "epoch": 0.32331115214437517,
+      "grad_norm": 0.7827639342952755,
+      "learning_rate": 5.510002570963465e-07,
+      "loss": 1.4477,
+      "step": 4640
+    },
+    {
+      "epoch": 0.32338083127199246,
+      "grad_norm": 0.6862468086728646,
+      "learning_rate": 5.50941648328203e-07,
+      "loss": 1.5549,
+      "step": 4641
+    },
+    {
+      "epoch": 0.3234505103996098,
+      "grad_norm": 0.7260160638503516,
+      "learning_rate": 5.508830316077952e-07,
+      "loss": 1.4935,
+      "step": 4642
+    },
+    {
+      "epoch": 0.3235201895272271,
+      "grad_norm": 0.6932870393561489,
+      "learning_rate": 5.508244069379321e-07,
+      "loss": 1.7095,
+      "step": 4643
+    },
+    {
+      "epoch": 0.32358986865484446,
+      "grad_norm": 0.7465469881600842,
+      "learning_rate": 5.507657743214228e-07,
+      "loss": 1.5113,
+      "step": 4644
+    },
+    {
+      "epoch": 0.32365954778246175,
+      "grad_norm": 0.724992133322777,
+      "learning_rate": 5.507071337610773e-07,
+      "loss": 1.6181,
+      "step": 4645
+    },
+    {
+      "epoch": 0.3237292269100791,
+      "grad_norm": 0.746052557847551,
+      "learning_rate": 5.506484852597058e-07,
+      "loss": 1.6747,
+      "step": 4646
+    },
+    {
+      "epoch": 0.3237989060376964,
+      "grad_norm": 0.7450971054334902,
+      "learning_rate": 5.505898288201188e-07,
+      "loss": 1.6415,
+      "step": 4647
+    },
+    {
+      "epoch": 0.32386858516531375,
+      "grad_norm": 0.7264260556915378,
+      "learning_rate": 5.505311644451272e-07,
+      "loss": 1.6594,
+      "step": 4648
+    },
+    {
+      "epoch": 0.32393826429293104,
+      "grad_norm": 0.7155597727480145,
+      "learning_rate": 5.504724921375425e-07,
+      "loss": 1.4577,
+      "step": 4649
+    },
+    {
+      "epoch": 0.3240079434205484,
+      "grad_norm": 0.7669193360173628,
+      "learning_rate": 5.504138119001761e-07,
+      "loss": 1.5443,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3240776225481657,
+      "grad_norm": 0.8326790168924357,
+      "learning_rate": 5.503551237358404e-07,
+      "loss": 1.7411,
+      "step": 4651
+    },
+    {
+      "epoch": 0.32414730167578304,
+      "grad_norm": 0.7134212535930974,
+      "learning_rate": 5.502964276473477e-07,
+      "loss": 1.5207,
+      "step": 4652
+    },
+    {
+      "epoch": 0.32421698080340033,
+      "grad_norm": 0.7164271081334321,
+      "learning_rate": 5.502377236375108e-07,
+      "loss": 1.563,
+      "step": 4653
+    },
+    {
+      "epoch": 0.3242866599310177,
+      "grad_norm": 0.7574513017908499,
+      "learning_rate": 5.501790117091429e-07,
+      "loss": 1.5626,
+      "step": 4654
+    },
+    {
+      "epoch": 0.324356339058635,
+      "grad_norm": 0.6950388343278109,
+      "learning_rate": 5.501202918650577e-07,
+      "loss": 1.5222,
+      "step": 4655
+    },
+    {
+      "epoch": 0.32442601818625233,
+      "grad_norm": 0.7125678417544166,
+      "learning_rate": 5.500615641080691e-07,
+      "loss": 1.6304,
+      "step": 4656
+    },
+    {
+      "epoch": 0.3244956973138696,
+      "grad_norm": 0.6998624739958945,
+      "learning_rate": 5.500028284409915e-07,
+      "loss": 1.5324,
+      "step": 4657
+    },
+    {
+      "epoch": 0.324565376441487,
+      "grad_norm": 0.7068349491161948,
+      "learning_rate": 5.499440848666395e-07,
+      "loss": 1.5863,
+      "step": 4658
+    },
+    {
+      "epoch": 0.32463505556910427,
+      "grad_norm": 0.661201049428343,
+      "learning_rate": 5.498853333878285e-07,
+      "loss": 1.4551,
+      "step": 4659
+    },
+    {
+      "epoch": 0.3247047346967216,
+      "grad_norm": 0.7269218043387352,
+      "learning_rate": 5.498265740073738e-07,
+      "loss": 1.5651,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3247744138243389,
+      "grad_norm": 0.749285337503175,
+      "learning_rate": 5.497678067280913e-07,
+      "loss": 1.7859,
+      "step": 4661
+    },
+    {
+      "epoch": 0.32484409295195626,
+      "grad_norm": 0.7480084626513137,
+      "learning_rate": 5.497090315527971e-07,
+      "loss": 1.5376,
+      "step": 4662
+    },
+    {
+      "epoch": 0.32491377207957356,
+      "grad_norm": 0.6844373648201154,
+      "learning_rate": 5.496502484843082e-07,
+      "loss": 1.4983,
+      "step": 4663
+    },
+    {
+      "epoch": 0.3249834512071909,
+      "grad_norm": 0.8226873965644259,
+      "learning_rate": 5.495914575254411e-07,
+      "loss": 1.7003,
+      "step": 4664
+    },
+    {
+      "epoch": 0.3250531303348082,
+      "grad_norm": 0.6728335863273786,
+      "learning_rate": 5.495326586790137e-07,
+      "loss": 1.5181,
+      "step": 4665
+    },
+    {
+      "epoch": 0.32512280946242555,
+      "grad_norm": 0.7469348771533575,
+      "learning_rate": 5.494738519478434e-07,
+      "loss": 1.6188,
+      "step": 4666
+    },
+    {
+      "epoch": 0.32519248859004285,
+      "grad_norm": 0.7149477160520245,
+      "learning_rate": 5.494150373347485e-07,
+      "loss": 1.5103,
+      "step": 4667
+    },
+    {
+      "epoch": 0.3252621677176602,
+      "grad_norm": 0.7543294043052631,
+      "learning_rate": 5.493562148425475e-07,
+      "loss": 1.531,
+      "step": 4668
+    },
+    {
+      "epoch": 0.3253318468452775,
+      "grad_norm": 0.8562125075394144,
+      "learning_rate": 5.492973844740592e-07,
+      "loss": 1.5215,
+      "step": 4669
+    },
+    {
+      "epoch": 0.32540152597289485,
+      "grad_norm": 0.7576430297954465,
+      "learning_rate": 5.492385462321028e-07,
+      "loss": 1.4913,
+      "step": 4670
+    },
+    {
+      "epoch": 0.32547120510051214,
+      "grad_norm": 0.693823950005331,
+      "learning_rate": 5.491797001194984e-07,
+      "loss": 1.5224,
+      "step": 4671
+    },
+    {
+      "epoch": 0.3255408842281295,
+      "grad_norm": 0.7327177721134271,
+      "learning_rate": 5.491208461390654e-07,
+      "loss": 1.4314,
+      "step": 4672
+    },
+    {
+      "epoch": 0.3256105633557468,
+      "grad_norm": 0.7566433006080883,
+      "learning_rate": 5.490619842936248e-07,
+      "loss": 1.6511,
+      "step": 4673
+    },
+    {
+      "epoch": 0.32568024248336414,
+      "grad_norm": 0.7885667972779885,
+      "learning_rate": 5.490031145859969e-07,
+      "loss": 1.7014,
+      "step": 4674
+    },
+    {
+      "epoch": 0.32574992161098143,
+      "grad_norm": 0.7768801147957153,
+      "learning_rate": 5.489442370190032e-07,
+      "loss": 1.5669,
+      "step": 4675
+    },
+    {
+      "epoch": 0.3258196007385988,
+      "grad_norm": 0.720878744425241,
+      "learning_rate": 5.488853515954651e-07,
+      "loss": 1.5863,
+      "step": 4676
+    },
+    {
+      "epoch": 0.3258892798662161,
+      "grad_norm": 0.6862498975781554,
+      "learning_rate": 5.488264583182043e-07,
+      "loss": 1.6059,
+      "step": 4677
+    },
+    {
+      "epoch": 0.32595895899383337,
+      "grad_norm": 0.7409461109224728,
+      "learning_rate": 5.487675571900435e-07,
+      "loss": 1.5253,
+      "step": 4678
+    },
+    {
+      "epoch": 0.3260286381214507,
+      "grad_norm": 0.7660810857821033,
+      "learning_rate": 5.48708648213805e-07,
+      "loss": 1.5677,
+      "step": 4679
+    },
+    {
+      "epoch": 0.326098317249068,
+      "grad_norm": 0.709707457530693,
+      "learning_rate": 5.486497313923121e-07,
+      "loss": 1.4906,
+      "step": 4680
+    },
+    {
+      "epoch": 0.32616799637668537,
+      "grad_norm": 0.7039157582721269,
+      "learning_rate": 5.48590806728388e-07,
+      "loss": 1.642,
+      "step": 4681
+    },
+    {
+      "epoch": 0.32623767550430266,
+      "grad_norm": 0.8055219249645671,
+      "learning_rate": 5.485318742248567e-07,
+      "loss": 1.5276,
+      "step": 4682
+    },
+    {
+      "epoch": 0.32630735463192,
+      "grad_norm": 0.7082044224744788,
+      "learning_rate": 5.484729338845422e-07,
+      "loss": 1.5328,
+      "step": 4683
+    },
+    {
+      "epoch": 0.3263770337595373,
+      "grad_norm": 0.7195194869776044,
+      "learning_rate": 5.484139857102691e-07,
+      "loss": 1.564,
+      "step": 4684
+    },
+    {
+      "epoch": 0.32644671288715466,
+      "grad_norm": 0.7260360051203139,
+      "learning_rate": 5.483550297048624e-07,
+      "loss": 1.5654,
+      "step": 4685
+    },
+    {
+      "epoch": 0.32651639201477195,
+      "grad_norm": 0.7095863416019655,
+      "learning_rate": 5.482960658711472e-07,
+      "loss": 1.6385,
+      "step": 4686
+    },
+    {
+      "epoch": 0.3265860711423893,
+      "grad_norm": 0.7089680351775876,
+      "learning_rate": 5.482370942119494e-07,
+      "loss": 1.5918,
+      "step": 4687
+    },
+    {
+      "epoch": 0.3266557502700066,
+      "grad_norm": 0.7754253458635536,
+      "learning_rate": 5.481781147300948e-07,
+      "loss": 1.5457,
+      "step": 4688
+    },
+    {
+      "epoch": 0.32672542939762395,
+      "grad_norm": 0.7133933783068345,
+      "learning_rate": 5.481191274284101e-07,
+      "loss": 1.4844,
+      "step": 4689
+    },
+    {
+      "epoch": 0.32679510852524124,
+      "grad_norm": 0.7290333114124878,
+      "learning_rate": 5.480601323097218e-07,
+      "loss": 1.6388,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3268647876528586,
+      "grad_norm": 0.7342510780107464,
+      "learning_rate": 5.480011293768572e-07,
+      "loss": 1.4667,
+      "step": 4691
+    },
+    {
+      "epoch": 0.3269344667804759,
+      "grad_norm": 0.691124208728813,
+      "learning_rate": 5.479421186326439e-07,
+      "loss": 1.4353,
+      "step": 4692
+    },
+    {
+      "epoch": 0.32700414590809324,
+      "grad_norm": 0.7409084935108537,
+      "learning_rate": 5.478831000799098e-07,
+      "loss": 1.4188,
+      "step": 4693
+    },
+    {
+      "epoch": 0.32707382503571053,
+      "grad_norm": 0.7581250592188127,
+      "learning_rate": 5.478240737214831e-07,
+      "loss": 1.4854,
+      "step": 4694
+    },
+    {
+      "epoch": 0.3271435041633279,
+      "grad_norm": 0.6916383712026993,
+      "learning_rate": 5.477650395601926e-07,
+      "loss": 1.6155,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3272131832909452,
+      "grad_norm": 0.7525082735725286,
+      "learning_rate": 5.477059975988671e-07,
+      "loss": 1.6562,
+      "step": 4696
+    },
+    {
+      "epoch": 0.3272828624185625,
+      "grad_norm": 0.6850325687097222,
+      "learning_rate": 5.476469478403363e-07,
+      "loss": 1.5095,
+      "step": 4697
+    },
+    {
+      "epoch": 0.3273525415461798,
+      "grad_norm": 0.6911370545439388,
+      "learning_rate": 5.475878902874298e-07,
+      "loss": 1.5014,
+      "step": 4698
+    },
+    {
+      "epoch": 0.32742222067379717,
+      "grad_norm": 0.6779251995747619,
+      "learning_rate": 5.475288249429777e-07,
+      "loss": 1.521,
+      "step": 4699
+    },
+    {
+      "epoch": 0.32749189980141447,
+      "grad_norm": 0.6777351655241748,
+      "learning_rate": 5.474697518098108e-07,
+      "loss": 1.5156,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3275615789290318,
+      "grad_norm": 0.6809433325320242,
+      "learning_rate": 5.4741067089076e-07,
+      "loss": 1.5013,
+      "step": 4701
+    },
+    {
+      "epoch": 0.3276312580566491,
+      "grad_norm": 0.7119144281537158,
+      "learning_rate": 5.47351582188656e-07,
+      "loss": 1.5415,
+      "step": 4702
+    },
+    {
+      "epoch": 0.32770093718426646,
+      "grad_norm": 0.6613717045837818,
+      "learning_rate": 5.472924857063311e-07,
+      "loss": 1.4926,
+      "step": 4703
+    },
+    {
+      "epoch": 0.32777061631188376,
+      "grad_norm": 0.7288042034986972,
+      "learning_rate": 5.472333814466173e-07,
+      "loss": 1.4969,
+      "step": 4704
+    },
+    {
+      "epoch": 0.3278402954395011,
+      "grad_norm": 0.6995801218556058,
+      "learning_rate": 5.471742694123465e-07,
+      "loss": 1.4725,
+      "step": 4705
+    },
+    {
+      "epoch": 0.3279099745671184,
+      "grad_norm": 0.6891604297815477,
+      "learning_rate": 5.471151496063519e-07,
+      "loss": 1.4852,
+      "step": 4706
+    },
+    {
+      "epoch": 0.32797965369473575,
+      "grad_norm": 0.7077304381998237,
+      "learning_rate": 5.470560220314666e-07,
+      "loss": 1.5236,
+      "step": 4707
+    },
+    {
+      "epoch": 0.32804933282235305,
+      "grad_norm": 0.6457256925404681,
+      "learning_rate": 5.469968866905239e-07,
+      "loss": 1.5301,
+      "step": 4708
+    },
+    {
+      "epoch": 0.3281190119499704,
+      "grad_norm": 0.7383356695325461,
+      "learning_rate": 5.469377435863577e-07,
+      "loss": 1.5252,
+      "step": 4709
+    },
+    {
+      "epoch": 0.3281886910775877,
+      "grad_norm": 0.7226731345944863,
+      "learning_rate": 5.468785927218026e-07,
+      "loss": 1.5984,
+      "step": 4710
+    },
+    {
+      "epoch": 0.32825837020520504,
+      "grad_norm": 0.7351157644434121,
+      "learning_rate": 5.468194340996929e-07,
+      "loss": 1.662,
+      "step": 4711
+    },
+    {
+      "epoch": 0.32832804933282234,
+      "grad_norm": 0.6957818071174824,
+      "learning_rate": 5.467602677228638e-07,
+      "loss": 1.4365,
+      "step": 4712
+    },
+    {
+      "epoch": 0.3283977284604397,
+      "grad_norm": 0.7196864910738779,
+      "learning_rate": 5.467010935941507e-07,
+      "loss": 1.5894,
+      "step": 4713
+    },
+    {
+      "epoch": 0.328467407588057,
+      "grad_norm": 0.7146083912133652,
+      "learning_rate": 5.466419117163889e-07,
+      "loss": 1.4484,
+      "step": 4714
+    },
+    {
+      "epoch": 0.32853708671567433,
+      "grad_norm": 0.8106128842589259,
+      "learning_rate": 5.465827220924151e-07,
+      "loss": 1.4741,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3286067658432916,
+      "grad_norm": 0.7103171764326403,
+      "learning_rate": 5.465235247250653e-07,
+      "loss": 1.508,
+      "step": 4716
+    },
+    {
+      "epoch": 0.328676444970909,
+      "grad_norm": 0.6812060402187873,
+      "learning_rate": 5.464643196171767e-07,
+      "loss": 1.4331,
+      "step": 4717
+    },
+    {
+      "epoch": 0.32874612409852627,
+      "grad_norm": 0.7177732946700219,
+      "learning_rate": 5.464051067715865e-07,
+      "loss": 1.4702,
+      "step": 4718
+    },
+    {
+      "epoch": 0.3288158032261436,
+      "grad_norm": 0.6958049083294975,
+      "learning_rate": 5.463458861911322e-07,
+      "loss": 1.4755,
+      "step": 4719
+    },
+    {
+      "epoch": 0.3288854823537609,
+      "grad_norm": 0.7325285859162157,
+      "learning_rate": 5.462866578786518e-07,
+      "loss": 1.4601,
+      "step": 4720
+    },
+    {
+      "epoch": 0.32895516148137827,
+      "grad_norm": 0.7582230823199709,
+      "learning_rate": 5.462274218369836e-07,
+      "loss": 1.4437,
+      "step": 4721
+    },
+    {
+      "epoch": 0.32902484060899556,
+      "grad_norm": 0.7210199222494685,
+      "learning_rate": 5.461681780689663e-07,
+      "loss": 1.5113,
+      "step": 4722
+    },
+    {
+      "epoch": 0.3290945197366129,
+      "grad_norm": 0.6795180257214272,
+      "learning_rate": 5.461089265774391e-07,
+      "loss": 1.4581,
+      "step": 4723
+    },
+    {
+      "epoch": 0.3291641988642302,
+      "grad_norm": 0.7355181900180564,
+      "learning_rate": 5.460496673652414e-07,
+      "loss": 1.5793,
+      "step": 4724
+    },
+    {
+      "epoch": 0.32923387799184756,
+      "grad_norm": 0.7418505821521018,
+      "learning_rate": 5.45990400435213e-07,
+      "loss": 1.5742,
+      "step": 4725
+    },
+    {
+      "epoch": 0.32930355711946485,
+      "grad_norm": 0.8148378311162706,
+      "learning_rate": 5.459311257901941e-07,
+      "loss": 1.5995,
+      "step": 4726
+    },
+    {
+      "epoch": 0.3293732362470822,
+      "grad_norm": 0.9894964109825274,
+      "learning_rate": 5.458718434330252e-07,
+      "loss": 1.4232,
+      "step": 4727
+    },
+    {
+      "epoch": 0.3294429153746995,
+      "grad_norm": 0.701725083846248,
+      "learning_rate": 5.458125533665475e-07,
+      "loss": 1.5014,
+      "step": 4728
+    },
+    {
+      "epoch": 0.32951259450231685,
+      "grad_norm": 0.7596048302356802,
+      "learning_rate": 5.45753255593602e-07,
+      "loss": 1.5457,
+      "step": 4729
+    },
+    {
+      "epoch": 0.32958227362993414,
+      "grad_norm": 0.6908157411025175,
+      "learning_rate": 5.456939501170304e-07,
+      "loss": 1.5558,
+      "step": 4730
+    },
+    {
+      "epoch": 0.3296519527575515,
+      "grad_norm": 0.7496144153051155,
+      "learning_rate": 5.456346369396747e-07,
+      "loss": 1.578,
+      "step": 4731
+    },
+    {
+      "epoch": 0.3297216318851688,
+      "grad_norm": 0.7164436063279995,
+      "learning_rate": 5.455753160643777e-07,
+      "loss": 1.6629,
+      "step": 4732
+    },
+    {
+      "epoch": 0.32979131101278614,
+      "grad_norm": 0.7559462022136434,
+      "learning_rate": 5.455159874939819e-07,
+      "loss": 1.4736,
+      "step": 4733
+    },
+    {
+      "epoch": 0.32986099014040343,
+      "grad_norm": 0.7600862407544624,
+      "learning_rate": 5.454566512313302e-07,
+      "loss": 1.4862,
+      "step": 4734
+    },
+    {
+      "epoch": 0.3299306692680208,
+      "grad_norm": 0.7004081297809731,
+      "learning_rate": 5.453973072792665e-07,
+      "loss": 1.5029,
+      "step": 4735
+    },
+    {
+      "epoch": 0.3300003483956381,
+      "grad_norm": 0.72591322289454,
+      "learning_rate": 5.453379556406344e-07,
+      "loss": 1.5313,
+      "step": 4736
+    },
+    {
+      "epoch": 0.3300700275232554,
+      "grad_norm": 0.7659965365810644,
+      "learning_rate": 5.452785963182786e-07,
+      "loss": 1.5111,
+      "step": 4737
+    },
+    {
+      "epoch": 0.3301397066508727,
+      "grad_norm": 0.7075696002326214,
+      "learning_rate": 5.452192293150432e-07,
+      "loss": 1.6228,
+      "step": 4738
+    },
+    {
+      "epoch": 0.33020938577849007,
+      "grad_norm": 0.7345339015777442,
+      "learning_rate": 5.451598546337734e-07,
+      "loss": 1.6043,
+      "step": 4739
+    },
+    {
+      "epoch": 0.33027906490610737,
+      "grad_norm": 0.7963071700771978,
+      "learning_rate": 5.451004722773148e-07,
+      "loss": 1.6095,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3303487440337247,
+      "grad_norm": 0.7525640730979843,
+      "learning_rate": 5.450410822485126e-07,
+      "loss": 1.4938,
+      "step": 4741
+    },
+    {
+      "epoch": 0.330418423161342,
+      "grad_norm": 0.6891761133669421,
+      "learning_rate": 5.449816845502132e-07,
+      "loss": 1.4215,
+      "step": 4742
+    },
+    {
+      "epoch": 0.33048810228895936,
+      "grad_norm": 0.7469893291500803,
+      "learning_rate": 5.449222791852631e-07,
+      "loss": 1.5721,
+      "step": 4743
+    },
+    {
+      "epoch": 0.33055778141657666,
+      "grad_norm": 0.7269189267880387,
+      "learning_rate": 5.448628661565092e-07,
+      "loss": 1.5862,
+      "step": 4744
+    },
+    {
+      "epoch": 0.330627460544194,
+      "grad_norm": 0.7483789383148026,
+      "learning_rate": 5.448034454667984e-07,
+      "loss": 1.6084,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3306971396718113,
+      "grad_norm": 0.7177898667623517,
+      "learning_rate": 5.447440171189784e-07,
+      "loss": 1.5663,
+      "step": 4746
+    },
+    {
+      "epoch": 0.33076681879942865,
+      "grad_norm": 0.7687765020088722,
+      "learning_rate": 5.446845811158973e-07,
+      "loss": 1.5396,
+      "step": 4747
+    },
+    {
+      "epoch": 0.33083649792704595,
+      "grad_norm": 0.7445506289197298,
+      "learning_rate": 5.446251374604032e-07,
+      "loss": 1.6004,
+      "step": 4748
+    },
+    {
+      "epoch": 0.3309061770546633,
+      "grad_norm": 0.7628599743997581,
+      "learning_rate": 5.445656861553449e-07,
+      "loss": 1.5991,
+      "step": 4749
+    },
+    {
+      "epoch": 0.3309758561822806,
+      "grad_norm": 0.7690295768355484,
+      "learning_rate": 5.44506227203571e-07,
+      "loss": 1.608,
+      "step": 4750
+    },
+    {
+      "epoch": 0.33104553530989794,
+      "grad_norm": 0.7066359165688519,
+      "learning_rate": 5.444467606079316e-07,
+      "loss": 1.553,
+      "step": 4751
+    },
+    {
+      "epoch": 0.33111521443751524,
+      "grad_norm": 0.7107118246361575,
+      "learning_rate": 5.443872863712759e-07,
+      "loss": 1.445,
+      "step": 4752
+    },
+    {
+      "epoch": 0.3311848935651326,
+      "grad_norm": 0.7189520145837976,
+      "learning_rate": 5.443278044964542e-07,
+      "loss": 1.5686,
+      "step": 4753
+    },
+    {
+      "epoch": 0.3312545726927499,
+      "grad_norm": 0.6934783281831515,
+      "learning_rate": 5.442683149863171e-07,
+      "loss": 1.4495,
+      "step": 4754
+    },
+    {
+      "epoch": 0.33132425182036723,
+      "grad_norm": 0.7524447895980935,
+      "learning_rate": 5.442088178437154e-07,
+      "loss": 1.6211,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3313939309479845,
+      "grad_norm": 0.7002606597250953,
+      "learning_rate": 5.441493130715002e-07,
+      "loss": 1.5382,
+      "step": 4756
+    },
+    {
+      "epoch": 0.3314636100756019,
+      "grad_norm": 0.7542379695219257,
+      "learning_rate": 5.440898006725234e-07,
+      "loss": 1.6374,
+      "step": 4757
+    },
+    {
+      "epoch": 0.3315332892032192,
+      "grad_norm": 0.7312472818227834,
+      "learning_rate": 5.440302806496365e-07,
+      "loss": 1.4442,
+      "step": 4758
+    },
+    {
+      "epoch": 0.3316029683308365,
+      "grad_norm": 0.6765765400729704,
+      "learning_rate": 5.439707530056922e-07,
+      "loss": 1.5379,
+      "step": 4759
+    },
+    {
+      "epoch": 0.3316726474584538,
+      "grad_norm": 0.7145871013427026,
+      "learning_rate": 5.43911217743543e-07,
+      "loss": 1.4743,
+      "step": 4760
+    },
+    {
+      "epoch": 0.33174232658607117,
+      "grad_norm": 0.7534020290273412,
+      "learning_rate": 5.438516748660421e-07,
+      "loss": 1.5155,
+      "step": 4761
+    },
+    {
+      "epoch": 0.33181200571368846,
+      "grad_norm": 0.701107642439635,
+      "learning_rate": 5.437921243760427e-07,
+      "loss": 1.5799,
+      "step": 4762
+    },
+    {
+      "epoch": 0.3318816848413058,
+      "grad_norm": 0.8062590359838392,
+      "learning_rate": 5.437325662763987e-07,
+      "loss": 1.6094,
+      "step": 4763
+    },
+    {
+      "epoch": 0.3319513639689231,
+      "grad_norm": 0.7264396881824076,
+      "learning_rate": 5.436730005699644e-07,
+      "loss": 1.5333,
+      "step": 4764
+    },
+    {
+      "epoch": 0.33202104309654046,
+      "grad_norm": 0.7164683710272938,
+      "learning_rate": 5.436134272595941e-07,
+      "loss": 1.643,
+      "step": 4765
+    },
+    {
+      "epoch": 0.33209072222415775,
+      "grad_norm": 0.7681061882158766,
+      "learning_rate": 5.435538463481427e-07,
+      "loss": 1.7029,
+      "step": 4766
+    },
+    {
+      "epoch": 0.3321604013517751,
+      "grad_norm": 0.68673542098374,
+      "learning_rate": 5.434942578384654e-07,
+      "loss": 1.4273,
+      "step": 4767
+    },
+    {
+      "epoch": 0.3322300804793924,
+      "grad_norm": 0.7630700252413264,
+      "learning_rate": 5.43434661733418e-07,
+      "loss": 1.6314,
+      "step": 4768
+    },
+    {
+      "epoch": 0.3322997596070097,
+      "grad_norm": 0.7378555507316659,
+      "learning_rate": 5.433750580358563e-07,
+      "loss": 1.5883,
+      "step": 4769
+    },
+    {
+      "epoch": 0.33236943873462704,
+      "grad_norm": 0.7184196608437186,
+      "learning_rate": 5.433154467486367e-07,
+      "loss": 1.553,
+      "step": 4770
+    },
+    {
+      "epoch": 0.33243911786224434,
+      "grad_norm": 0.7610107203023418,
+      "learning_rate": 5.43255827874616e-07,
+      "loss": 1.5474,
+      "step": 4771
+    },
+    {
+      "epoch": 0.3325087969898617,
+      "grad_norm": 0.6578353625178328,
+      "learning_rate": 5.43196201416651e-07,
+      "loss": 1.5826,
+      "step": 4772
+    },
+    {
+      "epoch": 0.332578476117479,
+      "grad_norm": 0.7332684613958692,
+      "learning_rate": 5.43136567377599e-07,
+      "loss": 1.5332,
+      "step": 4773
+    },
+    {
+      "epoch": 0.33264815524509633,
+      "grad_norm": 0.7403565072313969,
+      "learning_rate": 5.430769257603185e-07,
+      "loss": 1.6311,
+      "step": 4774
+    },
+    {
+      "epoch": 0.33271783437271363,
+      "grad_norm": 0.7207657489954281,
+      "learning_rate": 5.43017276567667e-07,
+      "loss": 1.6177,
+      "step": 4775
+    },
+    {
+      "epoch": 0.332787513500331,
+      "grad_norm": 0.7303985385444817,
+      "learning_rate": 5.429576198025032e-07,
+      "loss": 1.5329,
+      "step": 4776
+    },
+    {
+      "epoch": 0.3328571926279483,
+      "grad_norm": 0.7035495502689031,
+      "learning_rate": 5.428979554676861e-07,
+      "loss": 1.4916,
+      "step": 4777
+    },
+    {
