diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14026 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9995623769185085,
+  "eval_steps": 500,
+  "global_step": 1999,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001000281329123816,
+      "grad_norm": 0.9744666814804077,
+      "learning_rate": 4e-05,
+      "loss": 1.9866,
+      "step": 1
+    },
+    {
+      "epoch": 0.002000562658247632,
+      "grad_norm": 0.9364674687385559,
+      "learning_rate": 8e-05,
+      "loss": 1.9045,
+      "step": 2
+    },
+    {
+      "epoch": 0.003000843987371448,
+      "grad_norm": 0.9184039831161499,
+      "learning_rate": 0.00012,
+      "loss": 1.9518,
+      "step": 3
+    },
+    {
+      "epoch": 0.004001125316495264,
+      "grad_norm": 0.9024870991706848,
+      "learning_rate": 0.00016,
+      "loss": 1.8556,
+      "step": 4
+    },
+    {
+      "epoch": 0.005001406645619081,
+      "grad_norm": 0.952398955821991,
+      "learning_rate": 0.0002,
+      "loss": 1.7932,
+      "step": 5
+    },
+    {
+      "epoch": 0.006001687974742896,
+      "grad_norm": 0.8461670875549316,
+      "learning_rate": 0.0001999331550802139,
+      "loss": 1.4715,
+      "step": 6
+    },
+    {
+      "epoch": 0.007001969303866712,
+      "grad_norm": 1.0230799913406372,
+      "learning_rate": 0.00019986631016042782,
+      "loss": 1.2318,
+      "step": 7
+    },
+    {
+      "epoch": 0.008002250632990529,
+      "grad_norm": 0.8146515488624573,
+      "learning_rate": 0.00019979946524064174,
+      "loss": 1.0062,
+      "step": 8
+    },
+    {
+      "epoch": 0.009002531962114344,
+      "grad_norm": 0.8817946910858154,
+      "learning_rate": 0.00019973262032085563,
+      "loss": 0.8552,
+      "step": 9
+    },
+    {
+      "epoch": 0.010002813291238161,
+      "grad_norm": 0.7057645916938782,
+      "learning_rate": 0.00019966577540106952,
+      "loss": 0.7432,
+      "step": 10
+    },
+    {
+      "epoch": 0.011003094620361977,
+      "grad_norm": 0.7819643020629883,
+      "learning_rate": 0.00019959893048128344,
+      "loss": 0.6843,
+      "step": 11
+    },
+    {
+      "epoch": 0.012003375949485792,
+      "grad_norm": 0.9533947110176086,
+      "learning_rate": 0.00019953208556149733,
+      "loss": 0.697,
+      "step": 12
+    },
+    {
+      "epoch": 0.01300365727860961,
+      "grad_norm": 0.7520662546157837,
+      "learning_rate": 0.00019946524064171124,
+      "loss": 0.6386,
+      "step": 13
+    },
+    {
+      "epoch": 0.014003938607733425,
+      "grad_norm": 0.38872724771499634,
+      "learning_rate": 0.00019939839572192514,
+      "loss": 0.6368,
+      "step": 14
+    },
+    {
+      "epoch": 0.01500421993685724,
+      "grad_norm": 0.3418921232223511,
+      "learning_rate": 0.00019933155080213905,
+      "loss": 0.6726,
+      "step": 15
+    },
+    {
+      "epoch": 0.016004501265981057,
+      "grad_norm": 0.48597708344459534,
+      "learning_rate": 0.00019926470588235294,
+      "loss": 0.639,
+      "step": 16
+    },
+    {
+      "epoch": 0.017004782595104875,
+      "grad_norm": 0.23545189201831818,
+      "learning_rate": 0.00019919786096256686,
+      "loss": 0.6184,
+      "step": 17
+    },
+    {
+      "epoch": 0.01800506392422869,
+      "grad_norm": 0.21088548004627228,
+      "learning_rate": 0.00019913101604278075,
+      "loss": 0.6335,
+      "step": 18
+    },
+    {
+      "epoch": 0.019005345253352506,
+      "grad_norm": 0.216445192694664,
+      "learning_rate": 0.00019906417112299467,
+      "loss": 0.6259,
+      "step": 19
+    },
+    {
+      "epoch": 0.020005626582476323,
+      "grad_norm": 0.2561822235584259,
+      "learning_rate": 0.00019899732620320856,
+      "loss": 0.6099,
+      "step": 20
+    },
+    {
+      "epoch": 0.021005907911600136,
+      "grad_norm": 0.22108854353427887,
+      "learning_rate": 0.00019893048128342245,
+      "loss": 0.5994,
+      "step": 21
+    },
+    {
+      "epoch": 0.022006189240723954,
+      "grad_norm": 0.19824647903442383,
+      "learning_rate": 0.00019886363636363637,
+      "loss": 0.6077,
+      "step": 22
+    },
+    {
+      "epoch": 0.02300647056984777,
+      "grad_norm": 0.1788371354341507,
+      "learning_rate": 0.0001987967914438503,
+      "loss": 0.5628,
+      "step": 23
+    },
+    {
+      "epoch": 0.024006751898971584,
+      "grad_norm": 0.1805984526872635,
+      "learning_rate": 0.00019872994652406418,
+      "loss": 0.5898,
+      "step": 24
+    },
+    {
+      "epoch": 0.0250070332280954,
+      "grad_norm": 0.17772608995437622,
+      "learning_rate": 0.00019866310160427807,
+      "loss": 0.5735,
+      "step": 25
+    },
+    {
+      "epoch": 0.02600731455721922,
+      "grad_norm": 0.1585753858089447,
+      "learning_rate": 0.000198596256684492,
+      "loss": 0.5745,
+      "step": 26
+    },
+    {
+      "epoch": 0.027007595886343033,
+      "grad_norm": 0.16654984652996063,
+      "learning_rate": 0.0001985294117647059,
+      "loss": 0.5572,
+      "step": 27
+    },
+    {
+      "epoch": 0.02800787721546685,
+      "grad_norm": 0.1624041497707367,
+      "learning_rate": 0.0001984625668449198,
+      "loss": 0.5625,
+      "step": 28
+    },
+    {
+      "epoch": 0.029008158544590667,
+      "grad_norm": 0.17912402749061584,
+      "learning_rate": 0.0001983957219251337,
+      "loss": 0.5895,
+      "step": 29
+    },
+    {
+      "epoch": 0.03000843987371448,
+      "grad_norm": 0.16431595385074615,
+      "learning_rate": 0.0001983288770053476,
+      "loss": 0.5918,
+      "step": 30
+    },
+    {
+      "epoch": 0.031008721202838298,
+      "grad_norm": 0.1544601172208786,
+      "learning_rate": 0.00019826203208556152,
+      "loss": 0.5237,
+      "step": 31
+    },
+    {
+      "epoch": 0.032009002531962115,
+      "grad_norm": 0.17625634372234344,
+      "learning_rate": 0.00019819518716577541,
+      "loss": 0.5583,
+      "step": 32
+    },
+    {
+      "epoch": 0.03300928386108593,
+      "grad_norm": 0.15690909326076508,
+      "learning_rate": 0.0001981283422459893,
+      "loss": 0.5599,
+      "step": 33
+    },
+    {
+      "epoch": 0.03400956519020975,
+      "grad_norm": 0.15410259366035461,
+      "learning_rate": 0.00019806149732620322,
+      "loss": 0.5537,
+      "step": 34
+    },
+    {
+      "epoch": 0.03500984651933356,
+      "grad_norm": 0.1651620715856552,
+      "learning_rate": 0.0001979946524064171,
+      "loss": 0.542,
+      "step": 35
+    },
+    {
+      "epoch": 0.03601012784845738,
+      "grad_norm": 0.16008822619915009,
+      "learning_rate": 0.00019792780748663103,
+      "loss": 0.5579,
+      "step": 36
+    },
+    {
+      "epoch": 0.037010409177581194,
+      "grad_norm": 0.16451336443424225,
+      "learning_rate": 0.00019786096256684492,
+      "loss": 0.5494,
+      "step": 37
+    },
+    {
+      "epoch": 0.03801069050670501,
+      "grad_norm": 0.16137196123600006,
+      "learning_rate": 0.00019779411764705884,
+      "loss": 0.5308,
+      "step": 38
+    },
+    {
+      "epoch": 0.03901097183582883,
+      "grad_norm": 0.16389496624469757,
+      "learning_rate": 0.00019772727272727273,
+      "loss": 0.5785,
+      "step": 39
+    },
+    {
+      "epoch": 0.040011253164952645,
+      "grad_norm": 0.17476366460323334,
+      "learning_rate": 0.00019766042780748665,
+      "loss": 0.5455,
+      "step": 40
+    },
+    {
+      "epoch": 0.041011534494076456,
+      "grad_norm": 0.1790967434644699,
+      "learning_rate": 0.00019759358288770054,
+      "loss": 0.5334,
+      "step": 41
+    },
+    {
+      "epoch": 0.04201181582320027,
+      "grad_norm": 0.15343967080116272,
+      "learning_rate": 0.00019752673796791446,
+      "loss": 0.5078,
+      "step": 42
+    },
+    {
+      "epoch": 0.04301209715232409,
+      "grad_norm": 0.16167956590652466,
+      "learning_rate": 0.00019745989304812835,
+      "loss": 0.5199,
+      "step": 43
+    },
+    {
+      "epoch": 0.04401237848144791,
+      "grad_norm": 0.15342746675014496,
+      "learning_rate": 0.00019739304812834224,
+      "loss": 0.5214,
+      "step": 44
+    },
+    {
+      "epoch": 0.045012659810571724,
+      "grad_norm": 0.16108684241771698,
+      "learning_rate": 0.00019732620320855616,
+      "loss": 0.5073,
+      "step": 45
+    },
+    {
+      "epoch": 0.04601294113969554,
+      "grad_norm": 0.1659533828496933,
+      "learning_rate": 0.00019725935828877008,
+      "loss": 0.5285,
+      "step": 46
+    },
+    {
+      "epoch": 0.04701322246881936,
+      "grad_norm": 0.17120051383972168,
+      "learning_rate": 0.00019719251336898397,
+      "loss": 0.5175,
+      "step": 47
+    },
+    {
+      "epoch": 0.04801350379794317,
+      "grad_norm": 0.162173792719841,
+      "learning_rate": 0.00019712566844919786,
+      "loss": 0.5002,
+      "step": 48
+    },
+    {
+      "epoch": 0.049013785127066986,
+      "grad_norm": 0.17042332887649536,
+      "learning_rate": 0.00019705882352941177,
+      "loss": 0.5096,
+      "step": 49
+    },
+    {
+      "epoch": 0.0500140664561908,
+      "grad_norm": 0.16357912123203278,
+      "learning_rate": 0.0001969919786096257,
+      "loss": 0.5277,
+      "step": 50
+    },
+    {
+      "epoch": 0.05101434778531462,
+      "grad_norm": 0.18239375948905945,
+      "learning_rate": 0.00019692513368983958,
+      "loss": 0.5471,
+      "step": 51
+    },
+    {
+      "epoch": 0.05201462911443844,
+      "grad_norm": 0.16548025608062744,
+      "learning_rate": 0.00019685828877005347,
+      "loss": 0.5236,
+      "step": 52
+    },
+    {
+      "epoch": 0.053014910443562255,
+      "grad_norm": 0.16912941634655,
+      "learning_rate": 0.0001967914438502674,
+      "loss": 0.5052,
+      "step": 53
+    },
+    {
+      "epoch": 0.054015191772686065,
+      "grad_norm": 0.16741348803043365,
+      "learning_rate": 0.0001967245989304813,
+      "loss": 0.5248,
+      "step": 54
+    },
+    {
+      "epoch": 0.05501547310180988,
+      "grad_norm": 0.17378926277160645,
+      "learning_rate": 0.0001966577540106952,
+      "loss": 0.517,
+      "step": 55
+    },
+    {
+      "epoch": 0.0560157544309337,
+      "grad_norm": 0.17014221847057343,
+      "learning_rate": 0.0001965909090909091,
+      "loss": 0.5027,
+      "step": 56
+    },
+    {
+      "epoch": 0.05701603576005752,
+      "grad_norm": 0.17511023581027985,
+      "learning_rate": 0.000196524064171123,
+      "loss": 0.5288,
+      "step": 57
+    },
+    {
+      "epoch": 0.058016317089181334,
+      "grad_norm": 0.16974779963493347,
+      "learning_rate": 0.0001964572192513369,
+      "loss": 0.5045,
+      "step": 58
+    },
+    {
+      "epoch": 0.05901659841830515,
+      "grad_norm": 0.17105549573898315,
+      "learning_rate": 0.00019639037433155082,
+      "loss": 0.5224,
+      "step": 59
+    },
+    {
+      "epoch": 0.06001687974742896,
+      "grad_norm": 0.1767175942659378,
+      "learning_rate": 0.0001963235294117647,
+      "loss": 0.5308,
+      "step": 60
+    },
+    {
+      "epoch": 0.06101716107655278,
+      "grad_norm": 0.17822648584842682,
+      "learning_rate": 0.00019625668449197863,
+      "loss": 0.5035,
+      "step": 61
+    },
+    {
+      "epoch": 0.062017442405676595,
+      "grad_norm": 0.2005096822977066,
+      "learning_rate": 0.00019618983957219252,
+      "loss": 0.5319,
+      "step": 62
+    },
+    {
+      "epoch": 0.06301772373480041,
+      "grad_norm": 0.1792004257440567,
+      "learning_rate": 0.00019612299465240644,
+      "loss": 0.5202,
+      "step": 63
+    },
+    {
+      "epoch": 0.06401800506392423,
+      "grad_norm": 0.1759466826915741,
+      "learning_rate": 0.00019605614973262033,
+      "loss": 0.512,
+      "step": 64
+    },
+    {
+      "epoch": 0.06501828639304805,
+      "grad_norm": 0.18015748262405396,
+      "learning_rate": 0.00019598930481283424,
+      "loss": 0.5102,
+      "step": 65
+    },
+    {
+      "epoch": 0.06601856772217186,
+      "grad_norm": 0.16892513632774353,
+      "learning_rate": 0.00019592245989304814,
+      "loss": 0.5012,
+      "step": 66
+    },
+    {
+      "epoch": 0.06701884905129568,
+      "grad_norm": 0.20100218057632446,
+      "learning_rate": 0.00019585561497326203,
+      "loss": 0.5118,
+      "step": 67
+    },
+    {
+      "epoch": 0.0680191303804195,
+      "grad_norm": 0.19578860700130463,
+      "learning_rate": 0.00019578877005347594,
+      "loss": 0.5135,
+      "step": 68
+    },
+    {
+      "epoch": 0.06901941170954332,
+      "grad_norm": 0.17195583879947662,
+      "learning_rate": 0.00019572192513368986,
+      "loss": 0.5168,
+      "step": 69
+    },
+    {
+      "epoch": 0.07001969303866712,
+      "grad_norm": 0.17638365924358368,
+      "learning_rate": 0.00019565508021390375,
+      "loss": 0.4727,
+      "step": 70
+    },
+    {
+      "epoch": 0.07101997436779094,
+      "grad_norm": 0.18563459813594818,
+      "learning_rate": 0.00019558823529411764,
+      "loss": 0.5075,
+      "step": 71
+    },
+    {
+      "epoch": 0.07202025569691475,
+      "grad_norm": 0.18360653519630432,
+      "learning_rate": 0.00019552139037433156,
+      "loss": 0.491,
+      "step": 72
+    },
+    {
+      "epoch": 0.07302053702603857,
+      "grad_norm": 0.19104507565498352,
+      "learning_rate": 0.00019545454545454548,
+      "loss": 0.4864,
+      "step": 73
+    },
+    {
+      "epoch": 0.07402081835516239,
+      "grad_norm": 0.1746816188097,
+      "learning_rate": 0.00019538770053475937,
+      "loss": 0.5038,
+      "step": 74
+    },
+    {
+      "epoch": 0.0750210996842862,