+      "epoch": 0.3329268717555656,
+      "grad_norm": 0.7003566219294264,
+      "learning_rate": 5.428382835660746e-07,
+      "loss": 1.5497,
+      "step": 4778
+    },
+    {
+      "epoch": 0.3329965508831829,
+      "grad_norm": 0.7612676386996001,
+      "learning_rate": 5.427786041005286e-07,
+      "loss": 1.5206,
+      "step": 4779
+    },
+    {
+      "epoch": 0.33306623001080027,
+      "grad_norm": 0.7094601748268057,
+      "learning_rate": 5.42718917073908e-07,
+      "loss": 1.5688,
+      "step": 4780
+    },
+    {
+      "epoch": 0.33313590913841756,
+      "grad_norm": 0.728996130018037,
+      "learning_rate": 5.426592224890731e-07,
+      "loss": 1.5695,
+      "step": 4781
+    },
+    {
+      "epoch": 0.3332055882660349,
+      "grad_norm": 0.6826206661442296,
+      "learning_rate": 5.425995203488846e-07,
+      "loss": 1.5038,
+      "step": 4782
+    },
+    {
+      "epoch": 0.3332752673936522,
+      "grad_norm": 0.6850069680455925,
+      "learning_rate": 5.425398106562034e-07,
+      "loss": 1.4939,
+      "step": 4783
+    },
+    {
+      "epoch": 0.33334494652126956,
+      "grad_norm": 0.7198182130651017,
+      "learning_rate": 5.424800934138913e-07,
+      "loss": 1.6911,
+      "step": 4784
+    },
+    {
+      "epoch": 0.33341462564888685,
+      "grad_norm": 0.7031776478007508,
+      "learning_rate": 5.424203686248098e-07,
+      "loss": 1.3148,
+      "step": 4785
+    },
+    {
+      "epoch": 0.3334843047765042,
+      "grad_norm": 0.7033765004420741,
+      "learning_rate": 5.423606362918209e-07,
+      "loss": 1.5843,
+      "step": 4786
+    },
+    {
+      "epoch": 0.3335539839041215,
+      "grad_norm": 0.7277452680681038,
+      "learning_rate": 5.423008964177873e-07,
+      "loss": 1.6939,
+      "step": 4787
+    },
+    {
+      "epoch": 0.33362366303173885,
+      "grad_norm": 0.721356540917203,
+      "learning_rate": 5.422411490055717e-07,
+      "loss": 1.5653,
+      "step": 4788
+    },
+    {
+      "epoch": 0.33369334215935614,
+      "grad_norm": 0.7036008560433303,
+      "learning_rate": 5.421813940580377e-07,
+      "loss": 1.6107,
+      "step": 4789
+    },
+    {
+      "epoch": 0.3337630212869735,
+      "grad_norm": 0.7458747498855014,
+      "learning_rate": 5.421216315780484e-07,
+      "loss": 1.6373,
+      "step": 4790
+    },
+    {
+      "epoch": 0.3338327004145908,
+      "grad_norm": 0.7173976160596333,
+      "learning_rate": 5.420618615684681e-07,
+      "loss": 1.5497,
+      "step": 4791
+    },
+    {
+      "epoch": 0.33390237954220814,
+      "grad_norm": 0.7257908124277479,
+      "learning_rate": 5.420020840321608e-07,
+      "loss": 1.6506,
+      "step": 4792
+    },
+    {
+      "epoch": 0.33397205866982543,
+      "grad_norm": 0.7276705606366335,
+      "learning_rate": 5.419422989719914e-07,
+      "loss": 1.5947,
+      "step": 4793
+    },
+    {
+      "epoch": 0.3340417377974428,
+      "grad_norm": 0.7124819335777425,
+      "learning_rate": 5.418825063908247e-07,
+      "loss": 1.5129,
+      "step": 4794
+    },
+    {
+      "epoch": 0.3341114169250601,
+      "grad_norm": 0.6973185583977857,
+      "learning_rate": 5.418227062915263e-07,
+      "loss": 1.4769,
+      "step": 4795
+    },
+    {
+      "epoch": 0.33418109605267743,
+      "grad_norm": 0.7171103000694339,
+      "learning_rate": 5.417628986769621e-07,
+      "loss": 1.528,
+      "step": 4796
+    },
+    {
+      "epoch": 0.3342507751802947,
+      "grad_norm": 0.7212360242606413,
+      "learning_rate": 5.417030835499978e-07,
+      "loss": 1.4938,
+      "step": 4797
+    },
+    {
+      "epoch": 0.3343204543079121,
+      "grad_norm": 0.7517464170152247,
+      "learning_rate": 5.416432609135e-07,
+      "loss": 1.609,
+      "step": 4798
+    },
+    {
+      "epoch": 0.33439013343552937,
+      "grad_norm": 0.8652168233618581,
+      "learning_rate": 5.415834307703356e-07,
+      "loss": 1.5062,
+      "step": 4799
+    },
+    {
+      "epoch": 0.3344598125631467,
+      "grad_norm": 0.7218739889842337,
+      "learning_rate": 5.415235931233716e-07,
+      "loss": 1.4711,
+      "step": 4800
+    },
+    {
+      "epoch": 0.334529491690764,
+      "grad_norm": 0.7094515975723757,
+      "learning_rate": 5.414637479754757e-07,
+      "loss": 1.5788,
+      "step": 4801
+    },
+    {
+      "epoch": 0.33459917081838136,
+      "grad_norm": 0.7393768229184177,
+      "learning_rate": 5.414038953295158e-07,
+      "loss": 1.5629,
+      "step": 4802
+    },
+    {
+      "epoch": 0.33466884994599866,
+      "grad_norm": 0.7302062209177451,
+      "learning_rate": 5.413440351883602e-07,
+      "loss": 1.6523,
+      "step": 4803
+    },
+    {
+      "epoch": 0.334738529073616,
+      "grad_norm": 0.6845338065322232,
+      "learning_rate": 5.412841675548776e-07,
+      "loss": 1.522,
+      "step": 4804
+    },
+    {
+      "epoch": 0.3348082082012333,
+      "grad_norm": 0.7235354663832575,
+      "learning_rate": 5.412242924319366e-07,
+      "loss": 1.5268,
+      "step": 4805
+    },
+    {
+      "epoch": 0.33487788732885065,
+      "grad_norm": 0.7613419830759305,
+      "learning_rate": 5.411644098224069e-07,
+      "loss": 1.4734,
+      "step": 4806
+    },
+    {
+      "epoch": 0.33494756645646795,
+      "grad_norm": 0.6741423643970372,
+      "learning_rate": 5.411045197291581e-07,
+      "loss": 1.5251,
+      "step": 4807
+    },
+    {
+      "epoch": 0.3350172455840853,
+      "grad_norm": 0.7105937235840101,
+      "learning_rate": 5.410446221550603e-07,
+      "loss": 1.5714,
+      "step": 4808
+    },
+    {
+      "epoch": 0.3350869247117026,
+      "grad_norm": 0.7391867622954359,
+      "learning_rate": 5.409847171029837e-07,
+      "loss": 1.5494,
+      "step": 4809
+    },
+    {
+      "epoch": 0.33515660383931994,
+      "grad_norm": 0.697675218773251,
+      "learning_rate": 5.409248045757993e-07,
+      "loss": 1.5154,
+      "step": 4810
+    },
+    {
+      "epoch": 0.33522628296693724,
+      "grad_norm": 0.6927014456831924,
+      "learning_rate": 5.408648845763781e-07,
+      "loss": 1.5613,
+      "step": 4811
+    },
+    {
+      "epoch": 0.3352959620945546,
+      "grad_norm": 0.7007622265131535,
+      "learning_rate": 5.408049571075917e-07,
+      "loss": 1.515,
+      "step": 4812
+    },
+    {
+      "epoch": 0.3353656412221719,
+      "grad_norm": 0.6629006984199853,
+      "learning_rate": 5.40745022172312e-07,
+      "loss": 1.4397,
+      "step": 4813
+    },
+    {
+      "epoch": 0.33543532034978923,
+      "grad_norm": 0.7762079077753187,
+      "learning_rate": 5.406850797734109e-07,
+      "loss": 1.7242,
+      "step": 4814
+    },
+    {
+      "epoch": 0.33550499947740653,
+      "grad_norm": 0.705869108994762,
+      "learning_rate": 5.406251299137613e-07,
+      "loss": 1.5714,
+      "step": 4815
+    },
+    {
+      "epoch": 0.3355746786050239,
+      "grad_norm": 0.6860445057284045,
+      "learning_rate": 5.405651725962358e-07,
+      "loss": 1.5411,
+      "step": 4816
+    },
+    {
+      "epoch": 0.3356443577326412,
+      "grad_norm": 0.6967122024785454,
+      "learning_rate": 5.405052078237082e-07,
+      "loss": 1.6579,
+      "step": 4817
+    },
+    {
+      "epoch": 0.3357140368602585,
+      "grad_norm": 0.7647204058296255,
+      "learning_rate": 5.404452355990515e-07,
+      "loss": 1.5023,
+      "step": 4818
+    },
+    {
+      "epoch": 0.3357837159878758,
+      "grad_norm": 0.7108929741768184,
+      "learning_rate": 5.403852559251401e-07,
+      "loss": 1.5517,
+      "step": 4819
+    },
+    {
+      "epoch": 0.33585339511549317,
+      "grad_norm": 0.8140504410329782,
+      "learning_rate": 5.403252688048482e-07,
+      "loss": 1.5006,
+      "step": 4820
+    },
+    {
+      "epoch": 0.33592307424311046,
+      "grad_norm": 0.6860079696803548,
+      "learning_rate": 5.402652742410505e-07,
+      "loss": 1.6411,
+      "step": 4821
+    },
+    {
+      "epoch": 0.3359927533707278,
+      "grad_norm": 0.7650567940545888,
+      "learning_rate": 5.402052722366221e-07,
+      "loss": 1.3765,
+      "step": 4822
+    },
+    {
+      "epoch": 0.3360624324983451,
+      "grad_norm": 0.7242517343750577,
+      "learning_rate": 5.401452627944387e-07,
+      "loss": 1.4242,
+      "step": 4823
+    },
+    {
+      "epoch": 0.33613211162596246,
+      "grad_norm": 0.7431924617453302,
+      "learning_rate": 5.400852459173754e-07,
+      "loss": 1.586,
+      "step": 4824
+    },
+    {
+      "epoch": 0.33620179075357975,
+      "grad_norm": 0.698467639711488,
+      "learning_rate": 5.40025221608309e-07,
+      "loss": 1.5904,
+      "step": 4825
+    },
+    {
+      "epoch": 0.3362714698811971,
+      "grad_norm": 0.7824488496164661,
+      "learning_rate": 5.399651898701156e-07,
+      "loss": 1.4601,
+      "step": 4826
+    },
+    {
+      "epoch": 0.3363411490088144,
+      "grad_norm": 0.7380126909494652,
+      "learning_rate": 5.399051507056722e-07,
+      "loss": 1.421,
+      "step": 4827
+    },
+    {
+      "epoch": 0.33641082813643175,
+      "grad_norm": 0.7554036049745274,
+      "learning_rate": 5.39845104117856e-07,
+      "loss": 1.5047,
+      "step": 4828
+    },
+    {
+      "epoch": 0.33648050726404904,
+      "grad_norm": 0.6730427825408142,
+      "learning_rate": 5.397850501095445e-07,
+      "loss": 1.5169,
+      "step": 4829
+    },
+    {
+      "epoch": 0.3365501863916664,
+      "grad_norm": 0.7522968508915672,
+      "learning_rate": 5.397249886836155e-07,
+      "loss": 1.579,
+      "step": 4830
+    },
+    {
+      "epoch": 0.3366198655192837,
+      "grad_norm": 0.7656384123478277,
+      "learning_rate": 5.396649198429476e-07,
+      "loss": 1.5836,
+      "step": 4831
+    },
+    {
+      "epoch": 0.33668954464690104,
+      "grad_norm": 0.7095621969140741,
+      "learning_rate": 5.396048435904192e-07,
+      "loss": 1.5901,
+      "step": 4832
+    },
+    {
+      "epoch": 0.33675922377451833,
+      "grad_norm": 0.6984903216291455,
+      "learning_rate": 5.395447599289092e-07,
+      "loss": 1.5424,
+      "step": 4833
+    },
+    {
+      "epoch": 0.3368289029021357,
+      "grad_norm": 0.7296764226484795,
+      "learning_rate": 5.394846688612969e-07,
+      "loss": 1.5648,
+      "step": 4834
+    },
+    {
+      "epoch": 0.336898582029753,
+      "grad_norm": 0.6635513910837075,
+      "learning_rate": 5.394245703904623e-07,
+      "loss": 1.4931,
+      "step": 4835
+    },
+    {
+      "epoch": 0.33696826115737033,
+      "grad_norm": 0.7388466868414264,
+      "learning_rate": 5.393644645192853e-07,
+      "loss": 1.4121,
+      "step": 4836
+    },
+    {
+      "epoch": 0.3370379402849876,
+      "grad_norm": 0.7112377558521906,
+      "learning_rate": 5.393043512506462e-07,
+      "loss": 1.4546,
+      "step": 4837
+    },
+    {
+      "epoch": 0.337107619412605,
+      "grad_norm": 0.727315967629537,
+      "learning_rate": 5.392442305874258e-07,
+      "loss": 1.6098,
+      "step": 4838
+    },
+    {
+      "epoch": 0.33717729854022227,
+      "grad_norm": 0.8099460689341998,
+      "learning_rate": 5.391841025325051e-07,
+      "loss": 1.6303,
+      "step": 4839
+    },
+    {
+      "epoch": 0.3372469776678396,
+      "grad_norm": 0.7051935142489242,
+      "learning_rate": 5.391239670887659e-07,
+      "loss": 1.4497,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3373166567954569,
+      "grad_norm": 0.6783397479396385,
+      "learning_rate": 5.390638242590897e-07,
+      "loss": 1.5105,
+      "step": 4841
+    },
+    {
+      "epoch": 0.33738633592307427,
+      "grad_norm": 0.7347089211759783,
+      "learning_rate": 5.390036740463587e-07,
+      "loss": 1.6473,
+      "step": 4842
+    },
+    {
+      "epoch": 0.33745601505069156,
+      "grad_norm": 0.6732402289305454,
+      "learning_rate": 5.389435164534555e-07,
+      "loss": 1.4919,
+      "step": 4843
+    },
+    {
+      "epoch": 0.3375256941783089,
+      "grad_norm": 0.7148083262660647,
+      "learning_rate": 5.38883351483263e-07,
+      "loss": 1.5546,
+      "step": 4844
+    },
+    {
+      "epoch": 0.3375953733059262,
+      "grad_norm": 0.6836003305145799,
+      "learning_rate": 5.388231791386643e-07,
+      "loss": 1.4778,
+      "step": 4845
+    },
+    {
+      "epoch": 0.33766505243354356,
+      "grad_norm": 0.8084013190577042,
+      "learning_rate": 5.387629994225432e-07,
+      "loss": 1.6094,
+      "step": 4846
+    },
+    {
+      "epoch": 0.33773473156116085,
+      "grad_norm": 0.662801305334541,
+      "learning_rate": 5.387028123377832e-07,
+      "loss": 1.4713,
+      "step": 4847
+    },
+    {
+      "epoch": 0.3378044106887782,
+      "grad_norm": 0.6791407537391915,
+      "learning_rate": 5.386426178872692e-07,
+      "loss": 1.5404,
+      "step": 4848
+    },
+    {
+      "epoch": 0.3378740898163955,
+      "grad_norm": 0.7335026978467318,
+      "learning_rate": 5.385824160738854e-07,
+      "loss": 1.5911,
+      "step": 4849
+    },
+    {
+      "epoch": 0.33794376894401285,
+      "grad_norm": 0.6875140299288989,
+      "learning_rate": 5.385222069005169e-07,
+      "loss": 1.5747,
+      "step": 4850
+    },
+    {
+      "epoch": 0.33801344807163014,
+      "grad_norm": 0.6989430758131691,
+      "learning_rate": 5.384619903700492e-07,
+      "loss": 1.4246,
+      "step": 4851
+    },
+    {
+      "epoch": 0.3380831271992475,
+      "grad_norm": 0.6590948742130245,
+      "learning_rate": 5.384017664853677e-07,
+      "loss": 1.5146,
+      "step": 4852
+    },
+    {
+      "epoch": 0.3381528063268648,
+      "grad_norm": 0.706352782049481,
+      "learning_rate": 5.383415352493587e-07,
+      "loss": 1.571,
+      "step": 4853
+    },
+    {
+      "epoch": 0.33822248545448214,
+      "grad_norm": 0.7519551620740152,
+      "learning_rate": 5.382812966649086e-07,
+      "loss": 1.5446,
+      "step": 4854
+    },
+    {
+      "epoch": 0.33829216458209943,
+      "grad_norm": 0.7099021233898958,
+      "learning_rate": 5.38221050734904e-07,
+      "loss": 1.5529,
+      "step": 4855
+    },
+    {
+      "epoch": 0.3383618437097168,
+      "grad_norm": 0.7199442639569562,
+      "learning_rate": 5.38160797462232e-07,
+      "loss": 1.534,
+      "step": 4856
+    },
+    {
+      "epoch": 0.3384315228373341,
+      "grad_norm": 0.7144789348289255,
+      "learning_rate": 5.381005368497803e-07,
+      "loss": 1.4376,
+      "step": 4857
+    },
+    {
+      "epoch": 0.3385012019649514,
+      "grad_norm": 0.7403548398371584,
+      "learning_rate": 5.380402689004365e-07,
+      "loss": 1.5743,
+      "step": 4858
+    },
+    {
+      "epoch": 0.3385708810925687,
+      "grad_norm": 0.7627039850404922,
+      "learning_rate": 5.379799936170888e-07,
+      "loss": 1.5054,
+      "step": 4859
+    },
+    {
+      "epoch": 0.33864056022018607,
+      "grad_norm": 0.677840868673094,
+      "learning_rate": 5.379197110026258e-07,
+      "loss": 1.5203,
+      "step": 4860
+    },
+    {
+      "epoch": 0.33871023934780337,
+      "grad_norm": 0.7394880148402937,
+      "learning_rate": 5.378594210599363e-07,
+      "loss": 1.61,
+      "step": 4861
+    },
+    {
+      "epoch": 0.33877991847542066,
+      "grad_norm": 0.7543058613191993,
+      "learning_rate": 5.377991237919096e-07,
+      "loss": 1.5824,
+      "step": 4862
+    },
+    {
+      "epoch": 0.338849597603038,
+      "grad_norm": 0.665265338341848,
+      "learning_rate": 5.377388192014351e-07,
+      "loss": 1.4639,
+      "step": 4863
+    },
+    {
+      "epoch": 0.3389192767306553,
+      "grad_norm": 0.711462501878098,
+      "learning_rate": 5.376785072914029e-07,
+      "loss": 1.5846,
+      "step": 4864
+    },
+    {
+      "epoch": 0.33898895585827266,
+      "grad_norm": 0.7270116627711709,
+      "learning_rate": 5.37618188064703e-07,
+      "loss": 1.5409,
+      "step": 4865
+    },
+    {
+      "epoch": 0.33905863498588995,
+      "grad_norm": 0.6769787698247857,
+      "learning_rate": 5.375578615242263e-07,
+      "loss": 1.5406,
+      "step": 4866
+    },
+    {
+      "epoch": 0.3391283141135073,
+      "grad_norm": 0.7193369604108305,
+      "learning_rate": 5.374975276728638e-07,
+      "loss": 1.5786,
+      "step": 4867
+    },
+    {
+      "epoch": 0.3391979932411246,
+      "grad_norm": 0.7083441966840394,
+      "learning_rate": 5.374371865135067e-07,
+      "loss": 1.4636,
+      "step": 4868
+    },
+    {
+      "epoch": 0.33926767236874195,
+      "grad_norm": 0.6917558146241541,
+      "learning_rate": 5.373768380490466e-07,
+      "loss": 1.4435,
+      "step": 4869
+    },
+    {
+      "epoch": 0.33933735149635924,
+      "grad_norm": 0.714922470007537,
+      "learning_rate": 5.373164822823755e-07,
+      "loss": 1.4841,
+      "step": 4870
+    },
+    {
+      "epoch": 0.3394070306239766,
+      "grad_norm": 0.717829779405514,
+      "learning_rate": 5.37256119216386e-07,
+      "loss": 1.5644,
+      "step": 4871
+    },
+    {
+      "epoch": 0.3394767097515939,
+      "grad_norm": 0.7743959381513768,
+      "learning_rate": 5.371957488539706e-07,
+      "loss": 1.5185,
+      "step": 4872
+    },
+    {
+      "epoch": 0.33954638887921124,
+      "grad_norm": 0.8131617550648077,
+      "learning_rate": 5.371353711980225e-07,
+      "loss": 1.5316,
+      "step": 4873
+    },
+    {
+      "epoch": 0.33961606800682853,
+      "grad_norm": 0.7709352415864421,
+      "learning_rate": 5.370749862514352e-07,
+      "loss": 1.5742,
+      "step": 4874
+    },
+    {
+      "epoch": 0.3396857471344459,
+      "grad_norm": 0.7807457832872728,
+      "learning_rate": 5.370145940171022e-07,
+      "loss": 1.6104,
+      "step": 4875
+    },
+    {
+      "epoch": 0.3397554262620632,
+      "grad_norm": 0.7267607315291826,
+      "learning_rate": 5.369541944979178e-07,
+      "loss": 1.4927,
+      "step": 4876
+    },
+    {
+      "epoch": 0.3398251053896805,
+      "grad_norm": 0.7361558349717697,
+      "learning_rate": 5.368937876967765e-07,
+      "loss": 1.4753,
+      "step": 4877
+    },
+    {
+      "epoch": 0.3398947845172978,
+      "grad_norm": 0.6852432333809807,
+      "learning_rate": 5.36833373616573e-07,
+      "loss": 1.5538,
+      "step": 4878
+    },
+    {
+      "epoch": 0.33996446364491517,
+      "grad_norm": 0.7199238314597366,
+      "learning_rate": 5.367729522602026e-07,
+      "loss": 1.475,
+      "step": 4879
+    },
+    {
+      "epoch": 0.34003414277253247,
+      "grad_norm": 0.7732480003051626,
+      "learning_rate": 5.367125236305607e-07,
+      "loss": 1.6253,
+      "step": 4880
+    },
+    {
+      "epoch": 0.3401038219001498,
+      "grad_norm": 0.690556485623902,
+      "learning_rate": 5.366520877305433e-07,
+      "loss": 1.6229,
+      "step": 4881
+    },
+    {
+      "epoch": 0.3401735010277671,
+      "grad_norm": 0.7113155930404748,
+      "learning_rate": 5.365916445630464e-07,
+      "loss": 1.5381,
+      "step": 4882
+    },
+    {
+      "epoch": 0.34024318015538446,
+      "grad_norm": 0.6147503521616917,
+      "learning_rate": 5.365311941309667e-07,
+      "loss": 1.4189,
+      "step": 4883
+    },
+    {
+      "epoch": 0.34031285928300176,
+      "grad_norm": 0.714706834724729,
+      "learning_rate": 5.36470736437201e-07,
+      "loss": 1.5128,
+      "step": 4884
+    },
+    {
+      "epoch": 0.3403825384106191,
+      "grad_norm": 0.7150201217564616,
+      "learning_rate": 5.364102714846469e-07,
+      "loss": 1.5691,
+      "step": 4885
+    },
+    {
+      "epoch": 0.3404522175382364,
+      "grad_norm": 0.7304098295016737,
+      "learning_rate": 5.363497992762015e-07,
+      "loss": 1.432,
+      "step": 4886
+    },
+    {
+      "epoch": 0.34052189666585375,
+      "grad_norm": 0.7570206498289689,
+      "learning_rate": 5.362893198147631e-07,
+      "loss": 1.728,
+      "step": 4887
+    },
+    {
+      "epoch": 0.34059157579347105,
+      "grad_norm": 0.7143465756664162,
+      "learning_rate": 5.3622883310323e-07,
+      "loss": 1.5685,
+      "step": 4888
+    },
+    {
+      "epoch": 0.3406612549210884,
+      "grad_norm": 0.8122541734007522,
+      "learning_rate": 5.361683391445006e-07,
+      "loss": 1.5325,
+      "step": 4889
+    },
+    {
+      "epoch": 0.3407309340487057,
+      "grad_norm": 0.7734092568022763,
+      "learning_rate": 5.361078379414741e-07,
+      "loss": 1.4959,
+      "step": 4890
+    },
+    {
+      "epoch": 0.34080061317632304,
+      "grad_norm": 0.7251845358695838,
+      "learning_rate": 5.360473294970499e-07,
+      "loss": 1.5793,
+      "step": 4891
+    },
+    {
+      "epoch": 0.34087029230394034,
+      "grad_norm": 0.7678069911342846,
+      "learning_rate": 5.359868138141274e-07,
+      "loss": 1.4537,
+      "step": 4892
+    },
+    {
+      "epoch": 0.3409399714315577,
+      "grad_norm": 0.723545658098318,
+      "learning_rate": 5.359262908956068e-07,
+      "loss": 1.4833,
+      "step": 4893
+    },
+    {
+      "epoch": 0.341009650559175,
+      "grad_norm": 0.710062730643498,
+      "learning_rate": 5.358657607443887e-07,
+      "loss": 1.5221,
+      "step": 4894
+    },
+    {
+      "epoch": 0.34107932968679233,
+      "grad_norm": 0.7061472317372298,
+      "learning_rate": 5.358052233633734e-07,
+      "loss": 1.5833,
+      "step": 4895
+    },
+    {
+      "epoch": 0.3411490088144096,
+      "grad_norm": 0.7442311521043259,
+      "learning_rate": 5.357446787554623e-07,
+      "loss": 1.5627,
+      "step": 4896
+    },
+    {
+      "epoch": 0.341218687942027,
+      "grad_norm": 0.6977710712490137,
+      "learning_rate": 5.356841269235568e-07,
+      "loss": 1.5103,
+      "step": 4897
+    },
+    {
+      "epoch": 0.34128836706964427,
+      "grad_norm": 0.6829996598763791,
+      "learning_rate": 5.356235678705584e-07,
+      "loss": 1.539,
+      "step": 4898
+    },
+    {
+      "epoch": 0.3413580461972616,
+      "grad_norm": 0.7041108868124177,
+      "learning_rate": 5.355630015993696e-07,
+      "loss": 1.4318,
+      "step": 4899
+    },
+    {
+      "epoch": 0.3414277253248789,
+      "grad_norm": 0.739511445992145,
+      "learning_rate": 5.355024281128926e-07,
+      "loss": 1.6007,
+      "step": 4900
+    },
+    {
+      "epoch": 0.34149740445249627,
+      "grad_norm": 0.752034002439808,
+      "learning_rate": 5.354418474140302e-07,
+      "loss": 1.6381,
+      "step": 4901
+    },
+    {
+      "epoch": 0.34156708358011356,
+      "grad_norm": 0.7260501221496931,
+      "learning_rate": 5.353812595056856e-07,
+      "loss": 1.6835,
+      "step": 4902
+    },
+    {
+      "epoch": 0.3416367627077309,
+      "grad_norm": 0.6857671150292761,
+      "learning_rate": 5.353206643907624e-07,
+      "loss": 1.6081,
+      "step": 4903
+    },
+    {
+      "epoch": 0.3417064418353482,
+      "grad_norm": 0.7102916293714224,
+      "learning_rate": 5.352600620721644e-07,
+      "loss": 1.4413,
+      "step": 4904
+    },
+    {
+      "epoch": 0.34177612096296556,
+      "grad_norm": 0.6528079878598496,
+      "learning_rate": 5.351994525527957e-07,
+      "loss": 1.5078,
+      "step": 4905
+    },
+    {
+      "epoch": 0.34184580009058285,
+      "grad_norm": 0.765193785606818,
+      "learning_rate": 5.351388358355609e-07,
+      "loss": 1.4708,
+      "step": 4906
+    },
+    {
+      "epoch": 0.3419154792182002,
+      "grad_norm": 0.7233681669435182,
+      "learning_rate": 5.350782119233648e-07,
+      "loss": 1.5652,
+      "step": 4907
+    },
+    {
+      "epoch": 0.3419851583458175,
+      "grad_norm": 0.6973226611884625,
+      "learning_rate": 5.350175808191127e-07,
+      "loss": 1.5033,
+      "step": 4908
+    },
+    {
+      "epoch": 0.34205483747343485,
+      "grad_norm": 0.7136416717450348,
+      "learning_rate": 5.349569425257101e-07,
+      "loss": 1.4198,
+      "step": 4909
+    },
+    {
+      "epoch": 0.34212451660105214,
+      "grad_norm": 0.7693775797382657,
+      "learning_rate": 5.348962970460631e-07,
+      "loss": 1.409,
+      "step": 4910
+    },
+    {
+      "epoch": 0.3421941957286695,
+      "grad_norm": 0.7336409853595478,
+      "learning_rate": 5.348356443830777e-07,
+      "loss": 1.4816,
+      "step": 4911
+    },
+    {
+      "epoch": 0.3422638748562868,
+      "grad_norm": 0.6823577987189021,
+      "learning_rate": 5.347749845396606e-07,
+      "loss": 1.5111,
+      "step": 4912
+    },
+    {
+      "epoch": 0.34233355398390414,
+      "grad_norm": 0.7167651243573857,
+      "learning_rate": 5.347143175187188e-07,
+      "loss": 1.3348,
+      "step": 4913
+    },
+    {
+      "epoch": 0.34240323311152143,
+      "grad_norm": 0.7398765728960008,
+      "learning_rate": 5.346536433231596e-07,
+      "loss": 1.465,
+      "step": 4914
+    },
+    {
+      "epoch": 0.3424729122391388,
+      "grad_norm": 0.7291279264814856,
+      "learning_rate": 5.345929619558905e-07,
+      "loss": 1.4537,
+      "step": 4915
+    },
+    {
+      "epoch": 0.3425425913667561,
+      "grad_norm": 0.6960763279406099,
+      "learning_rate": 5.345322734198196e-07,
+      "loss": 1.5036,
+      "step": 4916
+    },
+    {
+      "epoch": 0.3426122704943734,
+      "grad_norm": 0.7419002594743389,
+      "learning_rate": 5.344715777178551e-07,
+      "loss": 1.6412,
+      "step": 4917
+    },
+    {
+      "epoch": 0.3426819496219907,
+      "grad_norm": 0.7595820712509183,
+      "learning_rate": 5.344108748529058e-07,
+      "loss": 1.6397,
+      "step": 4918
+    },
+    {
+      "epoch": 0.3427516287496081,
+      "grad_norm": 0.7694766737856703,
+      "learning_rate": 5.343501648278807e-07,
+      "loss": 1.6078,
+      "step": 4919
+    },
+    {
+      "epoch": 0.34282130787722537,
+      "grad_norm": 0.6807981058044563,
+      "learning_rate": 5.342894476456889e-07,
+      "loss": 1.4479,
+      "step": 4920
+    },
+    {
+      "epoch": 0.3428909870048427,
+      "grad_norm": 0.754974484289456,
+      "learning_rate": 5.342287233092405e-07,
+      "loss": 1.6514,
+      "step": 4921
+    },
+    {
+      "epoch": 0.34296066613246,
+      "grad_norm": 0.7535510628217905,
+      "learning_rate": 5.341679918214452e-07,
+      "loss": 1.5696,
+      "step": 4922
+    },
+    {
+      "epoch": 0.34303034526007736,
+      "grad_norm": 0.7357029495555406,
+      "learning_rate": 5.341072531852134e-07,
+      "loss": 1.521,
+      "step": 4923
+    },
+    {
+      "epoch": 0.34310002438769466,
+      "grad_norm": 0.7082023447167447,
+      "learning_rate": 5.340465074034562e-07,
+      "loss": 1.5007,
+      "step": 4924
+    },
+    {
+      "epoch": 0.343169703515312,
+      "grad_norm": 0.7097395892281676,
+      "learning_rate": 5.339857544790843e-07,
+      "loss": 1.4334,
+      "step": 4925
+    },
+    {
+      "epoch": 0.3432393826429293,
+      "grad_norm": 0.7437771128644068,
+      "learning_rate": 5.33924994415009e-07,
+      "loss": 1.6869,
+      "step": 4926
+    },
+    {
+      "epoch": 0.34330906177054665,
+      "grad_norm": 0.7222212332837278,
+      "learning_rate": 5.338642272141424e-07,
+      "loss": 1.6559,
+      "step": 4927
+    },
+    {
+      "epoch": 0.34337874089816395,
+      "grad_norm": 0.774011346047693,
+      "learning_rate": 5.338034528793963e-07,
+      "loss": 1.6342,
+      "step": 4928
+    },
+    {
+      "epoch": 0.3434484200257813,
+      "grad_norm": 0.6966346592207932,
+      "learning_rate": 5.337426714136832e-07,
+      "loss": 1.4734,
+      "step": 4929
+    },
+    {
+      "epoch": 0.3435180991533986,
+      "grad_norm": 0.788871599051374,
+      "learning_rate": 5.33681882819916e-07,
+      "loss": 1.5862,
+      "step": 4930
+    },
+    {
+      "epoch": 0.34358777828101594,
+      "grad_norm": 0.7397766374316244,
+      "learning_rate": 5.336210871010078e-07,
+      "loss": 1.6653,
+      "step": 4931
+    },
+    {
+      "epoch": 0.34365745740863324,
+      "grad_norm": 0.7180829427756388,
+      "learning_rate": 5.335602842598721e-07,
+      "loss": 1.5283,
+      "step": 4932
+    },
+    {
+      "epoch": 0.3437271365362506,
+      "grad_norm": 0.7348660111880068,
+      "learning_rate": 5.334994742994224e-07,
+      "loss": 1.618,
+      "step": 4933
+    },
+    {
+      "epoch": 0.3437968156638679,
+      "grad_norm": 0.7752690830193244,
+      "learning_rate": 5.33438657222573e-07,
+      "loss": 1.6631,
+      "step": 4934
+    },
+    {
+      "epoch": 0.34386649479148523,
+      "grad_norm": 0.7061573790825465,
+      "learning_rate": 5.333778330322386e-07,
+      "loss": 1.5554,
+      "step": 4935
+    },
+    {
+      "epoch": 0.34393617391910253,
+      "grad_norm": 0.6966452925169359,
+      "learning_rate": 5.333170017313336e-07,
+      "loss": 1.4936,
+      "step": 4936
+    },
+    {
+      "epoch": 0.3440058530467199,
+      "grad_norm": 0.7108489950765159,
+      "learning_rate": 5.332561633227736e-07,
+      "loss": 1.49,
+      "step": 4937
+    },
+    {
+      "epoch": 0.3440755321743372,
+      "grad_norm": 0.7360768060995024,
+      "learning_rate": 5.331953178094737e-07,
+      "loss": 1.5874,
+      "step": 4938
+    },
+    {
+      "epoch": 0.3441452113019545,
+      "grad_norm": 0.7061287583659489,
+      "learning_rate": 5.3313446519435e-07,
+      "loss": 1.5182,
+      "step": 4939
+    },
+    {
+      "epoch": 0.3442148904295718,
+      "grad_norm": 0.7531667234468807,
+      "learning_rate": 5.330736054803186e-07,
+      "loss": 1.5243,
+      "step": 4940
+    },
+    {
+      "epoch": 0.34428456955718917,
+      "grad_norm": 0.649115378193036,
+      "learning_rate": 5.330127386702962e-07,
+      "loss": 1.4874,
+      "step": 4941
+    },
+    {
+      "epoch": 0.34435424868480646,
+      "grad_norm": 0.6839245166901333,
+      "learning_rate": 5.329518647671992e-07,
+      "loss": 1.4666,
+      "step": 4942
+    },
+    {
+      "epoch": 0.3444239278124238,
+      "grad_norm": 0.6955071209132906,
+      "learning_rate": 5.328909837739454e-07,
+      "loss": 1.6033,
+      "step": 4943
+    },
+    {
+      "epoch": 0.3444936069400411,
+      "grad_norm": 0.7160416982988688,
+      "learning_rate": 5.328300956934519e-07,
+      "loss": 1.4956,
+      "step": 4944
+    },
+    {
+      "epoch": 0.34456328606765846,
+      "grad_norm": 0.6908646469509884,
+      "learning_rate": 5.327692005286366e-07,
+      "loss": 1.5294,
+      "step": 4945
+    },
+    {
+      "epoch": 0.34463296519527575,
+      "grad_norm": 0.7472362365311507,
+      "learning_rate": 5.327082982824181e-07,
+      "loss": 1.5895,
+      "step": 4946
+    },
+    {
+      "epoch": 0.3447026443228931,
+      "grad_norm": 0.7505746601450485,
+      "learning_rate": 5.326473889577145e-07,
+      "loss": 1.4966,
+      "step": 4947
+    },
+    {
+      "epoch": 0.3447723234505104,
+      "grad_norm": 0.6600104860006836,
+      "learning_rate": 5.325864725574451e-07,
+      "loss": 1.4753,
+      "step": 4948
+    },
+    {
+      "epoch": 0.34484200257812775,
+      "grad_norm": 0.7586810928023145,
+      "learning_rate": 5.325255490845287e-07,
+      "loss": 1.5685,
+      "step": 4949
+    },
+    {
+      "epoch": 0.34491168170574504,
+      "grad_norm": 0.7852011275224828,
+      "learning_rate": 5.324646185418853e-07,
+      "loss": 1.6254,
+      "step": 4950
+    },
+    {
+      "epoch": 0.3449813608333624,
+      "grad_norm": 0.7464220138774134,
+      "learning_rate": 5.324036809324347e-07,
+      "loss": 1.6366,
+      "step": 4951
+    },
+    {
+      "epoch": 0.3450510399609797,
+      "grad_norm": 0.707679820201954,
+      "learning_rate": 5.32342736259097e-07,
+      "loss": 1.5273,
+      "step": 4952
+    },
+    {
+      "epoch": 0.345120719088597,
+      "grad_norm": 0.6649346495566145,
+      "learning_rate": 5.322817845247929e-07,
+      "loss": 1.5056,
+      "step": 4953
+    },
+    {
+      "epoch": 0.34519039821621433,
+      "grad_norm": 0.7446126608675369,
+      "learning_rate": 5.322208257324433e-07,
+      "loss": 1.5291,
+      "step": 4954
+    },
+    {
+      "epoch": 0.34526007734383163,
+      "grad_norm": 0.7157920582125764,
+      "learning_rate": 5.321598598849695e-07,
+      "loss": 1.5463,
+      "step": 4955
+    },
+    {
+      "epoch": 0.345329756471449,
+      "grad_norm": 0.7130418175302066,
+      "learning_rate": 5.32098886985293e-07,
+      "loss": 1.5481,
+      "step": 4956
+    },
+    {
+      "epoch": 0.3453994355990663,
+      "grad_norm": 0.6549977386565085,
+      "learning_rate": 5.320379070363359e-07,
+      "loss": 1.38,
+      "step": 4957
+    },
+    {
+      "epoch": 0.3454691147266836,
+      "grad_norm": 0.6719591171989239,
+      "learning_rate": 5.319769200410205e-07,
+      "loss": 1.5743,
+      "step": 4958
+    },
+    {
+      "epoch": 0.3455387938543009,
+      "grad_norm": 0.7086929840190598,
+      "learning_rate": 5.319159260022694e-07,
+      "loss": 1.5321,
+      "step": 4959
+    },
+    {
+      "epoch": 0.34560847298191827,
+      "grad_norm": 0.7158822266473911,
+      "learning_rate": 5.318549249230055e-07,
+      "loss": 1.558,
+      "step": 4960
+    },
+    {
+      "epoch": 0.34567815210953556,
+      "grad_norm": 0.7338554111996759,
+      "learning_rate": 5.317939168061521e-07,
+      "loss": 1.6296,
+      "step": 4961
+    },
+    {
+      "epoch": 0.3457478312371529,
+      "grad_norm": 0.716688713173419,
+      "learning_rate": 5.317329016546326e-07,
+      "loss": 1.5146,
+      "step": 4962
+    },
+    {
+      "epoch": 0.3458175103647702,
+      "grad_norm": 0.6746045493486954,
+      "learning_rate": 5.316718794713716e-07,
+      "loss": 1.4125,
+      "step": 4963
+    },
+    {
+      "epoch": 0.34588718949238756,
+      "grad_norm": 0.7673016809121121,
+      "learning_rate": 5.316108502592928e-07,
+      "loss": 1.4457,
+      "step": 4964
+    },
+    {
+      "epoch": 0.34595686862000485,
+      "grad_norm": 0.6981807378933002,
+      "learning_rate": 5.315498140213211e-07,
+      "loss": 1.4776,
+      "step": 4965
+    },
+    {
+      "epoch": 0.3460265477476222,
+      "grad_norm": 0.7111768030048936,
+      "learning_rate": 5.314887707603814e-07,
+      "loss": 1.4599,
+      "step": 4966
+    },
+    {
+      "epoch": 0.3460962268752395,
+      "grad_norm": 0.7897503701287117,
+      "learning_rate": 5.314277204793992e-07,
+      "loss": 1.6354,
+      "step": 4967
+    },
+    {
+      "epoch": 0.34616590600285685,
+      "grad_norm": 0.7191071514005091,
+      "learning_rate": 5.313666631813e-07,
+      "loss": 1.5362,
+      "step": 4968
+    },
+    {
+      "epoch": 0.34623558513047414,
+      "grad_norm": 0.7216111690405148,
+      "learning_rate": 5.313055988690098e-07,
+      "loss": 1.4787,
+      "step": 4969
+    },
+    {
+      "epoch": 0.3463052642580915,
+      "grad_norm": 0.6989929154312301,
+      "learning_rate": 5.312445275454549e-07,
+      "loss": 1.5059,
+      "step": 4970
+    },
+    {
+      "epoch": 0.3463749433857088,
+      "grad_norm": 0.7180325961762156,
+      "learning_rate": 5.31183449213562e-07,
+      "loss": 1.5621,
+      "step": 4971
+    },
+    {
+      "epoch": 0.34644462251332614,
+      "grad_norm": 0.6972183647561123,
+      "learning_rate": 5.311223638762581e-07,
+      "loss": 1.5391,
+      "step": 4972
+    },
+    {
+      "epoch": 0.34651430164094343,
+      "grad_norm": 0.7371303123998308,
+      "learning_rate": 5.310612715364705e-07,
+      "loss": 1.4495,
+      "step": 4973
+    },
+    {
+      "epoch": 0.3465839807685608,
+      "grad_norm": 0.7235927075509028,
+      "learning_rate": 5.31000172197127e-07,
+      "loss": 1.66,
+      "step": 4974
+    },
+    {
+      "epoch": 0.3466536598961781,
+      "grad_norm": 0.6811158833110971,
+      "learning_rate": 5.309390658611555e-07,
+      "loss": 1.5143,
+      "step": 4975
+    },
+    {
+      "epoch": 0.34672333902379543,
+      "grad_norm": 0.7297291177884462,
+      "learning_rate": 5.308779525314844e-07,
+      "loss": 1.6434,
+      "step": 4976
+    },
+    {
+      "epoch": 0.3467930181514127,
+      "grad_norm": 0.6648605494038474,
+      "learning_rate": 5.308168322110423e-07,
+      "loss": 1.4774,
+      "step": 4977
+    },
+    {
+      "epoch": 0.3468626972790301,
+      "grad_norm": 0.777247005511412,
+      "learning_rate": 5.307557049027582e-07,
+      "loss": 1.4144,
+      "step": 4978
+    },
+    {
+      "epoch": 0.34693237640664737,
+      "grad_norm": 0.6929973859668674,
+      "learning_rate": 5.306945706095615e-07,
+      "loss": 1.5942,
+      "step": 4979
+    },
+    {
+      "epoch": 0.3470020555342647,
+      "grad_norm": 0.7283578679121835,
+      "learning_rate": 5.30633429334382e-07,
+      "loss": 1.5701,
+      "step": 4980
+    },
+    {
+      "epoch": 0.347071734661882,
+      "grad_norm": 0.7320143590031736,
+      "learning_rate": 5.305722810801493e-07,
+      "loss": 1.5392,
+      "step": 4981
+    },
+    {
+      "epoch": 0.34714141378949936,
+      "grad_norm": 0.6903230410445872,
+      "learning_rate": 5.305111258497943e-07,
+      "loss": 1.6587,
+      "step": 4982
+    },
+    {
+      "epoch": 0.34721109291711666,
+      "grad_norm": 0.6754268076389439,
+      "learning_rate": 5.304499636462473e-07,
+      "loss": 1.5187,
+      "step": 4983
+    },
+    {
+      "epoch": 0.347280772044734,
+      "grad_norm": 0.7071926402253579,
+      "learning_rate": 5.303887944724396e-07,
+      "loss": 1.5039,
+      "step": 4984
+    },
+    {
+      "epoch": 0.3473504511723513,
+      "grad_norm": 0.7314158959654293,
+      "learning_rate": 5.303276183313022e-07,
+      "loss": 1.4893,
+      "step": 4985
+    },
+    {
+      "epoch": 0.34742013029996865,
+      "grad_norm": 0.717579235163545,
+      "learning_rate": 5.30266435225767e-07,
+      "loss": 1.5723,
+      "step": 4986
+    },
+    {
+      "epoch": 0.34748980942758595,
+      "grad_norm": 0.670827472100565,
+      "learning_rate": 5.302052451587659e-07,
+      "loss": 1.4619,
+      "step": 4987
+    },
+    {
+      "epoch": 0.3475594885552033,
+      "grad_norm": 0.7071414740051439,
+      "learning_rate": 5.301440481332316e-07,
+      "loss": 1.6362,
+      "step": 4988
+    },
+    {
+      "epoch": 0.3476291676828206,
+      "grad_norm": 0.7195762826997261,
+      "learning_rate": 5.300828441520965e-07,
+      "loss": 1.5261,
+      "step": 4989
+    },
+    {
+      "epoch": 0.34769884681043794,
+      "grad_norm": 0.7269680457866935,
+      "learning_rate": 5.300216332182934e-07,
+      "loss": 1.5192,
+      "step": 4990
+    },
+    {
+      "epoch": 0.34776852593805524,
+      "grad_norm": 0.7167959844489366,
+      "learning_rate": 5.299604153347562e-07,
+      "loss": 1.5301,
+      "step": 4991
+    },
+    {
+      "epoch": 0.3478382050656726,
+      "grad_norm": 0.7521845229687258,
+      "learning_rate": 5.298991905044182e-07,
+      "loss": 1.5524,
+      "step": 4992
+    },
+    {
+      "epoch": 0.3479078841932899,
+      "grad_norm": 0.7313888512216582,
+      "learning_rate": 5.298379587302136e-07,
+      "loss": 1.5929,
+      "step": 4993
+    },
+    {
+      "epoch": 0.34797756332090723,
+      "grad_norm": 0.670917170186345,
+      "learning_rate": 5.297767200150765e-07,
+      "loss": 1.6268,
+      "step": 4994
+    },
+    {
+      "epoch": 0.34804724244852453,
+      "grad_norm": 0.7296243936097038,
+      "learning_rate": 5.297154743619418e-07,
+      "loss": 1.5761,
+      "step": 4995
+    },
+    {
+      "epoch": 0.3481169215761419,
+      "grad_norm": 0.6621500834968012,
+      "learning_rate": 5.296542217737445e-07,
+      "loss": 1.3975,
+      "step": 4996
+    },
+    {
+      "epoch": 0.3481866007037592,
+      "grad_norm": 0.6841736192653086,
+      "learning_rate": 5.2959296225342e-07,
+      "loss": 1.4708,
+      "step": 4997
+    },
+    {
+      "epoch": 0.3482562798313765,
+      "grad_norm": 0.7506006835561448,
+      "learning_rate": 5.295316958039038e-07,
+      "loss": 1.6007,
+      "step": 4998
+    },
+    {
+      "epoch": 0.3483259589589938,
+      "grad_norm": 0.7525641487586818,
+      "learning_rate": 5.294704224281321e-07,
+      "loss": 1.5337,
+      "step": 4999
+    },
+    {
+      "epoch": 0.34839563808661117,
+      "grad_norm": 0.8019951480668435,
+      "learning_rate": 5.294091421290412e-07,
+      "loss": 1.6959,
+      "step": 5000
+    },
+    {
+      "epoch": 0.34846531721422846,
+      "grad_norm": 0.9289139139040437,
+      "learning_rate": 5.293478549095676e-07,
+      "loss": 1.5733,
+      "step": 5001
+    },
+    {
+      "epoch": 0.3485349963418458,
+      "grad_norm": 0.7366925407109421,
+      "learning_rate": 5.292865607726485e-07,
+      "loss": 1.5289,
+      "step": 5002
+    },
+    {
+      "epoch": 0.3486046754694631,
+      "grad_norm": 0.6901044411339604,
+      "learning_rate": 5.292252597212212e-07,
+      "loss": 1.5065,
+      "step": 5003
+    },
+    {
+      "epoch": 0.34867435459708046,
+      "grad_norm": 0.7858548036787403,
+      "learning_rate": 5.291639517582235e-07,
+      "loss": 1.7448,
+      "step": 5004
+    },
+    {
+      "epoch": 0.34874403372469776,
+      "grad_norm": 0.6910046587719392,
+      "learning_rate": 5.291026368865932e-07,
+      "loss": 1.4958,
+      "step": 5005
+    },
+    {
+      "epoch": 0.3488137128523151,
+      "grad_norm": 0.7131421363913354,
+      "learning_rate": 5.290413151092685e-07,
+      "loss": 1.5104,
+      "step": 5006
+    },
+    {
+      "epoch": 0.3488833919799324,
+      "grad_norm": 0.678645451016774,
+      "learning_rate": 5.289799864291884e-07,
+      "loss": 1.6023,
+      "step": 5007
+    },
+    {
+      "epoch": 0.34895307110754975,
+      "grad_norm": 0.7135897875529243,
+      "learning_rate": 5.289186508492918e-07,
+      "loss": 1.5097,
+      "step": 5008
+    },
+    {
+      "epoch": 0.34902275023516705,
+      "grad_norm": 0.6925868792365488,
+      "learning_rate": 5.288573083725181e-07,
+      "loss": 1.4276,
+      "step": 5009
+    },
+    {
+      "epoch": 0.3490924293627844,
+      "grad_norm": 0.7355852311495054,
+      "learning_rate": 5.287959590018066e-07,
+      "loss": 1.6468,
+      "step": 5010
+    },
+    {
+      "epoch": 0.3491621084904017,
+      "grad_norm": 0.6591714591613197,
+      "learning_rate": 5.287346027400978e-07,
+      "loss": 1.4403,
+      "step": 5011
+    },
+    {
+      "epoch": 0.34923178761801904,
+      "grad_norm": 0.7384149162418628,
+      "learning_rate": 5.286732395903316e-07,
+      "loss": 1.5242,
+      "step": 5012
+    },
+    {
+      "epoch": 0.34930146674563634,
+      "grad_norm": 0.7951440907862842,
+      "learning_rate": 5.286118695554488e-07,
+      "loss": 1.5518,
+      "step": 5013
+    },
+    {
+      "epoch": 0.3493711458732537,
+      "grad_norm": 0.690403671698083,
+      "learning_rate": 5.285504926383904e-07,
+      "loss": 1.4066,
+      "step": 5014
+    },
+    {
+      "epoch": 0.349440825000871,
+      "grad_norm": 0.734429986958859,
+      "learning_rate": 5.284891088420977e-07,
+      "loss": 1.6652,
+      "step": 5015
+    },
+    {
+      "epoch": 0.34951050412848833,
+      "grad_norm": 0.6791322665246944,
+      "learning_rate": 5.284277181695124e-07,
+      "loss": 1.4876,
+      "step": 5016
+    },
+    {
+      "epoch": 0.3495801832561056,
+      "grad_norm": 0.7333763237125314,
+      "learning_rate": 5.283663206235762e-07,
+      "loss": 1.558,
+      "step": 5017
+    },
+    {
+      "epoch": 0.349649862383723,
+      "grad_norm": 0.7894035079650024,
+      "learning_rate": 5.283049162072316e-07,
+      "loss": 1.471,
+      "step": 5018
+    },
+    {
+      "epoch": 0.34971954151134027,
+      "grad_norm": 0.7444643304296316,
+      "learning_rate": 5.282435049234214e-07,
+      "loss": 1.522,
+      "step": 5019
+    },
+    {
+      "epoch": 0.3497892206389576,
+      "grad_norm": 0.6643970018185514,
+      "learning_rate": 5.281820867750883e-07,
+      "loss": 1.5659,
+      "step": 5020
+    },
+    {
+      "epoch": 0.3498588997665749,
+      "grad_norm": 0.7232823075509024,
+      "learning_rate": 5.281206617651756e-07,
+      "loss": 1.5947,
+      "step": 5021
+    },
+    {
+      "epoch": 0.34992857889419227,
+      "grad_norm": 0.7206660363783639,
+      "learning_rate": 5.280592298966271e-07,
+      "loss": 1.5868,
+      "step": 5022
+    },
+    {
+      "epoch": 0.34999825802180956,
+      "grad_norm": 0.7874148940637059,
+      "learning_rate": 5.279977911723866e-07,
+      "loss": 1.6011,
+      "step": 5023
+    },
+    {
+      "epoch": 0.3500679371494269,
+      "grad_norm": 0.7463976753010189,
+      "learning_rate": 5.279363455953982e-07,
+      "loss": 1.6795,
+      "step": 5024
+    },
+    {
+      "epoch": 0.3501376162770442,
+      "grad_norm": 0.7140576677032477,
+      "learning_rate": 5.278748931686068e-07,
+      "loss": 1.4726,
+      "step": 5025
+    },
+    {
+      "epoch": 0.35020729540466156,
+      "grad_norm": 0.8032460802640788,
+      "learning_rate": 5.278134338949572e-07,
+      "loss": 1.5765,
+      "step": 5026
+    },
+    {
+      "epoch": 0.35027697453227885,
+      "grad_norm": 0.736073108144398,
+      "learning_rate": 5.277519677773946e-07,
+      "loss": 1.6496,
+      "step": 5027
+    },
+    {
+      "epoch": 0.3503466536598962,
+      "grad_norm": 0.7433157359642669,
+      "learning_rate": 5.276904948188647e-07,
+      "loss": 1.607,
+      "step": 5028
+    },
+    {
+      "epoch": 0.3504163327875135,
+      "grad_norm": 0.7302026377592102,
+      "learning_rate": 5.276290150223133e-07,
+      "loss": 1.5698,
+      "step": 5029
+    },
+    {
+      "epoch": 0.35048601191513085,
+      "grad_norm": 0.7156186997183115,
+      "learning_rate": 5.275675283906867e-07,
+      "loss": 1.6206,
+      "step": 5030
+    },
+    {
+      "epoch": 0.35055569104274814,
+      "grad_norm": 0.684118999147647,
+      "learning_rate": 5.275060349269315e-07,
+      "loss": 1.5142,
+      "step": 5031
+    },
+    {
+      "epoch": 0.3506253701703655,
+      "grad_norm": 0.7293360774488382,
+      "learning_rate": 5.274445346339945e-07,
+      "loss": 1.5432,
+      "step": 5032
+    },
+    {
+      "epoch": 0.3506950492979828,
+      "grad_norm": 0.7374356798123075,
+      "learning_rate": 5.273830275148231e-07,
+      "loss": 1.547,
+      "step": 5033
+    },
+    {
+      "epoch": 0.35076472842560014,
+      "grad_norm": 0.6984990019219187,
+      "learning_rate": 5.273215135723644e-07,
+      "loss": 1.5509,
+      "step": 5034
+    },
+    {
+      "epoch": 0.35083440755321743,
+      "grad_norm": 0.6984423015837139,
+      "learning_rate": 5.272599928095669e-07,
+      "loss": 1.6198,
+      "step": 5035
+    },
+    {
+      "epoch": 0.3509040866808348,
+      "grad_norm": 0.7221225311495126,
+      "learning_rate": 5.271984652293784e-07,
+      "loss": 1.5365,
+      "step": 5036
+    },
+    {
+      "epoch": 0.3509737658084521,
+      "grad_norm": 0.6832259001328025,
+      "learning_rate": 5.271369308347475e-07,
+      "loss": 1.5944,
+      "step": 5037
+    },
+    {
+      "epoch": 0.3510434449360694,
+      "grad_norm": 0.7288757483748387,
+      "learning_rate": 5.27075389628623e-07,
+      "loss": 1.565,
+      "step": 5038
+    },
+    {
+      "epoch": 0.3511131240636867,
+      "grad_norm": 0.773464821300105,
+      "learning_rate": 5.270138416139543e-07,
+      "loss": 1.5954,
+      "step": 5039
+    },
+    {
+      "epoch": 0.35118280319130407,
+      "grad_norm": 0.719342206905609,
+      "learning_rate": 5.269522867936905e-07,
+      "loss": 1.4334,
+      "step": 5040
+    },
+    {
+      "epoch": 0.35125248231892137,
+      "grad_norm": 0.7003743350586481,
+      "learning_rate": 5.268907251707821e-07,
+      "loss": 1.5568,
+      "step": 5041
+    },
+    {
+      "epoch": 0.3513221614465387,
+      "grad_norm": 0.6756700295397962,
+      "learning_rate": 5.268291567481786e-07,
+      "loss": 1.4975,
+      "step": 5042
+    },
+    {
+      "epoch": 0.351391840574156,
+      "grad_norm": 0.7101359625406757,
+      "learning_rate": 5.267675815288307e-07,
+      "loss": 1.5818,
+      "step": 5043
+    },
+    {
+      "epoch": 0.3514615197017733,
+      "grad_norm": 0.7656785077162607,
+      "learning_rate": 5.267059995156894e-07,
+      "loss": 1.4525,
+      "step": 5044
+    },
+    {
+      "epoch": 0.35153119882939066,
+      "grad_norm": 0.7206756017968566,
+      "learning_rate": 5.266444107117056e-07,
+      "loss": 1.5277,
+      "step": 5045
+    },
+    {
+      "epoch": 0.35160087795700795,
+      "grad_norm": 0.7893166371200515,
+      "learning_rate": 5.265828151198307e-07,
+      "loss": 1.7029,
+      "step": 5046
+    },
+    {
+      "epoch": 0.3516705570846253,
+      "grad_norm": 0.7090526225797813,
+      "learning_rate": 5.265212127430169e-07,
+      "loss": 1.5572,
+      "step": 5047
+    },
+    {
+      "epoch": 0.3517402362122426,
+      "grad_norm": 0.7587655099716046,
+      "learning_rate": 5.264596035842158e-07,
+      "loss": 1.4987,
+      "step": 5048
+    },
+    {
+      "epoch": 0.35180991533985995,
+      "grad_norm": 0.6922136341394148,
+      "learning_rate": 5.263979876463804e-07,
+      "loss": 1.4961,
+      "step": 5049
+    },
+    {
+      "epoch": 0.35187959446747724,
+      "grad_norm": 0.7018301647037124,
+      "learning_rate": 5.263363649324629e-07,
+      "loss": 1.5998,
+      "step": 5050
+    },
+    {
+      "epoch": 0.3519492735950946,
+      "grad_norm": 0.7100948833017958,
+      "learning_rate": 5.262747354454167e-07,
+      "loss": 1.6238,
+      "step": 5051
+    },
+    {
+      "epoch": 0.3520189527227119,
+      "grad_norm": 0.6878432194296483,
+      "learning_rate": 5.262130991881952e-07,
+      "loss": 1.5418,
+      "step": 5052
+    },
+    {
+      "epoch": 0.35208863185032924,
+      "grad_norm": 0.739369196023084,
+      "learning_rate": 5.26151456163752e-07,
+      "loss": 1.5088,
+      "step": 5053
+    },
+    {
+      "epoch": 0.35215831097794653,