+      "grad_norm": 0.17216187715530396,
+      "learning_rate": 0.00019532085561497326,
+      "loss": 0.4995,
+      "step": 75
+    },
+    {
+      "epoch": 0.07602138101341002,
+      "grad_norm": 0.19356052577495575,
+      "learning_rate": 0.00019525401069518718,
+      "loss": 0.4995,
+      "step": 76
+    },
+    {
+      "epoch": 0.07702166234253384,
+      "grad_norm": 0.17926344275474548,
+      "learning_rate": 0.00019518716577540107,
+      "loss": 0.4734,
+      "step": 77
+    },
+    {
+      "epoch": 0.07802194367165766,
+      "grad_norm": 0.18571226298809052,
+      "learning_rate": 0.000195120320855615,
+      "loss": 0.5089,
+      "step": 78
+    },
+    {
+      "epoch": 0.07902222500078147,
+      "grad_norm": 0.18840278685092926,
+      "learning_rate": 0.00019505347593582888,
+      "loss": 0.4905,
+      "step": 79
+    },
+    {
+      "epoch": 0.08002250632990529,
+      "grad_norm": 0.18464742600917816,
+      "learning_rate": 0.0001949866310160428,
+      "loss": 0.5073,
+      "step": 80
+    },
+    {
+      "epoch": 0.08102278765902911,
+      "grad_norm": 0.1761467307806015,
+      "learning_rate": 0.0001949197860962567,
+      "loss": 0.4837,
+      "step": 81
+    },
+    {
+      "epoch": 0.08202306898815291,
+      "grad_norm": 0.19362109899520874,
+      "learning_rate": 0.0001948529411764706,
+      "loss": 0.4807,
+      "step": 82
+    },
+    {
+      "epoch": 0.08302335031727673,
+      "grad_norm": 0.17530831694602966,
+      "learning_rate": 0.0001947860962566845,
+      "loss": 0.4904,
+      "step": 83
+    },
+    {
+      "epoch": 0.08402363164640055,
+      "grad_norm": 0.1803610920906067,
+      "learning_rate": 0.00019471925133689841,
+      "loss": 0.4909,
+      "step": 84
+    },
+    {
+      "epoch": 0.08502391297552436,
+      "grad_norm": 0.19175425171852112,
+      "learning_rate": 0.0001946524064171123,
+      "loss": 0.4839,
+      "step": 85
+    },
+    {
+      "epoch": 0.08602419430464818,
+      "grad_norm": 0.17780834436416626,
+      "learning_rate": 0.0001945855614973262,
+      "loss": 0.4627,
+      "step": 86
+    },
+    {
+      "epoch": 0.087024475633772,
+      "grad_norm": 0.18597464263439178,
+      "learning_rate": 0.0001945187165775401,
+      "loss": 0.4939,
+      "step": 87
+    },
+    {
+      "epoch": 0.08802475696289581,
+      "grad_norm": 0.1801149696111679,
+      "learning_rate": 0.00019445187165775403,
+      "loss": 0.4737,
+      "step": 88
+    },
+    {
+      "epoch": 0.08902503829201963,
+      "grad_norm": 0.18139155209064484,
+      "learning_rate": 0.00019438502673796792,
+      "loss": 0.4891,
+      "step": 89
+    },
+    {
+      "epoch": 0.09002531962114345,
+      "grad_norm": 0.1813085973262787,
+      "learning_rate": 0.0001943181818181818,
+      "loss": 0.4949,
+      "step": 90
+    },
+    {
+      "epoch": 0.09102560095026727,
+      "grad_norm": 0.1852802038192749,
+      "learning_rate": 0.00019425133689839573,
+      "loss": 0.4898,
+      "step": 91
+    },
+    {
+      "epoch": 0.09202588227939108,
+      "grad_norm": 0.19154123961925507,
+      "learning_rate": 0.00019418449197860965,
+      "loss": 0.4524,
+      "step": 92
+    },
+    {
+      "epoch": 0.0930261636085149,
+      "grad_norm": 0.1693875789642334,
+      "learning_rate": 0.00019411764705882354,
+      "loss": 0.4771,
+      "step": 93
+    },
+    {
+      "epoch": 0.09402644493763872,
+      "grad_norm": 0.1709897220134735,
+      "learning_rate": 0.00019405080213903743,
+      "loss": 0.5056,
+      "step": 94
+    },
+    {
+      "epoch": 0.09502672626676252,
+      "grad_norm": 0.19264625012874603,
+      "learning_rate": 0.00019398395721925135,
+      "loss": 0.5083,
+      "step": 95
+    },
+    {
+      "epoch": 0.09602700759588634,
+      "grad_norm": 0.18185105919837952,
+      "learning_rate": 0.00019391711229946527,
+      "loss": 0.4955,
+      "step": 96
+    },
+    {
+      "epoch": 0.09702728892501016,
+      "grad_norm": 0.18781577050685883,
+      "learning_rate": 0.00019385026737967916,
+      "loss": 0.5029,
+      "step": 97
+    },
+    {
+      "epoch": 0.09802757025413397,
+      "grad_norm": 0.18748541176319122,
+      "learning_rate": 0.00019378342245989305,
+      "loss": 0.5011,
+      "step": 98
+    },
+    {
+      "epoch": 0.09902785158325779,
+      "grad_norm": 0.17571064829826355,
+      "learning_rate": 0.00019371657754010697,
+      "loss": 0.4586,
+      "step": 99
+    },
+    {
+      "epoch": 0.1000281329123816,
+      "grad_norm": 0.1871766448020935,
+      "learning_rate": 0.00019364973262032086,
+      "loss": 0.4766,
+      "step": 100
+    },
+    {
+      "epoch": 0.10102841424150542,
+      "grad_norm": 0.1704852133989334,
+      "learning_rate": 0.00019358288770053477,
+      "loss": 0.4922,
+      "step": 101
+    },
+    {
+      "epoch": 0.10202869557062924,
+      "grad_norm": 0.18065835535526276,
+      "learning_rate": 0.00019351604278074867,
+      "loss": 0.4902,
+      "step": 102
+    },
+    {
+      "epoch": 0.10302897689975306,
+      "grad_norm": 0.18194100260734558,
+      "learning_rate": 0.00019344919786096258,
+      "loss": 0.4736,
+      "step": 103
+    },
+    {
+      "epoch": 0.10402925822887688,
+      "grad_norm": 0.1811995506286621,
+      "learning_rate": 0.00019338235294117647,
+      "loss": 0.486,
+      "step": 104
+    },
+    {
+      "epoch": 0.10502953955800069,
+      "grad_norm": 0.194939985871315,
+      "learning_rate": 0.0001933155080213904,
+      "loss": 0.4971,
+      "step": 105
+    },
+    {
+      "epoch": 0.10602982088712451,
+      "grad_norm": 0.18497943878173828,
+      "learning_rate": 0.00019324866310160428,
+      "loss": 0.4832,
+      "step": 106
+    },
+    {
+      "epoch": 0.10703010221624833,
+      "grad_norm": 0.1584584265947342,
+      "learning_rate": 0.0001931818181818182,
+      "loss": 0.4884,
+      "step": 107
+    },
+    {
+      "epoch": 0.10803038354537213,
+      "grad_norm": 0.1716640740633011,
+      "learning_rate": 0.0001931149732620321,
+      "loss": 0.5107,
+      "step": 108
+    },
+    {
+      "epoch": 0.10903066487449595,
+      "grad_norm": 0.17184816300868988,
+      "learning_rate": 0.00019304812834224598,
+      "loss": 0.4865,
+      "step": 109
+    },
+    {
+      "epoch": 0.11003094620361976,
+      "grad_norm": 0.1761111319065094,
+      "learning_rate": 0.0001929812834224599,
+      "loss": 0.4739,
+      "step": 110
+    },
+    {
+      "epoch": 0.11103122753274358,
+      "grad_norm": 0.17194178700447083,
+      "learning_rate": 0.00019291443850267382,
+      "loss": 0.4706,
+      "step": 111
+    },
+    {
+      "epoch": 0.1120315088618674,
+      "grad_norm": 0.18889540433883667,
+      "learning_rate": 0.0001928475935828877,
+      "loss": 0.4582,
+      "step": 112
+    },
+    {
+      "epoch": 0.11303179019099122,
+      "grad_norm": 0.20257653295993805,
+      "learning_rate": 0.0001927807486631016,
+      "loss": 0.4735,
+      "step": 113
+    },
+    {
+      "epoch": 0.11403207152011503,
+      "grad_norm": 0.17461931705474854,
+      "learning_rate": 0.00019271390374331552,
+      "loss": 0.4529,
+      "step": 114
+    },
+    {
+      "epoch": 0.11503235284923885,
+      "grad_norm": 0.1717446744441986,
+      "learning_rate": 0.00019264705882352944,
+      "loss": 0.4538,
+      "step": 115
+    },
+    {
+      "epoch": 0.11603263417836267,
+      "grad_norm": 0.1839551031589508,
+      "learning_rate": 0.00019258021390374333,
+      "loss": 0.4822,
+      "step": 116
+    },
+    {
+      "epoch": 0.11703291550748648,
+      "grad_norm": 0.19117318093776703,
+      "learning_rate": 0.00019251336898395722,
+      "loss": 0.4741,
+      "step": 117
+    },
+    {
+      "epoch": 0.1180331968366103,
+      "grad_norm": 0.17903394997119904,
+      "learning_rate": 0.00019244652406417114,
+      "loss": 0.4755,
+      "step": 118
+    },
+    {
+      "epoch": 0.11903347816573412,
+      "grad_norm": 0.18842382729053497,
+      "learning_rate": 0.00019237967914438503,
+      "loss": 0.4995,
+      "step": 119
+    },
+    {
+      "epoch": 0.12003375949485792,
+      "grad_norm": 0.18437859416007996,
+      "learning_rate": 0.00019231283422459894,
+      "loss": 0.5088,
+      "step": 120
+    },
+    {
+      "epoch": 0.12103404082398174,
+      "grad_norm": 0.17745694518089294,
+      "learning_rate": 0.00019224598930481283,
+      "loss": 0.4565,
+      "step": 121
+    },
+    {
+      "epoch": 0.12203432215310556,
+      "grad_norm": 0.17243270576000214,
+      "learning_rate": 0.00019217914438502675,
+      "loss": 0.4642,
+      "step": 122
+    },
+    {
+      "epoch": 0.12303460348222937,
+      "grad_norm": 0.2010149359703064,
+      "learning_rate": 0.00019211229946524064,
+      "loss": 0.4872,
+      "step": 123
+    },
+    {
+      "epoch": 0.12403488481135319,
+      "grad_norm": 0.19740119576454163,
+      "learning_rate": 0.00019204545454545456,
+      "loss": 0.5071,
+      "step": 124
+    },
+    {
+      "epoch": 0.12503516614047702,
+      "grad_norm": 0.17282332479953766,
+      "learning_rate": 0.00019197860962566845,
+      "loss": 0.4748,
+      "step": 125
+    },
+    {
+      "epoch": 0.12603544746960083,
+      "grad_norm": 0.2022191882133484,
+      "learning_rate": 0.00019191176470588237,
+      "loss": 0.4593,
+      "step": 126
+    },
+    {
+      "epoch": 0.12703572879872463,
+      "grad_norm": 0.20872025191783905,
+      "learning_rate": 0.00019184491978609626,
+      "loss": 0.4573,
+      "step": 127
+    },
+    {
+      "epoch": 0.12803601012784846,
+      "grad_norm": 0.17185282707214355,
+      "learning_rate": 0.00019177807486631015,
+      "loss": 0.4589,
+      "step": 128
+    },
+    {
+      "epoch": 0.12903629145697226,
+      "grad_norm": 0.20027658343315125,
+      "learning_rate": 0.00019171122994652407,
+      "loss": 0.4501,
+      "step": 129
+    },
+    {
+      "epoch": 0.1300365727860961,
+      "grad_norm": 0.20652532577514648,
+      "learning_rate": 0.000191644385026738,
+      "loss": 0.4961,
+      "step": 130
+    },
+    {
+      "epoch": 0.1310368541152199,
+      "grad_norm": 0.17151765525341034,
+      "learning_rate": 0.00019157754010695188,
+      "loss": 0.455,
+      "step": 131
+    },
+    {
+      "epoch": 0.13203713544434373,
+      "grad_norm": 0.1886252760887146,
+      "learning_rate": 0.00019151069518716577,
+      "loss": 0.486,
+      "step": 132
+    },
+    {
+      "epoch": 0.13303741677346753,
+      "grad_norm": 0.22897754609584808,
+      "learning_rate": 0.0001914438502673797,
+      "loss": 0.4953,
+      "step": 133
+    },
+    {
+      "epoch": 0.13403769810259136,
+      "grad_norm": 0.18379279971122742,
+      "learning_rate": 0.0001913770053475936,
+      "loss": 0.4931,
+      "step": 134
+    },
+    {
+      "epoch": 0.13503797943171517,
+      "grad_norm": 0.18525294959545135,
+      "learning_rate": 0.0001913101604278075,
+      "loss": 0.4543,
+      "step": 135
+    },
+    {
+      "epoch": 0.136038260760839,
+      "grad_norm": 0.17013776302337646,
+      "learning_rate": 0.0001912433155080214,
+      "loss": 0.4787,
+      "step": 136
+    },
+    {
+      "epoch": 0.1370385420899628,
+      "grad_norm": 0.1702297180891037,
+      "learning_rate": 0.0001911764705882353,
+      "loss": 0.4459,
+      "step": 137
+    },
+    {
+      "epoch": 0.13803882341908663,
+      "grad_norm": 0.17295871675014496,
+      "learning_rate": 0.00019110962566844922,
+      "loss": 0.4694,
+      "step": 138
+    },
+    {
+      "epoch": 0.13903910474821043,
+      "grad_norm": 0.16555902361869812,
+      "learning_rate": 0.0001910427807486631,
+      "loss": 0.4486,
+      "step": 139
+    },
+    {
+      "epoch": 0.14003938607733424,
+      "grad_norm": 0.1774980127811432,
+      "learning_rate": 0.000190975935828877,
+      "loss": 0.4728,
+      "step": 140
+    },
+    {
+      "epoch": 0.14103966740645807,
+      "grad_norm": 0.19104857742786407,
+      "learning_rate": 0.00019090909090909092,
+      "loss": 0.4583,
+      "step": 141
+    },
+    {
+      "epoch": 0.14203994873558187,
+      "grad_norm": 0.194419264793396,
+      "learning_rate": 0.0001908422459893048,
+      "loss": 0.4473,
+      "step": 142
+    },
+    {
+      "epoch": 0.1430402300647057,
+      "grad_norm": 0.18389761447906494,
+      "learning_rate": 0.00019077540106951873,
+      "loss": 0.5063,
+      "step": 143
+    },
+    {
+      "epoch": 0.1440405113938295,
+      "grad_norm": 0.1690511256456375,
+      "learning_rate": 0.00019070855614973262,
+      "loss": 0.4406,
+      "step": 144
+    },
+    {
+      "epoch": 0.14504079272295334,
+      "grad_norm": 0.18417517840862274,
+      "learning_rate": 0.00019064171122994654,
+      "loss": 0.5078,
+      "step": 145
+    },
+    {
+      "epoch": 0.14604107405207714,
+      "grad_norm": 0.1713017076253891,
+      "learning_rate": 0.00019057486631016043,
+      "loss": 0.493,
+      "step": 146
+    },
+    {
+      "epoch": 0.14704135538120097,
+      "grad_norm": 0.18398739397525787,
+      "learning_rate": 0.00019050802139037435,
+      "loss": 0.4594,
+      "step": 147
+    },
+    {
+      "epoch": 0.14804163671032478,
+      "grad_norm": 0.1759570688009262,
+      "learning_rate": 0.00019044117647058824,
+      "loss": 0.4714,
+      "step": 148
+    },
+    {
+      "epoch": 0.1490419180394486,
+      "grad_norm": 0.17486704885959625,
+      "learning_rate": 0.00019037433155080216,
+      "loss": 0.4855,
+      "step": 149
+    },
+    {
+      "epoch": 0.1500421993685724,