+      "grad_norm": 0.7327168663945245,
+      "learning_rate": 5.260898063750413e-07,
+      "loss": 1.5453,
+      "step": 5054
+    },
+    {
+      "epoch": 0.3522279901055639,
+      "grad_norm": 0.7494770476194448,
+      "learning_rate": 5.260281498250174e-07,
+      "loss": 1.5667,
+      "step": 5055
+    },
+    {
+      "epoch": 0.3522976692331812,
+      "grad_norm": 0.7003511050728188,
+      "learning_rate": 5.25966486516635e-07,
+      "loss": 1.4696,
+      "step": 5056
+    },
+    {
+      "epoch": 0.3523673483607985,
+      "grad_norm": 0.682006424136004,
+      "learning_rate": 5.259048164528492e-07,
+      "loss": 1.5554,
+      "step": 5057
+    },
+    {
+      "epoch": 0.3524370274884158,
+      "grad_norm": 0.7261598036764687,
+      "learning_rate": 5.258431396366154e-07,
+      "loss": 1.5927,
+      "step": 5058
+    },
+    {
+      "epoch": 0.35250670661603317,
+      "grad_norm": 0.7304805813607402,
+      "learning_rate": 5.257814560708891e-07,
+      "loss": 1.5012,
+      "step": 5059
+    },
+    {
+      "epoch": 0.35257638574365047,
+      "grad_norm": 0.701653516106193,
+      "learning_rate": 5.257197657586264e-07,
+      "loss": 1.4971,
+      "step": 5060
+    },
+    {
+      "epoch": 0.3526460648712678,
+      "grad_norm": 0.7665366559302987,
+      "learning_rate": 5.256580687027837e-07,
+      "loss": 1.5559,
+      "step": 5061
+    },
+    {
+      "epoch": 0.3527157439988851,
+      "grad_norm": 0.6825767430611114,
+      "learning_rate": 5.255963649063176e-07,
+      "loss": 1.6248,
+      "step": 5062
+    },
+    {
+      "epoch": 0.35278542312650246,
+      "grad_norm": 0.7166776263768039,
+      "learning_rate": 5.255346543721849e-07,
+      "loss": 1.5674,
+      "step": 5063
+    },
+    {
+      "epoch": 0.35285510225411976,
+      "grad_norm": 0.6688400141973269,
+      "learning_rate": 5.254729371033433e-07,
+      "loss": 1.5386,
+      "step": 5064
+    },
+    {
+      "epoch": 0.3529247813817371,
+      "grad_norm": 0.7341742905821448,
+      "learning_rate": 5.2541121310275e-07,
+      "loss": 1.5394,
+      "step": 5065
+    },
+    {
+      "epoch": 0.3529944605093544,
+      "grad_norm": 0.7139855856855144,
+      "learning_rate": 5.25349482373363e-07,
+      "loss": 1.5032,
+      "step": 5066
+    },
+    {
+      "epoch": 0.35306413963697175,
+      "grad_norm": 0.7377531592463131,
+      "learning_rate": 5.252877449181409e-07,
+      "loss": 1.6183,
+      "step": 5067
+    },
+    {
+      "epoch": 0.35313381876458905,
+      "grad_norm": 0.6992765237751483,
+      "learning_rate": 5.25226000740042e-07,
+      "loss": 1.5182,
+      "step": 5068
+    },
+    {
+      "epoch": 0.3532034978922064,
+      "grad_norm": 0.8121552367033361,
+      "learning_rate": 5.251642498420253e-07,
+      "loss": 1.5733,
+      "step": 5069
+    },
+    {
+      "epoch": 0.3532731770198237,
+      "grad_norm": 0.7225162064445212,
+      "learning_rate": 5.251024922270498e-07,
+      "loss": 1.6745,
+      "step": 5070
+    },
+    {
+      "epoch": 0.35334285614744104,
+      "grad_norm": 0.7220584431415679,
+      "learning_rate": 5.250407278980753e-07,
+      "loss": 1.6101,
+      "step": 5071
+    },
+    {
+      "epoch": 0.35341253527505834,
+      "grad_norm": 0.7338355042291222,
+      "learning_rate": 5.249789568580618e-07,
+      "loss": 1.522,
+      "step": 5072
+    },
+    {
+      "epoch": 0.3534822144026757,
+      "grad_norm": 0.6781864282217835,
+      "learning_rate": 5.249171791099692e-07,
+      "loss": 1.4984,
+      "step": 5073
+    },
+    {
+      "epoch": 0.353551893530293,
+      "grad_norm": 0.7579599868943648,
+      "learning_rate": 5.248553946567581e-07,
+      "loss": 1.6135,
+      "step": 5074
+    },
+    {
+      "epoch": 0.35362157265791033,
+      "grad_norm": 0.740562375568996,
+      "learning_rate": 5.247936035013895e-07,
+      "loss": 1.5535,
+      "step": 5075
+    },
+    {
+      "epoch": 0.3536912517855276,
+      "grad_norm": 0.6684099288907697,
+      "learning_rate": 5.247318056468243e-07,
+      "loss": 1.5008,
+      "step": 5076
+    },
+    {
+      "epoch": 0.353760930913145,
+      "grad_norm": 0.6961783128265661,
+      "learning_rate": 5.246700010960242e-07,
+      "loss": 1.3987,
+      "step": 5077
+    },
+    {
+      "epoch": 0.3538306100407623,
+      "grad_norm": 0.7742250152885914,
+      "learning_rate": 5.246081898519508e-07,
+      "loss": 1.4998,
+      "step": 5078
+    },
+    {
+      "epoch": 0.3539002891683796,
+      "grad_norm": 0.6974290524914329,
+      "learning_rate": 5.245463719175663e-07,
+      "loss": 1.4631,
+      "step": 5079
+    },
+    {
+      "epoch": 0.3539699682959969,
+      "grad_norm": 0.7394144947732425,
+      "learning_rate": 5.244845472958334e-07,
+      "loss": 1.6719,
+      "step": 5080
+    },
+    {
+      "epoch": 0.35403964742361427,
+      "grad_norm": 0.7306324314278728,
+      "learning_rate": 5.244227159897145e-07,
+      "loss": 1.4915,
+      "step": 5081
+    },
+    {
+      "epoch": 0.35410932655123156,
+      "grad_norm": 0.742438727557881,
+      "learning_rate": 5.243608780021729e-07,
+      "loss": 1.5282,
+      "step": 5082
+    },
+    {
+      "epoch": 0.3541790056788489,
+      "grad_norm": 0.7491554405441359,
+      "learning_rate": 5.242990333361718e-07,
+      "loss": 1.4513,
+      "step": 5083
+    },
+    {
+      "epoch": 0.3542486848064662,
+      "grad_norm": 0.7174024426449469,
+      "learning_rate": 5.242371819946751e-07,
+      "loss": 1.4468,
+      "step": 5084
+    },
+    {
+      "epoch": 0.35431836393408356,
+      "grad_norm": 0.6926765842468269,
+      "learning_rate": 5.241753239806468e-07,
+      "loss": 1.6201,
+      "step": 5085
+    },
+    {
+      "epoch": 0.35438804306170085,
+      "grad_norm": 0.6878017168512975,
+      "learning_rate": 5.241134592970512e-07,
+      "loss": 1.4277,
+      "step": 5086
+    },
+    {
+      "epoch": 0.3544577221893182,
+      "grad_norm": 0.7167457202174242,
+      "learning_rate": 5.24051587946853e-07,
+      "loss": 1.4938,
+      "step": 5087
+    },
+    {
+      "epoch": 0.3545274013169355,
+      "grad_norm": 0.6964352455294602,
+      "learning_rate": 5.239897099330175e-07,
+      "loss": 1.6145,
+      "step": 5088
+    },
+    {
+      "epoch": 0.35459708044455285,
+      "grad_norm": 0.6974430427662163,
+      "learning_rate": 5.239278252585096e-07,
+      "loss": 1.4977,
+      "step": 5089
+    },
+    {
+      "epoch": 0.35466675957217014,
+      "grad_norm": 0.6855351540727951,
+      "learning_rate": 5.23865933926295e-07,
+      "loss": 1.5572,
+      "step": 5090
+    },
+    {
+      "epoch": 0.3547364386997875,
+      "grad_norm": 0.7201666934289629,
+      "learning_rate": 5.238040359393399e-07,
+      "loss": 1.6433,
+      "step": 5091
+    },
+    {
+      "epoch": 0.3548061178274048,
+      "grad_norm": 0.7277989750215804,
+      "learning_rate": 5.237421313006103e-07,
+      "loss": 1.536,
+      "step": 5092
+    },
+    {
+      "epoch": 0.35487579695502214,
+      "grad_norm": 0.6634341796241707,
+      "learning_rate": 5.236802200130731e-07,
+      "loss": 1.4078,
+      "step": 5093
+    },
+    {
+      "epoch": 0.35494547608263943,
+      "grad_norm": 0.7052333618869744,
+      "learning_rate": 5.23618302079695e-07,
+      "loss": 1.4558,
+      "step": 5094
+    },
+    {
+      "epoch": 0.3550151552102568,
+      "grad_norm": 0.7597778563956493,
+      "learning_rate": 5.235563775034431e-07,
+      "loss": 1.5985,
+      "step": 5095
+    },
+    {
+      "epoch": 0.3550848343378741,
+      "grad_norm": 0.7706519972191562,
+      "learning_rate": 5.234944462872853e-07,
+      "loss": 1.4622,
+      "step": 5096
+    },
+    {
+      "epoch": 0.35515451346549143,
+      "grad_norm": 0.7272603173122851,
+      "learning_rate": 5.234325084341893e-07,
+      "loss": 1.4587,
+      "step": 5097
+    },
+    {
+      "epoch": 0.3552241925931087,
+      "grad_norm": 0.7338825767979815,
+      "learning_rate": 5.233705639471233e-07,
+      "loss": 1.7211,
+      "step": 5098
+    },
+    {
+      "epoch": 0.3552938717207261,
+      "grad_norm": 0.7069141949193741,
+      "learning_rate": 5.233086128290559e-07,
+      "loss": 1.4935,
+      "step": 5099
+    },
+    {
+      "epoch": 0.35536355084834337,
+      "grad_norm": 0.6440283358521028,
+      "learning_rate": 5.232466550829557e-07,
+      "loss": 1.391,
+      "step": 5100
+    },
+    {
+      "epoch": 0.3554332299759607,
+      "grad_norm": 0.7555141567575172,
+      "learning_rate": 5.231846907117919e-07,
+      "loss": 1.6087,
+      "step": 5101
+    },
+    {
+      "epoch": 0.355502909103578,
+      "grad_norm": 0.7162415478121203,
+      "learning_rate": 5.231227197185342e-07,
+      "loss": 1.5383,
+      "step": 5102
+    },
+    {
+      "epoch": 0.35557258823119536,
+      "grad_norm": 0.731361586378461,
+      "learning_rate": 5.230607421061522e-07,
+      "loss": 1.4528,
+      "step": 5103
+    },
+    {
+      "epoch": 0.35564226735881266,
+      "grad_norm": 0.7519791657504464,
+      "learning_rate": 5.22998757877616e-07,
+      "loss": 1.5298,
+      "step": 5104
+    },
+    {
+      "epoch": 0.35571194648643,
+      "grad_norm": 0.7158025830973831,
+      "learning_rate": 5.22936767035896e-07,
+      "loss": 1.494,
+      "step": 5105
+    },
+    {
+      "epoch": 0.3557816256140473,
+      "grad_norm": 0.723748809683638,
+      "learning_rate": 5.228747695839628e-07,
+      "loss": 1.6235,
+      "step": 5106
+    },
+    {
+      "epoch": 0.35585130474166465,
+      "grad_norm": 0.690675333061282,
+      "learning_rate": 5.228127655247878e-07,
+      "loss": 1.5151,
+      "step": 5107
+    },
+    {
+      "epoch": 0.35592098386928195,
+      "grad_norm": 0.7119021276041587,
+      "learning_rate": 5.227507548613421e-07,
+      "loss": 1.5736,
+      "step": 5108
+    },
+    {
+      "epoch": 0.3559906629968993,
+      "grad_norm": 0.7377717095815133,
+      "learning_rate": 5.226887375965974e-07,
+      "loss": 1.6437,
+      "step": 5109
+    },
+    {
+      "epoch": 0.3560603421245166,
+      "grad_norm": 0.6893478182531352,
+      "learning_rate": 5.226267137335256e-07,
+      "loss": 1.4684,
+      "step": 5110
+    },
+    {
+      "epoch": 0.35613002125213394,
+      "grad_norm": 0.7497735677098524,
+      "learning_rate": 5.225646832750993e-07,
+      "loss": 1.6747,
+      "step": 5111
+    },
+    {
+      "epoch": 0.35619970037975124,
+      "grad_norm": 0.7102877627281015,
+      "learning_rate": 5.225026462242909e-07,
+      "loss": 1.5834,
+      "step": 5112
+    },
+    {
+      "epoch": 0.3562693795073686,
+      "grad_norm": 0.7598662761974709,
+      "learning_rate": 5.224406025840734e-07,
+      "loss": 1.5745,
+      "step": 5113
+    },
+    {
+      "epoch": 0.3563390586349859,
+      "grad_norm": 0.690830365671091,
+      "learning_rate": 5.223785523574201e-07,
+      "loss": 1.5295,
+      "step": 5114
+    },
+    {
+      "epoch": 0.35640873776260323,
+      "grad_norm": 0.7349684708292723,
+      "learning_rate": 5.223164955473045e-07,
+      "loss": 1.739,
+      "step": 5115
+    },
+    {
+      "epoch": 0.35647841689022053,
+      "grad_norm": 0.7050326698210355,
+      "learning_rate": 5.222544321567006e-07,
+      "loss": 1.4603,
+      "step": 5116
+    },
+    {
+      "epoch": 0.3565480960178379,
+      "grad_norm": 0.744957758056315,
+      "learning_rate": 5.221923621885824e-07,
+      "loss": 1.6169,
+      "step": 5117
+    },
+    {
+      "epoch": 0.3566177751454552,
+      "grad_norm": 0.7146331562092086,
+      "learning_rate": 5.221302856459247e-07,
+      "loss": 1.5036,
+      "step": 5118
+    },
+    {
+      "epoch": 0.3566874542730725,
+      "grad_norm": 0.7084897847748508,
+      "learning_rate": 5.220682025317022e-07,
+      "loss": 1.5428,
+      "step": 5119
+    },
+    {
+      "epoch": 0.3567571334006898,
+      "grad_norm": 0.768911401979755,
+      "learning_rate": 5.220061128488898e-07,
+      "loss": 1.4999,
+      "step": 5120
+    },
+    {
+      "epoch": 0.35682681252830717,
+      "grad_norm": 0.7568365121723222,
+      "learning_rate": 5.219440166004635e-07,
+      "loss": 1.48,
+      "step": 5121
+    },
+    {
+      "epoch": 0.35689649165592446,
+      "grad_norm": 0.754288265747282,
+      "learning_rate": 5.218819137893987e-07,
+      "loss": 1.6032,
+      "step": 5122
+    },
+    {
+      "epoch": 0.3569661707835418,
+      "grad_norm": 0.7313637670227049,
+      "learning_rate": 5.218198044186714e-07,
+      "loss": 1.528,
+      "step": 5123
+    },
+    {
+      "epoch": 0.3570358499111591,
+      "grad_norm": 0.7476164464044086,
+      "learning_rate": 5.217576884912583e-07,
+      "loss": 1.41,
+      "step": 5124
+    },
+    {
+      "epoch": 0.35710552903877646,
+      "grad_norm": 0.7170855090682579,
+      "learning_rate": 5.216955660101362e-07,
+      "loss": 1.4268,
+      "step": 5125
+    },
+    {
+      "epoch": 0.35717520816639375,
+      "grad_norm": 0.7142896772084316,
+      "learning_rate": 5.216334369782816e-07,
+      "loss": 1.5978,
+      "step": 5126
+    },
+    {
+      "epoch": 0.3572448872940111,
+      "grad_norm": 0.7358415994670768,
+      "learning_rate": 5.215713013986725e-07,
+      "loss": 1.4269,
+      "step": 5127
+    },
+    {
+      "epoch": 0.3573145664216284,
+      "grad_norm": 0.7153570108585909,
+      "learning_rate": 5.215091592742861e-07,
+      "loss": 1.5301,
+      "step": 5128
+    },
+    {
+      "epoch": 0.35738424554924575,
+      "grad_norm": 0.7447373193332415,
+      "learning_rate": 5.214470106081006e-07,
+      "loss": 1.5988,
+      "step": 5129
+    },
+    {
+      "epoch": 0.35745392467686304,
+      "grad_norm": 0.7478977331229874,
+      "learning_rate": 5.213848554030942e-07,
+      "loss": 1.5723,
+      "step": 5130
+    },
+    {
+      "epoch": 0.3575236038044804,
+      "grad_norm": 0.6830375549123991,
+      "learning_rate": 5.213226936622456e-07,
+      "loss": 1.5305,
+      "step": 5131
+    },
+    {
+      "epoch": 0.3575932829320977,
+      "grad_norm": 0.7308156057281666,
+      "learning_rate": 5.212605253885334e-07,
+      "loss": 1.5576,
+      "step": 5132
+    },
+    {
+      "epoch": 0.35766296205971504,
+      "grad_norm": 0.6936867476883012,
+      "learning_rate": 5.211983505849374e-07,
+      "loss": 1.4978,
+      "step": 5133
+    },
+    {
+      "epoch": 0.35773264118733233,
+      "grad_norm": 0.7707040281173906,
+      "learning_rate": 5.211361692544366e-07,
+      "loss": 1.5393,
+      "step": 5134
+    },
+    {
+      "epoch": 0.3578023203149497,
+      "grad_norm": 0.7024155180503071,
+      "learning_rate": 5.210739814000112e-07,
+      "loss": 1.5802,
+      "step": 5135
+    },
+    {
+      "epoch": 0.357871999442567,
+      "grad_norm": 0.7394402935977779,
+      "learning_rate": 5.210117870246413e-07,
+      "loss": 1.5669,
+      "step": 5136
+    },
+    {
+      "epoch": 0.3579416785701843,
+      "grad_norm": 0.8847985010981289,
+      "learning_rate": 5.209495861313073e-07,
+      "loss": 1.6315,
+      "step": 5137
+    },
+    {
+      "epoch": 0.3580113576978016,
+      "grad_norm": 0.6932795155767438,
+      "learning_rate": 5.208873787229901e-07,
+      "loss": 1.5393,
+      "step": 5138
+    },
+    {
+      "epoch": 0.3580810368254189,
+      "grad_norm": 0.7381773345843451,
+      "learning_rate": 5.208251648026706e-07,
+      "loss": 1.4413,
+      "step": 5139
+    },
+    {
+      "epoch": 0.35815071595303627,
+      "grad_norm": 0.7475210673169992,
+      "learning_rate": 5.207629443733305e-07,
+      "loss": 1.5271,
+      "step": 5140
+    },
+    {
+      "epoch": 0.35822039508065356,
+      "grad_norm": 0.6821769502467249,
+      "learning_rate": 5.207007174379512e-07,
+      "loss": 1.3818,
+      "step": 5141
+    },
+    {
+      "epoch": 0.3582900742082709,
+      "grad_norm": 0.7540507964274135,
+      "learning_rate": 5.206384839995151e-07,
+      "loss": 1.6267,
+      "step": 5142
+    },
+    {
+      "epoch": 0.3583597533358882,
+      "grad_norm": 0.7189622914498229,
+      "learning_rate": 5.205762440610043e-07,
+      "loss": 1.465,
+      "step": 5143
+    },
+    {
+      "epoch": 0.35842943246350556,
+      "grad_norm": 0.7217356184482514,
+      "learning_rate": 5.205139976254017e-07,
+      "loss": 1.3787,
+      "step": 5144
+    },
+    {
+      "epoch": 0.35849911159112285,
+      "grad_norm": 0.727465508665904,
+      "learning_rate": 5.204517446956899e-07,
+      "loss": 1.5033,
+      "step": 5145
+    },
+    {
+      "epoch": 0.3585687907187402,
+      "grad_norm": 0.7359901406990895,
+      "learning_rate": 5.203894852748525e-07,
+      "loss": 1.593,
+      "step": 5146
+    },
+    {
+      "epoch": 0.3586384698463575,
+      "grad_norm": 0.7027153744924136,
+      "learning_rate": 5.203272193658731e-07,
+      "loss": 1.5325,
+      "step": 5147
+    },
+    {
+      "epoch": 0.35870814897397485,
+      "grad_norm": 0.7257847487940218,
+      "learning_rate": 5.202649469717355e-07,
+      "loss": 1.6293,
+      "step": 5148
+    },
+    {
+      "epoch": 0.35877782810159214,
+      "grad_norm": 0.7370632429557675,
+      "learning_rate": 5.202026680954239e-07,
+      "loss": 1.5385,
+      "step": 5149
+    },
+    {
+      "epoch": 0.3588475072292095,
+      "grad_norm": 0.6892655930676715,
+      "learning_rate": 5.201403827399229e-07,
+      "loss": 1.4966,
+      "step": 5150
+    },
+    {
+      "epoch": 0.3589171863568268,
+      "grad_norm": 0.6988742124124329,
+      "learning_rate": 5.200780909082172e-07,
+      "loss": 1.4691,
+      "step": 5151
+    },
+    {
+      "epoch": 0.35898686548444414,
+      "grad_norm": 0.760811010008508,
+      "learning_rate": 5.200157926032923e-07,
+      "loss": 1.6241,
+      "step": 5152
+    },
+    {
+      "epoch": 0.35905654461206143,
+      "grad_norm": 0.7212598757194405,
+      "learning_rate": 5.199534878281334e-07,
+      "loss": 1.5798,
+      "step": 5153
+    },
+    {
+      "epoch": 0.3591262237396788,
+      "grad_norm": 0.7060841907531608,
+      "learning_rate": 5.198911765857262e-07,
+      "loss": 1.5388,
+      "step": 5154
+    },
+    {
+      "epoch": 0.3591959028672961,
+      "grad_norm": 0.7837864246835402,
+      "learning_rate": 5.198288588790569e-07,
+      "loss": 1.5413,
+      "step": 5155
+    },
+    {
+      "epoch": 0.35926558199491343,
+      "grad_norm": 0.7216877555907196,
+      "learning_rate": 5.197665347111119e-07,
+      "loss": 1.5262,
+      "step": 5156
+    },
+    {
+      "epoch": 0.3593352611225307,
+      "grad_norm": 0.804075556812053,
+      "learning_rate": 5.19704204084878e-07,
+      "loss": 1.5869,
+      "step": 5157
+    },
+    {
+      "epoch": 0.3594049402501481,
+      "grad_norm": 0.7217025913661967,
+      "learning_rate": 5.19641867003342e-07,
+      "loss": 1.5733,
+      "step": 5158
+    },
+    {
+      "epoch": 0.35947461937776537,
+      "grad_norm": 0.6582452744365979,
+      "learning_rate": 5.195795234694912e-07,
+      "loss": 1.6061,
+      "step": 5159
+    },
+    {
+      "epoch": 0.3595442985053827,
+      "grad_norm": 0.7737437639055407,
+      "learning_rate": 5.195171734863135e-07,
+      "loss": 1.5443,
+      "step": 5160
+    },
+    {
+      "epoch": 0.359613977633,
+      "grad_norm": 0.7250214544500239,
+      "learning_rate": 5.194548170567967e-07,
+      "loss": 1.5808,
+      "step": 5161
+    },
+    {
+      "epoch": 0.35968365676061737,
+      "grad_norm": 0.6857607962575639,
+      "learning_rate": 5.19392454183929e-07,
+      "loss": 1.4723,
+      "step": 5162
+    },
+    {
+      "epoch": 0.35975333588823466,
+      "grad_norm": 0.6663403161373356,
+      "learning_rate": 5.19330084870699e-07,
+      "loss": 1.4604,
+      "step": 5163
+    },
+    {
+      "epoch": 0.359823015015852,
+      "grad_norm": 0.7785922329800811,
+      "learning_rate": 5.192677091200955e-07,
+      "loss": 1.5168,
+      "step": 5164
+    },
+    {
+      "epoch": 0.3598926941434693,
+      "grad_norm": 0.69477412998931,
+      "learning_rate": 5.192053269351078e-07,
+      "loss": 1.5047,
+      "step": 5165
+    },
+    {
+      "epoch": 0.35996237327108666,
+      "grad_norm": 0.7145866167341475,
+      "learning_rate": 5.191429383187252e-07,
+      "loss": 1.7196,
+      "step": 5166
+    },
+    {
+      "epoch": 0.36003205239870395,
+      "grad_norm": 0.8216013176843149,
+      "learning_rate": 5.190805432739378e-07,
+      "loss": 1.6393,
+      "step": 5167
+    },
+    {
+      "epoch": 0.3601017315263213,
+      "grad_norm": 0.7794476392517875,
+      "learning_rate": 5.190181418037354e-07,
+      "loss": 1.4782,
+      "step": 5168
+    },
+    {
+      "epoch": 0.3601714106539386,
+      "grad_norm": 0.6724797600349622,
+      "learning_rate": 5.189557339111084e-07,
+      "loss": 1.4706,
+      "step": 5169
+    },
+    {
+      "epoch": 0.36024108978155595,
+      "grad_norm": 0.7334712166799854,
+      "learning_rate": 5.188933195990478e-07,
+      "loss": 1.6242,
+      "step": 5170
+    },
+    {
+      "epoch": 0.36031076890917324,
+      "grad_norm": 0.7208270609468629,
+      "learning_rate": 5.188308988705443e-07,
+      "loss": 1.5976,
+      "step": 5171
+    },
+    {
+      "epoch": 0.3603804480367906,
+      "grad_norm": 0.8019105316309917,
+      "learning_rate": 5.187684717285897e-07,
+      "loss": 1.5433,
+      "step": 5172
+    },
+    {
+      "epoch": 0.3604501271644079,
+      "grad_norm": 0.7227317280229342,
+      "learning_rate": 5.187060381761749e-07,
+      "loss": 1.6401,
+      "step": 5173
+    },
+    {
+      "epoch": 0.36051980629202524,
+      "grad_norm": 0.7261853178782461,
+      "learning_rate": 5.186435982162924e-07,
+      "loss": 1.6069,
+      "step": 5174
+    },
+    {
+      "epoch": 0.36058948541964253,
+      "grad_norm": 0.731713234324846,
+      "learning_rate": 5.185811518519344e-07,
+      "loss": 1.4512,
+      "step": 5175
+    },
+    {
+      "epoch": 0.3606591645472599,
+      "grad_norm": 0.6683510801178882,
+      "learning_rate": 5.185186990860932e-07,
+      "loss": 1.4891,
+      "step": 5176
+    },
+    {
+      "epoch": 0.3607288436748772,
+      "grad_norm": 0.7496938877958745,
+      "learning_rate": 5.184562399217621e-07,
+      "loss": 1.4466,
+      "step": 5177
+    },
+    {
+      "epoch": 0.3607985228024945,
+      "grad_norm": 0.733168705759652,
+      "learning_rate": 5.183937743619337e-07,
+      "loss": 1.5857,
+      "step": 5178
+    },
+    {
+      "epoch": 0.3608682019301118,
+      "grad_norm": 0.7430949705009791,
+      "learning_rate": 5.18331302409602e-07,
+      "loss": 1.5315,
+      "step": 5179
+    },
+    {
+      "epoch": 0.36093788105772917,
+      "grad_norm": 0.7773572657707958,
+      "learning_rate": 5.182688240677605e-07,
+      "loss": 1.5201,
+      "step": 5180
+    },
+    {
+      "epoch": 0.36100756018534647,
+      "grad_norm": 0.7536228512940022,
+      "learning_rate": 5.182063393394033e-07,
+      "loss": 1.5811,
+      "step": 5181
+    },
+    {
+      "epoch": 0.3610772393129638,
+      "grad_norm": 0.6928947809986236,
+      "learning_rate": 5.181438482275249e-07,
+      "loss": 1.4702,
+      "step": 5182
+    },
+    {
+      "epoch": 0.3611469184405811,
+      "grad_norm": 0.7192975853115967,
+      "learning_rate": 5.1808135073512e-07,
+      "loss": 1.451,
+      "step": 5183
+    },
+    {
+      "epoch": 0.36121659756819846,
+      "grad_norm": 0.693654126443667,
+      "learning_rate": 5.180188468651835e-07,
+      "loss": 1.4676,
+      "step": 5184
+    },
+    {
+      "epoch": 0.36128627669581576,
+      "grad_norm": 0.7256757119654305,
+      "learning_rate": 5.179563366207108e-07,
+      "loss": 1.6045,
+      "step": 5185
+    },
+    {
+      "epoch": 0.3613559558234331,
+      "grad_norm": 0.6875294331327294,
+      "learning_rate": 5.178938200046974e-07,
+      "loss": 1.513,
+      "step": 5186
+    },
+    {
+      "epoch": 0.3614256349510504,
+      "grad_norm": 0.739944485913754,
+      "learning_rate": 5.178312970201394e-07,
+      "loss": 1.4669,
+      "step": 5187
+    },
+    {
+      "epoch": 0.36149531407866775,
+      "grad_norm": 0.6832797355399127,
+      "learning_rate": 5.17768767670033e-07,
+      "loss": 1.4283,
+      "step": 5188
+    },
+    {
+      "epoch": 0.36156499320628505,
+      "grad_norm": 0.6876615223406541,
+      "learning_rate": 5.177062319573746e-07,
+      "loss": 1.5305,
+      "step": 5189
+    },
+    {
+      "epoch": 0.3616346723339024,
+      "grad_norm": 0.701091374244738,
+      "learning_rate": 5.176436898851611e-07,
+      "loss": 1.4617,
+      "step": 5190
+    },
+    {
+      "epoch": 0.3617043514615197,
+      "grad_norm": 0.723571517664063,
+      "learning_rate": 5.175811414563897e-07,
+      "loss": 1.5976,
+      "step": 5191
+    },
+    {
+      "epoch": 0.36177403058913704,
+      "grad_norm": 0.7259446504081081,
+      "learning_rate": 5.17518586674058e-07,
+      "loss": 1.6222,
+      "step": 5192
+    },
+    {
+      "epoch": 0.36184370971675434,
+      "grad_norm": 0.7117002981964684,
+      "learning_rate": 5.174560255411634e-07,
+      "loss": 1.5056,
+      "step": 5193
+    },
+    {
+      "epoch": 0.3619133888443717,
+      "grad_norm": 0.7279042262192607,
+      "learning_rate": 5.173934580607041e-07,
+      "loss": 1.6258,
+      "step": 5194
+    },
+    {
+      "epoch": 0.361983067971989,
+      "grad_norm": 0.7104886630481225,
+      "learning_rate": 5.173308842356785e-07,
+      "loss": 1.5851,
+      "step": 5195
+    },
+    {
+      "epoch": 0.36205274709960633,
+      "grad_norm": 0.6871358125170827,
+      "learning_rate": 5.172683040690853e-07,
+      "loss": 1.5191,
+      "step": 5196
+    },
+    {
+      "epoch": 0.3621224262272236,
+      "grad_norm": 0.7192980784986389,
+      "learning_rate": 5.172057175639234e-07,
+      "loss": 1.6021,
+      "step": 5197
+    },
+    {
+      "epoch": 0.362192105354841,
+      "grad_norm": 0.6654693323072031,
+      "learning_rate": 5.171431247231921e-07,
+      "loss": 1.5323,
+      "step": 5198
+    },
+    {
+      "epoch": 0.36226178448245827,
+      "grad_norm": 0.7188297136953314,
+      "learning_rate": 5.17080525549891e-07,
+      "loss": 1.6149,
+      "step": 5199
+    },
+    {
+      "epoch": 0.3623314636100756,
+      "grad_norm": 0.7728183408864958,
+      "learning_rate": 5.170179200470199e-07,
+      "loss": 1.6164,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3624011427376929,
+      "grad_norm": 0.7382453104563416,
+      "learning_rate": 5.169553082175792e-07,
+      "loss": 1.5468,
+      "step": 5201
+    },
+    {
+      "epoch": 0.36247082186531027,
+      "grad_norm": 0.7664313372593619,
+      "learning_rate": 5.16892690064569e-07,
+      "loss": 1.5538,
+      "step": 5202
+    },
+    {
+      "epoch": 0.36254050099292756,
+      "grad_norm": 0.7403909821642547,
+      "learning_rate": 5.168300655909905e-07,
+      "loss": 1.5414,
+      "step": 5203
+    },
+    {
+      "epoch": 0.3626101801205449,
+      "grad_norm": 0.7610636702682715,
+      "learning_rate": 5.167674347998446e-07,
+      "loss": 1.6746,
+      "step": 5204
+    },
+    {
+      "epoch": 0.3626798592481622,
+      "grad_norm": 0.7008708794040567,
+      "learning_rate": 5.167047976941327e-07,
+      "loss": 1.4871,
+      "step": 5205
+    },
+    {
+      "epoch": 0.36274953837577956,
+      "grad_norm": 0.6827672523246823,
+      "learning_rate": 5.166421542768564e-07,
+      "loss": 1.5816,
+      "step": 5206
+    },
+    {
+      "epoch": 0.36281921750339685,
+      "grad_norm": 0.7860676509331375,
+      "learning_rate": 5.165795045510179e-07,
+      "loss": 1.5773,
+      "step": 5207
+    },
+    {
+      "epoch": 0.3628888966310142,
+      "grad_norm": 0.7178660737374489,
+      "learning_rate": 5.165168485196194e-07,
+      "loss": 1.466,
+      "step": 5208
+    },
+    {
+      "epoch": 0.3629585757586315,
+      "grad_norm": 0.7093634056240848,
+      "learning_rate": 5.164541861856636e-07,
+      "loss": 1.3871,
+      "step": 5209
+    },
+    {
+      "epoch": 0.36302825488624885,
+      "grad_norm": 0.6972079077503902,
+      "learning_rate": 5.163915175521532e-07,
+      "loss": 1.449,
+      "step": 5210
+    },
+    {
+      "epoch": 0.36309793401386614,
+      "grad_norm": 0.747745907435438,
+      "learning_rate": 5.163288426220918e-07,
+      "loss": 1.5072,
+      "step": 5211
+    },
+    {
+      "epoch": 0.3631676131414835,
+      "grad_norm": 0.733427906317132,
+      "learning_rate": 5.162661613984823e-07,
+      "loss": 1.5162,
+      "step": 5212
+    },
+    {
+      "epoch": 0.3632372922691008,
+      "grad_norm": 0.7233838960156477,
+      "learning_rate": 5.162034738843291e-07,
+      "loss": 1.66,
+      "step": 5213
+    },
+    {
+      "epoch": 0.36330697139671814,
+      "grad_norm": 0.7357691997602305,
+      "learning_rate": 5.161407800826359e-07,
+      "loss": 1.6048,
+      "step": 5214
+    },
+    {
+      "epoch": 0.36337665052433543,
+      "grad_norm": 0.7419119044198975,
+      "learning_rate": 5.160780799964074e-07,
+      "loss": 1.5819,
+      "step": 5215
+    },
+    {
+      "epoch": 0.3634463296519528,
+      "grad_norm": 0.7333257219084258,
+      "learning_rate": 5.160153736286481e-07,
+      "loss": 1.5136,
+      "step": 5216
+    },
+    {
+      "epoch": 0.3635160087795701,
+      "grad_norm": 0.7561954088937832,
+      "learning_rate": 5.159526609823633e-07,
+      "loss": 1.532,
+      "step": 5217
+    },
+    {
+      "epoch": 0.3635856879071874,
+      "grad_norm": 0.7266071236575034,
+      "learning_rate": 5.15889942060558e-07,
+      "loss": 1.6245,
+      "step": 5218
+    },
+    {
+      "epoch": 0.3636553670348047,
+      "grad_norm": 0.7365917391566773,
+      "learning_rate": 5.15827216866238e-07,
+      "loss": 1.6322,
+      "step": 5219
+    },
+    {
+      "epoch": 0.36372504616242207,
+      "grad_norm": 0.7040763190740109,
+      "learning_rate": 5.157644854024093e-07,
+      "loss": 1.49,
+      "step": 5220
+    },
+    {
+      "epoch": 0.36379472529003937,
+      "grad_norm": 0.7142262462047746,
+      "learning_rate": 5.15701747672078e-07,
+      "loss": 1.5597,
+      "step": 5221
+    },
+    {
+      "epoch": 0.3638644044176567,
+      "grad_norm": 0.6858746531213153,
+      "learning_rate": 5.156390036782504e-07,
+      "loss": 1.5639,
+      "step": 5222
+    },
+    {
+      "epoch": 0.363934083545274,
+      "grad_norm": 0.6759937525398312,
+      "learning_rate": 5.155762534239337e-07,
+      "loss": 1.5822,
+      "step": 5223
+    },
+    {
+      "epoch": 0.36400376267289136,
+      "grad_norm": 0.7509133714833661,
+      "learning_rate": 5.155134969121349e-07,
+      "loss": 1.5394,
+      "step": 5224
+    },
+    {
+      "epoch": 0.36407344180050866,
+      "grad_norm": 0.7170835709751526,
+      "learning_rate": 5.154507341458613e-07,
+      "loss": 1.5826,
+      "step": 5225
+    },
+    {
+      "epoch": 0.364143120928126,
+      "grad_norm": 0.7065609504484022,
+      "learning_rate": 5.153879651281208e-07,
+      "loss": 1.5208,
+      "step": 5226
+    },
+    {
+      "epoch": 0.3642128000557433,
+      "grad_norm": 0.6860528121460813,
+      "learning_rate": 5.153251898619212e-07,
+      "loss": 1.5345,
+      "step": 5227
+    },
+    {
+      "epoch": 0.3642824791833606,
+      "grad_norm": 0.8400742404674034,
+      "learning_rate": 5.152624083502711e-07,
+      "loss": 1.4601,
+      "step": 5228
+    },
+    {
+      "epoch": 0.36435215831097795,
+      "grad_norm": 0.7068731002179544,
+      "learning_rate": 5.151996205961789e-07,
+      "loss": 1.5634,
+      "step": 5229
+    },
+    {
+      "epoch": 0.36442183743859524,
+      "grad_norm": 0.682862701091438,
+      "learning_rate": 5.151368266026535e-07,
+      "loss": 1.5034,
+      "step": 5230
+    },
+    {
+      "epoch": 0.3644915165662126,
+      "grad_norm": 0.740431488301115,
+      "learning_rate": 5.150740263727043e-07,
+      "loss": 1.6257,
+      "step": 5231
+    },
+    {
+      "epoch": 0.3645611956938299,
+      "grad_norm": 0.6693329964335967,
+      "learning_rate": 5.150112199093407e-07,
+      "loss": 1.4643,
+      "step": 5232
+    },
+    {
+      "epoch": 0.36463087482144724,
+      "grad_norm": 0.7470365567305248,
+      "learning_rate": 5.149484072155725e-07,
+      "loss": 1.5566,
+      "step": 5233
+    },
+    {
+      "epoch": 0.36470055394906453,
+      "grad_norm": 0.7473497564826113,
+      "learning_rate": 5.1488558829441e-07,
+      "loss": 1.4871,
+      "step": 5234
+    },
+    {
+      "epoch": 0.3647702330766819,
+      "grad_norm": 0.6911032507873867,
+      "learning_rate": 5.148227631488632e-07,
+      "loss": 1.4324,
+      "step": 5235
+    },
+    {
+      "epoch": 0.3648399122042992,
+      "grad_norm": 0.6816650784423868,
+      "learning_rate": 5.147599317819434e-07,
+      "loss": 1.5158,
+      "step": 5236
+    },
+    {
+      "epoch": 0.3649095913319165,
+      "grad_norm": 0.7410861631268022,
+      "learning_rate": 5.14697094196661e-07,
+      "loss": 1.5026,
+      "step": 5237
+    },
+    {
+      "epoch": 0.3649792704595338,
+      "grad_norm": 0.6936484383081856,
+      "learning_rate": 5.146342503960276e-07,
+      "loss": 1.5705,
+      "step": 5238
+    },
+    {
+      "epoch": 0.3650489495871512,
+      "grad_norm": 0.6786652114843855,
+      "learning_rate": 5.14571400383055e-07,
+      "loss": 1.5508,
+      "step": 5239
+    },
+    {
+      "epoch": 0.36511862871476847,
+      "grad_norm": 0.7273301299834638,
+      "learning_rate": 5.145085441607548e-07,
+      "loss": 1.4541,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3651883078423858,
+      "grad_norm": 0.7934277087687172,
+      "learning_rate": 5.144456817321391e-07,
+      "loss": 1.5811,
+      "step": 5241
+    },
+    {
+      "epoch": 0.3652579869700031,
+      "grad_norm": 0.7129202993259002,
+      "learning_rate": 5.143828131002207e-07,
+      "loss": 1.6266,
+      "step": 5242
+    },
+    {
+      "epoch": 0.36532766609762046,
+      "grad_norm": 0.662891493603992,
+      "learning_rate": 5.143199382680122e-07,
+      "loss": 1.4635,
+      "step": 5243
+    },
+    {
+      "epoch": 0.36539734522523776,
+      "grad_norm": 0.7154475325660717,
+      "learning_rate": 5.14257057238527e-07,
+      "loss": 1.5247,
+      "step": 5244
+    },
+    {
+      "epoch": 0.3654670243528551,
+      "grad_norm": 0.7066260380609907,
+      "learning_rate": 5.141941700147782e-07,
+      "loss": 1.5311,
+      "step": 5245
+    },
+    {
+      "epoch": 0.3655367034804724,
+      "grad_norm": 0.6989214833213997,
+      "learning_rate": 5.141312765997793e-07,
+      "loss": 1.2925,
+      "step": 5246
+    },
+    {
+      "epoch": 0.36560638260808975,
+      "grad_norm": 0.7044441524364802,
+      "learning_rate": 5.140683769965448e-07,
+      "loss": 1.61,
+      "step": 5247
+    },
+    {
+      "epoch": 0.36567606173570705,
+      "grad_norm": 0.6607704920750958,
+      "learning_rate": 5.140054712080886e-07,
+      "loss": 1.444,
+      "step": 5248
+    },
+    {
+      "epoch": 0.3657457408633244,
+      "grad_norm": 0.7357333603325759,
+      "learning_rate": 5.139425592374253e-07,
+      "loss": 1.5409,
+      "step": 5249
+    },
+    {
+      "epoch": 0.3658154199909417,
+      "grad_norm": 0.6514266738332384,
+      "learning_rate": 5.138796410875699e-07,
+      "loss": 1.4404,
+      "step": 5250
+    },
+    {
+      "epoch": 0.36588509911855904,
+      "grad_norm": 0.6992053094386647,
+      "learning_rate": 5.138167167615376e-07,
+      "loss": 1.5441,
+      "step": 5251
+    },
+    {
+      "epoch": 0.36595477824617634,
+      "grad_norm": 0.7186567293557066,
+      "learning_rate": 5.137537862623437e-07,
+      "loss": 1.5629,
+      "step": 5252
+    },
+    {
+      "epoch": 0.3660244573737937,
+      "grad_norm": 0.7064096116081595,
+      "learning_rate": 5.136908495930041e-07,
+      "loss": 1.353,
+      "step": 5253
+    },
+    {
+      "epoch": 0.366094136501411,
+      "grad_norm": 0.694342282292615,
+      "learning_rate": 5.136279067565347e-07,
+      "loss": 1.6093,
+      "step": 5254
+    },
+    {
+      "epoch": 0.36616381562902833,
+      "grad_norm": 0.695106825117824,
+      "learning_rate": 5.135649577559519e-07,
+      "loss": 1.53,
+      "step": 5255
+    },
+    {
+      "epoch": 0.36623349475664563,
+      "grad_norm": 0.6721407293211693,
+      "learning_rate": 5.135020025942726e-07,
+      "loss": 1.5514,
+      "step": 5256
+    },
+    {
+      "epoch": 0.366303173884263,
+      "grad_norm": 0.7433143801168884,
+      "learning_rate": 5.134390412745134e-07,
+      "loss": 1.6664,
+      "step": 5257
+    },
+    {
+      "epoch": 0.3663728530118803,
+      "grad_norm": 0.7759049168554641,
+      "learning_rate": 5.133760737996916e-07,
+      "loss": 1.6917,
+      "step": 5258
+    },
+    {
+      "epoch": 0.3664425321394976,
+      "grad_norm": 0.7287911458113379,
+      "learning_rate": 5.133131001728248e-07,
+      "loss": 1.6547,
+      "step": 5259
+    },
+    {
+      "epoch": 0.3665122112671149,
+      "grad_norm": 0.7351234554888396,
+      "learning_rate": 5.132501203969309e-07,
+      "loss": 1.5529,
+      "step": 5260
+    },
+    {
+      "epoch": 0.36658189039473227,
+      "grad_norm": 0.7453393320312598,
+      "learning_rate": 5.131871344750279e-07,
+      "loss": 1.5328,
+      "step": 5261
+    },
+    {
+      "epoch": 0.36665156952234956,
+      "grad_norm": 0.6918766579616535,
+      "learning_rate": 5.131241424101342e-07,
+      "loss": 1.5461,
+      "step": 5262
+    },
+    {
+      "epoch": 0.3667212486499669,
+      "grad_norm": 0.7194318357060998,
+      "learning_rate": 5.130611442052686e-07,
+      "loss": 1.6265,
+      "step": 5263
+    },
+    {
+      "epoch": 0.3667909277775842,
+      "grad_norm": 0.7300076987686653,
+      "learning_rate": 5.129981398634502e-07,
+      "loss": 1.5558,
+      "step": 5264
+    },
+    {
+      "epoch": 0.36686060690520156,
+      "grad_norm": 0.7009279205137522,
+      "learning_rate": 5.129351293876979e-07,
+      "loss": 1.4367,
+      "step": 5265
+    },
+    {
+      "epoch": 0.36693028603281885,
+      "grad_norm": 0.8297625632738989,
+      "learning_rate": 5.128721127810318e-07,
+      "loss": 1.7773,
+      "step": 5266
+    },
+    {
+      "epoch": 0.3669999651604362,
+      "grad_norm": 0.6939962844924814,
+      "learning_rate": 5.128090900464713e-07,
+      "loss": 1.5233,
+      "step": 5267
+    },
+    {
+      "epoch": 0.3670696442880535,
+      "grad_norm": 0.7366202778643759,
+      "learning_rate": 5.12746061187037e-07,
+      "loss": 1.6129,
+      "step": 5268
+    },
+    {
+      "epoch": 0.36713932341567085,
+      "grad_norm": 0.7558579968863058,
+      "learning_rate": 5.126830262057491e-07,
+      "loss": 1.5278,
+      "step": 5269
+    },
+    {
+      "epoch": 0.36720900254328814,
+      "grad_norm": 0.718857236429046,
+      "learning_rate": 5.126199851056285e-07,
+      "loss": 1.4735,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3672786816709055,
+      "grad_norm": 0.7972347547964148,
+      "learning_rate": 5.125569378896962e-07,
+      "loss": 1.6827,
+      "step": 5271
+    },
+    {
+      "epoch": 0.3673483607985228,
+      "grad_norm": 0.6899496063426461,
+      "learning_rate": 5.124938845609736e-07,
+      "loss": 1.58,
+      "step": 5272
+    },
+    {
+      "epoch": 0.36741803992614014,
+      "grad_norm": 0.7818174794056979,
+      "learning_rate": 5.124308251224824e-07,
+      "loss": 1.5645,
+      "step": 5273
+    },
+    {
+      "epoch": 0.36748771905375743,
+      "grad_norm": 0.7124659133872496,
+      "learning_rate": 5.123677595772443e-07,
+      "loss": 1.5349,
+      "step": 5274
+    },
+    {
+      "epoch": 0.3675573981813748,
+      "grad_norm": 0.7511031303630104,
+      "learning_rate": 5.123046879282817e-07,
+      "loss": 1.5803,
+      "step": 5275
+    },
+    {
+      "epoch": 0.3676270773089921,
+      "grad_norm": 0.6851223267795419,
+      "learning_rate": 5.122416101786171e-07,
+      "loss": 1.5072,
+      "step": 5276
+    },
+    {
+      "epoch": 0.36769675643660943,
+      "grad_norm": 0.806850869793742,
+      "learning_rate": 5.121785263312734e-07,
+      "loss": 1.6434,
+      "step": 5277
+    },
+    {
+      "epoch": 0.3677664355642267,
+      "grad_norm": 0.7254644946888305,
+      "learning_rate": 5.121154363892735e-07,
+      "loss": 1.5579,
+      "step": 5278
+    },
+    {
+      "epoch": 0.3678361146918441,
+      "grad_norm": 0.7252493972658083,
+      "learning_rate": 5.12052340355641e-07,
+      "loss": 1.5835,
+      "step": 5279
+    },
+    {
+      "epoch": 0.36790579381946137,
+      "grad_norm": 0.7662299719465424,
+      "learning_rate": 5.119892382333996e-07,
+      "loss": 1.5148,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3679754729470787,
+      "grad_norm": 0.7165871659616716,
+      "learning_rate": 5.119261300255731e-07,
+      "loss": 1.5447,
+      "step": 5281
+    },
+    {
+      "epoch": 0.368045152074696,
+      "grad_norm": 0.6639950924743402,
+      "learning_rate": 5.118630157351859e-07,
+      "loss": 1.554,
+      "step": 5282
+    },
+    {
+      "epoch": 0.36811483120231336,
+      "grad_norm": 0.7302210295216659,
+      "learning_rate": 5.117998953652625e-07,
+      "loss": 1.5355,
+      "step": 5283
+    },
+    {
+      "epoch": 0.36818451032993066,
+      "grad_norm": 0.7491105839545834,
+      "learning_rate": 5.117367689188279e-07,
+      "loss": 1.6475,
+      "step": 5284
+    },
+    {
+      "epoch": 0.368254189457548,
+      "grad_norm": 0.7506643071993163,
+      "learning_rate": 5.11673636398907e-07,
+      "loss": 1.6575,
+      "step": 5285
+    },
+    {
+      "epoch": 0.3683238685851653,
+      "grad_norm": 0.6812138419849934,
+      "learning_rate": 5.116104978085254e-07,
+      "loss": 1.4553,
+      "step": 5286
+    },
+    {
+      "epoch": 0.36839354771278265,
+      "grad_norm": 0.7854900925563119,
+      "learning_rate": 5.11547353150709e-07,
+      "loss": 1.5519,
+      "step": 5287
+    },
+    {
+      "epoch": 0.36846322684039995,
+      "grad_norm": 0.7203783971396994,
+      "learning_rate": 5.114842024284834e-07,
+      "loss": 1.6115,
+      "step": 5288
+    },
+    {
+      "epoch": 0.3685329059680173,
+      "grad_norm": 0.7243845386114148,
+      "learning_rate": 5.114210456448753e-07,
+      "loss": 1.5813,
+      "step": 5289
+    },
+    {
+      "epoch": 0.3686025850956346,
+      "grad_norm": 0.713318574120179,
+      "learning_rate": 5.113578828029111e-07,
+      "loss": 1.5824,
+      "step": 5290
+    },
+    {
+      "epoch": 0.36867226422325194,
+      "grad_norm": 0.7536484451835456,
+      "learning_rate": 5.112947139056177e-07,
+      "loss": 1.5102,
+      "step": 5291
+    },
+    {
+      "epoch": 0.36874194335086924,
+      "grad_norm": 0.7081609936285251,
+      "learning_rate": 5.112315389560226e-07,
+      "loss": 1.4296,
+      "step": 5292
+    },
+    {
+      "epoch": 0.3688116224784866,
+      "grad_norm": 0.6799279862368306,
+      "learning_rate": 5.111683579571528e-07,
+      "loss": 1.5083,
+      "step": 5293
+    },
+    {
+      "epoch": 0.3688813016061039,
+      "grad_norm": 0.890442651971021,
+      "learning_rate": 5.111051709120361e-07,
+      "loss": 1.6717,
+      "step": 5294
+    },
+    {
+      "epoch": 0.36895098073372123,
+      "grad_norm": 0.6787398243474793,
+      "learning_rate": 5.11041977823701e-07,
+      "loss": 1.5357,
+      "step": 5295
+    },
+    {
+      "epoch": 0.36902065986133853,
+      "grad_norm": 0.6934921564234446,
+      "learning_rate": 5.109787786951755e-07,
+      "loss": 1.5096,
+      "step": 5296
+    },
+    {
+      "epoch": 0.3690903389889559,
+      "grad_norm": 0.6968589312259752,
+      "learning_rate": 5.109155735294882e-07,
+      "loss": 1.4032,
+      "step": 5297
+    },
+    {
+      "epoch": 0.3691600181165732,
+      "grad_norm": 0.819070605787555,
+      "learning_rate": 5.108523623296682e-07,
+      "loss": 1.5379,
+      "step": 5298
+    },
+    {
+      "epoch": 0.3692296972441905,
+      "grad_norm": 0.7192037379184335,
+      "learning_rate": 5.107891450987445e-07,
+      "loss": 1.5684,
+      "step": 5299
+    },
+    {
+      "epoch": 0.3692993763718078,
+      "grad_norm": 0.7157852409802509,
+      "learning_rate": 5.107259218397469e-07,
+      "loss": 1.5831,
+      "step": 5300
+    },
+    {
+      "epoch": 0.36936905549942517,
+      "grad_norm": 0.7492510878850005,
+      "learning_rate": 5.10662692555705e-07,
+      "loss": 1.5437,
+      "step": 5301
+    },
+    {
+      "epoch": 0.36943873462704246,
+      "grad_norm": 0.7286317846247485,
+      "learning_rate": 5.105994572496488e-07,
+      "loss": 1.5993,
+      "step": 5302
+    },
+    {
+      "epoch": 0.3695084137546598,
+      "grad_norm": 0.7648274962276388,
+      "learning_rate": 5.105362159246089e-07,
+      "loss": 1.6396,
+      "step": 5303
+    },
+    {
+      "epoch": 0.3695780928822771,
+      "grad_norm": 0.7420079599949096,
+      "learning_rate": 5.104729685836156e-07,
+      "loss": 1.6214,
+      "step": 5304
+    },
+    {
+      "epoch": 0.36964777200989446,
+      "grad_norm": 0.692661841317161,
+      "learning_rate": 5.104097152297002e-07,
+      "loss": 1.502,
+      "step": 5305
+    },
+    {
+      "epoch": 0.36971745113751175,
+      "grad_norm": 0.6671121171209912,
+      "learning_rate": 5.103464558658937e-07,
+      "loss": 1.484,
+      "step": 5306
+    },
+    {
+      "epoch": 0.3697871302651291,
+      "grad_norm": 0.698266599735423,
+      "learning_rate": 5.102831904952278e-07,
+      "loss": 1.5238,
+      "step": 5307
+    },
+    {
+      "epoch": 0.3698568093927464,
+      "grad_norm": 0.721316187644341,
+      "learning_rate": 5.102199191207341e-07,
+      "loss": 1.5747,
+      "step": 5308
+    },
+    {
+      "epoch": 0.36992648852036375,
+      "grad_norm": 0.7555051596431045,
+      "learning_rate": 5.10156641745445e-07,
+      "loss": 1.4998,
+      "step": 5309
+    },
+    {
+      "epoch": 0.36999616764798104,
+      "grad_norm": 0.7378153963986179,
+      "learning_rate": 5.100933583723924e-07,
+      "loss": 1.5894,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3700658467755984,
+      "grad_norm": 0.7299552421467299,
+      "learning_rate": 5.100300690046095e-07,
+      "loss": 1.4737,
+      "step": 5311
+    },
+    {
+      "epoch": 0.3701355259032157,
+      "grad_norm": 0.7222764229223232,
+      "learning_rate": 5.099667736451289e-07,
+      "loss": 1.534,
+      "step": 5312
+    },
+    {
+      "epoch": 0.37020520503083304,
+      "grad_norm": 0.6816801189726973,
+      "learning_rate": 5.099034722969839e-07,
+      "loss": 1.473,
+      "step": 5313
+    },
+    {
+      "epoch": 0.37027488415845033,
+      "grad_norm": 0.6548112913225141,
+      "learning_rate": 5.098401649632083e-07,
+      "loss": 1.4501,
+      "step": 5314
+    },
+    {
+      "epoch": 0.3703445632860677,
+      "grad_norm": 0.6909094462608453,
+      "learning_rate": 5.097768516468354e-07,
+      "loss": 1.6325,
+      "step": 5315
+    },
+    {
+      "epoch": 0.370414242413685,
+      "grad_norm": 0.745114108887267,
+      "learning_rate": 5.097135323508998e-07,
+      "loss": 1.4192,
+      "step": 5316
+    },
+    {
+      "epoch": 0.37048392154130233,
+      "grad_norm": 0.7094163660025467,
+      "learning_rate": 5.096502070784358e-07,
+      "loss": 1.5063,
+      "step": 5317
+    },
+    {
+      "epoch": 0.3705536006689196,
+      "grad_norm": 0.6850285597934616,
+      "learning_rate": 5.095868758324777e-07,
+      "loss": 1.5345,
+      "step": 5318
+    },
+    {
+      "epoch": 0.3706232797965369,
+      "grad_norm": 0.689281312430016,
+      "learning_rate": 5.095235386160609e-07,
+      "loss": 1.4471,
+      "step": 5319
+    },
+    {
+      "epoch": 0.37069295892415427,
+      "grad_norm": 0.6915170853407353,
+      "learning_rate": 5.094601954322205e-07,
+      "loss": 1.5534,
+      "step": 5320
+    },
+    {
+      "epoch": 0.37076263805177156,
+      "grad_norm": 0.7047793583439973,
+      "learning_rate": 5.09396846283992e-07,
+      "loss": 1.4784,
+      "step": 5321
+    },
+    {
+      "epoch": 0.3708323171793889,
+      "grad_norm": 0.6663159334094043,
+      "learning_rate": 5.093334911744112e-07,
+      "loss": 1.4606,
+      "step": 5322
+    },
+    {
+      "epoch": 0.3709019963070062,
+      "grad_norm": 0.7150983337368044,
+      "learning_rate": 5.092701301065143e-07,
+      "loss": 1.5738,
+      "step": 5323
+    },
+    {
+      "epoch": 0.37097167543462356,
+      "grad_norm": 0.7746113045884914,
+      "learning_rate": 5.092067630833376e-07,
+      "loss": 1.5458,
+      "step": 5324
+    },
+    {
+      "epoch": 0.37104135456224085,
+      "grad_norm": 0.726557554264369,
+      "learning_rate": 5.091433901079178e-07,
+      "loss": 1.5695,
+      "step": 5325
+    },
+    {
+      "epoch": 0.3711110336898582,
+      "grad_norm": 0.7372150181593171,
+      "learning_rate": 5.09080011183292e-07,
+      "loss": 1.5355,
+      "step": 5326
+    },
+    {
+      "epoch": 0.3711807128174755,
+      "grad_norm": 0.6940705110176282,
+      "learning_rate": 5.090166263124972e-07,
+      "loss": 1.6131,
+      "step": 5327
+    },
+    {
+      "epoch": 0.37125039194509285,
+      "grad_norm": 0.6706521663019233,
+      "learning_rate": 5.089532354985712e-07,
+      "loss": 1.4164,
+      "step": 5328
+    },
+    {
+      "epoch": 0.37132007107271015,
+      "grad_norm": 0.7522207421573274,
+      "learning_rate": 5.088898387445516e-07,
+      "loss": 1.6635,
+      "step": 5329
+    },
+    {
+      "epoch": 0.3713897502003275,