+      "grad_norm": 0.16296333074569702,
+      "learning_rate": 0.00019030748663101605,
+      "loss": 0.4766,
+      "step": 150
+    },
+    {
+      "epoch": 0.15104248069769624,
+      "grad_norm": 0.176128551363945,
+      "learning_rate": 0.00019024064171122994,
+      "loss": 0.4495,
+      "step": 151
+    },
+    {
+      "epoch": 0.15204276202682004,
+      "grad_norm": 0.191524475812912,
+      "learning_rate": 0.00019017379679144388,
+      "loss": 0.5134,
+      "step": 152
+    },
+    {
+      "epoch": 0.15304304335594385,
+      "grad_norm": 0.1700233370065689,
+      "learning_rate": 0.00019010695187165777,
+      "loss": 0.4567,
+      "step": 153
+    },
+    {
+      "epoch": 0.15404332468506768,
+      "grad_norm": 0.20025531947612762,
+      "learning_rate": 0.00019004010695187167,
+      "loss": 0.4878,
+      "step": 154
+    },
+    {
+      "epoch": 0.15504360601419148,
+      "grad_norm": 0.19142255187034607,
+      "learning_rate": 0.00018997326203208556,
+      "loss": 0.5049,
+      "step": 155
+    },
+    {
+      "epoch": 0.1560438873433153,
+      "grad_norm": 0.18157531321048737,
+      "learning_rate": 0.00018990641711229947,
+      "loss": 0.4678,
+      "step": 156
+    },
+    {
+      "epoch": 0.15704416867243912,
+      "grad_norm": 0.1831781417131424,
+      "learning_rate": 0.0001898395721925134,
+      "loss": 0.4737,
+      "step": 157
+    },
+    {
+      "epoch": 0.15804445000156295,
+      "grad_norm": 0.18358385562896729,
+      "learning_rate": 0.00018977272727272728,
+      "loss": 0.4813,
+      "step": 158
+    },
+    {
+      "epoch": 0.15904473133068675,
+      "grad_norm": 0.18451449275016785,
+      "learning_rate": 0.00018970588235294117,
+      "loss": 0.4726,
+      "step": 159
+    },
+    {
+      "epoch": 0.16004501265981058,
+      "grad_norm": 0.17466507852077484,
+      "learning_rate": 0.0001896390374331551,
+      "loss": 0.4616,
+      "step": 160
+    },
+    {
+      "epoch": 0.16104529398893438,
+      "grad_norm": 0.17851275205612183,
+      "learning_rate": 0.000189572192513369,
+      "loss": 0.4755,
+      "step": 161
+    },
+    {
+      "epoch": 0.16204557531805822,
+      "grad_norm": 0.17984986305236816,
+      "learning_rate": 0.0001895053475935829,
+      "loss": 0.4842,
+      "step": 162
+    },
+    {
+      "epoch": 0.16304585664718202,
+      "grad_norm": 0.1726839244365692,
+      "learning_rate": 0.0001894385026737968,
+      "loss": 0.4654,
+      "step": 163
+    },
+    {
+      "epoch": 0.16404613797630582,
+      "grad_norm": 0.1909494400024414,
+      "learning_rate": 0.0001893716577540107,
+      "loss": 0.466,
+      "step": 164
+    },
+    {
+      "epoch": 0.16504641930542965,
+      "grad_norm": 0.16181348264217377,
+      "learning_rate": 0.0001893048128342246,
+      "loss": 0.444,
+      "step": 165
+    },
+    {
+      "epoch": 0.16604670063455346,
+      "grad_norm": 0.17508991062641144,
+      "learning_rate": 0.00018923796791443852,
+      "loss": 0.451,
+      "step": 166
+    },
+    {
+      "epoch": 0.1670469819636773,
+      "grad_norm": 0.1983264833688736,
+      "learning_rate": 0.0001891711229946524,
+      "loss": 0.4763,
+      "step": 167
+    },
+    {
+      "epoch": 0.1680472632928011,
+      "grad_norm": 0.16887639462947845,
+      "learning_rate": 0.00018910427807486633,
+      "loss": 0.4481,
+      "step": 168
+    },
+    {
+      "epoch": 0.16904754462192492,
+      "grad_norm": 0.18914102017879486,
+      "learning_rate": 0.00018903743315508022,
+      "loss": 0.4888,
+      "step": 169
+    },
+    {
+      "epoch": 0.17004782595104873,
+      "grad_norm": 0.1777648627758026,
+      "learning_rate": 0.00018897058823529413,
+      "loss": 0.4541,
+      "step": 170
+    },
+    {
+      "epoch": 0.17104810728017256,
+      "grad_norm": 0.17672370374202728,
+      "learning_rate": 0.00018890374331550803,
+      "loss": 0.4518,
+      "step": 171
+    },
+    {
+      "epoch": 0.17204838860929636,
+      "grad_norm": 0.19094176590442657,
+      "learning_rate": 0.00018883689839572194,
+      "loss": 0.4672,
+      "step": 172
+    },
+    {
+      "epoch": 0.1730486699384202,
+      "grad_norm": 0.18507908284664154,
+      "learning_rate": 0.00018877005347593583,
+      "loss": 0.465,
+      "step": 173
+    },
+    {
+      "epoch": 0.174048951267544,
+      "grad_norm": 0.17785289883613586,
+      "learning_rate": 0.00018870320855614973,
+      "loss": 0.4934,
+      "step": 174
+    },
+    {
+      "epoch": 0.17504923259666783,
+      "grad_norm": 0.1623869389295578,
+      "learning_rate": 0.00018863636363636364,
+      "loss": 0.4614,
+      "step": 175
+    },
+    {
+      "epoch": 0.17604951392579163,
+      "grad_norm": 0.1781870424747467,
+      "learning_rate": 0.00018856951871657756,
+      "loss": 0.4428,
+      "step": 176
+    },
+    {
+      "epoch": 0.17704979525491543,
+      "grad_norm": 0.1758892834186554,
+      "learning_rate": 0.00018850267379679145,
+      "loss": 0.459,
+      "step": 177
+    },
+    {
+      "epoch": 0.17805007658403926,
+      "grad_norm": 0.16301238536834717,
+      "learning_rate": 0.00018843582887700534,
+      "loss": 0.4655,
+      "step": 178
+    },
+    {
+      "epoch": 0.17905035791316307,
+      "grad_norm": 0.1834089159965515,
+      "learning_rate": 0.00018836898395721926,
+      "loss": 0.4371,
+      "step": 179
+    },
+    {
+      "epoch": 0.1800506392422869,
+      "grad_norm": 0.19042764604091644,
+      "learning_rate": 0.00018830213903743318,
+      "loss": 0.443,
+      "step": 180
+    },
+    {
+      "epoch": 0.1810509205714107,
+      "grad_norm": 0.18761961162090302,
+      "learning_rate": 0.00018823529411764707,
+      "loss": 0.4698,
+      "step": 181
+    },
+    {
+      "epoch": 0.18205120190053453,
+      "grad_norm": 0.18886564671993256,
+      "learning_rate": 0.00018816844919786096,
+      "loss": 0.4751,
+      "step": 182
+    },
+    {
+      "epoch": 0.18305148322965833,
+      "grad_norm": 0.17123280465602875,
+      "learning_rate": 0.00018810160427807488,
+      "loss": 0.4504,
+      "step": 183
+    },
+    {
+      "epoch": 0.18405176455878217,
+      "grad_norm": 0.19330959022045135,
+      "learning_rate": 0.00018803475935828877,
+      "loss": 0.4919,
+      "step": 184
+    },
+    {
+      "epoch": 0.18505204588790597,
+      "grad_norm": 0.18290665745735168,
+      "learning_rate": 0.0001879679144385027,
+      "loss": 0.4474,
+      "step": 185
+    },
+    {
+      "epoch": 0.1860523272170298,
+      "grad_norm": 0.18846595287322998,
+      "learning_rate": 0.00018790106951871658,
+      "loss": 0.4569,
+      "step": 186
+    },
+    {
+      "epoch": 0.1870526085461536,
+      "grad_norm": 0.1864161491394043,
+      "learning_rate": 0.0001878342245989305,
+      "loss": 0.4618,
+      "step": 187
+    },
+    {
+      "epoch": 0.18805288987527743,
+      "grad_norm": 0.1784975528717041,
+      "learning_rate": 0.00018776737967914439,
+      "loss": 0.4541,
+      "step": 188
+    },
+    {
+      "epoch": 0.18905317120440124,
+      "grad_norm": 0.18015554547309875,
+      "learning_rate": 0.0001877005347593583,
+      "loss": 0.4566,
+      "step": 189
+    },
+    {
+      "epoch": 0.19005345253352504,
+      "grad_norm": 0.17766129970550537,
+      "learning_rate": 0.0001876336898395722,
+      "loss": 0.4825,
+      "step": 190
+    },
+    {
+      "epoch": 0.19105373386264887,
+      "grad_norm": 0.173505499958992,
+      "learning_rate": 0.0001875668449197861,
+      "loss": 0.4657,
+      "step": 191
+    },
+    {
+      "epoch": 0.19205401519177268,
+      "grad_norm": 0.17310304939746857,
+      "learning_rate": 0.0001875,
+      "loss": 0.4356,
+      "step": 192
+    },
+    {
+      "epoch": 0.1930542965208965,
+      "grad_norm": 0.19145220518112183,
+      "learning_rate": 0.0001874331550802139,
+      "loss": 0.4787,
+      "step": 193
+    },
+    {
+      "epoch": 0.1940545778500203,
+      "grad_norm": 0.17684215307235718,
+      "learning_rate": 0.00018736631016042784,
+      "loss": 0.4627,
+      "step": 194
+    },
+    {
+      "epoch": 0.19505485917914414,
+      "grad_norm": 0.18631704151630402,
+      "learning_rate": 0.00018729946524064173,
+      "loss": 0.447,
+      "step": 195
+    },
+    {
+      "epoch": 0.19605514050826794,
+      "grad_norm": 0.19067594408988953,
+      "learning_rate": 0.00018723262032085562,
+      "loss": 0.4646,
+      "step": 196
+    },
+    {
+      "epoch": 0.19705542183739178,
+      "grad_norm": 0.17477108538150787,
+      "learning_rate": 0.0001871657754010695,
+      "loss": 0.4656,
+      "step": 197
+    },
+    {
+      "epoch": 0.19805570316651558,
+      "grad_norm": 0.18243315815925598,
+      "learning_rate": 0.00018709893048128343,
+      "loss": 0.4692,
+      "step": 198
+    },
+    {
+      "epoch": 0.1990559844956394,
+      "grad_norm": 0.17050983011722565,
+      "learning_rate": 0.00018703208556149735,
+      "loss": 0.4406,
+      "step": 199
+    },
+    {
+      "epoch": 0.2000562658247632,
+      "grad_norm": 0.17637942731380463,
+      "learning_rate": 0.00018696524064171124,
+      "loss": 0.455,
+      "step": 200
+    },
+    {
+      "epoch": 0.20105654715388704,
+      "grad_norm": 0.19478300213813782,
+      "learning_rate": 0.00018689839572192513,
+      "loss": 0.44,
+      "step": 201
+    },
+    {
+      "epoch": 0.20205682848301085,
+      "grad_norm": 0.17724725604057312,
+      "learning_rate": 0.00018683155080213905,
+      "loss": 0.4577,
+      "step": 202
+    },
+    {
+      "epoch": 0.20305710981213465,
+      "grad_norm": 0.18279367685317993,
+      "learning_rate": 0.00018676470588235297,
+      "loss": 0.4457,
+      "step": 203
+    },
+    {
+      "epoch": 0.20405739114125848,
+      "grad_norm": 0.19004391133785248,
+      "learning_rate": 0.00018669786096256686,
+      "loss": 0.4398,
+      "step": 204
+    },
+    {
+      "epoch": 0.20505767247038229,
+      "grad_norm": 0.20774249732494354,
+      "learning_rate": 0.00018663101604278075,
+      "loss": 0.4504,
+      "step": 205
+    },
+    {
+      "epoch": 0.20605795379950612,
+      "grad_norm": 0.18987542390823364,
+      "learning_rate": 0.00018656417112299467,
+      "loss": 0.4714,
+      "step": 206
+    },
+    {
+      "epoch": 0.20705823512862992,
+      "grad_norm": 0.2253541797399521,
+      "learning_rate": 0.00018649732620320856,
+      "loss": 0.4674,
+      "step": 207
+    },
+    {
+      "epoch": 0.20805851645775375,
+      "grad_norm": 0.19410820305347443,
+      "learning_rate": 0.00018643048128342247,
+      "loss": 0.4561,
+      "step": 208
+    },
+    {
+      "epoch": 0.20905879778687755,
+      "grad_norm": 0.17445877194404602,
+      "learning_rate": 0.00018636363636363636,
+      "loss": 0.4272,
+      "step": 209
+    },
+    {
+      "epoch": 0.21005907911600138,
+      "grad_norm": 0.18313977122306824,
+      "learning_rate": 0.00018629679144385028,
+      "loss": 0.4982,
+      "step": 210
+    },
+    {
+      "epoch": 0.2110593604451252,
+      "grad_norm": 0.1821029633283615,
+      "learning_rate": 0.00018622994652406417,
+      "loss": 0.4556,
+      "step": 211
+    },
+    {
+      "epoch": 0.21205964177424902,
+      "grad_norm": 0.1718428134918213,
+      "learning_rate": 0.0001861631016042781,
+      "loss": 0.4434,
+      "step": 212
+    },
+    {
+      "epoch": 0.21305992310337282,
+      "grad_norm": 0.17503920197486877,
+      "learning_rate": 0.000186096256684492,
+      "loss": 0.4518,
+      "step": 213
+    },
+    {
+      "epoch": 0.21406020443249665,
+      "grad_norm": 0.19129222631454468,
+      "learning_rate": 0.0001860294117647059,
+      "loss": 0.4337,
+      "step": 214
+    },
+    {
+      "epoch": 0.21506048576162046,
+      "grad_norm": 0.18414440751075745,
+      "learning_rate": 0.0001859625668449198,
+      "loss": 0.4476,
+      "step": 215
+    },
+    {
+      "epoch": 0.21606076709074426,
+      "grad_norm": 0.1825624704360962,
+      "learning_rate": 0.00018589572192513368,
+      "loss": 0.4602,
+      "step": 216
+    },
+    {
+      "epoch": 0.2170610484198681,
+      "grad_norm": 0.20369993150234222,
+      "learning_rate": 0.0001858288770053476,
+      "loss": 0.449,
+      "step": 217
+    },
+    {
+      "epoch": 0.2180613297489919,
+      "grad_norm": 0.18359532952308655,
+      "learning_rate": 0.00018576203208556152,
+      "loss": 0.4717,
+      "step": 218
+    },
+    {
+      "epoch": 0.21906161107811573,
+      "grad_norm": 0.1811211258172989,
+      "learning_rate": 0.0001856951871657754,
+      "loss": 0.4502,
+      "step": 219
+    },
+    {
+      "epoch": 0.22006189240723953,
+      "grad_norm": 0.1943439245223999,
+      "learning_rate": 0.0001856283422459893,
+      "loss": 0.4358,
+      "step": 220
+    },
+    {
+      "epoch": 0.22106217373636336,
+      "grad_norm": 0.17112450301647186,
+      "learning_rate": 0.00018556149732620322,
+      "loss": 0.4379,
+      "step": 221
+    },
+    {
+      "epoch": 0.22206245506548716,
+      "grad_norm": 0.17877739667892456,
+      "learning_rate": 0.00018549465240641713,
+      "loss": 0.4666,
+      "step": 222
+    },
+    {
+      "epoch": 0.223062736394611,
+      "grad_norm": 0.1775282621383667,
+      "learning_rate": 0.00018542780748663103,
+      "loss": 0.4614,
+      "step": 223
+    },
+    {
+      "epoch": 0.2240630177237348,
+      "grad_norm": 0.2399931252002716,
+      "learning_rate": 0.00018536096256684492,
+      "loss": 0.469,
+      "step": 224
+    },
+    {