+      "grad_norm": 0.7628608894232416,
+      "learning_rate": 5.088264360534766e-07,
+      "loss": 1.4664,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3714594293279448,
+      "grad_norm": 0.7187722963470405,
+      "learning_rate": 5.087630274283846e-07,
+      "loss": 1.4531,
+      "step": 5331
+    },
+    {
+      "epoch": 0.37152910845556214,
+      "grad_norm": 0.6801481781174705,
+      "learning_rate": 5.086996128723142e-07,
+      "loss": 1.4878,
+      "step": 5332
+    },
+    {
+      "epoch": 0.37159878758317944,
+      "grad_norm": 0.7645501547843065,
+      "learning_rate": 5.086361923883045e-07,
+      "loss": 1.4992,
+      "step": 5333
+    },
+    {
+      "epoch": 0.3716684667107968,
+      "grad_norm": 0.711836763750871,
+      "learning_rate": 5.085727659793944e-07,
+      "loss": 1.4532,
+      "step": 5334
+    },
+    {
+      "epoch": 0.3717381458384141,
+      "grad_norm": 0.712627458827013,
+      "learning_rate": 5.085093336486239e-07,
+      "loss": 1.5743,
+      "step": 5335
+    },
+    {
+      "epoch": 0.37180782496603143,
+      "grad_norm": 0.6921801368603905,
+      "learning_rate": 5.084458953990325e-07,
+      "loss": 1.5584,
+      "step": 5336
+    },
+    {
+      "epoch": 0.3718775040936487,
+      "grad_norm": 0.7227212206111542,
+      "learning_rate": 5.083824512336604e-07,
+      "loss": 1.4999,
+      "step": 5337
+    },
+    {
+      "epoch": 0.3719471832212661,
+      "grad_norm": 0.7828431028231742,
+      "learning_rate": 5.083190011555478e-07,
+      "loss": 1.6068,
+      "step": 5338
+    },
+    {
+      "epoch": 0.37201686234888337,
+      "grad_norm": 0.6628573302677304,
+      "learning_rate": 5.082555451677356e-07,
+      "loss": 1.5878,
+      "step": 5339
+    },
+    {
+      "epoch": 0.3720865414765007,
+      "grad_norm": 0.6802179570076206,
+      "learning_rate": 5.081920832732647e-07,
+      "loss": 1.4912,
+      "step": 5340
+    },
+    {
+      "epoch": 0.372156220604118,
+      "grad_norm": 0.728418405218629,
+      "learning_rate": 5.081286154751763e-07,
+      "loss": 1.5246,
+      "step": 5341
+    },
+    {
+      "epoch": 0.37222589973173537,
+      "grad_norm": 0.697994481641133,
+      "learning_rate": 5.080651417765117e-07,
+      "loss": 1.4452,
+      "step": 5342
+    },
+    {
+      "epoch": 0.37229557885935266,
+      "grad_norm": 0.7055931388578118,
+      "learning_rate": 5.080016621803128e-07,
+      "loss": 1.4946,
+      "step": 5343
+    },
+    {
+      "epoch": 0.37236525798697,
+      "grad_norm": 0.7065869550826593,
+      "learning_rate": 5.079381766896219e-07,
+      "loss": 1.6107,
+      "step": 5344
+    },
+    {
+      "epoch": 0.3724349371145873,
+      "grad_norm": 0.7410511984098631,
+      "learning_rate": 5.078746853074811e-07,
+      "loss": 1.5031,
+      "step": 5345
+    },
+    {
+      "epoch": 0.37250461624220466,
+      "grad_norm": 0.7901975002588718,
+      "learning_rate": 5.078111880369331e-07,
+      "loss": 1.7529,
+      "step": 5346
+    },
+    {
+      "epoch": 0.37257429536982195,
+      "grad_norm": 0.6930274846449783,
+      "learning_rate": 5.07747684881021e-07,
+      "loss": 1.649,
+      "step": 5347
+    },
+    {
+      "epoch": 0.3726439744974393,
+      "grad_norm": 0.8041954663566083,
+      "learning_rate": 5.076841758427877e-07,
+      "loss": 1.6462,
+      "step": 5348
+    },
+    {
+      "epoch": 0.3727136536250566,
+      "grad_norm": 0.7305914510607037,
+      "learning_rate": 5.07620660925277e-07,
+      "loss": 1.4538,
+      "step": 5349
+    },
+    {
+      "epoch": 0.37278333275267395,
+      "grad_norm": 0.7158559747595667,
+      "learning_rate": 5.075571401315321e-07,
+      "loss": 1.4756,
+      "step": 5350
+    },
+    {
+      "epoch": 0.37285301188029124,
+      "grad_norm": 0.7073775288961177,
+      "learning_rate": 5.074936134645978e-07,
+      "loss": 1.5127,
+      "step": 5351
+    },
+    {
+      "epoch": 0.3729226910079086,
+      "grad_norm": 0.7430699639842753,
+      "learning_rate": 5.074300809275181e-07,
+      "loss": 1.5496,
+      "step": 5352
+    },
+    {
+      "epoch": 0.3729923701355259,
+      "grad_norm": 0.7641833104167214,
+      "learning_rate": 5.073665425233374e-07,
+      "loss": 1.4541,
+      "step": 5353
+    },
+    {
+      "epoch": 0.37306204926314324,
+      "grad_norm": 0.7180671749293043,
+      "learning_rate": 5.073029982551009e-07,
+      "loss": 1.5463,
+      "step": 5354
+    },
+    {
+      "epoch": 0.37313172839076053,
+      "grad_norm": 0.6599128583126527,
+      "learning_rate": 5.072394481258535e-07,
+      "loss": 1.3976,
+      "step": 5355
+    },
+    {
+      "epoch": 0.3732014075183779,
+      "grad_norm": 0.7688131761567236,
+      "learning_rate": 5.071758921386409e-07,
+      "loss": 1.5714,
+      "step": 5356
+    },
+    {
+      "epoch": 0.3732710866459952,
+      "grad_norm": 0.7072828006472733,
+      "learning_rate": 5.071123302965086e-07,
+      "loss": 1.5208,
+      "step": 5357
+    },
+    {
+      "epoch": 0.3733407657736125,
+      "grad_norm": 0.7187426792443632,
+      "learning_rate": 5.070487626025027e-07,
+      "loss": 1.6456,
+      "step": 5358
+    },
+    {
+      "epoch": 0.3734104449012298,
+      "grad_norm": 0.7180845966367558,
+      "learning_rate": 5.069851890596698e-07,
+      "loss": 1.5107,
+      "step": 5359
+    },
+    {
+      "epoch": 0.37348012402884717,
+      "grad_norm": 0.6708545748911989,
+      "learning_rate": 5.06921609671056e-07,
+      "loss": 1.4611,
+      "step": 5360
+    },
+    {
+      "epoch": 0.37354980315646447,
+      "grad_norm": 0.7028485609419693,
+      "learning_rate": 5.068580244397085e-07,
+      "loss": 1.4579,
+      "step": 5361
+    },
+    {
+      "epoch": 0.3736194822840818,
+      "grad_norm": 0.6713681496765652,
+      "learning_rate": 5.067944333686743e-07,
+      "loss": 1.4134,
+      "step": 5362
+    },
+    {
+      "epoch": 0.3736891614116991,
+      "grad_norm": 0.6680925713652732,
+      "learning_rate": 5.067308364610006e-07,
+      "loss": 1.5172,
+      "step": 5363
+    },
+    {
+      "epoch": 0.37375884053931646,
+      "grad_norm": 0.7842330472219948,
+      "learning_rate": 5.066672337197354e-07,
+      "loss": 1.5071,
+      "step": 5364
+    },
+    {
+      "epoch": 0.37382851966693376,
+      "grad_norm": 0.7628791637111324,
+      "learning_rate": 5.066036251479266e-07,
+      "loss": 1.5212,
+      "step": 5365
+    },
+    {
+      "epoch": 0.3738981987945511,
+      "grad_norm": 0.6813664665615757,
+      "learning_rate": 5.065400107486226e-07,
+      "loss": 1.5094,
+      "step": 5366
+    },
+    {
+      "epoch": 0.3739678779221684,
+      "grad_norm": 0.7181594596254843,
+      "learning_rate": 5.064763905248716e-07,
+      "loss": 1.4813,
+      "step": 5367
+    },
+    {
+      "epoch": 0.37403755704978575,
+      "grad_norm": 0.7657310789918401,
+      "learning_rate": 5.064127644797227e-07,
+      "loss": 1.6721,
+      "step": 5368
+    },
+    {
+      "epoch": 0.37410723617740305,
+      "grad_norm": 0.7271926630736468,
+      "learning_rate": 5.063491326162248e-07,
+      "loss": 1.5984,
+      "step": 5369
+    },
+    {
+      "epoch": 0.3741769153050204,
+      "grad_norm": 0.7156351904619168,
+      "learning_rate": 5.062854949374273e-07,
+      "loss": 1.6321,
+      "step": 5370
+    },
+    {
+      "epoch": 0.3742465944326377,
+      "grad_norm": 0.6973806824549705,
+      "learning_rate": 5.062218514463799e-07,
+      "loss": 1.5303,
+      "step": 5371
+    },
+    {
+      "epoch": 0.37431627356025504,
+      "grad_norm": 0.7268627075391219,
+      "learning_rate": 5.061582021461325e-07,
+      "loss": 1.6009,
+      "step": 5372
+    },
+    {
+      "epoch": 0.37438595268787234,
+      "grad_norm": 0.7435202270708391,
+      "learning_rate": 5.060945470397354e-07,
+      "loss": 1.5263,
+      "step": 5373
+    },
+    {
+      "epoch": 0.3744556318154897,
+      "grad_norm": 0.7061226191508438,
+      "learning_rate": 5.060308861302389e-07,
+      "loss": 1.5312,
+      "step": 5374
+    },
+    {
+      "epoch": 0.374525310943107,
+      "grad_norm": 0.6935908957909679,
+      "learning_rate": 5.059672194206938e-07,
+      "loss": 1.4977,
+      "step": 5375
+    },
+    {
+      "epoch": 0.37459499007072433,
+      "grad_norm": 0.8044608362031292,
+      "learning_rate": 5.059035469141515e-07,
+      "loss": 1.5661,
+      "step": 5376
+    },
+    {
+      "epoch": 0.3746646691983416,
+      "grad_norm": 0.6649746679278241,
+      "learning_rate": 5.058398686136628e-07,
+      "loss": 1.5039,
+      "step": 5377
+    },
+    {
+      "epoch": 0.374734348325959,
+      "grad_norm": 0.7565675239565676,
+      "learning_rate": 5.057761845222795e-07,
+      "loss": 1.5466,
+      "step": 5378
+    },
+    {
+      "epoch": 0.37480402745357627,
+      "grad_norm": 0.6988450148686387,
+      "learning_rate": 5.057124946430535e-07,
+      "loss": 1.524,
+      "step": 5379
+    },
+    {
+      "epoch": 0.3748737065811936,
+      "grad_norm": 0.6789677269566424,
+      "learning_rate": 5.056487989790371e-07,
+      "loss": 1.5258,
+      "step": 5380
+    },
+    {
+      "epoch": 0.3749433857088109,
+      "grad_norm": 0.7113732310328157,
+      "learning_rate": 5.055850975332823e-07,
+      "loss": 1.5414,
+      "step": 5381
+    },
+    {
+      "epoch": 0.37501306483642827,
+      "grad_norm": 0.6976581810589285,
+      "learning_rate": 5.055213903088424e-07,
+      "loss": 1.5575,
+      "step": 5382
+    },
+    {
+      "epoch": 0.37508274396404556,
+      "grad_norm": 0.6854275337810213,
+      "learning_rate": 5.054576773087697e-07,
+      "loss": 1.5032,
+      "step": 5383
+    },
+    {
+      "epoch": 0.3751524230916629,
+      "grad_norm": 0.7527863161147542,
+      "learning_rate": 5.05393958536118e-07,
+      "loss": 1.5894,
+      "step": 5384
+    },
+    {
+      "epoch": 0.3752221022192802,
+      "grad_norm": 0.6747886537395279,
+      "learning_rate": 5.053302339939407e-07,
+      "loss": 1.4749,
+      "step": 5385
+    },
+    {
+      "epoch": 0.37529178134689756,
+      "grad_norm": 0.6837579373025371,
+      "learning_rate": 5.052665036852914e-07,
+      "loss": 1.5364,
+      "step": 5386
+    },
+    {
+      "epoch": 0.37536146047451485,
+      "grad_norm": 0.7873639942673621,
+      "learning_rate": 5.052027676132245e-07,
+      "loss": 1.7117,
+      "step": 5387
+    },
+    {
+      "epoch": 0.3754311396021322,
+      "grad_norm": 0.7919474795693494,
+      "learning_rate": 5.051390257807941e-07,
+      "loss": 1.6618,
+      "step": 5388
+    },
+    {
+      "epoch": 0.3755008187297495,
+      "grad_norm": 0.7103425061197322,
+      "learning_rate": 5.05075278191055e-07,
+      "loss": 1.6743,
+      "step": 5389
+    },
+    {
+      "epoch": 0.37557049785736685,
+      "grad_norm": 0.7161819553104973,
+      "learning_rate": 5.05011524847062e-07,
+      "loss": 1.566,
+      "step": 5390
+    },
+    {
+      "epoch": 0.37564017698498414,
+      "grad_norm": 0.7334385643113612,
+      "learning_rate": 5.049477657518704e-07,
+      "loss": 1.5287,
+      "step": 5391
+    },
+    {
+      "epoch": 0.3757098561126015,
+      "grad_norm": 0.7572351651031275,
+      "learning_rate": 5.048840009085356e-07,
+      "loss": 1.5927,
+      "step": 5392
+    },
+    {
+      "epoch": 0.3757795352402188,
+      "grad_norm": 0.7124113602854409,
+      "learning_rate": 5.048202303201134e-07,
+      "loss": 1.5382,
+      "step": 5393
+    },
+    {
+      "epoch": 0.37584921436783614,
+      "grad_norm": 0.6855279268109093,
+      "learning_rate": 5.047564539896597e-07,
+      "loss": 1.489,
+      "step": 5394
+    },
+    {
+      "epoch": 0.37591889349545343,
+      "grad_norm": 0.7122062077638511,
+      "learning_rate": 5.046926719202309e-07,
+      "loss": 1.4946,
+      "step": 5395
+    },
+    {
+      "epoch": 0.3759885726230708,
+      "grad_norm": 0.7091079314088186,
+      "learning_rate": 5.046288841148835e-07,
+      "loss": 1.6135,
+      "step": 5396
+    },
+    {
+      "epoch": 0.3760582517506881,
+      "grad_norm": 0.6740079747894211,
+      "learning_rate": 5.045650905766745e-07,
+      "loss": 1.553,
+      "step": 5397
+    },
+    {
+      "epoch": 0.3761279308783054,
+      "grad_norm": 0.6702936126540127,
+      "learning_rate": 5.045012913086607e-07,
+      "loss": 1.4964,
+      "step": 5398
+    },
+    {
+      "epoch": 0.3761976100059227,
+      "grad_norm": 0.6985064459949539,
+      "learning_rate": 5.044374863138998e-07,
+      "loss": 1.4901,
+      "step": 5399
+    },
+    {
+      "epoch": 0.3762672891335401,
+      "grad_norm": 0.7301075792224525,
+      "learning_rate": 5.043736755954493e-07,
+      "loss": 1.5574,
+      "step": 5400
+    },
+    {
+      "epoch": 0.37633696826115737,
+      "grad_norm": 0.749172825026058,
+      "learning_rate": 5.043098591563673e-07,
+      "loss": 1.5362,
+      "step": 5401
+    },
+    {
+      "epoch": 0.3764066473887747,
+      "grad_norm": 0.7094520771114469,
+      "learning_rate": 5.042460369997119e-07,
+      "loss": 1.5059,
+      "step": 5402
+    },
+    {
+      "epoch": 0.376476326516392,
+      "grad_norm": 0.7207987192677225,
+      "learning_rate": 5.041822091285415e-07,
+      "loss": 1.4182,
+      "step": 5403
+    },
+    {
+      "epoch": 0.37654600564400936,
+      "grad_norm": 0.7707547850189875,
+      "learning_rate": 5.041183755459151e-07,
+      "loss": 1.619,
+      "step": 5404
+    },
+    {
+      "epoch": 0.37661568477162666,
+      "grad_norm": 0.6830636923143021,
+      "learning_rate": 5.040545362548915e-07,
+      "loss": 1.5678,
+      "step": 5405
+    },
+    {
+      "epoch": 0.376685363899244,
+      "grad_norm": 0.732725900627606,
+      "learning_rate": 5.039906912585302e-07,
+      "loss": 1.5759,
+      "step": 5406
+    },
+    {
+      "epoch": 0.3767550430268613,
+      "grad_norm": 0.7422161056576552,
+      "learning_rate": 5.039268405598906e-07,
+      "loss": 1.4853,
+      "step": 5407
+    },
+    {
+      "epoch": 0.37682472215447865,
+      "grad_norm": 0.7256022665248492,
+      "learning_rate": 5.038629841620327e-07,
+      "loss": 1.5363,
+      "step": 5408
+    },
+    {
+      "epoch": 0.37689440128209595,
+      "grad_norm": 0.6892575042707771,
+      "learning_rate": 5.037991220680166e-07,
+      "loss": 1.6261,
+      "step": 5409
+    },
+    {
+      "epoch": 0.3769640804097133,
+      "grad_norm": 0.6804116027311646,
+      "learning_rate": 5.037352542809025e-07,
+      "loss": 1.5511,
+      "step": 5410
+    },
+    {
+      "epoch": 0.3770337595373306,
+      "grad_norm": 0.7689597898444267,
+      "learning_rate": 5.036713808037515e-07,
+      "loss": 1.5802,
+      "step": 5411
+    },
+    {
+      "epoch": 0.3771034386649479,
+      "grad_norm": 0.6964237941047433,
+      "learning_rate": 5.036075016396242e-07,
+      "loss": 1.4965,
+      "step": 5412
+    },
+    {
+      "epoch": 0.37717311779256524,
+      "grad_norm": 0.7435068845298077,
+      "learning_rate": 5.035436167915819e-07,
+      "loss": 1.5212,
+      "step": 5413
+    },
+    {
+      "epoch": 0.37724279692018253,
+      "grad_norm": 0.7349257406525216,
+      "learning_rate": 5.034797262626861e-07,
+      "loss": 1.4656,
+      "step": 5414
+    },
+    {
+      "epoch": 0.3773124760477999,
+      "grad_norm": 0.6882927821800732,
+      "learning_rate": 5.034158300559986e-07,
+      "loss": 1.5055,
+      "step": 5415
+    },
+    {
+      "epoch": 0.3773821551754172,
+      "grad_norm": 0.7189670232247855,
+      "learning_rate": 5.033519281745815e-07,
+      "loss": 1.6791,
+      "step": 5416
+    },
+    {
+      "epoch": 0.37745183430303453,
+      "grad_norm": 0.7508315288135808,
+      "learning_rate": 5.032880206214968e-07,
+      "loss": 1.5512,
+      "step": 5417
+    },
+    {
+      "epoch": 0.3775215134306518,
+      "grad_norm": 0.7366512193101076,
+      "learning_rate": 5.032241073998076e-07,
+      "loss": 1.5568,
+      "step": 5418
+    },
+    {
+      "epoch": 0.3775911925582692,
+      "grad_norm": 0.6640423914093597,
+      "learning_rate": 5.031601885125763e-07,
+      "loss": 1.487,
+      "step": 5419
+    },
+    {
+      "epoch": 0.37766087168588647,
+      "grad_norm": 0.6553708894293232,
+      "learning_rate": 5.030962639628663e-07,
+      "loss": 1.3905,
+      "step": 5420
+    },
+    {
+      "epoch": 0.3777305508135038,
+      "grad_norm": 0.738643876142048,
+      "learning_rate": 5.030323337537408e-07,
+      "loss": 1.5277,
+      "step": 5421
+    },
+    {
+      "epoch": 0.3778002299411211,
+      "grad_norm": 0.6964778026790148,
+      "learning_rate": 5.029683978882635e-07,
+      "loss": 1.4764,
+      "step": 5422
+    },
+    {
+      "epoch": 0.37786990906873846,
+      "grad_norm": 0.7305960630310033,
+      "learning_rate": 5.029044563694985e-07,
+      "loss": 1.5922,
+      "step": 5423
+    },
+    {
+      "epoch": 0.37793958819635576,
+      "grad_norm": 0.7575001757201274,
+      "learning_rate": 5.028405092005098e-07,
+      "loss": 1.5197,
+      "step": 5424
+    },
+    {
+      "epoch": 0.3780092673239731,
+      "grad_norm": 0.6845732719679103,
+      "learning_rate": 5.02776556384362e-07,
+      "loss": 1.6039,
+      "step": 5425
+    },
+    {
+      "epoch": 0.3780789464515904,
+      "grad_norm": 0.7388090099334985,
+      "learning_rate": 5.0271259792412e-07,
+      "loss": 1.4985,
+      "step": 5426
+    },
+    {
+      "epoch": 0.37814862557920775,
+      "grad_norm": 0.6893090869935965,
+      "learning_rate": 5.026486338228484e-07,
+      "loss": 1.5317,
+      "step": 5427
+    },
+    {
+      "epoch": 0.37821830470682505,
+      "grad_norm": 0.7481567809018613,
+      "learning_rate": 5.025846640836129e-07,
+      "loss": 1.5365,
+      "step": 5428
+    },
+    {
+      "epoch": 0.3782879838344424,
+      "grad_norm": 0.7477294337521692,
+      "learning_rate": 5.025206887094789e-07,
+      "loss": 1.6067,
+      "step": 5429
+    },
+    {
+      "epoch": 0.3783576629620597,
+      "grad_norm": 0.702638609667168,
+      "learning_rate": 5.02456707703512e-07,
+      "loss": 1.5787,
+      "step": 5430
+    },
+    {
+      "epoch": 0.37842734208967704,
+      "grad_norm": 0.6619744765586045,
+      "learning_rate": 5.023927210687788e-07,
+      "loss": 1.5176,
+      "step": 5431
+    },
+    {
+      "epoch": 0.37849702121729434,
+      "grad_norm": 0.6782202426529567,
+      "learning_rate": 5.023287288083453e-07,
+      "loss": 1.3905,
+      "step": 5432
+    },
+    {
+      "epoch": 0.3785667003449117,
+      "grad_norm": 0.7090382071085164,
+      "learning_rate": 5.022647309252783e-07,
+      "loss": 1.548,
+      "step": 5433
+    },
+    {
+      "epoch": 0.378636379472529,
+      "grad_norm": 0.8299623002044955,
+      "learning_rate": 5.022007274226446e-07,
+      "loss": 1.6024,
+      "step": 5434
+    },
+    {
+      "epoch": 0.37870605860014633,
+      "grad_norm": 0.7610548390668402,
+      "learning_rate": 5.021367183035114e-07,
+      "loss": 1.6044,
+      "step": 5435
+    },
+    {
+      "epoch": 0.37877573772776363,
+      "grad_norm": 0.7751485981500597,
+      "learning_rate": 5.020727035709463e-07,
+      "loss": 1.5679,
+      "step": 5436
+    },
+    {
+      "epoch": 0.378845416855381,
+      "grad_norm": 0.7063453389555776,
+      "learning_rate": 5.020086832280168e-07,
+      "loss": 1.5361,
+      "step": 5437
+    },
+    {
+      "epoch": 0.3789150959829983,
+      "grad_norm": 0.6902667458465007,
+      "learning_rate": 5.019446572777909e-07,
+      "loss": 1.5486,
+      "step": 5438
+    },
+    {
+      "epoch": 0.3789847751106156,
+      "grad_norm": 0.6808435381311487,
+      "learning_rate": 5.018806257233372e-07,
+      "loss": 1.5548,
+      "step": 5439
+    },
+    {
+      "epoch": 0.3790544542382329,
+      "grad_norm": 0.715213686595762,
+      "learning_rate": 5.018165885677238e-07,
+      "loss": 1.4849,
+      "step": 5440
+    },
+    {
+      "epoch": 0.37912413336585027,
+      "grad_norm": 0.6856158512743733,
+      "learning_rate": 5.017525458140196e-07,
+      "loss": 1.4843,
+      "step": 5441
+    },
+    {
+      "epoch": 0.37919381249346756,
+      "grad_norm": 0.7319426863719979,
+      "learning_rate": 5.016884974652937e-07,
+      "loss": 1.5292,
+      "step": 5442
+    },
+    {
+      "epoch": 0.3792634916210849,
+      "grad_norm": 0.7207250460074514,
+      "learning_rate": 5.016244435246157e-07,
+      "loss": 1.563,
+      "step": 5443
+    },
+    {
+      "epoch": 0.3793331707487022,
+      "grad_norm": 0.7791614239074859,
+      "learning_rate": 5.015603839950547e-07,
+      "loss": 1.5604,
+      "step": 5444
+    },
+    {
+      "epoch": 0.37940284987631956,
+      "grad_norm": 0.7509861266693079,
+      "learning_rate": 5.014963188796808e-07,
+      "loss": 1.6997,
+      "step": 5445
+    },
+    {
+      "epoch": 0.37947252900393685,
+      "grad_norm": 0.7855757249315607,
+      "learning_rate": 5.014322481815643e-07,
+      "loss": 1.3949,
+      "step": 5446
+    },
+    {
+      "epoch": 0.3795422081315542,
+      "grad_norm": 0.7625165655696754,
+      "learning_rate": 5.013681719037753e-07,
+      "loss": 1.6439,
+      "step": 5447
+    },
+    {
+      "epoch": 0.3796118872591715,
+      "grad_norm": 0.7524121252006589,
+      "learning_rate": 5.013040900493848e-07,
+      "loss": 1.5756,
+      "step": 5448
+    },
+    {
+      "epoch": 0.37968156638678885,
+      "grad_norm": 0.7053803505432666,
+      "learning_rate": 5.012400026214633e-07,
+      "loss": 1.409,
+      "step": 5449
+    },
+    {
+      "epoch": 0.37975124551440614,
+      "grad_norm": 0.7577344283390356,
+      "learning_rate": 5.011759096230823e-07,
+      "loss": 1.6402,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3798209246420235,
+      "grad_norm": 0.6913724292869309,
+      "learning_rate": 5.011118110573133e-07,
+      "loss": 1.5397,
+      "step": 5451
+    },
+    {
+      "epoch": 0.3798906037696408,
+      "grad_norm": 0.656322013225874,
+      "learning_rate": 5.010477069272278e-07,
+      "loss": 1.4744,
+      "step": 5452
+    },
+    {
+      "epoch": 0.37996028289725814,
+      "grad_norm": 0.6962733401306395,
+      "learning_rate": 5.009835972358981e-07,
+      "loss": 1.6503,
+      "step": 5453
+    },
+    {
+      "epoch": 0.38002996202487543,
+      "grad_norm": 0.7618166591802068,
+      "learning_rate": 5.009194819863962e-07,
+      "loss": 1.5865,
+      "step": 5454
+    },
+    {
+      "epoch": 0.3800996411524928,
+      "grad_norm": 0.7562107443197053,
+      "learning_rate": 5.008553611817948e-07,
+      "loss": 1.5157,
+      "step": 5455
+    },
+    {
+      "epoch": 0.3801693202801101,
+      "grad_norm": 0.7074325420993474,
+      "learning_rate": 5.007912348251666e-07,
+      "loss": 1.4497,
+      "step": 5456
+    },
+    {
+      "epoch": 0.38023899940772743,
+      "grad_norm": 0.7242519999159883,
+      "learning_rate": 5.007271029195848e-07,
+      "loss": 1.4,
+      "step": 5457
+    },
+    {
+      "epoch": 0.3803086785353447,
+      "grad_norm": 0.6342168243414719,
+      "learning_rate": 5.006629654681224e-07,
+      "loss": 1.3542,
+      "step": 5458
+    },
+    {
+      "epoch": 0.3803783576629621,
+      "grad_norm": 0.7126075656711214,
+      "learning_rate": 5.005988224738535e-07,
+      "loss": 1.4551,
+      "step": 5459
+    },
+    {
+      "epoch": 0.38044803679057937,
+      "grad_norm": 0.7270437470510306,
+      "learning_rate": 5.005346739398517e-07,
+      "loss": 1.5397,
+      "step": 5460
+    },
+    {
+      "epoch": 0.3805177159181967,
+      "grad_norm": 0.7624831018762762,
+      "learning_rate": 5.00470519869191e-07,
+      "loss": 1.6236,
+      "step": 5461
+    },
+    {
+      "epoch": 0.380587395045814,
+      "grad_norm": 0.6992172062429126,
+      "learning_rate": 5.00406360264946e-07,
+      "loss": 1.4604,
+      "step": 5462
+    },
+    {
+      "epoch": 0.38065707417343136,
+      "grad_norm": 0.6566970192760084,
+      "learning_rate": 5.003421951301914e-07,
+      "loss": 1.4452,
+      "step": 5463
+    },
+    {
+      "epoch": 0.38072675330104866,
+      "grad_norm": 0.6998814130100979,
+      "learning_rate": 5.002780244680018e-07,
+      "loss": 1.5508,
+      "step": 5464
+    },
+    {
+      "epoch": 0.380796432428666,
+      "grad_norm": 0.7732539204937761,
+      "learning_rate": 5.002138482814528e-07,
+      "loss": 1.7181,
+      "step": 5465
+    },
+    {
+      "epoch": 0.3808661115562833,
+      "grad_norm": 0.6567647406485111,
+      "learning_rate": 5.001496665736194e-07,
+      "loss": 1.5205,
+      "step": 5466
+    },
+    {
+      "epoch": 0.38093579068390065,
+      "grad_norm": 0.7193635427827528,
+      "learning_rate": 5.000854793475778e-07,
+      "loss": 1.6246,
+      "step": 5467
+    },
+    {
+      "epoch": 0.38100546981151795,
+      "grad_norm": 0.6951657989880955,
+      "learning_rate": 5.000212866064038e-07,
+      "loss": 1.6509,
+      "step": 5468
+    },
+    {
+      "epoch": 0.3810751489391353,
+      "grad_norm": 0.7213532599547762,
+      "learning_rate": 4.999570883531735e-07,
+      "loss": 1.4985,
+      "step": 5469
+    },
+    {
+      "epoch": 0.3811448280667526,
+      "grad_norm": 0.754431340305897,
+      "learning_rate": 4.998928845909635e-07,
+      "loss": 1.5693,
+      "step": 5470
+    },
+    {
+      "epoch": 0.38121450719436994,
+      "grad_norm": 0.6989133775278445,
+      "learning_rate": 4.998286753228507e-07,
+      "loss": 1.5648,
+      "step": 5471
+    },
+    {
+      "epoch": 0.38128418632198724,
+      "grad_norm": 0.7707962594955189,
+      "learning_rate": 4.99764460551912e-07,
+      "loss": 1.6039,
+      "step": 5472
+    },
+    {
+      "epoch": 0.3813538654496046,
+      "grad_norm": 0.7054262814931245,
+      "learning_rate": 4.997002402812248e-07,
+      "loss": 1.4979,
+      "step": 5473
+    },
+    {
+      "epoch": 0.3814235445772219,
+      "grad_norm": 0.7334950363402816,
+      "learning_rate": 4.996360145138664e-07,
+      "loss": 1.5707,
+      "step": 5474
+    },
+    {
+      "epoch": 0.38149322370483923,
+      "grad_norm": 0.7876729845840897,
+      "learning_rate": 4.99571783252915e-07,
+      "loss": 1.7612,
+      "step": 5475
+    },
+    {
+      "epoch": 0.38156290283245653,
+      "grad_norm": 0.658733357304968,
+      "learning_rate": 4.995075465014486e-07,
+      "loss": 1.4626,
+      "step": 5476
+    },
+    {
+      "epoch": 0.3816325819600739,
+      "grad_norm": 0.7316933855652641,
+      "learning_rate": 4.994433042625454e-07,
+      "loss": 1.5544,
+      "step": 5477
+    },
+    {
+      "epoch": 0.3817022610876912,
+      "grad_norm": 0.6762053571770205,
+      "learning_rate": 4.99379056539284e-07,
+      "loss": 1.4745,
+      "step": 5478
+    },
+    {
+      "epoch": 0.3817719402153085,
+      "grad_norm": 0.7084236910699324,
+      "learning_rate": 4.993148033347437e-07,
+      "loss": 1.5001,
+      "step": 5479
+    },
+    {
+      "epoch": 0.3818416193429258,
+      "grad_norm": 0.7271727538874175,
+      "learning_rate": 4.992505446520031e-07,
+      "loss": 1.5476,
+      "step": 5480
+    },
+    {
+      "epoch": 0.38191129847054317,
+      "grad_norm": 0.8292686404246161,
+      "learning_rate": 4.991862804941419e-07,
+      "loss": 1.5489,
+      "step": 5481
+    },
+    {
+      "epoch": 0.38198097759816046,
+      "grad_norm": 0.7133473578558076,
+      "learning_rate": 4.991220108642397e-07,
+      "loss": 1.5058,
+      "step": 5482
+    },
+    {
+      "epoch": 0.3820506567257778,
+      "grad_norm": 0.8080700827951963,
+      "learning_rate": 4.990577357653765e-07,
+      "loss": 1.566,
+      "step": 5483
+    },
+    {
+      "epoch": 0.3821203358533951,
+      "grad_norm": 0.6708760556900807,
+      "learning_rate": 4.989934552006323e-07,
+      "loss": 1.5028,
+      "step": 5484
+    },
+    {
+      "epoch": 0.38219001498101246,
+      "grad_norm": 0.7097547017774847,
+      "learning_rate": 4.989291691730879e-07,
+      "loss": 1.4198,
+      "step": 5485
+    },
+    {
+      "epoch": 0.38225969410862976,
+      "grad_norm": 0.7369934710042176,
+      "learning_rate": 4.988648776858237e-07,
+      "loss": 1.4532,
+      "step": 5486
+    },
+    {
+      "epoch": 0.3823293732362471,
+      "grad_norm": 0.6939493609803657,
+      "learning_rate": 4.98800580741921e-07,
+      "loss": 1.4274,
+      "step": 5487
+    },
+    {
+      "epoch": 0.3823990523638644,
+      "grad_norm": 0.6749067587849705,
+      "learning_rate": 4.987362783444606e-07,
+      "loss": 1.5632,
+      "step": 5488
+    },
+    {
+      "epoch": 0.38246873149148175,
+      "grad_norm": 0.6991470780685725,
+      "learning_rate": 4.986719704965244e-07,
+      "loss": 1.4716,
+      "step": 5489
+    },
+    {
+      "epoch": 0.38253841061909905,
+      "grad_norm": 0.7311252973539579,
+      "learning_rate": 4.98607657201194e-07,
+      "loss": 1.5337,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3826080897467164,
+      "grad_norm": 0.746727609531956,
+      "learning_rate": 4.985433384615513e-07,
+      "loss": 1.563,
+      "step": 5491
+    },
+    {
+      "epoch": 0.3826777688743337,
+      "grad_norm": 0.6725022234464829,
+      "learning_rate": 4.984790142806788e-07,
+      "loss": 1.4125,
+      "step": 5492
+    },
+    {
+      "epoch": 0.38274744800195104,
+      "grad_norm": 0.7290834568894429,
+      "learning_rate": 4.98414684661659e-07,
+      "loss": 1.4797,
+      "step": 5493
+    },
+    {
+      "epoch": 0.38281712712956834,
+      "grad_norm": 0.8232159581467848,
+      "learning_rate": 4.983503496075746e-07,
+      "loss": 1.5075,
+      "step": 5494
+    },
+    {
+      "epoch": 0.3828868062571857,
+      "grad_norm": 0.7540441799845059,
+      "learning_rate": 4.982860091215087e-07,
+      "loss": 1.6719,
+      "step": 5495
+    },
+    {
+      "epoch": 0.382956485384803,
+      "grad_norm": 0.6928721393975205,
+      "learning_rate": 4.982216632065449e-07,
+      "loss": 1.4631,
+      "step": 5496
+    },
+    {
+      "epoch": 0.38302616451242033,
+      "grad_norm": 0.6788751573672828,
+      "learning_rate": 4.981573118657665e-07,
+      "loss": 1.3759,
+      "step": 5497
+    },
+    {
+      "epoch": 0.3830958436400376,
+      "grad_norm": 0.7214898466660512,
+      "learning_rate": 4.980929551022572e-07,
+      "loss": 1.6067,
+      "step": 5498
+    },
+    {
+      "epoch": 0.383165522767655,
+      "grad_norm": 0.7308153122274459,
+      "learning_rate": 4.980285929191015e-07,
+      "loss": 1.4594,
+      "step": 5499
+    },
+    {
+      "epoch": 0.38323520189527227,
+      "grad_norm": 0.6670540192625501,
+      "learning_rate": 4.979642253193835e-07,
+      "loss": 1.4535,
+      "step": 5500
+    },
+    {
+      "epoch": 0.3833048810228896,
+      "grad_norm": 0.7395483755910071,
+      "learning_rate": 4.978998523061879e-07,
+      "loss": 1.539,
+      "step": 5501
+    },
+    {
+      "epoch": 0.3833745601505069,
+      "grad_norm": 0.7695363996801382,
+      "learning_rate": 4.978354738825996e-07,
+      "loss": 1.4483,
+      "step": 5502
+    },
+    {
+      "epoch": 0.3834442392781242,
+      "grad_norm": 0.718100408377757,
+      "learning_rate": 4.977710900517039e-07,
+      "loss": 1.6256,
+      "step": 5503
+    },
+    {
+      "epoch": 0.38351391840574156,
+      "grad_norm": 0.69109645152382,
+      "learning_rate": 4.977067008165859e-07,
+      "loss": 1.5024,
+      "step": 5504
+    },
+    {
+      "epoch": 0.38358359753335886,
+      "grad_norm": 0.7116426387946586,
+      "learning_rate": 4.976423061803315e-07,
+      "loss": 1.6046,
+      "step": 5505
+    },
+    {
+      "epoch": 0.3836532766609762,
+      "grad_norm": 0.7852170766850769,
+      "learning_rate": 4.975779061460264e-07,
+      "loss": 1.7033,
+      "step": 5506
+    },
+    {
+      "epoch": 0.3837229557885935,
+      "grad_norm": 0.7689308983734927,
+      "learning_rate": 4.97513500716757e-07,
+      "loss": 1.557,
+      "step": 5507
+    },
+    {
+      "epoch": 0.38379263491621085,
+      "grad_norm": 0.7233222068190561,
+      "learning_rate": 4.974490898956097e-07,
+      "loss": 1.5805,
+      "step": 5508
+    },
+    {
+      "epoch": 0.38386231404382815,
+      "grad_norm": 0.7177217070825982,
+      "learning_rate": 4.973846736856711e-07,
+      "loss": 1.5573,
+      "step": 5509
+    },
+    {
+      "epoch": 0.3839319931714455,
+      "grad_norm": 0.7513702094061874,
+      "learning_rate": 4.973202520900282e-07,
+      "loss": 1.5063,
+      "step": 5510
+    },
+    {
+      "epoch": 0.3840016722990628,
+      "grad_norm": 0.7713853038917127,
+      "learning_rate": 4.972558251117684e-07,
+      "loss": 1.5505,
+      "step": 5511
+    },
+    {
+      "epoch": 0.38407135142668014,
+      "grad_norm": 0.7090530504996623,
+      "learning_rate": 4.971913927539787e-07,
+      "loss": 1.4762,
+      "step": 5512
+    },
+    {
+      "epoch": 0.38414103055429744,
+      "grad_norm": 0.7231100039969972,
+      "learning_rate": 4.971269550197474e-07,
+      "loss": 1.5707,
+      "step": 5513
+    },
+    {
+      "epoch": 0.3842107096819148,
+      "grad_norm": 0.705309164694651,
+      "learning_rate": 4.970625119121621e-07,
+      "loss": 1.5038,
+      "step": 5514
+    },
+    {
+      "epoch": 0.3842803888095321,
+      "grad_norm": 0.6751102981403432,
+      "learning_rate": 4.969980634343112e-07,
+      "loss": 1.6372,
+      "step": 5515
+    },
+    {
+      "epoch": 0.38435006793714943,
+      "grad_norm": 0.7449367266167386,
+      "learning_rate": 4.969336095892832e-07,
+      "loss": 1.5536,
+      "step": 5516
+    },
+    {
+      "epoch": 0.3844197470647667,
+      "grad_norm": 0.8001794393337616,
+      "learning_rate": 4.968691503801668e-07,
+      "loss": 1.5196,
+      "step": 5517
+    },
+    {
+      "epoch": 0.3844894261923841,
+      "grad_norm": 0.8006640674300962,
+      "learning_rate": 4.968046858100511e-07,
+      "loss": 1.5362,
+      "step": 5518
+    },
+    {
+      "epoch": 0.38455910532000137,
+      "grad_norm": 0.7143062739796038,
+      "learning_rate": 4.967402158820253e-07,
+      "loss": 1.5016,
+      "step": 5519
+    },
+    {
+      "epoch": 0.3846287844476187,
+      "grad_norm": 0.7395199641524862,
+      "learning_rate": 4.966757405991789e-07,
+      "loss": 1.5336,
+      "step": 5520
+    },
+    {
+      "epoch": 0.384698463575236,
+      "grad_norm": 0.7173154576633266,
+      "learning_rate": 4.966112599646018e-07,
+      "loss": 1.473,
+      "step": 5521
+    },
+    {
+      "epoch": 0.38476814270285337,
+      "grad_norm": 0.6862498887428835,
+      "learning_rate": 4.965467739813839e-07,
+      "loss": 1.5229,
+      "step": 5522
+    },
+    {
+      "epoch": 0.38483782183047066,
+      "grad_norm": 0.7432309321652748,
+      "learning_rate": 4.964822826526156e-07,
+      "loss": 1.4813,
+      "step": 5523
+    },
+    {
+      "epoch": 0.384907500958088,
+      "grad_norm": 0.744188463216907,
+      "learning_rate": 4.964177859813877e-07,
+      "loss": 1.4979,
+      "step": 5524
+    },
+    {
+      "epoch": 0.3849771800857053,
+      "grad_norm": 0.6818187295241545,
+      "learning_rate": 4.963532839707905e-07,
+      "loss": 1.5359,
+      "step": 5525
+    },
+    {
+      "epoch": 0.38504685921332266,
+      "grad_norm": 0.7178209922409182,
+      "learning_rate": 4.962887766239153e-07,
+      "loss": 1.6431,
+      "step": 5526
+    },
+    {
+      "epoch": 0.38511653834093995,
+      "grad_norm": 0.7332062253347716,
+      "learning_rate": 4.962242639438536e-07,
+      "loss": 1.5583,
+      "step": 5527
+    },
+    {
+      "epoch": 0.3851862174685573,
+      "grad_norm": 0.6889685000442761,
+      "learning_rate": 4.961597459336968e-07,
+      "loss": 1.5706,
+      "step": 5528
+    },
+    {
+      "epoch": 0.3852558965961746,
+      "grad_norm": 0.8052842084542119,
+      "learning_rate": 4.960952225965369e-07,
+      "loss": 1.5433,
+      "step": 5529
+    },
+    {
+      "epoch": 0.38532557572379195,
+      "grad_norm": 0.766408402464758,
+      "learning_rate": 4.960306939354656e-07,
+      "loss": 1.5682,
+      "step": 5530
+    },
+    {
+      "epoch": 0.38539525485140924,
+      "grad_norm": 0.7484794652400308,
+      "learning_rate": 4.959661599535756e-07,
+      "loss": 1.5582,
+      "step": 5531
+    },
+    {
+      "epoch": 0.3854649339790266,
+      "grad_norm": 0.7820263373509483,
+      "learning_rate": 4.959016206539595e-07,
+      "loss": 1.7689,
+      "step": 5532
+    },
+    {
+      "epoch": 0.3855346131066439,
+      "grad_norm": 0.7259786737233994,
+      "learning_rate": 4.9583707603971e-07,
+      "loss": 1.6318,
+      "step": 5533
+    },
+    {
+      "epoch": 0.38560429223426124,
+      "grad_norm": 0.7676985737076506,
+      "learning_rate": 4.957725261139201e-07,
+      "loss": 1.6155,
+      "step": 5534
+    },
+    {
+      "epoch": 0.38567397136187853,
+      "grad_norm": 0.7069887072622391,
+      "learning_rate": 4.957079708796835e-07,
+      "loss": 1.3996,
+      "step": 5535
+    },
+    {
+      "epoch": 0.3857436504894959,
+      "grad_norm": 0.7030376049130003,
+      "learning_rate": 4.956434103400937e-07,
+      "loss": 1.5679,
+      "step": 5536
+    },
+    {
+      "epoch": 0.3858133296171132,
+      "grad_norm": 0.6855415717927498,
+      "learning_rate": 4.955788444982444e-07,
+      "loss": 1.5555,
+      "step": 5537
+    },
+    {
+      "epoch": 0.3858830087447305,
+      "grad_norm": 0.7614205686727051,
+      "learning_rate": 4.9551427335723e-07,
+      "loss": 1.6741,
+      "step": 5538
+    },
+    {
+      "epoch": 0.3859526878723478,
+      "grad_norm": 0.6989562304165375,
+      "learning_rate": 4.954496969201446e-07,
+      "loss": 1.6391,
+      "step": 5539
+    },
+    {
+      "epoch": 0.38602236699996517,
+      "grad_norm": 0.695698775967819,
+      "learning_rate": 4.953851151900831e-07,
+      "loss": 1.5328,
+      "step": 5540
+    },
+    {
+      "epoch": 0.38609204612758247,
+      "grad_norm": 0.7417673904195509,
+      "learning_rate": 4.953205281701402e-07,
+      "loss": 1.6011,
+      "step": 5541
+    },
+    {
+      "epoch": 0.3861617252551998,
+      "grad_norm": 0.6907559477471068,
+      "learning_rate": 4.95255935863411e-07,
+      "loss": 1.4323,
+      "step": 5542
+    },
+    {
+      "epoch": 0.3862314043828171,
+      "grad_norm": 0.6779292317473243,
+      "learning_rate": 4.951913382729913e-07,
+      "loss": 1.563,
+      "step": 5543
+    },
+    {
+      "epoch": 0.38630108351043446,
+      "grad_norm": 0.7723058973909813,
+      "learning_rate": 4.951267354019762e-07,
+      "loss": 1.5341,
+      "step": 5544
+    },
+    {
+      "epoch": 0.38637076263805176,
+      "grad_norm": 0.7137990992337757,
+      "learning_rate": 4.95062127253462e-07,
+      "loss": 1.5635,
+      "step": 5545
+    },
+    {
+      "epoch": 0.3864404417656691,
+      "grad_norm": 0.704918015158124,
+      "learning_rate": 4.949975138305446e-07,
+      "loss": 1.4581,
+      "step": 5546
+    },
+    {
+      "epoch": 0.3865101208932864,
+      "grad_norm": 0.7155688218885253,
+      "learning_rate": 4.949328951363204e-07,
+      "loss": 1.5916,
+      "step": 5547
+    },
+    {
+      "epoch": 0.38657980002090375,
+      "grad_norm": 0.7932583910439622,
+      "learning_rate": 4.948682711738865e-07,
+      "loss": 1.5528,
+      "step": 5548
+    },
+    {
+      "epoch": 0.38664947914852105,
+      "grad_norm": 0.8083960035017587,
+      "learning_rate": 4.948036419463393e-07,
+      "loss": 1.5914,
+      "step": 5549
+    },
+    {
+      "epoch": 0.3867191582761384,
+      "grad_norm": 0.667764867736382,
+      "learning_rate": 4.947390074567761e-07,
+      "loss": 1.5833,
+      "step": 5550
+    },
+    {
+      "epoch": 0.3867888374037557,
+      "grad_norm": 0.7419549291652704,
+      "learning_rate": 4.946743677082945e-07,
+      "loss": 1.5124,
+      "step": 5551
+    },
+    {
+      "epoch": 0.38685851653137304,
+      "grad_norm": 0.6870025872836284,
+      "learning_rate": 4.946097227039921e-07,
+      "loss": 1.5114,
+      "step": 5552
+    },
+    {
+      "epoch": 0.38692819565899034,
+      "grad_norm": 0.6755742379070345,
+      "learning_rate": 4.945450724469665e-07,
+      "loss": 1.5062,
+      "step": 5553
+    },
+    {
+      "epoch": 0.3869978747866077,
+      "grad_norm": 0.7192605944563015,
+      "learning_rate": 4.944804169403164e-07,
+      "loss": 1.5299,
+      "step": 5554
+    },
+    {
+      "epoch": 0.387067553914225,
+      "grad_norm": 0.7074686788750786,
+      "learning_rate": 4.944157561871397e-07,
+      "loss": 1.4861,
+      "step": 5555
+    },
+    {
+      "epoch": 0.38713723304184233,
+      "grad_norm": 0.7303712587132282,
+      "learning_rate": 4.943510901905356e-07,
+      "loss": 1.5789,
+      "step": 5556
+    },
+    {
+      "epoch": 0.3872069121694596,
+      "grad_norm": 0.70092718611266,
+      "learning_rate": 4.942864189536027e-07,
+      "loss": 1.5269,
+      "step": 5557
+    },
+    {
+      "epoch": 0.387276591297077,
+      "grad_norm": 0.7749207884282524,
+      "learning_rate": 4.942217424794401e-07,
+      "loss": 1.5306,
+      "step": 5558
+    },
+    {
+      "epoch": 0.3873462704246943,
+      "grad_norm": 0.7612048856830113,
+      "learning_rate": 4.941570607711474e-07,
+      "loss": 1.6048,
+      "step": 5559
+    },
+    {
+      "epoch": 0.3874159495523116,
+      "grad_norm": 0.7291029938509181,
+      "learning_rate": 4.940923738318245e-07,
+      "loss": 1.4984,
+      "step": 5560
+    },
+    {
+      "epoch": 0.3874856286799289,
+      "grad_norm": 0.6677895182838004,
+      "learning_rate": 4.940276816645708e-07,
+      "loss": 1.4915,
+      "step": 5561
+    },
+    {
+      "epoch": 0.38755530780754627,
+      "grad_norm": 0.7735121442224963,
+      "learning_rate": 4.939629842724868e-07,
+      "loss": 1.5627,
+      "step": 5562
+    },
+    {
+      "epoch": 0.38762498693516356,
+      "grad_norm": 0.7790175853606958,
+      "learning_rate": 4.93898281658673e-07,
+      "loss": 1.4801,
+      "step": 5563
+    },
+    {
+      "epoch": 0.3876946660627809,
+      "grad_norm": 0.7145681800827799,
+      "learning_rate": 4.9383357382623e-07,
+      "loss": 1.4691,
+      "step": 5564
+    },
+    {
+      "epoch": 0.3877643451903982,
+      "grad_norm": 0.7432846857234464,
+      "learning_rate": 4.937688607782586e-07,
+      "loss": 1.5028,
+      "step": 5565
+    },
+    {
+      "epoch": 0.38783402431801556,
+      "grad_norm": 0.7159183915911248,
+      "learning_rate": 4.937041425178601e-07,
+      "loss": 1.5037,
+      "step": 5566
+    },
+    {
+      "epoch": 0.38790370344563285,
+      "grad_norm": 0.7269668278784979,
+      "learning_rate": 4.936394190481361e-07,
+      "loss": 1.4085,
+      "step": 5567
+    },
+    {
+      "epoch": 0.3879733825732502,
+      "grad_norm": 0.6866858671417458,
+      "learning_rate": 4.93574690372188e-07,
+      "loss": 1.4342,
+      "step": 5568
+    },
+    {
+      "epoch": 0.3880430617008675,
+      "grad_norm": 0.7145490114822929,
+      "learning_rate": 4.935099564931177e-07,
+      "loss": 1.5746,
+      "step": 5569
+    },
+    {
+      "epoch": 0.38811274082848485,
+      "grad_norm": 0.7138045405297827,
+      "learning_rate": 4.934452174140276e-07,
+      "loss": 1.5896,
+      "step": 5570
+    },
+    {
+      "epoch": 0.38818241995610214,
+      "grad_norm": 0.7095926528479751,
+      "learning_rate": 4.933804731380201e-07,
+      "loss": 1.5292,
+      "step": 5571
+    },
+    {
+      "epoch": 0.3882520990837195,
+      "grad_norm": 0.7129684749761618,
+      "learning_rate": 4.933157236681978e-07,
+      "loss": 1.5525,
+      "step": 5572
+    },
+    {
+      "epoch": 0.3883217782113368,
+      "grad_norm": 0.7328684505244791,
+      "learning_rate": 4.932509690076638e-07,
+      "loss": 1.7226,
+      "step": 5573
+    },
+    {
+      "epoch": 0.38839145733895414,
+      "grad_norm": 0.7339828615577176,
+      "learning_rate": 4.931862091595208e-07,
+      "loss": 1.6035,
+      "step": 5574
+    },
+    {
+      "epoch": 0.38846113646657143,
+      "grad_norm": 0.6450537560923328,
+      "learning_rate": 4.931214441268729e-07,
+      "loss": 1.5124,
+      "step": 5575
+    },
+    {
+      "epoch": 0.3885308155941888,
+      "grad_norm": 0.7300985149976399,
+      "learning_rate": 4.930566739128232e-07,
+      "loss": 1.4279,
+      "step": 5576
+    },
+    {
+      "epoch": 0.3886004947218061,
+      "grad_norm": 0.6980864377031916,
+      "learning_rate": 4.92991898520476e-07,
+      "loss": 1.5553,
+      "step": 5577
+    },
+    {
+      "epoch": 0.38867017384942343,
+      "grad_norm": 0.7027638402706293,
+      "learning_rate": 4.929271179529351e-07,
+      "loss": 1.5512,
+      "step": 5578
+    },
+    {
+      "epoch": 0.3887398529770407,
+      "grad_norm": 0.7548467168888858,
+      "learning_rate": 4.928623322133055e-07,
+      "loss": 1.5324,
+      "step": 5579
+    },
+    {
+      "epoch": 0.3888095321046581,
+      "grad_norm": 0.7409166404373027,
+      "learning_rate": 4.927975413046912e-07,
+      "loss": 1.6207,
+      "step": 5580
+    },
+    {
+      "epoch": 0.38887921123227537,
+      "grad_norm": 0.6605626587391256,
+      "learning_rate": 4.927327452301975e-07,
+      "loss": 1.502,
+      "step": 5581
+    },
+    {
+      "epoch": 0.3889488903598927,
+      "grad_norm": 0.6880962749504176,
+      "learning_rate": 4.926679439929295e-07,
+      "loss": 1.4707,
+      "step": 5582
+    },
+    {
+      "epoch": 0.38901856948751,
+      "grad_norm": 0.7802737438340347,
+      "learning_rate": 4.926031375959926e-07,
+      "loss": 1.5877,
+      "step": 5583
+    },
+    {
+      "epoch": 0.38908824861512736,
+      "grad_norm": 0.7257829551630415,
+      "learning_rate": 4.925383260424924e-07,
+      "loss": 1.5467,
+      "step": 5584
+    },
+    {
+      "epoch": 0.38915792774274466,
+      "grad_norm": 0.6993590557963282,
+      "learning_rate": 4.924735093355348e-07,
+      "loss": 1.5517,
+      "step": 5585
+    },
+    {
+      "epoch": 0.389227606870362,
+      "grad_norm": 0.7438495635411839,
+      "learning_rate": 4.92408687478226e-07,
+      "loss": 1.5874,
+      "step": 5586
+    },
+    {
+      "epoch": 0.3892972859979793,
+      "grad_norm": 0.733088731946248,
+      "learning_rate": 4.923438604736725e-07,
+      "loss": 1.5717,
+      "step": 5587
+    },
+    {
+      "epoch": 0.38936696512559665,
+      "grad_norm": 0.7083801837424353,
+      "learning_rate": 4.922790283249808e-07,
+      "loss": 1.3843,
+      "step": 5588
+    },
+    {
+      "epoch": 0.38943664425321395,
+      "grad_norm": 0.717862741840073,
+      "learning_rate": 4.922141910352578e-07,
+      "loss": 1.5743,
+      "step": 5589
+    },
+    {
+      "epoch": 0.3895063233808313,
+      "grad_norm": 0.7021056241225214,
+      "learning_rate": 4.921493486076106e-07,
+      "loss": 1.4911,
+      "step": 5590
+    },
+    {
+      "epoch": 0.3895760025084486,
+      "grad_norm": 0.7326892044390457,
+      "learning_rate": 4.920845010451468e-07,
+      "loss": 1.5485,
+      "step": 5591
+    },
+    {
+      "epoch": 0.38964568163606594,
+      "grad_norm": 0.7167763296787694,
+      "learning_rate": 4.920196483509737e-07,
+      "loss": 1.5422,
+      "step": 5592
+    },
+    {
+      "epoch": 0.38971536076368324,
+      "grad_norm": 0.6716757792080215,
+      "learning_rate": 4.919547905281994e-07,
+      "loss": 1.6012,
+      "step": 5593
+    },
+    {
+      "epoch": 0.38978503989130053,
+      "grad_norm": 0.7156461246251185,
+      "learning_rate": 4.91889927579932e-07,
+      "loss": 1.5866,
+      "step": 5594
+    },
+    {
+      "epoch": 0.3898547190189179,
+      "grad_norm": 0.7419782612066664,
+      "learning_rate": 4.918250595092798e-07,
+      "loss": 1.4934,
+      "step": 5595
+    },
+    {
+      "epoch": 0.3899243981465352,
+      "grad_norm": 0.7476547747400746,
+      "learning_rate": 4.917601863193514e-07,
+      "loss": 1.6338,
+      "step": 5596
+    },
+    {
+      "epoch": 0.38999407727415253,
+      "grad_norm": 0.728083014760731,
+      "learning_rate": 4.916953080132558e-07,
+      "loss": 1.5645,
+      "step": 5597
+    },
+    {
+      "epoch": 0.3900637564017698,
+      "grad_norm": 0.7735797984284867,
+      "learning_rate": 4.916304245941018e-07,
+      "loss": 1.6132,
+      "step": 5598
+    },
+    {
+      "epoch": 0.3901334355293872,
+      "grad_norm": 0.7034034881934533,
+      "learning_rate": 4.915655360649992e-07,
+      "loss": 1.5384,
+      "step": 5599
+    },
+    {
+      "epoch": 0.39020311465700447,
+      "grad_norm": 0.7489804917635291,
+      "learning_rate": 4.915006424290572e-07,
+      "loss": 1.5392,
+      "step": 5600
+    },
+    {
+      "epoch": 0.3902727937846218,
+      "grad_norm": 0.7456332541983953,
+      "learning_rate": 4.914357436893857e-07,
+      "loss": 1.5891,
+      "step": 5601
+    },
+    {
+      "epoch": 0.3903424729122391,
+      "grad_norm": 0.7036216749370718,
+      "learning_rate": 4.91370839849095e-07,
+      "loss": 1.6004,
+      "step": 5602
+    },
+    {
+      "epoch": 0.39041215203985646,
+      "grad_norm": 0.7193251436793048,
+      "learning_rate": 4.913059309112952e-07,
+      "loss": 1.529,
+      "step": 5603
+    },
+    {
+      "epoch": 0.39048183116747376,
+      "grad_norm": 0.7729473570738046,
+      "learning_rate": 4.91241016879097e-07,
+      "loss": 1.5915,
+      "step": 5604
+    },
+    {
+      "epoch": 0.3905515102950911,
+      "grad_norm": 0.7334388646784056,
+      "learning_rate": 4.911760977556112e-07,
+      "loss": 1.5761,
+      "step": 5605
+    },
+    {
+      "epoch": 0.3906211894227084,