+      "epoch": 0.22506329905285863,
+      "grad_norm": 0.18117383122444153,
+      "learning_rate": 0.00018529411764705883,
+      "loss": 0.4533,
+      "step": 225
+    },
+    {
+      "epoch": 0.22606358038198243,
+      "grad_norm": 0.17575931549072266,
+      "learning_rate": 0.00018522727272727273,
+      "loss": 0.4488,
+      "step": 226
+    },
+    {
+      "epoch": 0.22706386171110624,
+      "grad_norm": 0.1864219754934311,
+      "learning_rate": 0.00018516042780748664,
+      "loss": 0.4783,
+      "step": 227
+    },
+    {
+      "epoch": 0.22806414304023007,
+      "grad_norm": 0.1910594254732132,
+      "learning_rate": 0.00018509358288770053,
+      "loss": 0.457,
+      "step": 228
+    },
+    {
+      "epoch": 0.22906442436935387,
+      "grad_norm": 0.19613304734230042,
+      "learning_rate": 0.00018502673796791445,
+      "loss": 0.4257,
+      "step": 229
+    },
+    {
+      "epoch": 0.2300647056984777,
+      "grad_norm": 0.18423783779144287,
+      "learning_rate": 0.00018495989304812834,
+      "loss": 0.4526,
+      "step": 230
+    },
+    {
+      "epoch": 0.2310649870276015,
+      "grad_norm": 0.17522463202476501,
+      "learning_rate": 0.00018489304812834226,
+      "loss": 0.4154,
+      "step": 231
+    },
+    {
+      "epoch": 0.23206526835672533,
+      "grad_norm": 0.1805536448955536,
+      "learning_rate": 0.00018482620320855615,
+      "loss": 0.4538,
+      "step": 232
+    },
+    {
+      "epoch": 0.23306554968584914,
+      "grad_norm": 0.1862708181142807,
+      "learning_rate": 0.00018475935828877007,
+      "loss": 0.4379,
+      "step": 233
+    },
+    {
+      "epoch": 0.23406583101497297,
+      "grad_norm": 0.1823825240135193,
+      "learning_rate": 0.00018469251336898396,
+      "loss": 0.4555,
+      "step": 234
+    },
+    {
+      "epoch": 0.23506611234409677,
+      "grad_norm": 0.19339920580387115,
+      "learning_rate": 0.00018462566844919785,
+      "loss": 0.4434,
+      "step": 235
+    },
+    {
+      "epoch": 0.2360663936732206,
+      "grad_norm": 0.18247473239898682,
+      "learning_rate": 0.0001845588235294118,
+      "loss": 0.4502,
+      "step": 236
+    },
+    {
+      "epoch": 0.2370666750023444,
+      "grad_norm": 0.18202941119670868,
+      "learning_rate": 0.0001844919786096257,
+      "loss": 0.4523,
+      "step": 237
+    },
+    {
+      "epoch": 0.23806695633146824,
+      "grad_norm": 0.16963312029838562,
+      "learning_rate": 0.00018442513368983958,
+      "loss": 0.472,
+      "step": 238
+    },
+    {
+      "epoch": 0.23906723766059204,
+      "grad_norm": 0.186141699552536,
+      "learning_rate": 0.00018435828877005347,
+      "loss": 0.4555,
+      "step": 239
+    },
+    {
+      "epoch": 0.24006751898971584,
+      "grad_norm": 0.19084033370018005,
+      "learning_rate": 0.00018429144385026739,
+      "loss": 0.4571,
+      "step": 240
+    },
+    {
+      "epoch": 0.24106780031883968,
+      "grad_norm": 0.1802280694246292,
+      "learning_rate": 0.0001842245989304813,
+      "loss": 0.4394,
+      "step": 241
+    },
+    {
+      "epoch": 0.24206808164796348,
+      "grad_norm": 0.17549753189086914,
+      "learning_rate": 0.0001841577540106952,
+      "loss": 0.4621,
+      "step": 242
+    },
+    {
+      "epoch": 0.2430683629770873,
+      "grad_norm": 0.16524949669837952,
+      "learning_rate": 0.00018409090909090909,
+      "loss": 0.4437,
+      "step": 243
+    },
+    {
+      "epoch": 0.2440686443062111,
+      "grad_norm": 0.18108424544334412,
+      "learning_rate": 0.000184024064171123,
+      "loss": 0.4701,
+      "step": 244
+    },
+    {
+      "epoch": 0.24506892563533494,
+      "grad_norm": 0.17770162224769592,
+      "learning_rate": 0.00018395721925133692,
+      "loss": 0.4459,
+      "step": 245
+    },
+    {
+      "epoch": 0.24606920696445875,
+      "grad_norm": 0.180390864610672,
+      "learning_rate": 0.0001838903743315508,
+      "loss": 0.4381,
+      "step": 246
+    },
+    {
+      "epoch": 0.24706948829358258,
+      "grad_norm": 0.17641372978687286,
+      "learning_rate": 0.0001838235294117647,
+      "loss": 0.4242,
+      "step": 247
+    },
+    {
+      "epoch": 0.24806976962270638,
+      "grad_norm": 0.1810798943042755,
+      "learning_rate": 0.00018375668449197862,
+      "loss": 0.4744,
+      "step": 248
+    },
+    {
+      "epoch": 0.2490700509518302,
+      "grad_norm": 0.18043872714042664,
+      "learning_rate": 0.0001836898395721925,
+      "loss": 0.4639,
+      "step": 249
+    },
+    {
+      "epoch": 0.25007033228095404,
+      "grad_norm": 0.18257585167884827,
+      "learning_rate": 0.00018362299465240643,
+      "loss": 0.4434,
+      "step": 250
+    },
+    {
+      "epoch": 0.2510706136100778,
+      "grad_norm": 0.1873181015253067,
+      "learning_rate": 0.00018355614973262032,
+      "loss": 0.4416,
+      "step": 251
+    },
+    {
+      "epoch": 0.25207089493920165,
+      "grad_norm": 0.19883769750595093,
+      "learning_rate": 0.00018348930481283424,
+      "loss": 0.4484,
+      "step": 252
+    },
+    {
+      "epoch": 0.2530711762683255,
+      "grad_norm": 0.19196969270706177,
+      "learning_rate": 0.00018342245989304813,
+      "loss": 0.4466,
+      "step": 253
+    },
+    {
+      "epoch": 0.25407145759744926,
+      "grad_norm": 0.1771295964717865,
+      "learning_rate": 0.00018335561497326205,
+      "loss": 0.4452,
+      "step": 254
+    },
+    {
+      "epoch": 0.2550717389265731,
+      "grad_norm": 0.17763710021972656,
+      "learning_rate": 0.00018328877005347597,
+      "loss": 0.4485,
+      "step": 255
+    },
+    {
+      "epoch": 0.2560720202556969,
+      "grad_norm": 0.18344314396381378,
+      "learning_rate": 0.00018322192513368986,
+      "loss": 0.4499,
+      "step": 256
+    },
+    {
+      "epoch": 0.25707230158482075,
+      "grad_norm": 0.1725986897945404,
+      "learning_rate": 0.00018315508021390375,
+      "loss": 0.4582,
+      "step": 257
+    },
+    {
+      "epoch": 0.2580725829139445,
+      "grad_norm": 0.179547518491745,
+      "learning_rate": 0.00018308823529411764,
+      "loss": 0.4522,
+      "step": 258
+    },
+    {
+      "epoch": 0.25907286424306836,
+      "grad_norm": 0.17802830040454865,
+      "learning_rate": 0.00018302139037433158,
+      "loss": 0.451,
+      "step": 259
+    },
+    {
+      "epoch": 0.2600731455721922,
+      "grad_norm": 0.1829269975423813,
+      "learning_rate": 0.00018295454545454547,
+      "loss": 0.494,
+      "step": 260
+    },
+    {
+      "epoch": 0.261073426901316,
+      "grad_norm": 0.17635449767112732,
+      "learning_rate": 0.00018288770053475936,
+      "loss": 0.4434,
+      "step": 261
+    },
+    {
+      "epoch": 0.2620737082304398,
+      "grad_norm": 0.1768188774585724,
+      "learning_rate": 0.00018282085561497326,
+      "loss": 0.4262,
+      "step": 262
+    },
+    {
+      "epoch": 0.2630739895595636,
+      "grad_norm": 0.16978022456169128,
+      "learning_rate": 0.00018275401069518717,
+      "loss": 0.4131,
+      "step": 263
+    },
+    {
+      "epoch": 0.26407427088868746,
+      "grad_norm": 0.1829555630683899,
+      "learning_rate": 0.0001826871657754011,
+      "loss": 0.4738,
+      "step": 264
+    },
+    {
+      "epoch": 0.26507455221781123,
+      "grad_norm": 0.1790144294500351,
+      "learning_rate": 0.00018262032085561498,
+      "loss": 0.4591,
+      "step": 265
+    },
+    {
+      "epoch": 0.26607483354693506,
+      "grad_norm": 0.16714085638523102,
+      "learning_rate": 0.00018255347593582887,
+      "loss": 0.45,
+      "step": 266
+    },
+    {
+      "epoch": 0.2670751148760589,
+      "grad_norm": 0.1914021223783493,
+      "learning_rate": 0.0001824866310160428,
+      "loss": 0.4452,
+      "step": 267
+    },
+    {
+      "epoch": 0.2680753962051827,
+      "grad_norm": 0.1649700105190277,
+      "learning_rate": 0.0001824197860962567,
+      "loss": 0.412,
+      "step": 268
+    },
+    {
+      "epoch": 0.2690756775343065,
+      "grad_norm": 0.1769041270017624,
+      "learning_rate": 0.0001823529411764706,
+      "loss": 0.4507,
+      "step": 269
+    },
+    {
+      "epoch": 0.27007595886343033,
+      "grad_norm": 0.17330801486968994,
+      "learning_rate": 0.0001822860962566845,
+      "loss": 0.434,
+      "step": 270
+    },
+    {
+      "epoch": 0.27107624019255416,
+      "grad_norm": 0.17363230884075165,
+      "learning_rate": 0.0001822192513368984,
+      "loss": 0.4284,
+      "step": 271
+    },
+    {
+      "epoch": 0.272076521521678,
+      "grad_norm": 0.1773332804441452,
+      "learning_rate": 0.0001821524064171123,
+      "loss": 0.4501,
+      "step": 272
+    },
+    {
+      "epoch": 0.27307680285080177,
+      "grad_norm": 0.18393732607364655,
+      "learning_rate": 0.00018208556149732622,
+      "loss": 0.4541,
+      "step": 273
+    },
+    {
+      "epoch": 0.2740770841799256,
+      "grad_norm": 0.1747572124004364,
+      "learning_rate": 0.0001820187165775401,
+      "loss": 0.4458,
+      "step": 274
+    },
+    {
+      "epoch": 0.27507736550904943,
+      "grad_norm": 0.17748300731182098,
+      "learning_rate": 0.00018195187165775403,
+      "loss": 0.4446,
+      "step": 275
+    },
+    {
+      "epoch": 0.27607764683817326,
+      "grad_norm": 0.16832216084003448,
+      "learning_rate": 0.00018188502673796792,
+      "loss": 0.4394,
+      "step": 276
+    },
+    {
+      "epoch": 0.27707792816729704,
+      "grad_norm": 0.1661766767501831,
+      "learning_rate": 0.00018181818181818183,
+      "loss": 0.468,
+      "step": 277
+    },
+    {
+      "epoch": 0.27807820949642087,
+      "grad_norm": 0.17143194377422333,
+      "learning_rate": 0.00018175133689839575,
+      "loss": 0.4391,
+      "step": 278
+    },
+    {
+      "epoch": 0.2790784908255447,
+      "grad_norm": 0.18259485065937042,
+      "learning_rate": 0.00018168449197860964,
+      "loss": 0.4466,
+      "step": 279
+    },
+    {
+      "epoch": 0.2800787721546685,
+      "grad_norm": 0.17107920348644257,
+      "learning_rate": 0.00018161764705882353,
+      "loss": 0.4306,
+      "step": 280
+    },
+    {
+      "epoch": 0.2810790534837923,
+      "grad_norm": 0.17599613964557648,
+      "learning_rate": 0.00018155080213903742,
+      "loss": 0.4699,
+      "step": 281
+    },
+    {
+      "epoch": 0.28207933481291614,
+      "grad_norm": 0.1756802499294281,
+      "learning_rate": 0.00018148395721925134,
+      "loss": 0.4634,
+      "step": 282
+    },
+    {
+      "epoch": 0.28307961614203997,
+      "grad_norm": 0.17890664935112,
+      "learning_rate": 0.00018141711229946526,
+      "loss": 0.4705,
+      "step": 283
+    },
+    {
+      "epoch": 0.28407989747116374,
+      "grad_norm": 0.17714929580688477,
+      "learning_rate": 0.00018135026737967915,
+      "loss": 0.4599,
+      "step": 284
+    },
+    {
+      "epoch": 0.2850801788002876,
+      "grad_norm": 0.16694459319114685,
+      "learning_rate": 0.00018128342245989304,
+      "loss": 0.4342,
+      "step": 285
+    },
+    {
+      "epoch": 0.2860804601294114,
+      "grad_norm": 0.1781182587146759,
+      "learning_rate": 0.00018121657754010696,
+      "loss": 0.457,
+      "step": 286
+    },
+    {
+      "epoch": 0.28708074145853524,
+      "grad_norm": 0.17375454306602478,
+      "learning_rate": 0.00018114973262032088,
+      "loss": 0.4485,
+      "step": 287
+    },
+    {
+      "epoch": 0.288081022787659,
+      "grad_norm": 0.17337113618850708,
+      "learning_rate": 0.00018108288770053477,
+      "loss": 0.4567,
+      "step": 288
+    },
+    {
+      "epoch": 0.28908130411678284,
+      "grad_norm": 0.18448792397975922,
+      "learning_rate": 0.00018101604278074866,
+      "loss": 0.4494,
+      "step": 289
+    },
+    {
+      "epoch": 0.2900815854459067,
+      "grad_norm": 0.17240051925182343,
+      "learning_rate": 0.00018094919786096258,
+      "loss": 0.4401,
+      "step": 290
+    },
+    {
+      "epoch": 0.29108186677503045,
+      "grad_norm": 0.16623741388320923,
+      "learning_rate": 0.00018088235294117647,
+      "loss": 0.4243,
+      "step": 291
+    },
+    {
+      "epoch": 0.2920821481041543,
+      "grad_norm": 0.17461100220680237,
+      "learning_rate": 0.00018081550802139039,
+      "loss": 0.4369,
+      "step": 292
+    },
+    {
+      "epoch": 0.2930824294332781,
+      "grad_norm": 0.16360893845558167,
+      "learning_rate": 0.00018074866310160428,
+      "loss": 0.4424,
+      "step": 293
+    },
+    {
+      "epoch": 0.29408271076240194,
+      "grad_norm": 0.17972786724567413,
+      "learning_rate": 0.0001806818181818182,
+      "loss": 0.4625,
+      "step": 294
+    },
+    {
+      "epoch": 0.2950829920915257,
+      "grad_norm": 0.17173191905021667,
+      "learning_rate": 0.00018061497326203209,
+      "loss": 0.4336,
+      "step": 295
+    },
+    {
+      "epoch": 0.29608327342064955,
+      "grad_norm": 0.18299077451229095,
+      "learning_rate": 0.000180548128342246,
+      "loss": 0.4417,
+      "step": 296
+    },
+    {
+      "epoch": 0.2970835547497734,
+      "grad_norm": 0.1699686497449875,
+      "learning_rate": 0.00018048128342245992,
+      "loss": 0.4414,
+      "step": 297
+    },
+    {
+      "epoch": 0.2980838360788972,
+      "grad_norm": 0.1805889904499054,
+      "learning_rate": 0.0001804144385026738,
+      "loss": 0.4583,
+      "step": 298
+    },
+    {
+      "epoch": 0.299084117408021,
+      "grad_norm": 0.17077642679214478,
+      "learning_rate": 0.0001803475935828877,
+      "loss": 0.4176,
+      "step": 299
+    },
+    {
+      "epoch": 0.3000843987371448,
+      "grad_norm": 0.18295934796333313,
+      "learning_rate": 0.0001802807486631016,
+      "loss": 0.4667,
+      "step": 300
+    },
+    {
+      "epoch": 0.30108468006626865,
+      "grad_norm": 0.17283697426319122,
+      "learning_rate": 0.00018021390374331554,
+      "loss": 0.4537,
+      "step": 301
+    },
+    {
+      "epoch": 0.3020849613953925,
+      "grad_norm": 0.1809026598930359,
+      "learning_rate": 0.00018014705882352943,
+      "loss": 0.4606,
+      "step": 302
+    },
+    {
+      "epoch": 0.30308524272451626,
+      "grad_norm": 0.18748724460601807,
+      "learning_rate": 0.00018008021390374332,
+      "loss": 0.4711,
+      "step": 303
+    },
+    {
+      "epoch": 0.3040855240536401,
+      "grad_norm": 0.1709994077682495,
+      "learning_rate": 0.0001800133689839572,
+      "loss": 0.4234,
+      "step": 304
+    },
+    {
+      "epoch": 0.3050858053827639,
+      "grad_norm": 0.17212459444999695,
+      "learning_rate": 0.00017994652406417113,
+      "loss": 0.4399,
+      "step": 305
+    },
+    {
+      "epoch": 0.3060860867118877,
+      "grad_norm": 0.18580351769924164,
+      "learning_rate": 0.00017987967914438505,
+      "loss": 0.4528,
+      "step": 306
+    },
+    {
+      "epoch": 0.3070863680410115,
+      "grad_norm": 0.1647576242685318,
+      "learning_rate": 0.00017981283422459894,
+      "loss": 0.4306,
+      "step": 307
+    },
+    {
+      "epoch": 0.30808664937013536,
+      "grad_norm": 0.1691340208053589,
+      "learning_rate": 0.00017974598930481283,
+      "loss": 0.4366,
+      "step": 308
+    },
+    {
+      "epoch": 0.3090869306992592,
+      "grad_norm": 0.16741327941417694,
+      "learning_rate": 0.00017967914438502675,
+      "loss": 0.4165,
+      "step": 309
+    },
+    {
+      "epoch": 0.31008721202838296,
+      "grad_norm": 0.1716233640909195,
+      "learning_rate": 0.00017961229946524066,
+      "loss": 0.4187,
+      "step": 310
+    },
+    {
+      "epoch": 0.3110874933575068,
+      "grad_norm": 0.17521385848522186,
+      "learning_rate": 0.00017954545454545456,
+      "loss": 0.4236,
+      "step": 311
+    },
+    {
+      "epoch": 0.3120877746866306,
+      "grad_norm": 0.1715785562992096,
+      "learning_rate": 0.00017947860962566845,
+      "loss": 0.4273,
+      "step": 312
+    },
+    {
+      "epoch": 0.31308805601575446,
+      "grad_norm": 0.18492746353149414,
+      "learning_rate": 0.00017941176470588236,
+      "loss": 0.4489,
+      "step": 313
+    },
+    {
+      "epoch": 0.31408833734487823,
+      "grad_norm": 0.17954295873641968,
+      "learning_rate": 0.00017934491978609626,
+      "loss": 0.4094,
+      "step": 314
+    },
+    {
+      "epoch": 0.31508861867400206,
+      "grad_norm": 0.17855146527290344,
+      "learning_rate": 0.00017927807486631017,
+      "loss": 0.4233,
+      "step": 315
+    },
+    {
+      "epoch": 0.3160889000031259,
+      "grad_norm": 0.17364732921123505,
+      "learning_rate": 0.00017921122994652406,
+      "loss": 0.4335,
+      "step": 316
+    },
+    {
+      "epoch": 0.31708918133224967,
+      "grad_norm": 0.17259429395198822,
+      "learning_rate": 0.00017914438502673798,
+      "loss": 0.4449,
+      "step": 317
+    },
+    {
+      "epoch": 0.3180894626613735,
+      "grad_norm": 0.17149266600608826,
+      "learning_rate": 0.00017907754010695187,
+      "loss": 0.456,
+      "step": 318
+    },
+    {
+      "epoch": 0.31908974399049733,
+      "grad_norm": 0.1876770406961441,
+      "learning_rate": 0.0001790106951871658,
+      "loss": 0.4284,
+      "step": 319
+    },
+    {
+      "epoch": 0.32009002531962116,
+      "grad_norm": 0.20950675010681152,
+      "learning_rate": 0.0001789438502673797,
+      "loss": 0.4351,
+      "step": 320
+    },
+    {
+      "epoch": 0.32109030664874494,
+      "grad_norm": 0.16910240054130554,
+      "learning_rate": 0.0001788770053475936,
+      "loss": 0.4298,
+      "step": 321
+    },
+    {
+      "epoch": 0.32209058797786877,
+      "grad_norm": 0.18063318729400635,
+      "learning_rate": 0.0001788101604278075,
+      "loss": 0.4493,
+      "step": 322
+    },
+    {
+      "epoch": 0.3230908693069926,
+      "grad_norm": 0.23141996562480927,
+      "learning_rate": 0.00017874331550802138,
+      "loss": 0.4466,
+      "step": 323
+    },
+    {
+      "epoch": 0.32409115063611643,
+      "grad_norm": 0.18119311332702637,
+      "learning_rate": 0.0001786764705882353,
+      "loss": 0.4255,
+      "step": 324
+    },
+    {
+      "epoch": 0.3250914319652402,
+      "grad_norm": 0.1791078895330429,
+      "learning_rate": 0.00017860962566844922,
+      "loss": 0.4235,
+      "step": 325
+    },
+    {
+      "epoch": 0.32609171329436404,
+      "grad_norm": 0.1803896129131317,
+      "learning_rate": 0.0001785427807486631,
+      "loss": 0.4035,
+      "step": 326
+    },
+    {
+      "epoch": 0.32709199462348787,
+      "grad_norm": 0.19173625111579895,
+      "learning_rate": 0.000178475935828877,
+      "loss": 0.4676,
+      "step": 327
+    },
+    {
+      "epoch": 0.32809227595261165,
+      "grad_norm": 0.17149046063423157,
+      "learning_rate": 0.00017840909090909092,
+      "loss": 0.4358,
+      "step": 328
+    },
+    {
+      "epoch": 0.3290925572817355,
+      "grad_norm": 0.17859594523906708,
+      "learning_rate": 0.00017834224598930483,
+      "loss": 0.426,
+      "step": 329
+    },
+    {
+      "epoch": 0.3300928386108593,
+      "grad_norm": 0.1794040948152542,
+      "learning_rate": 0.00017827540106951872,
+      "loss": 0.415,
+      "step": 330
+    },
+    {
+      "epoch": 0.33109311993998314,
+      "grad_norm": 0.19304677844047546,
+      "learning_rate": 0.00017820855614973262,
+      "loss": 0.4618,
+      "step": 331
+    },
+    {
+      "epoch": 0.3320934012691069,
+      "grad_norm": 0.17321263253688812,
+      "learning_rate": 0.00017814171122994653,
+      "loss": 0.4568,
+      "step": 332
+    },
+    {
+      "epoch": 0.33309368259823074,
+      "grad_norm": 0.1944608837366104,
+      "learning_rate": 0.00017807486631016042,
+      "loss": 0.4331,
+      "step": 333
+    },
+    {
+      "epoch": 0.3340939639273546,
+      "grad_norm": 0.17461110651493073,
+      "learning_rate": 0.00017800802139037434,
+      "loss": 0.4561,
+      "step": 334
+    },
+    {
+      "epoch": 0.3350942452564784,
+      "grad_norm": 0.19736169278621674,
+      "learning_rate": 0.00017794117647058823,
+      "loss": 0.4541,
+      "step": 335
+    },
+    {
+      "epoch": 0.3360945265856022,
+      "grad_norm": 0.17889437079429626,
+      "learning_rate": 0.00017787433155080215,
+      "loss": 0.4124,
+      "step": 336
+    },
+    {
+      "epoch": 0.337094807914726,
+      "grad_norm": 0.17399004101753235,
+      "learning_rate": 0.00017780748663101604,
+      "loss": 0.4532,
+      "step": 337
+    },
+    {
+      "epoch": 0.33809508924384984,
+      "grad_norm": 0.18272772431373596,
+      "learning_rate": 0.00017774064171122996,
+      "loss": 0.4232,
+      "step": 338
+    },
+    {
+      "epoch": 0.3390953705729737,
+      "grad_norm": 0.19176781177520752,
+      "learning_rate": 0.00017767379679144388,
+      "loss": 0.4596,
+      "step": 339
+    },
+    {
+      "epoch": 0.34009565190209745,
+      "grad_norm": 0.18667066097259521,
+      "learning_rate": 0.00017760695187165777,
+      "loss": 0.4456,
+      "step": 340
+    },
+    {
+      "epoch": 0.3410959332312213,
+      "grad_norm": 0.17947258055210114,
+      "learning_rate": 0.00017754010695187166,
+      "loss": 0.4425,
+      "step": 341
+    },
+    {
+      "epoch": 0.3420962145603451,
+      "grad_norm": 0.176508367061615,
+      "learning_rate": 0.00017747326203208555,
+      "loss": 0.4376,
+      "step": 342
+    },
+    {
+      "epoch": 0.3430964958894689,
+      "grad_norm": 0.1660817414522171,
+      "learning_rate": 0.0001774064171122995,
+      "loss": 0.4451,
+      "step": 343
+    },
+    {
+      "epoch": 0.3440967772185927,
+      "grad_norm": 0.17813463509082794,
+      "learning_rate": 0.00017733957219251339,
+      "loss": 0.4719,
+      "step": 344
+    },
+    {
+      "epoch": 0.34509705854771655,
+      "grad_norm": 0.1736927628517151,
+      "learning_rate": 0.00017727272727272728,
+      "loss": 0.4535,
+      "step": 345
+    },
+    {
+      "epoch": 0.3460973398768404,
+      "grad_norm": 0.17061454057693481,
+      "learning_rate": 0.00017720588235294117,
+      "loss": 0.425,
+      "step": 346
+    },
+    {
+      "epoch": 0.34709762120596416,
+      "grad_norm": 0.18123167753219604,
+      "learning_rate": 0.00017713903743315509,
+      "loss": 0.4473,
+      "step": 347
+    },
+    {
+      "epoch": 0.348097902535088,
+      "grad_norm": 0.1735294610261917,
+      "learning_rate": 0.000177072192513369,
+      "loss": 0.4456,
+      "step": 348
+    },
+    {
+      "epoch": 0.3490981838642118,
+      "grad_norm": 0.16325166821479797,
+      "learning_rate": 0.0001770053475935829,
+      "loss": 0.4368,
+      "step": 349
+    },
+    {
+      "epoch": 0.35009846519333565,
+      "grad_norm": 0.16635017096996307,
+      "learning_rate": 0.00017693850267379679,
+      "loss": 0.4358,
+      "step": 350
+    },
+    {
+      "epoch": 0.3510987465224594,
+      "grad_norm": 0.16072800755500793,
+      "learning_rate": 0.0001768716577540107,
+      "loss": 0.4159,
+      "step": 351
+    },
+    {
+      "epoch": 0.35209902785158326,
+      "grad_norm": 0.17705607414245605,
+      "learning_rate": 0.00017680481283422462,
+      "loss": 0.4338,
+      "step": 352
+    },
+    {
+      "epoch": 0.3530993091807071,
+      "grad_norm": 0.19407153129577637,
+      "learning_rate": 0.0001767379679144385,
+      "loss": 0.4487,
+      "step": 353
+    },
+    {
+      "epoch": 0.35409959050983086,
+      "grad_norm": 0.16685500741004944,
+      "learning_rate": 0.0001766711229946524,
+      "loss": 0.4335,
+      "step": 354
+    },
+    {
+      "epoch": 0.3550998718389547,
+      "grad_norm": 0.1735745668411255,
+      "learning_rate": 0.00017660427807486632,
+      "loss": 0.4456,
+      "step": 355
+    },
+    {
+      "epoch": 0.3561001531680785,
+      "grad_norm": 0.16689899563789368,
+      "learning_rate": 0.0001765374331550802,
+      "loss": 0.4432,
+      "step": 356
+    },
+    {
+      "epoch": 0.35710043449720236,
+      "grad_norm": 0.16612571477890015,
+      "learning_rate": 0.00017647058823529413,
+      "loss": 0.4451,
+      "step": 357
+    },
+    {
+      "epoch": 0.35810071582632613,
+      "grad_norm": 0.17355671525001526,
+      "learning_rate": 0.00017640374331550802,
+      "loss": 0.428,
+      "step": 358
+    },
+    {
+      "epoch": 0.35910099715544996,
+      "grad_norm": 0.16404558718204498,
+      "learning_rate": 0.00017633689839572194,
+      "loss": 0.4178,
+      "step": 359
+    },
+    {
+      "epoch": 0.3601012784845738,
+      "grad_norm": 0.16635802388191223,
+      "learning_rate": 0.00017627005347593583,
+      "loss": 0.4171,
+      "step": 360
+    },
+    {
+      "epoch": 0.3611015598136976,
+      "grad_norm": 0.17166917026042938,
+      "learning_rate": 0.00017620320855614975,
+      "loss": 0.4354,
+      "step": 361
+    },
+    {
+      "epoch": 0.3621018411428214,
+      "grad_norm": 0.1794259250164032,
+      "learning_rate": 0.00017613636363636366,
+      "loss": 0.4425,
+      "step": 362
+    },
+    {
+      "epoch": 0.36310212247194523,
+      "grad_norm": 0.1819561868906021,
+      "learning_rate": 0.00017606951871657756,
+      "loss": 0.453,
+      "step": 363
+    },
+    {
+      "epoch": 0.36410240380106906,
+      "grad_norm": 0.178354874253273,
+      "learning_rate": 0.00017600267379679145,
+      "loss": 0.4436,
+      "step": 364
+    },
+    {
+      "epoch": 0.3651026851301929,
+      "grad_norm": 0.17876660823822021,
+      "learning_rate": 0.00017593582887700534,
+      "loss": 0.4525,
+      "step": 365
+    },
+    {
+      "epoch": 0.36610296645931667,
+      "grad_norm": 0.18647396564483643,
+      "learning_rate": 0.00017586898395721928,
+      "loss": 0.458,
+      "step": 366
+    },
+    {
+      "epoch": 0.3671032477884405,
+      "grad_norm": 0.16759748756885529,
+      "learning_rate": 0.00017580213903743317,
+      "loss": 0.422,
+      "step": 367
+    },
+    {
+      "epoch": 0.36810352911756433,
+      "grad_norm": 0.18720442056655884,
+      "learning_rate": 0.00017573529411764706,
+      "loss": 0.4482,
+      "step": 368
+    },
+    {
+      "epoch": 0.3691038104466881,
+      "grad_norm": 0.17658625543117523,
+      "learning_rate": 0.00017566844919786095,
+      "loss": 0.4303,
+      "step": 369
+    },
+    {
+      "epoch": 0.37010409177581194,
+      "grad_norm": 0.17517362534999847,
+      "learning_rate": 0.00017560160427807487,
+      "loss": 0.4577,
+      "step": 370
+    },
+    {
+      "epoch": 0.37110437310493577,
+      "grad_norm": 0.17305578291416168,
+      "learning_rate": 0.0001755347593582888,
+      "loss": 0.4162,
+      "step": 371
+    },
+    {
+      "epoch": 0.3721046544340596,
+      "grad_norm": 0.18148252367973328,
+      "learning_rate": 0.00017546791443850268,
+      "loss": 0.4323,
+      "step": 372
+    },
+    {
+      "epoch": 0.3731049357631834,
+      "grad_norm": 0.18457911908626556,
+      "learning_rate": 0.00017540106951871657,
+      "loss": 0.4302,
+      "step": 373
+    },
+    {
+      "epoch": 0.3741052170923072,
+      "grad_norm": 0.18693894147872925,
+      "learning_rate": 0.0001753342245989305,
+      "loss": 0.4458,