+      "grad_norm": 0.7335809025498045,
+      "learning_rate": 4.911111735439487e-07,
+      "loss": 1.5482,
+      "step": 5606
+    },
+    {
+      "epoch": 0.39069086855032575,
+      "grad_norm": 0.7207635907762502,
+      "learning_rate": 4.91046244247221e-07,
+      "loss": 1.4662,
+      "step": 5607
+    },
+    {
+      "epoch": 0.39076054767794305,
+      "grad_norm": 0.7091201638463493,
+      "learning_rate": 4.909813098685395e-07,
+      "loss": 1.5795,
+      "step": 5608
+    },
+    {
+      "epoch": 0.3908302268055604,
+      "grad_norm": 0.7013771838761281,
+      "learning_rate": 4.909163704110161e-07,
+      "loss": 1.544,
+      "step": 5609
+    },
+    {
+      "epoch": 0.3908999059331777,
+      "grad_norm": 0.7074049818275138,
+      "learning_rate": 4.908514258777628e-07,
+      "loss": 1.6141,
+      "step": 5610
+    },
+    {
+      "epoch": 0.39096958506079504,
+      "grad_norm": 0.6972315835236457,
+      "learning_rate": 4.907864762718918e-07,
+      "loss": 1.4784,
+      "step": 5611
+    },
+    {
+      "epoch": 0.39103926418841234,
+      "grad_norm": 0.715516245329465,
+      "learning_rate": 4.907215215965157e-07,
+      "loss": 1.5359,
+      "step": 5612
+    },
+    {
+      "epoch": 0.3911089433160297,
+      "grad_norm": 0.7574553672743075,
+      "learning_rate": 4.906565618547475e-07,
+      "loss": 1.7375,
+      "step": 5613
+    },
+    {
+      "epoch": 0.391178622443647,
+      "grad_norm": 0.6502857177607614,
+      "learning_rate": 4.905915970496996e-07,
+      "loss": 1.5163,
+      "step": 5614
+    },
+    {
+      "epoch": 0.39124830157126433,
+      "grad_norm": 0.6938368096085746,
+      "learning_rate": 4.905266271844857e-07,
+      "loss": 1.5624,
+      "step": 5615
+    },
+    {
+      "epoch": 0.39131798069888163,
+      "grad_norm": 0.6712918492602664,
+      "learning_rate": 4.904616522622193e-07,
+      "loss": 1.5201,
+      "step": 5616
+    },
+    {
+      "epoch": 0.391387659826499,
+      "grad_norm": 0.7988195643434735,
+      "learning_rate": 4.903966722860139e-07,
+      "loss": 1.5929,
+      "step": 5617
+    },
+    {
+      "epoch": 0.3914573389541163,
+      "grad_norm": 0.7057183780225315,
+      "learning_rate": 4.903316872589836e-07,
+      "loss": 1.648,
+      "step": 5618
+    },
+    {
+      "epoch": 0.3915270180817336,
+      "grad_norm": 0.6757523132570389,
+      "learning_rate": 4.902666971842426e-07,
+      "loss": 1.5514,
+      "step": 5619
+    },
+    {
+      "epoch": 0.3915966972093509,
+      "grad_norm": 0.6980300975678734,
+      "learning_rate": 4.902017020649053e-07,
+      "loss": 1.6303,
+      "step": 5620
+    },
+    {
+      "epoch": 0.39166637633696827,
+      "grad_norm": 0.7461011434866858,
+      "learning_rate": 4.901367019040866e-07,
+      "loss": 1.4192,
+      "step": 5621
+    },
+    {
+      "epoch": 0.39173605546458556,
+      "grad_norm": 0.7287549638649695,
+      "learning_rate": 4.900716967049011e-07,
+      "loss": 1.5044,
+      "step": 5622
+    },
+    {
+      "epoch": 0.3918057345922029,
+      "grad_norm": 0.6884426807253309,
+      "learning_rate": 4.900066864704644e-07,
+      "loss": 1.4788,
+      "step": 5623
+    },
+    {
+      "epoch": 0.3918754137198202,
+      "grad_norm": 0.698307728524486,
+      "learning_rate": 4.899416712038918e-07,
+      "loss": 1.5435,
+      "step": 5624
+    },
+    {
+      "epoch": 0.39194509284743756,
+      "grad_norm": 0.6789008299948818,
+      "learning_rate": 4.898766509082986e-07,
+      "loss": 1.5289,
+      "step": 5625
+    },
+    {
+      "epoch": 0.39201477197505485,
+      "grad_norm": 0.7970341094255471,
+      "learning_rate": 4.89811625586801e-07,
+      "loss": 1.7357,
+      "step": 5626
+    },
+    {
+      "epoch": 0.3920844511026722,
+      "grad_norm": 0.7188738277514154,
+      "learning_rate": 4.897465952425153e-07,
+      "loss": 1.4774,
+      "step": 5627
+    },
+    {
+      "epoch": 0.3921541302302895,
+      "grad_norm": 0.8373284142960566,
+      "learning_rate": 4.896815598785576e-07,
+      "loss": 1.6099,
+      "step": 5628
+    },
+    {
+      "epoch": 0.39222380935790685,
+      "grad_norm": 0.7088823337013,
+      "learning_rate": 4.896165194980447e-07,
+      "loss": 1.529,
+      "step": 5629
+    },
+    {
+      "epoch": 0.39229348848552414,
+      "grad_norm": 0.7223001653765114,
+      "learning_rate": 4.895514741040933e-07,
+      "loss": 1.5641,
+      "step": 5630
+    },
+    {
+      "epoch": 0.3923631676131415,
+      "grad_norm": 0.6914220741166752,
+      "learning_rate": 4.894864236998208e-07,
+      "loss": 1.5283,
+      "step": 5631
+    },
+    {
+      "epoch": 0.3924328467407588,
+      "grad_norm": 0.7147757332932779,
+      "learning_rate": 4.894213682883443e-07,
+      "loss": 1.521,
+      "step": 5632
+    },
+    {
+      "epoch": 0.39250252586837614,
+      "grad_norm": 0.764680925124264,
+      "learning_rate": 4.893563078727815e-07,
+      "loss": 1.5515,
+      "step": 5633
+    },
+    {
+      "epoch": 0.39257220499599343,
+      "grad_norm": 0.7822852931923628,
+      "learning_rate": 4.892912424562501e-07,
+      "loss": 1.5652,
+      "step": 5634
+    },
+    {
+      "epoch": 0.3926418841236108,
+      "grad_norm": 0.7929478109855593,
+      "learning_rate": 4.892261720418682e-07,
+      "loss": 1.7101,
+      "step": 5635
+    },
+    {
+      "epoch": 0.3927115632512281,
+      "grad_norm": 0.7218548026309808,
+      "learning_rate": 4.891610966327543e-07,
+      "loss": 1.652,
+      "step": 5636
+    },
+    {
+      "epoch": 0.39278124237884543,
+      "grad_norm": 0.7335248098012389,
+      "learning_rate": 4.890960162320267e-07,
+      "loss": 1.5682,
+      "step": 5637
+    },
+    {
+      "epoch": 0.3928509215064627,
+      "grad_norm": 0.7143306188937307,
+      "learning_rate": 4.890309308428044e-07,
+      "loss": 1.5508,
+      "step": 5638
+    },
+    {
+      "epoch": 0.3929206006340801,
+      "grad_norm": 0.6688063423914016,
+      "learning_rate": 4.889658404682062e-07,
+      "loss": 1.5865,
+      "step": 5639
+    },
+    {
+      "epoch": 0.39299027976169737,
+      "grad_norm": 0.6915994442437262,
+      "learning_rate": 4.889007451113515e-07,
+      "loss": 1.4609,
+      "step": 5640
+    },
+    {
+      "epoch": 0.3930599588893147,
+      "grad_norm": 0.7061342396943753,
+      "learning_rate": 4.8883564477536e-07,
+      "loss": 1.5257,
+      "step": 5641
+    },
+    {
+      "epoch": 0.393129638016932,
+      "grad_norm": 0.7330023613356081,
+      "learning_rate": 4.88770539463351e-07,
+      "loss": 1.5863,
+      "step": 5642
+    },
+    {
+      "epoch": 0.39319931714454937,
+      "grad_norm": 0.6569236724030605,
+      "learning_rate": 4.887054291784448e-07,
+      "loss": 1.5345,
+      "step": 5643
+    },
+    {
+      "epoch": 0.39326899627216666,
+      "grad_norm": 0.6997962715930213,
+      "learning_rate": 4.886403139237615e-07,
+      "loss": 1.6101,
+      "step": 5644
+    },
+    {
+      "epoch": 0.393338675399784,
+      "grad_norm": 0.6660182692895334,
+      "learning_rate": 4.885751937024216e-07,
+      "loss": 1.4714,
+      "step": 5645
+    },
+    {
+      "epoch": 0.3934083545274013,
+      "grad_norm": 0.7775772452706329,
+      "learning_rate": 4.885100685175459e-07,
+      "loss": 1.6419,
+      "step": 5646
+    },
+    {
+      "epoch": 0.39347803365501866,
+      "grad_norm": 0.7940753653334305,
+      "learning_rate": 4.88444938372255e-07,
+      "loss": 1.4888,
+      "step": 5647
+    },
+    {
+      "epoch": 0.39354771278263595,
+      "grad_norm": 0.6889526655071125,
+      "learning_rate": 4.883798032696704e-07,
+      "loss": 1.5522,
+      "step": 5648
+    },
+    {
+      "epoch": 0.3936173919102533,
+      "grad_norm": 0.7216018333181632,
+      "learning_rate": 4.883146632129135e-07,
+      "loss": 1.424,
+      "step": 5649
+    },
+    {
+      "epoch": 0.3936870710378706,
+      "grad_norm": 0.7383631566837524,
+      "learning_rate": 4.882495182051056e-07,
+      "loss": 1.5577,
+      "step": 5650
+    },
+    {
+      "epoch": 0.39375675016548795,
+      "grad_norm": 0.6910122209800914,
+      "learning_rate": 4.88184368249369e-07,
+      "loss": 1.651,
+      "step": 5651
+    },
+    {
+      "epoch": 0.39382642929310524,
+      "grad_norm": 0.7071475670558016,
+      "learning_rate": 4.881192133488256e-07,
+      "loss": 1.4289,
+      "step": 5652
+    },
+    {
+      "epoch": 0.3938961084207226,
+      "grad_norm": 0.7940647359092496,
+      "learning_rate": 4.880540535065978e-07,
+      "loss": 1.5281,
+      "step": 5653
+    },
+    {
+      "epoch": 0.3939657875483399,
+      "grad_norm": 0.7424024603209955,
+      "learning_rate": 4.879888887258082e-07,
+      "loss": 1.7073,
+      "step": 5654
+    },
+    {
+      "epoch": 0.39403546667595724,
+      "grad_norm": 0.7040686230275012,
+      "learning_rate": 4.879237190095795e-07,
+      "loss": 1.6446,
+      "step": 5655
+    },
+    {
+      "epoch": 0.39410514580357453,
+      "grad_norm": 0.6804339815316579,
+      "learning_rate": 4.878585443610351e-07,
+      "loss": 1.5443,
+      "step": 5656
+    },
+    {
+      "epoch": 0.3941748249311919,
+      "grad_norm": 0.7694443687860312,
+      "learning_rate": 4.877933647832978e-07,
+      "loss": 1.6059,
+      "step": 5657
+    },
+    {
+      "epoch": 0.3942445040588092,
+      "grad_norm": 0.7095308678114916,
+      "learning_rate": 4.877281802794917e-07,
+      "loss": 1.5783,
+      "step": 5658
+    },
+    {
+      "epoch": 0.3943141831864265,
+      "grad_norm": 0.6834728695375323,
+      "learning_rate": 4.876629908527402e-07,
+      "loss": 1.4167,
+      "step": 5659
+    },
+    {
+      "epoch": 0.3943838623140438,
+      "grad_norm": 0.732453498933405,
+      "learning_rate": 4.875977965061674e-07,
+      "loss": 1.5803,
+      "step": 5660
+    },
+    {
+      "epoch": 0.39445354144166117,
+      "grad_norm": 0.8493170677261569,
+      "learning_rate": 4.875325972428976e-07,
+      "loss": 1.5688,
+      "step": 5661
+    },
+    {
+      "epoch": 0.39452322056927847,
+      "grad_norm": 0.7366435447689659,
+      "learning_rate": 4.874673930660551e-07,
+      "loss": 1.6815,
+      "step": 5662
+    },
+    {
+      "epoch": 0.3945928996968958,
+      "grad_norm": 0.6972247175402578,
+      "learning_rate": 4.87402183978765e-07,
+      "loss": 1.3911,
+      "step": 5663
+    },
+    {
+      "epoch": 0.3946625788245131,
+      "grad_norm": 0.6724755543519445,
+      "learning_rate": 4.87336969984152e-07,
+      "loss": 1.553,
+      "step": 5664
+    },
+    {
+      "epoch": 0.39473225795213046,
+      "grad_norm": 0.7443101644041815,
+      "learning_rate": 4.872717510853411e-07,
+      "loss": 1.4836,
+      "step": 5665
+    },
+    {
+      "epoch": 0.39480193707974776,
+      "grad_norm": 0.7105185151933725,
+      "learning_rate": 4.872065272854581e-07,
+      "loss": 1.604,
+      "step": 5666
+    },
+    {
+      "epoch": 0.3948716162073651,
+      "grad_norm": 0.7224894163248697,
+      "learning_rate": 4.871412985876283e-07,
+      "loss": 1.5213,
+      "step": 5667
+    },
+    {
+      "epoch": 0.3949412953349824,
+      "grad_norm": 0.7109761043079036,
+      "learning_rate": 4.87076064994978e-07,
+      "loss": 1.5168,
+      "step": 5668
+    },
+    {
+      "epoch": 0.39501097446259975,
+      "grad_norm": 0.8068347274295018,
+      "learning_rate": 4.870108265106329e-07,
+      "loss": 1.7317,
+      "step": 5669
+    },
+    {
+      "epoch": 0.39508065359021705,
+      "grad_norm": 0.7051206301189206,
+      "learning_rate": 4.869455831377196e-07,
+      "loss": 1.5398,
+      "step": 5670
+    },
+    {
+      "epoch": 0.3951503327178344,
+      "grad_norm": 0.7227348729953125,
+      "learning_rate": 4.868803348793649e-07,
+      "loss": 1.5696,
+      "step": 5671
+    },
+    {
+      "epoch": 0.3952200118454517,
+      "grad_norm": 0.7597406001763358,
+      "learning_rate": 4.868150817386952e-07,
+      "loss": 1.4969,
+      "step": 5672
+    },
+    {
+      "epoch": 0.39528969097306904,
+      "grad_norm": 0.8029844792030686,
+      "learning_rate": 4.867498237188378e-07,
+      "loss": 1.5866,
+      "step": 5673
+    },
+    {
+      "epoch": 0.39535937010068634,
+      "grad_norm": 0.7330628123186286,
+      "learning_rate": 4.8668456082292e-07,
+      "loss": 1.5821,
+      "step": 5674
+    },
+    {
+      "epoch": 0.3954290492283037,
+      "grad_norm": 0.7313312062625753,
+      "learning_rate": 4.866192930540692e-07,
+      "loss": 1.6279,
+      "step": 5675
+    },
+    {
+      "epoch": 0.395498728355921,
+      "grad_norm": 0.8043239241069365,
+      "learning_rate": 4.865540204154133e-07,
+      "loss": 1.544,
+      "step": 5676
+    },
+    {
+      "epoch": 0.39556840748353833,
+      "grad_norm": 0.7187432972009813,
+      "learning_rate": 4.864887429100803e-07,
+      "loss": 1.4934,
+      "step": 5677
+    },
+    {
+      "epoch": 0.3956380866111556,
+      "grad_norm": 0.7673309697309961,
+      "learning_rate": 4.864234605411983e-07,
+      "loss": 1.5899,
+      "step": 5678
+    },
+    {
+      "epoch": 0.395707765738773,
+      "grad_norm": 0.7442189649830544,
+      "learning_rate": 4.86358173311896e-07,
+      "loss": 1.5012,
+      "step": 5679
+    },
+    {
+      "epoch": 0.39577744486639027,
+      "grad_norm": 0.7633515183291592,
+      "learning_rate": 4.862928812253018e-07,
+      "loss": 1.6551,
+      "step": 5680
+    },
+    {
+      "epoch": 0.3958471239940076,
+      "grad_norm": 0.6976117552063694,
+      "learning_rate": 4.862275842845448e-07,
+      "loss": 1.5045,
+      "step": 5681
+    },
+    {
+      "epoch": 0.3959168031216249,
+      "grad_norm": 0.7490951654378956,
+      "learning_rate": 4.861622824927543e-07,
+      "loss": 1.5752,
+      "step": 5682
+    },
+    {
+      "epoch": 0.39598648224924227,
+      "grad_norm": 0.6767327539185277,
+      "learning_rate": 4.860969758530593e-07,
+      "loss": 1.497,
+      "step": 5683
+    },
+    {
+      "epoch": 0.39605616137685956,
+      "grad_norm": 0.6841274638774705,
+      "learning_rate": 4.860316643685898e-07,
+      "loss": 1.5006,
+      "step": 5684
+    },
+    {
+      "epoch": 0.3961258405044769,
+      "grad_norm": 0.7178087195736019,
+      "learning_rate": 4.859663480424755e-07,
+      "loss": 1.6026,
+      "step": 5685
+    },
+    {
+      "epoch": 0.3961955196320942,
+      "grad_norm": 0.7031296552917827,
+      "learning_rate": 4.859010268778465e-07,
+      "loss": 1.6191,
+      "step": 5686
+    },
+    {
+      "epoch": 0.3962651987597115,
+      "grad_norm": 0.70942369444354,
+      "learning_rate": 4.858357008778333e-07,
+      "loss": 1.6875,
+      "step": 5687
+    },
+    {
+      "epoch": 0.39633487788732885,
+      "grad_norm": 0.7593496749721073,
+      "learning_rate": 4.857703700455662e-07,
+      "loss": 1.4996,
+      "step": 5688
+    },
+    {
+      "epoch": 0.39640455701494615,
+      "grad_norm": 0.7200832749609688,
+      "learning_rate": 4.85705034384176e-07,
+      "loss": 1.5551,
+      "step": 5689
+    },
+    {
+      "epoch": 0.3964742361425635,
+      "grad_norm": 0.6920762660051075,
+      "learning_rate": 4.856396938967939e-07,
+      "loss": 1.5067,
+      "step": 5690
+    },
+    {
+      "epoch": 0.3965439152701808,
+      "grad_norm": 0.7669411676412027,
+      "learning_rate": 4.855743485865511e-07,
+      "loss": 1.4865,
+      "step": 5691
+    },
+    {
+      "epoch": 0.39661359439779814,
+      "grad_norm": 0.7247852762545683,
+      "learning_rate": 4.85508998456579e-07,
+      "loss": 1.5194,
+      "step": 5692
+    },
+    {
+      "epoch": 0.39668327352541544,
+      "grad_norm": 0.6844608804865571,
+      "learning_rate": 4.854436435100093e-07,
+      "loss": 1.5579,
+      "step": 5693
+    },
+    {
+      "epoch": 0.3967529526530328,
+      "grad_norm": 0.8003618335097139,
+      "learning_rate": 4.85378283749974e-07,
+      "loss": 1.5606,
+      "step": 5694
+    },
+    {
+      "epoch": 0.3968226317806501,
+      "grad_norm": 0.6849811665507215,
+      "learning_rate": 4.853129191796053e-07,
+      "loss": 1.3802,
+      "step": 5695
+    },
+    {
+      "epoch": 0.39689231090826743,
+      "grad_norm": 0.6944625667383478,
+      "learning_rate": 4.852475498020355e-07,
+      "loss": 1.4661,
+      "step": 5696
+    },
+    {
+      "epoch": 0.3969619900358847,
+      "grad_norm": 0.7215389063076278,
+      "learning_rate": 4.851821756203975e-07,
+      "loss": 1.5935,
+      "step": 5697
+    },
+    {
+      "epoch": 0.3970316691635021,
+      "grad_norm": 0.7034503235896736,
+      "learning_rate": 4.851167966378238e-07,
+      "loss": 1.6066,
+      "step": 5698
+    },
+    {
+      "epoch": 0.39710134829111937,
+      "grad_norm": 0.7325183000161554,
+      "learning_rate": 4.850514128574478e-07,
+      "loss": 1.5436,
+      "step": 5699
+    },
+    {
+      "epoch": 0.3971710274187367,
+      "grad_norm": 0.7186466732509661,
+      "learning_rate": 4.849860242824026e-07,
+      "loss": 1.4569,
+      "step": 5700
+    },
+    {
+      "epoch": 0.397240706546354,
+      "grad_norm": 0.69751502384842,
+      "learning_rate": 4.84920630915822e-07,
+      "loss": 1.4414,
+      "step": 5701
+    },
+    {
+      "epoch": 0.39731038567397137,
+      "grad_norm": 0.7034771671737267,
+      "learning_rate": 4.848552327608393e-07,
+      "loss": 1.4511,
+      "step": 5702
+    },
+    {
+      "epoch": 0.39738006480158866,
+      "grad_norm": 0.7062965342234896,
+      "learning_rate": 4.847898298205892e-07,
+      "loss": 1.5185,
+      "step": 5703
+    },
+    {
+      "epoch": 0.397449743929206,
+      "grad_norm": 0.7433825843030339,
+      "learning_rate": 4.847244220982053e-07,
+      "loss": 1.5063,
+      "step": 5704
+    },
+    {
+      "epoch": 0.3975194230568233,
+      "grad_norm": 1.0650085692078823,
+      "learning_rate": 4.846590095968226e-07,
+      "loss": 1.5949,
+      "step": 5705
+    },
+    {
+      "epoch": 0.39758910218444066,
+      "grad_norm": 0.7328515507980331,
+      "learning_rate": 4.845935923195755e-07,
+      "loss": 1.543,
+      "step": 5706
+    },
+    {
+      "epoch": 0.39765878131205795,
+      "grad_norm": 0.7022710312029437,
+      "learning_rate": 4.845281702695989e-07,
+      "loss": 1.6136,
+      "step": 5707
+    },
+    {
+      "epoch": 0.3977284604396753,
+      "grad_norm": 0.712743673405011,
+      "learning_rate": 4.844627434500282e-07,
+      "loss": 1.4946,
+      "step": 5708
+    },
+    {
+      "epoch": 0.3977981395672926,
+      "grad_norm": 0.7208989189265302,
+      "learning_rate": 4.843973118639986e-07,
+      "loss": 1.5949,
+      "step": 5709
+    },
+    {
+      "epoch": 0.39786781869490995,
+      "grad_norm": 0.6990871375396196,
+      "learning_rate": 4.843318755146456e-07,
+      "loss": 1.5536,
+      "step": 5710
+    },
+    {
+      "epoch": 0.39793749782252724,
+      "grad_norm": 0.7238622642027562,
+      "learning_rate": 4.842664344051053e-07,
+      "loss": 1.4782,
+      "step": 5711
+    },
+    {
+      "epoch": 0.3980071769501446,
+      "grad_norm": 0.7130800444416632,
+      "learning_rate": 4.842009885385137e-07,
+      "loss": 1.4871,
+      "step": 5712
+    },
+    {
+      "epoch": 0.3980768560777619,
+      "grad_norm": 0.660704364407987,
+      "learning_rate": 4.841355379180071e-07,
+      "loss": 1.4377,
+      "step": 5713
+    },
+    {
+      "epoch": 0.39814653520537924,
+      "grad_norm": 0.7016372786912725,
+      "learning_rate": 4.840700825467219e-07,
+      "loss": 1.5465,
+      "step": 5714
+    },
+    {
+      "epoch": 0.39821621433299653,
+      "grad_norm": 0.8233394310985324,
+      "learning_rate": 4.84004622427795e-07,
+      "loss": 1.5596,
+      "step": 5715
+    },
+    {
+      "epoch": 0.3982858934606139,
+      "grad_norm": 0.7431768553667599,
+      "learning_rate": 4.839391575643634e-07,
+      "loss": 1.5784,
+      "step": 5716
+    },
+    {
+      "epoch": 0.3983555725882312,
+      "grad_norm": 0.7286515546228858,
+      "learning_rate": 4.838736879595643e-07,
+      "loss": 1.5407,
+      "step": 5717
+    },
+    {
+      "epoch": 0.3984252517158485,
+      "grad_norm": 0.6803185223474442,
+      "learning_rate": 4.838082136165349e-07,
+      "loss": 1.5144,
+      "step": 5718
+    },
+    {
+      "epoch": 0.3984949308434658,
+      "grad_norm": 0.729071303691758,
+      "learning_rate": 4.837427345384132e-07,
+      "loss": 1.4729,
+      "step": 5719
+    },
+    {
+      "epoch": 0.3985646099710832,
+      "grad_norm": 0.7505050057733356,
+      "learning_rate": 4.836772507283369e-07,
+      "loss": 1.6377,
+      "step": 5720
+    },
+    {
+      "epoch": 0.39863428909870047,
+      "grad_norm": 0.6596317365173326,
+      "learning_rate": 4.836117621894442e-07,
+      "loss": 1.4454,
+      "step": 5721
+    },
+    {
+      "epoch": 0.3987039682263178,
+      "grad_norm": 0.747012650690718,
+      "learning_rate": 4.835462689248733e-07,
+      "loss": 1.4649,
+      "step": 5722
+    },
+    {
+      "epoch": 0.3987736473539351,
+      "grad_norm": 0.7566351309737823,
+      "learning_rate": 4.83480770937763e-07,
+      "loss": 1.6035,
+      "step": 5723
+    },
+    {
+      "epoch": 0.39884332648155246,
+      "grad_norm": 0.7442740560150342,
+      "learning_rate": 4.83415268231252e-07,
+      "loss": 1.4965,
+      "step": 5724
+    },
+    {
+      "epoch": 0.39891300560916976,
+      "grad_norm": 0.7856195341062227,
+      "learning_rate": 4.833497608084793e-07,
+      "loss": 1.5433,
+      "step": 5725
+    },
+    {
+      "epoch": 0.3989826847367871,
+      "grad_norm": 0.69840242548124,
+      "learning_rate": 4.83284248672584e-07,
+      "loss": 1.5195,
+      "step": 5726
+    },
+    {
+      "epoch": 0.3990523638644044,
+      "grad_norm": 0.6809083026173022,
+      "learning_rate": 4.832187318267059e-07,
+      "loss": 1.4451,
+      "step": 5727
+    },
+    {
+      "epoch": 0.39912204299202175,
+      "grad_norm": 0.6750686878569807,
+      "learning_rate": 4.831532102739844e-07,
+      "loss": 1.5211,
+      "step": 5728
+    },
+    {
+      "epoch": 0.39919172211963905,
+      "grad_norm": 0.7334443597087285,
+      "learning_rate": 4.830876840175596e-07,
+      "loss": 1.6082,
+      "step": 5729
+    },
+    {
+      "epoch": 0.3992614012472564,
+      "grad_norm": 0.68852786930121,
+      "learning_rate": 4.830221530605716e-07,
+      "loss": 1.6109,
+      "step": 5730
+    },
+    {
+      "epoch": 0.3993310803748737,
+      "grad_norm": 0.7544369302712011,
+      "learning_rate": 4.829566174061609e-07,
+      "loss": 1.5038,
+      "step": 5731
+    },
+    {
+      "epoch": 0.39940075950249104,
+      "grad_norm": 0.7049817535947259,
+      "learning_rate": 4.828910770574679e-07,
+      "loss": 1.5423,
+      "step": 5732
+    },
+    {
+      "epoch": 0.39947043863010834,
+      "grad_norm": 0.7385275567962578,
+      "learning_rate": 4.828255320176336e-07,
+      "loss": 1.5723,
+      "step": 5733
+    },
+    {
+      "epoch": 0.3995401177577257,
+      "grad_norm": 0.759768274010817,
+      "learning_rate": 4.827599822897988e-07,
+      "loss": 1.5713,
+      "step": 5734
+    },
+    {
+      "epoch": 0.399609796885343,
+      "grad_norm": 0.8286752447322427,
+      "learning_rate": 4.826944278771051e-07,
+      "loss": 1.6021,
+      "step": 5735
+    },
+    {
+      "epoch": 0.39967947601296033,
+      "grad_norm": 0.7238465828951179,
+      "learning_rate": 4.826288687826939e-07,
+      "loss": 1.6626,
+      "step": 5736
+    },
+    {
+      "epoch": 0.39974915514057763,
+      "grad_norm": 0.7770751932847518,
+      "learning_rate": 4.825633050097067e-07,
+      "loss": 1.6797,
+      "step": 5737
+    },
+    {
+      "epoch": 0.399818834268195,
+      "grad_norm": 0.7711030533066128,
+      "learning_rate": 4.824977365612857e-07,
+      "loss": 1.6196,
+      "step": 5738
+    },
+    {
+      "epoch": 0.3998885133958123,
+      "grad_norm": 0.6875827946985658,
+      "learning_rate": 4.82432163440573e-07,
+      "loss": 1.5406,
+      "step": 5739
+    },
+    {
+      "epoch": 0.3999581925234296,
+      "grad_norm": 0.7335197989377081,
+      "learning_rate": 4.82366585650711e-07,
+      "loss": 1.6377,
+      "step": 5740
+    },
+    {
+      "epoch": 0.4000278716510469,
+      "grad_norm": 0.6941435526661424,
+      "learning_rate": 4.823010031948425e-07,
+      "loss": 1.6344,
+      "step": 5741
+    },
+    {
+      "epoch": 0.40009755077866427,
+      "grad_norm": 0.6797310835665473,
+      "learning_rate": 4.8223541607611e-07,
+      "loss": 1.4352,
+      "step": 5742
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 14351,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2871,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3010137962840064.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}