+      "step": 374
+    },
+    {
+      "epoch": 0.37510549842143104,
+      "grad_norm": 0.17343318462371826,
+      "learning_rate": 0.0001752673796791444,
+      "loss": 0.432,
+      "step": 375
+    },
+    {
+      "epoch": 0.37610577975055487,
+      "grad_norm": 0.170964777469635,
+      "learning_rate": 0.0001752005347593583,
+      "loss": 0.4302,
+      "step": 376
+    },
+    {
+      "epoch": 0.37710606107967864,
+      "grad_norm": 0.1706034541130066,
+      "learning_rate": 0.0001751336898395722,
+      "loss": 0.4528,
+      "step": 377
+    },
+    {
+      "epoch": 0.3781063424088025,
+      "grad_norm": 0.19066473841667175,
+      "learning_rate": 0.0001750668449197861,
+      "loss": 0.4529,
+      "step": 378
+    },
+    {
+      "epoch": 0.3791066237379263,
+      "grad_norm": 0.18090970814228058,
+      "learning_rate": 0.000175,
+      "loss": 0.4485,
+      "step": 379
+    },
+    {
+      "epoch": 0.3801069050670501,
+      "grad_norm": 0.18235009908676147,
+      "learning_rate": 0.00017493315508021392,
+      "loss": 0.4482,
+      "step": 380
+    },
+    {
+      "epoch": 0.3811071863961739,
+      "grad_norm": 0.17675139009952545,
+      "learning_rate": 0.00017486631016042783,
+      "loss": 0.4108,
+      "step": 381
+    },
+    {
+      "epoch": 0.38210746772529774,
+      "grad_norm": 0.17461447417736053,
+      "learning_rate": 0.00017479946524064172,
+      "loss": 0.4445,
+      "step": 382
+    },
+    {
+      "epoch": 0.3831077490544216,
+      "grad_norm": 0.1662529855966568,
+      "learning_rate": 0.00017473262032085562,
+      "loss": 0.406,
+      "step": 383
+    },
+    {
+      "epoch": 0.38410803038354535,
+      "grad_norm": 0.1768869012594223,
+      "learning_rate": 0.00017466577540106953,
+      "loss": 0.4205,
+      "step": 384
+    },
+    {
+      "epoch": 0.3851083117126692,
+      "grad_norm": 0.18918122351169586,
+      "learning_rate": 0.00017459893048128345,
+      "loss": 0.4807,
+      "step": 385
+    },
+    {
+      "epoch": 0.386108593041793,
+      "grad_norm": 0.18057717382907867,
+      "learning_rate": 0.00017453208556149734,
+      "loss": 0.4359,
+      "step": 386
+    },
+    {
+      "epoch": 0.38710887437091684,
+      "grad_norm": 0.17287477850914001,
+      "learning_rate": 0.00017446524064171123,
+      "loss": 0.4233,
+      "step": 387
+    },
+    {
+      "epoch": 0.3881091557000406,
+      "grad_norm": 0.1814192831516266,
+      "learning_rate": 0.00017439839572192512,
+      "loss": 0.4592,
+      "step": 388
+    },
+    {
+      "epoch": 0.38910943702916445,
+      "grad_norm": 0.17137302458286285,
+      "learning_rate": 0.00017433155080213904,
+      "loss": 0.4102,
+      "step": 389
+    },
+    {
+      "epoch": 0.3901097183582883,
+      "grad_norm": 0.17359258234500885,
+      "learning_rate": 0.00017426470588235296,
+      "loss": 0.4134,
+      "step": 390
+    },
+    {
+      "epoch": 0.39110999968741206,
+      "grad_norm": 0.17649279534816742,
+      "learning_rate": 0.00017419786096256685,
+      "loss": 0.4224,
+      "step": 391
+    },
+    {
+      "epoch": 0.3921102810165359,
+      "grad_norm": 0.1819760948419571,
+      "learning_rate": 0.00017413101604278074,
+      "loss": 0.4128,
+      "step": 392
+    },
+    {
+      "epoch": 0.3931105623456597,
+      "grad_norm": 0.17651373147964478,
+      "learning_rate": 0.00017406417112299466,
+      "loss": 0.4424,
+      "step": 393
+    },
+    {
+      "epoch": 0.39411084367478355,
+      "grad_norm": 0.18870307505130768,
+      "learning_rate": 0.00017399732620320858,
+      "loss": 0.4475,
+      "step": 394
+    },
+    {
+      "epoch": 0.3951111250039073,
+      "grad_norm": 0.17702838778495789,
+      "learning_rate": 0.00017393048128342247,
+      "loss": 0.4412,
+      "step": 395
+    },
+    {
+      "epoch": 0.39611140633303116,
+      "grad_norm": 0.17925161123275757,
+      "learning_rate": 0.00017386363636363636,
+      "loss": 0.4643,
+      "step": 396
+    },
+    {
+      "epoch": 0.397111687662155,
+      "grad_norm": 0.1825191229581833,
+      "learning_rate": 0.00017379679144385028,
+      "loss": 0.4309,
+      "step": 397
+    },
+    {
+      "epoch": 0.3981119689912788,
+      "grad_norm": 0.18673871457576752,
+      "learning_rate": 0.00017372994652406417,
+      "loss": 0.4395,
+      "step": 398
+    },
+    {
+      "epoch": 0.3991122503204026,
+      "grad_norm": 0.1835436075925827,
+      "learning_rate": 0.00017366310160427809,
+      "loss": 0.414,
+      "step": 399
+    },
+    {
+      "epoch": 0.4001125316495264,
+      "grad_norm": 0.17538726329803467,
+      "learning_rate": 0.000173596256684492,
+      "loss": 0.4544,
+      "step": 400
+    },
+    {
+      "epoch": 0.40111281297865026,
+      "grad_norm": 0.19464363157749176,
+      "learning_rate": 0.0001735294117647059,
+      "loss": 0.4288,
+      "step": 401
+    },
+    {
+      "epoch": 0.4021130943077741,
+      "grad_norm": 0.1836351454257965,
+      "learning_rate": 0.00017346256684491979,
+      "loss": 0.4205,
+      "step": 402
+    },
+    {
+      "epoch": 0.40311337563689786,
+      "grad_norm": 0.16461104154586792,
+      "learning_rate": 0.0001733957219251337,
+      "loss": 0.4642,
+      "step": 403
+    },
+    {
+      "epoch": 0.4041136569660217,
+      "grad_norm": 0.1738802194595337,
+      "learning_rate": 0.00017332887700534762,
+      "loss": 0.4529,
+      "step": 404
+    },
+    {
+      "epoch": 0.4051139382951455,
+      "grad_norm": 0.17489774525165558,
+      "learning_rate": 0.0001732620320855615,
+      "loss": 0.4581,
+      "step": 405
+    },
+    {
+      "epoch": 0.4061142196242693,
+      "grad_norm": 0.17276865243911743,
+      "learning_rate": 0.0001731951871657754,
+      "loss": 0.4546,
+      "step": 406
+    },
+    {
+      "epoch": 0.40711450095339313,
+      "grad_norm": 0.17556218802928925,
+      "learning_rate": 0.0001731283422459893,
+      "loss": 0.4435,
+      "step": 407
+    },
+    {
+      "epoch": 0.40811478228251696,
+      "grad_norm": 0.1763908714056015,
+      "learning_rate": 0.00017306149732620324,
+      "loss": 0.4279,
+      "step": 408
+    },
+    {
+      "epoch": 0.4091150636116408,
+      "grad_norm": 0.18071916699409485,
+      "learning_rate": 0.00017299465240641713,
+      "loss": 0.4058,
+      "step": 409
+    },
+    {
+      "epoch": 0.41011534494076457,
+      "grad_norm": 0.17551599442958832,
+      "learning_rate": 0.00017292780748663102,
+      "loss": 0.4368,
+      "step": 410
+    },
+    {
+      "epoch": 0.4111156262698884,
+      "grad_norm": 0.18373918533325195,
+      "learning_rate": 0.0001728609625668449,
+      "loss": 0.4376,
+      "step": 411
+    },
+    {
+      "epoch": 0.41211590759901223,
+      "grad_norm": 0.19964686036109924,
+      "learning_rate": 0.00017279411764705883,
+      "loss": 0.4556,
+      "step": 412
+    },
+    {
+      "epoch": 0.41311618892813606,
+      "grad_norm": 0.20226283371448517,
+      "learning_rate": 0.00017272727272727275,
+      "loss": 0.4216,
+      "step": 413
+    },
+    {
+      "epoch": 0.41411647025725984,
+      "grad_norm": 0.17443345487117767,
+      "learning_rate": 0.00017266042780748664,
+      "loss": 0.4299,
+      "step": 414
+    },
+    {
+      "epoch": 0.41511675158638367,
+      "grad_norm": 0.16941744089126587,
+      "learning_rate": 0.00017259358288770053,
+      "loss": 0.4352,
+      "step": 415
+    },
+    {
+      "epoch": 0.4161170329155075,
+      "grad_norm": 0.2127050906419754,
+      "learning_rate": 0.00017252673796791445,
+      "loss": 0.4499,
+      "step": 416
+    },
+    {
+      "epoch": 0.4171173142446313,
+      "grad_norm": 0.19113492965698242,
+      "learning_rate": 0.00017245989304812836,
+      "loss": 0.4335,
+      "step": 417
+    },
+    {
+      "epoch": 0.4181175955737551,
+      "grad_norm": 0.17924468219280243,
+      "learning_rate": 0.00017239304812834225,
+      "loss": 0.4076,
+      "step": 418
+    },
+    {
+      "epoch": 0.41911787690287894,
+      "grad_norm": 0.1749243140220642,
+      "learning_rate": 0.00017232620320855615,
+      "loss": 0.466,
+      "step": 419
+    },
+    {
+      "epoch": 0.42011815823200277,
+      "grad_norm": 0.18644505739212036,
+      "learning_rate": 0.00017225935828877006,
+      "loss": 0.4352,
+      "step": 420
+    },
+    {
+      "epoch": 0.42111843956112655,
+      "grad_norm": 0.17397917807102203,
+      "learning_rate": 0.00017219251336898395,
+      "loss": 0.4388,
+      "step": 421
+    },
+    {
+      "epoch": 0.4221187208902504,
+      "grad_norm": 0.17085954546928406,
+      "learning_rate": 0.00017212566844919787,
+      "loss": 0.4515,
+      "step": 422
+    },
+    {
+      "epoch": 0.4231190022193742,
+      "grad_norm": 0.16099511086940765,
+      "learning_rate": 0.0001720588235294118,
+      "loss": 0.4076,
+      "step": 423
+    },
+    {
+      "epoch": 0.42411928354849804,
+      "grad_norm": 0.17795774340629578,
+      "learning_rate": 0.00017199197860962568,
+      "loss": 0.4389,
+      "step": 424
+    },
+    {
+      "epoch": 0.4251195648776218,
+      "grad_norm": 0.18014943599700928,
+      "learning_rate": 0.00017192513368983957,
+      "loss": 0.4256,
+      "step": 425
+    },
+    {
+      "epoch": 0.42611984620674564,
+      "grad_norm": 0.18704374134540558,
+      "learning_rate": 0.0001718582887700535,
+      "loss": 0.4492,
+      "step": 426
+    },
+    {
+      "epoch": 0.4271201275358695,
+      "grad_norm": 0.16948561370372772,
+      "learning_rate": 0.0001717914438502674,
+      "loss": 0.4482,
+      "step": 427
+    },
+    {
+      "epoch": 0.4281204088649933,
+      "grad_norm": 0.17665095627307892,
+      "learning_rate": 0.0001717245989304813,
+      "loss": 0.4234,
+      "step": 428
+    },
+    {
+      "epoch": 0.4291206901941171,
+      "grad_norm": 0.19772249460220337,
+      "learning_rate": 0.0001716577540106952,
+      "loss": 0.4612,
+      "step": 429
+    },
+    {
+      "epoch": 0.4301209715232409,
+      "grad_norm": 0.17998537421226501,
+      "learning_rate": 0.00017159090909090908,
+      "loss": 0.4551,
+      "step": 430
+    },
+    {
+      "epoch": 0.43112125285236474,
+      "grad_norm": 0.17477943003177643,
+      "learning_rate": 0.000171524064171123,
+      "loss": 0.4464,
+      "step": 431
+    },
+    {
+      "epoch": 0.4321215341814885,
+      "grad_norm": 0.16711243987083435,
+      "learning_rate": 0.00017145721925133692,
+      "loss": 0.4215,
+      "step": 432
+    },
+    {
+      "epoch": 0.43312181551061235,
+      "grad_norm": 0.17099756002426147,
+      "learning_rate": 0.0001713903743315508,
+      "loss": 0.4178,
+      "step": 433
+    },
+    {
+      "epoch": 0.4341220968397362,
+      "grad_norm": 0.17479564249515533,
+      "learning_rate": 0.0001713235294117647,
+      "loss": 0.412,
+      "step": 434
+    },
+    {
+      "epoch": 0.43512237816886,
+      "grad_norm": 0.1868622601032257,
+      "learning_rate": 0.00017125668449197862,
+      "loss": 0.4246,
+      "step": 435
+    },
+    {
+      "epoch": 0.4361226594979838,
+      "grad_norm": 0.18481655418872833,
+      "learning_rate": 0.00017118983957219253,
+      "loss": 0.4274,
+      "step": 436
+    },
+    {
+      "epoch": 0.4371229408271076,
+      "grad_norm": 0.17644591629505157,
+      "learning_rate": 0.00017112299465240642,
+      "loss": 0.4475,
+      "step": 437
+    },
+    {
+      "epoch": 0.43812322215623145,
+      "grad_norm": 0.18062442541122437,
+      "learning_rate": 0.00017105614973262032,
+      "loss": 0.4413,
+      "step": 438
+    },
+    {
+      "epoch": 0.4391235034853553,
+      "grad_norm": 0.18078821897506714,
+      "learning_rate": 0.00017098930481283423,
+      "loss": 0.4315,
+      "step": 439
+    },
+    {
+      "epoch": 0.44012378481447906,
+      "grad_norm": 0.17297616600990295,
+      "learning_rate": 0.00017092245989304812,
+      "loss": 0.41,
+      "step": 440
+    },
+    {
+      "epoch": 0.4411240661436029,
+      "grad_norm": 0.18727439641952515,
+      "learning_rate": 0.00017085561497326204,
+      "loss": 0.4345,
+      "step": 441
+    },
+    {
+      "epoch": 0.4421243474727267,
+      "grad_norm": 0.18139448761940002,
+      "learning_rate": 0.00017078877005347596,
+      "loss": 0.4126,
+      "step": 442
+    },
+    {
+      "epoch": 0.4431246288018505,
+      "grad_norm": 0.17658720910549164,
+      "learning_rate": 0.00017072192513368985,
+      "loss": 0.4343,
+      "step": 443
+    },
+    {
+      "epoch": 0.4441249101309743,
+      "grad_norm": 0.17612750828266144,
+      "learning_rate": 0.00017065508021390374,
+      "loss": 0.4179,
+      "step": 444
+    },
+    {
+      "epoch": 0.44512519146009816,
+      "grad_norm": 0.19309662282466888,
+      "learning_rate": 0.00017058823529411766,
+      "loss": 0.4469,
+      "step": 445
+    },
+    {
+      "epoch": 0.446125472789222,
+      "grad_norm": 0.17980198562145233,
+      "learning_rate": 0.00017052139037433158,
+      "loss": 0.4114,
+      "step": 446
+    },
+    {
+      "epoch": 0.44712575411834576,
+      "grad_norm": 0.17348790168762207,
+      "learning_rate": 0.00017045454545454547,
+      "loss": 0.4234,
+      "step": 447
+    },
+    {
+      "epoch": 0.4481260354474696,
+      "grad_norm": 0.1644233614206314,
+      "learning_rate": 0.00017038770053475936,
+      "loss": 0.4137,
+      "step": 448
+    },
+    {
+      "epoch": 0.4491263167765934,
+      "grad_norm": 0.18324849009513855,
+      "learning_rate": 0.00017032085561497325,
+      "loss": 0.4148,
+      "step": 449
+    },
+    {
+      "epoch": 0.45012659810571726,
+      "grad_norm": 0.18216699361801147,
+      "learning_rate": 0.0001702540106951872,
+      "loss": 0.4248,
+      "step": 450
+    },
+    {
+      "epoch": 0.45112687943484103,
+      "grad_norm": 0.1735803186893463,
+      "learning_rate": 0.00017018716577540109,
+      "loss": 0.4042,
+      "step": 451
+    },
+    {
+      "epoch": 0.45212716076396486,
+      "grad_norm": 0.1867716908454895,
+      "learning_rate": 0.00017012032085561498,
+      "loss": 0.4005,
+      "step": 452
+    },
+    {
+      "epoch": 0.4531274420930887,
+      "grad_norm": 0.1966044008731842,
+      "learning_rate": 0.00017005347593582887,
+      "loss": 0.4372,
+      "step": 453
+    },
+    {
+      "epoch": 0.45412772342221247,
+      "grad_norm": 0.20946615934371948,
+      "learning_rate": 0.00016998663101604278,
+      "loss": 0.4619,
+      "step": 454
+    },
+    {
+      "epoch": 0.4551280047513363,
+      "grad_norm": 0.16564136743545532,
+      "learning_rate": 0.0001699197860962567,
+      "loss": 0.4189,
+      "step": 455
+    },
+    {
+      "epoch": 0.45612828608046013,
+      "grad_norm": 0.17898212373256683,
+      "learning_rate": 0.0001698529411764706,
+      "loss": 0.4326,
+      "step": 456
+    },
+    {
+      "epoch": 0.45712856740958396,
+      "grad_norm": 0.16770191490650177,
+      "learning_rate": 0.00016978609625668448,
+      "loss": 0.4216,
+      "step": 457
+    },
+    {
+      "epoch": 0.45812884873870774,
+      "grad_norm": 0.1793423444032669,
+      "learning_rate": 0.0001697192513368984,
+      "loss": 0.4035,
+      "step": 458
+    },
+    {
+      "epoch": 0.45912913006783157,
+      "grad_norm": 0.17702436447143555,
+      "learning_rate": 0.00016965240641711232,
+      "loss": 0.4351,
+      "step": 459
+    },
+    {
+      "epoch": 0.4601294113969554,
+      "grad_norm": 0.16669146716594696,
+      "learning_rate": 0.0001695855614973262,
+      "loss": 0.4261,
+      "step": 460
+    },
+    {
+      "epoch": 0.46112969272607923,
+      "grad_norm": 0.1851879209280014,
+      "learning_rate": 0.0001695187165775401,
+      "loss": 0.432,
+      "step": 461
+    },
+    {
+      "epoch": 0.462129974055203,
+      "grad_norm": 0.18338064849376678,
+      "learning_rate": 0.00016945187165775402,
+      "loss": 0.4399,
+      "step": 462
+    },
+    {
+      "epoch": 0.46313025538432684,
+      "grad_norm": 0.1746056079864502,
+      "learning_rate": 0.0001693850267379679,
+      "loss": 0.4243,
+      "step": 463
+    },
+    {
+      "epoch": 0.46413053671345067,
+      "grad_norm": 0.18128280341625214,
+      "learning_rate": 0.00016931818181818183,
+      "loss": 0.458,
+      "step": 464
+    },
+    {
+      "epoch": 0.4651308180425745,
+      "grad_norm": 0.17172598838806152,
+      "learning_rate": 0.00016925133689839575,
+      "loss": 0.4352,
+      "step": 465
+    },
+    {
+      "epoch": 0.4661310993716983,
+      "grad_norm": 0.16762162744998932,
+      "learning_rate": 0.00016918449197860964,
+      "loss": 0.4287,
+      "step": 466
+    },
+    {
+      "epoch": 0.4671313807008221,
+      "grad_norm": 0.17377620935440063,
+      "learning_rate": 0.00016911764705882353,
+      "loss": 0.3961,
+      "step": 467
+    },
+    {
+      "epoch": 0.46813166202994594,
+      "grad_norm": 0.16892342269420624,
+      "learning_rate": 0.00016905080213903745,
+      "loss": 0.42,
+      "step": 468
+    },
+    {
+      "epoch": 0.4691319433590697,
+      "grad_norm": 0.1723669320344925,
+      "learning_rate": 0.00016898395721925136,
+      "loss": 0.4261,
+      "step": 469
+    },
+    {
+      "epoch": 0.47013222468819355,
+      "grad_norm": 0.17638066411018372,
+      "learning_rate": 0.00016891711229946525,
+      "loss": 0.423,
+      "step": 470
+    },
+    {
+      "epoch": 0.4711325060173174,
+      "grad_norm": 0.17456413805484772,
+      "learning_rate": 0.00016885026737967915,
+      "loss": 0.4443,
+      "step": 471
+    },
+    {
+      "epoch": 0.4721327873464412,
+      "grad_norm": 0.1629389524459839,
+      "learning_rate": 0.00016878342245989304,
+      "loss": 0.4,
+      "step": 472
+    },
+    {
+      "epoch": 0.473133068675565,
+      "grad_norm": 0.17360949516296387,
+      "learning_rate": 0.00016871657754010698,
+      "loss": 0.4235,
+      "step": 473
+    },
+    {
+      "epoch": 0.4741333500046888,
+      "grad_norm": 0.17233876883983612,
+      "learning_rate": 0.00016864973262032087,
+      "loss": 0.429,
+      "step": 474
+    },
+    {
+      "epoch": 0.47513363133381264,
+      "grad_norm": 0.17649391293525696,
+      "learning_rate": 0.00016858288770053476,
+      "loss": 0.4278,
+      "step": 475
+    },
+    {
+      "epoch": 0.4761339126629365,
+      "grad_norm": 0.17380329966545105,
+      "learning_rate": 0.00016851604278074865,
+      "loss": 0.4399,
+      "step": 476
+    },
+    {
+      "epoch": 0.47713419399206025,
+      "grad_norm": 0.1689458042383194,
+      "learning_rate": 0.00016844919786096257,
+      "loss": 0.4163,
+      "step": 477
+    },
+    {
+      "epoch": 0.4781344753211841,
+      "grad_norm": 0.16643808782100677,
+      "learning_rate": 0.0001683823529411765,
+      "loss": 0.4297,
+      "step": 478
+    },
+    {
+      "epoch": 0.4791347566503079,
+      "grad_norm": 0.16798333823680878,
+      "learning_rate": 0.00016831550802139038,
+      "loss": 0.4328,
+      "step": 479
+    },
+    {
+      "epoch": 0.4801350379794317,
+      "grad_norm": 0.16696397960186005,
+      "learning_rate": 0.00016824866310160427,
+      "loss": 0.4374,
+      "step": 480
+    },
+    {
+      "epoch": 0.4811353193085555,
+      "grad_norm": 0.16460229456424713,
+      "learning_rate": 0.0001681818181818182,
+      "loss": 0.4474,
+      "step": 481
+    },
+    {
+      "epoch": 0.48213560063767935,
+      "grad_norm": 0.17829792201519012,
+      "learning_rate": 0.0001681149732620321,
+      "loss": 0.457,
+      "step": 482
+    },
+    {
+      "epoch": 0.4831358819668032,
+      "grad_norm": 0.16649176180362701,
+      "learning_rate": 0.000168048128342246,
+      "loss": 0.4198,
+      "step": 483
+    },
+    {
+      "epoch": 0.48413616329592696,
+      "grad_norm": 0.17545387148857117,
+      "learning_rate": 0.00016798128342245992,
+      "loss": 0.4472,
+      "step": 484
+    },
+    {
+      "epoch": 0.4851364446250508,
+      "grad_norm": 0.17688940465450287,
+      "learning_rate": 0.0001679144385026738,
+      "loss": 0.4283,
+      "step": 485
+    },
+    {
+      "epoch": 0.4861367259541746,
+      "grad_norm": 0.17023973166942596,
+      "learning_rate": 0.0001678475935828877,
+      "loss": 0.4163,
+      "step": 486
+    },
+    {
+      "epoch": 0.48713700728329845,
+      "grad_norm": 0.17296594381332397,
+      "learning_rate": 0.00016778074866310162,
+      "loss": 0.4154,
+      "step": 487
+    },
+    {
+      "epoch": 0.4881372886124222,
+      "grad_norm": 0.17880389094352722,
+      "learning_rate": 0.00016771390374331553,
+      "loss": 0.4237,
+      "step": 488
+    },
+    {
+      "epoch": 0.48913756994154606,
+      "grad_norm": 0.17438553273677826,
+      "learning_rate": 0.00016764705882352942,
+      "loss": 0.4164,
+      "step": 489
+    },
+    {
+      "epoch": 0.4901378512706699,
+      "grad_norm": 0.1780180037021637,
+      "learning_rate": 0.00016758021390374331,
+      "loss": 0.418,
+      "step": 490
+    },
+    {
+      "epoch": 0.4911381325997937,
+      "grad_norm": 0.18408729135990143,
+      "learning_rate": 0.00016751336898395723,
+      "loss": 0.4103,
+      "step": 491
+    },
+    {
+      "epoch": 0.4921384139289175,
+      "grad_norm": 0.17175129055976868,
+      "learning_rate": 0.00016744652406417115,
+      "loss": 0.4385,
+      "step": 492
+    },
+    {
+      "epoch": 0.4931386952580413,
+      "grad_norm": 0.1726345419883728,
+      "learning_rate": 0.00016737967914438504,
+      "loss": 0.3934,
+      "step": 493
+    },
+    {
+      "epoch": 0.49413897658716516,
+      "grad_norm": 0.1681516170501709,
+      "learning_rate": 0.00016731283422459893,
+      "loss": 0.4063,
+      "step": 494
+    },
+    {
+      "epoch": 0.49513925791628893,
+      "grad_norm": 0.1690843254327774,
+      "learning_rate": 0.00016724598930481282,
+      "loss": 0.4309,
+      "step": 495
+    },
+    {
+      "epoch": 0.49613953924541276,
+      "grad_norm": 0.1773832142353058,
+      "learning_rate": 0.00016717914438502674,
+      "loss": 0.4512,
+      "step": 496
+    },
+    {
+      "epoch": 0.4971398205745366,
+      "grad_norm": 0.16625097393989563,
+      "learning_rate": 0.00016711229946524066,
+      "loss": 0.4025,
+      "step": 497
+    },
+    {
+      "epoch": 0.4981401019036604,
+      "grad_norm": 0.17904406785964966,
+      "learning_rate": 0.00016704545454545455,
+      "loss": 0.4398,
+      "step": 498
+    },
+    {
+      "epoch": 0.4991403832327842,
+      "grad_norm": 0.16707228124141693,
+      "learning_rate": 0.00016697860962566844,
+      "loss": 0.4309,
+      "step": 499
+    },
+    {
+      "epoch": 0.5001406645619081,
+      "grad_norm": 0.16499480605125427,
+      "learning_rate": 0.00016691176470588236,
+      "loss": 0.4105,
+      "step": 500
+    },
+    {
+      "epoch": 0.5011409458910319,
+      "grad_norm": 0.1724577099084854,
+      "learning_rate": 0.00016684491978609628,
+      "loss": 0.4278,
+      "step": 501
+    },
+    {
+      "epoch": 0.5021412272201556,
+      "grad_norm": 0.18185541033744812,
+      "learning_rate": 0.00016677807486631017,
+      "loss": 0.4176,
+      "step": 502
+    },
+    {
+      "epoch": 0.5031415085492795,
+      "grad_norm": 0.1753641813993454,
+      "learning_rate": 0.00016671122994652406,
+      "loss": 0.4153,
+      "step": 503
+    },
+    {
+      "epoch": 0.5041417898784033,
+      "grad_norm": 0.17050912976264954,
+      "learning_rate": 0.00016664438502673798,
+      "loss": 0.4111,
+      "step": 504
+    },
+    {
+      "epoch": 0.5051420712075271,
+      "grad_norm": 0.17663219571113586,
+      "learning_rate": 0.00016657754010695187,
+      "loss": 0.4226,
+      "step": 505
+    },
+    {
+      "epoch": 0.506142352536651,
+      "grad_norm": 0.1756785362958908,
+      "learning_rate": 0.00016651069518716578,
+      "loss": 0.4269,
+      "step": 506
+    },
+    {
+      "epoch": 0.5071426338657747,
+      "grad_norm": 0.1745450794696808,
+      "learning_rate": 0.0001664438502673797,
+      "loss": 0.3996,
+      "step": 507
+    },
+    {
+      "epoch": 0.5081429151948985,
+      "grad_norm": 0.17337395250797272,
+      "learning_rate": 0.0001663770053475936,
+      "loss": 0.401,
+      "step": 508
+    },
+    {
+      "epoch": 0.5091431965240224,
+      "grad_norm": 0.18209293484687805,
+      "learning_rate": 0.00016631016042780748,
+      "loss": 0.4115,
+      "step": 509
+    },
+    {
+      "epoch": 0.5101434778531462,
+      "grad_norm": 0.18765206634998322,
+      "learning_rate": 0.0001662433155080214,
+      "loss": 0.4352,
+      "step": 510
+    },
+    {
+      "epoch": 0.5111437591822701,
+      "grad_norm": 0.1839272677898407,
+      "learning_rate": 0.00016617647058823532,
+      "loss": 0.3999,
+      "step": 511
+    },
+    {
+      "epoch": 0.5121440405113938,
+      "grad_norm": 0.18588875234127045,
+      "learning_rate": 0.0001661096256684492,
+      "loss": 0.4352,
+      "step": 512
+    },
+    {
+      "epoch": 0.5131443218405176,
+      "grad_norm": 0.175035297870636,
+      "learning_rate": 0.0001660427807486631,
+      "loss": 0.4256,
+      "step": 513
+    },
+    {
+      "epoch": 0.5141446031696415,
+      "grad_norm": 0.19355317950248718,
+      "learning_rate": 0.000165975935828877,
+      "loss": 0.4157,
+      "step": 514
+    },
+    {
+      "epoch": 0.5151448844987653,
+      "grad_norm": 0.18002134561538696,
+      "learning_rate": 0.00016590909090909094,
+      "loss": 0.426,
+      "step": 515
+    },
+    {
+      "epoch": 0.516145165827889,
+      "grad_norm": 0.16922691464424133,
+      "learning_rate": 0.00016584224598930483,
+      "loss": 0.421,
+      "step": 516
+    },
+    {
+      "epoch": 0.5171454471570129,
+      "grad_norm": 0.16930259764194489,
+      "learning_rate": 0.00016577540106951872,
+      "loss": 0.4082,
+      "step": 517
+    },
+    {
+      "epoch": 0.5181457284861367,
+      "grad_norm": 0.19758272171020508,
+      "learning_rate": 0.0001657085561497326,
+      "loss": 0.3815,
+      "step": 518
+    },
+    {
+      "epoch": 0.5191460098152605,
+      "grad_norm": 0.18100370466709137,
+      "learning_rate": 0.00016564171122994653,
+      "loss": 0.3952,
+      "step": 519
+    },
+    {
+      "epoch": 0.5201462911443844,
+      "grad_norm": 0.17010869085788727,
+      "learning_rate": 0.00016557486631016045,
+      "loss": 0.4332,
+      "step": 520
+    },
+    {
+      "epoch": 0.5211465724735082,
+      "grad_norm": 0.17811369895935059,
+      "learning_rate": 0.00016550802139037434,
+      "loss": 0.4138,
+      "step": 521
+    },
+    {
+      "epoch": 0.522146853802632,
+      "grad_norm": 0.16998940706253052,
+      "learning_rate": 0.00016544117647058823,
+      "loss": 0.4176,
+      "step": 522
+    },
+    {
+      "epoch": 0.5231471351317558,
+      "grad_norm": 0.16401931643486023,
+      "learning_rate": 0.00016537433155080215,
+      "loss": 0.4115,
+      "step": 523
+    },
+    {
+      "epoch": 0.5241474164608796,
+      "grad_norm": 0.16618947684764862,
+      "learning_rate": 0.00016530748663101606,
+      "loss": 0.4224,
+      "step": 524
+    },
+    {
+      "epoch": 0.5251476977900035,
+      "grad_norm": 0.16931407153606415,
+      "learning_rate": 0.00016524064171122995,
+      "loss": 0.4226,
+      "step": 525
+    },
+    {
+      "epoch": 0.5261479791191273,
+      "grad_norm": 0.16492187976837158,
+      "learning_rate": 0.00016517379679144387,
+      "loss": 0.4126,
+      "step": 526
+    },
+    {
+      "epoch": 0.527148260448251,
+      "grad_norm": 0.16903190314769745,
+      "learning_rate": 0.00016510695187165776,
+      "loss": 0.4212,
+      "step": 527
+    },
+    {
+      "epoch": 0.5281485417773749,
+      "grad_norm": 0.1670171618461609,
+      "learning_rate": 0.00016504010695187165,
+      "loss": 0.4071,
+      "step": 528
+    },
+    {
+      "epoch": 0.5291488231064987,
+      "grad_norm": 0.1678818166255951,
+      "learning_rate": 0.00016497326203208557,
+      "loss": 0.4588,
+      "step": 529
+    },
+    {
+      "epoch": 0.5301491044356225,
+      "grad_norm": 0.16249947249889374,
+      "learning_rate": 0.0001649064171122995,
+      "loss": 0.3993,
+      "step": 530
+    },
+    {
+      "epoch": 0.5311493857647464,
+      "grad_norm": 0.17632173001766205,
+      "learning_rate": 0.00016483957219251338,
+      "loss": 0.418,
+      "step": 531
+    },
+    {
+      "epoch": 0.5321496670938701,
+      "grad_norm": 0.17249086499214172,
+      "learning_rate": 0.00016477272727272727,
+      "loss": 0.409,
+      "step": 532
+    },
+    {
+      "epoch": 0.533149948422994,
+      "grad_norm": 0.1681407243013382,
+      "learning_rate": 0.0001647058823529412,
+      "loss": 0.4019,
+      "step": 533
+    },
+    {
+      "epoch": 0.5341502297521178,
+      "grad_norm": 0.16458339989185333,
+      "learning_rate": 0.0001646390374331551,
+      "loss": 0.4104,
+      "step": 534
+    },
+    {
+      "epoch": 0.5351505110812416,
+      "grad_norm": 0.17575211822986603,
+      "learning_rate": 0.000164572192513369,
+      "loss": 0.4006,
+      "step": 535
+    },
+    {
+      "epoch": 0.5361507924103655,
+      "grad_norm": 0.17218567430973053,
+      "learning_rate": 0.0001645053475935829,
+      "loss": 0.422,
+      "step": 536
+    },
+    {
+      "epoch": 0.5371510737394892,
+      "grad_norm": 0.1740548014640808,
+      "learning_rate": 0.00016443850267379678,
+      "loss": 0.4138,
+      "step": 537
+    },
+    {
+      "epoch": 0.538151355068613,
+      "grad_norm": 0.18501082062721252,
+      "learning_rate": 0.0001643716577540107,
+      "loss": 0.4371,
+      "step": 538
+    },
+    {
+      "epoch": 0.5391516363977369,
+      "grad_norm": 0.17697785794734955,
+      "learning_rate": 0.00016430481283422462,
+      "loss": 0.4163,
+      "step": 539
+    },
+    {
+      "epoch": 0.5401519177268607,
+      "grad_norm": 0.1623258739709854,
+      "learning_rate": 0.0001642379679144385,
+      "loss": 0.3901,
+      "step": 540
+    },
+    {
+      "epoch": 0.5411521990559846,
+      "grad_norm": 0.1694764792919159,
+      "learning_rate": 0.0001641711229946524,
+      "loss": 0.4227,
+      "step": 541
+    },
+    {
+      "epoch": 0.5421524803851083,
+      "grad_norm": 0.16928663849830627,
+      "learning_rate": 0.00016410427807486631,
+      "loss": 0.3991,
+      "step": 542
+    },
+    {
+      "epoch": 0.5431527617142321,
+      "grad_norm": 0.17353345453739166,
+      "learning_rate": 0.00016403743315508023,
+      "loss": 0.396,
+      "step": 543
+    },
+    {
+      "epoch": 0.544153043043356,
+      "grad_norm": 0.19051139056682587,
+      "learning_rate": 0.00016397058823529412,
+      "loss": 0.4305,
+      "step": 544
+    },
+    {
+      "epoch": 0.5451533243724798,
+      "grad_norm": 0.18180570006370544,
+      "learning_rate": 0.00016390374331550801,
+      "loss": 0.4121,
+      "step": 545
+    },
+    {
+      "epoch": 0.5461536057016035,
+      "grad_norm": 0.1944774091243744,
+      "learning_rate": 0.00016383689839572193,
+      "loss": 0.4219,
+      "step": 546
+    },
+    {
+      "epoch": 0.5471538870307274,
+      "grad_norm": 0.1738865077495575,
+      "learning_rate": 0.00016377005347593582,
+      "loss": 0.4152,
+      "step": 547
+    },
+    {
+      "epoch": 0.5481541683598512,
+      "grad_norm": 0.17925100028514862,
+      "learning_rate": 0.00016370320855614974,
+      "loss": 0.4306,
+      "step": 548
+    },
+    {
+      "epoch": 0.549154449688975,
+      "grad_norm": 0.1889563798904419,
+      "learning_rate": 0.00016363636363636366,
+      "loss": 0.429,
+      "step": 549
+    },
+    {
+      "epoch": 0.5501547310180989,
+      "grad_norm": 0.19363461434841156,
+      "learning_rate": 0.00016356951871657755,
+      "loss": 0.431,
+      "step": 550
+    },
+    {
+      "epoch": 0.5511550123472226,
+      "grad_norm": 0.17269189655780792,
+      "learning_rate": 0.00016350267379679144,
+      "loss": 0.4177,
+      "step": 551
+    },
+    {
+      "epoch": 0.5521552936763465,
+      "grad_norm": 0.1760244518518448,
+      "learning_rate": 0.00016343582887700536,
+      "loss": 0.4369,
+      "step": 552
+    },
+    {
+      "epoch": 0.5531555750054703,
+      "grad_norm": 0.1785270869731903,
+      "learning_rate": 0.00016336898395721928,
+      "loss": 0.42,
+      "step": 553
+    },
+    {
+      "epoch": 0.5541558563345941,
+      "grad_norm": 0.18411394953727722,
+      "learning_rate": 0.00016330213903743317,
+      "loss": 0.4166,
+      "step": 554
+    },
+    {
+      "epoch": 0.555156137663718,
+      "grad_norm": 0.17528562247753143,
+      "learning_rate": 0.00016323529411764706,
+      "loss": 0.4415,
+      "step": 555
+    },
+    {
+      "epoch": 0.5561564189928417,
+      "grad_norm": 0.1735825389623642,
+      "learning_rate": 0.00016316844919786095,
+      "loss": 0.4038,
+      "step": 556
+    },
+    {
+      "epoch": 0.5571567003219655,
+      "grad_norm": 0.1755310446023941,
+      "learning_rate": 0.0001631016042780749,
+      "loss": 0.4119,
+      "step": 557
+    },
+    {
+      "epoch": 0.5581569816510894,
+      "grad_norm": 0.18940363824367523,
+      "learning_rate": 0.00016303475935828878,
+      "loss": 0.417,
+      "step": 558
+    },
+    {
+      "epoch": 0.5591572629802132,
+      "grad_norm": 0.18386200070381165,
+      "learning_rate": 0.00016296791443850268,
+      "loss": 0.4166,
+      "step": 559
+    },
+    {
+      "epoch": 0.560157544309337,
+      "grad_norm": 0.17416515946388245,
+      "learning_rate": 0.00016290106951871657,
+      "loss": 0.4006,
+      "step": 560
+    },
+    {
+      "epoch": 0.5611578256384608,
+      "grad_norm": 0.1847766488790512,
+      "learning_rate": 0.00016283422459893048,
+      "loss": 0.4399,
+      "step": 561
+    },
+    {
+      "epoch": 0.5621581069675846,
+      "grad_norm": 0.1844712197780609,
+      "learning_rate": 0.0001627673796791444,
+      "loss": 0.4153,
+      "step": 562
+    },
+    {
+      "epoch": 0.5631583882967085,
+      "grad_norm": 0.18717245757579803,
+      "learning_rate": 0.0001627005347593583,
+      "loss": 0.4587,
+      "step": 563
+    },
+    {
+      "epoch": 0.5641586696258323,
+      "grad_norm": 0.1745343953371048,
+      "learning_rate": 0.00016263368983957218,
+      "loss": 0.4271,
+      "step": 564
+    },
+    {
+      "epoch": 0.565158950954956,
+      "grad_norm": 0.17303511500358582,
+      "learning_rate": 0.0001625668449197861,
+      "loss": 0.4232,
+      "step": 565
+    },
+    {
+      "epoch": 0.5661592322840799,
+      "grad_norm": 0.18203690648078918,
+      "learning_rate": 0.00016250000000000002,
+      "loss": 0.4285,
+      "step": 566
+    },
+    {
+      "epoch": 0.5671595136132037,
+      "grad_norm": 0.179796501994133,
+      "learning_rate": 0.0001624331550802139,
+      "loss": 0.4043,
+      "step": 567
+    },
+    {
+      "epoch": 0.5681597949423275,
+      "grad_norm": 0.16504420340061188,
+      "learning_rate": 0.00016236631016042783,
+      "loss": 0.3814,
+      "step": 568
+    },
+    {
+      "epoch": 0.5691600762714514,
+      "grad_norm": 0.1814291626214981,
+      "learning_rate": 0.00016229946524064172,
+      "loss": 0.4174,
+      "step": 569
+    },
+    {
+      "epoch": 0.5701603576005752,
+      "grad_norm": 0.18535339832305908,
+      "learning_rate": 0.0001622326203208556,
+      "loss": 0.4347,
+      "step": 570
+    },
+    {
+      "epoch": 0.5711606389296989,
+      "grad_norm": 0.17747478187084198,
+      "learning_rate": 0.00016216577540106953,
+      "loss": 0.4079,
+      "step": 571
+    },
+    {
+      "epoch": 0.5721609202588228,
+      "grad_norm": 0.1741618514060974,
+      "learning_rate": 0.00016209893048128345,
+      "loss": 0.4227,
+      "step": 572
+    },
+    {
+      "epoch": 0.5731612015879466,
+      "grad_norm": 0.17553062736988068,
+      "learning_rate": 0.00016203208556149734,
+      "loss": 0.429,
+      "step": 573
+    },
+    {
+      "epoch": 0.5741614829170705,
+      "grad_norm": 0.19394700229167938,
+      "learning_rate": 0.00016196524064171123,
+      "loss": 0.4382,
+      "step": 574
+    },
+    {
+      "epoch": 0.5751617642461943,
+      "grad_norm": 0.17279013991355896,
+      "learning_rate": 0.00016189839572192515,
+      "loss": 0.4438,
+      "step": 575
+    },
+    {
+      "epoch": 0.576162045575318,
+      "grad_norm": 0.1781262457370758,
+      "learning_rate": 0.00016183155080213906,
+      "loss": 0.4095,
+      "step": 576
+    },
+    {
+      "epoch": 0.5771623269044419,
+      "grad_norm": 0.16312651336193085,
+      "learning_rate": 0.00016176470588235295,
+      "loss": 0.4036,
+      "step": 577
+    },
+    {
+      "epoch": 0.5781626082335657,
+      "grad_norm": 0.17451059818267822,
+      "learning_rate": 0.00016169786096256684,
+      "loss": 0.3971,
+      "step": 578
+    },
+    {
+      "epoch": 0.5791628895626895,
+      "grad_norm": 0.185821533203125,
+      "learning_rate": 0.00016163101604278074,
+      "loss": 0.4266,
+      "step": 579
+    },
+    {
+      "epoch": 0.5801631708918134,
+      "grad_norm": 0.1637738049030304,
+      "learning_rate": 0.00016156417112299468,
+      "loss": 0.4039,
+      "step": 580
+    },
+    {
+      "epoch": 0.5811634522209371,
+      "grad_norm": 0.17572863399982452,
+      "learning_rate": 0.00016149732620320857,
+      "loss": 0.4154,
+      "step": 581
+    },
+    {
+      "epoch": 0.5821637335500609,
+      "grad_norm": 0.17879387736320496,
+      "learning_rate": 0.00016143048128342246,
+      "loss": 0.4208,
+      "step": 582
+    },
+    {
+      "epoch": 0.5831640148791848,
+      "grad_norm": 0.17145629227161407,
+      "learning_rate": 0.00016136363636363635,
+      "loss": 0.4061,
+      "step": 583
+    },
+    {
+      "epoch": 0.5841642962083086,
+      "grad_norm": 0.17503105103969574,
+      "learning_rate": 0.00016129679144385027,
+      "loss": 0.4173,
+      "step": 584
+    },
+    {
+      "epoch": 0.5851645775374325,
+      "grad_norm": 0.17581413686275482,
+      "learning_rate": 0.0001612299465240642,
+      "loss": 0.4212,
+      "step": 585
+    },
+    {
+      "epoch": 0.5861648588665562,
+      "grad_norm": 0.17574447393417358,
+      "learning_rate": 0.00016116310160427808,
+      "loss": 0.4023,
+      "step": 586
+    },
+    {
+      "epoch": 0.58716514019568,
+      "grad_norm": 0.17156046628952026,
+      "learning_rate": 0.000161096256684492,
+      "loss": 0.4115,
+      "step": 587
+    },
+    {
+      "epoch": 0.5881654215248039,
+      "grad_norm": 0.1805231273174286,
+      "learning_rate": 0.0001610294117647059,
+      "loss": 0.4235,
+      "step": 588
+    },
+    {
+      "epoch": 0.5891657028539277,
+      "grad_norm": 0.17837534844875336,
+      "learning_rate": 0.0001609625668449198,
+      "loss": 0.4429,
+      "step": 589
+    },
+    {
+      "epoch": 0.5901659841830514,
+      "grad_norm": 0.16711211204528809,
+      "learning_rate": 0.0001608957219251337,
+      "loss": 0.4129,
+      "step": 590
+    },
+    {
+      "epoch": 0.5911662655121753,
+      "grad_norm": 0.182355597615242,
+      "learning_rate": 0.00016082887700534762,
+      "loss": 0.4417,
+      "step": 591
+    },
+    {
+      "epoch": 0.5921665468412991,
+      "grad_norm": 0.17298145592212677,
+      "learning_rate": 0.0001607620320855615,
+      "loss": 0.4206,
+      "step": 592
+    },
+    {
+      "epoch": 0.5931668281704229,
+      "grad_norm": 0.16272862255573273,
+      "learning_rate": 0.0001606951871657754,
+      "loss": 0.4032,
+      "step": 593
+    },
+    {
+      "epoch": 0.5941671094995468,
+      "grad_norm": 0.170026957988739,
+      "learning_rate": 0.00016062834224598931,
+      "loss": 0.399,
+      "step": 594
+    },
+    {
+      "epoch": 0.5951673908286705,
+      "grad_norm": 0.1715223640203476,
+      "learning_rate": 0.00016056149732620323,
+      "loss": 0.4024,
+      "step": 595
+    },
+    {
+      "epoch": 0.5961676721577944,
+      "grad_norm": 0.18015970289707184,
+      "learning_rate": 0.00016049465240641712,
+      "loss": 0.4189,
+      "step": 596
+    },
+    {
+      "epoch": 0.5971679534869182,
+      "grad_norm": 0.16714516282081604,
+      "learning_rate": 0.00016042780748663101,
+      "loss": 0.4135,
+      "step": 597
+    },
+    {
+      "epoch": 0.598168234816042,
+      "grad_norm": 0.17524467408657074,
+      "learning_rate": 0.00016036096256684493,
+      "loss": 0.3866,
+      "step": 598
+    },
+    {
+      "epoch": 0.5991685161451659,
+      "grad_norm": 0.1763477623462677,
+      "learning_rate": 0.00016029411764705885,
+      "loss": 0.3972,
+      "step": 599
+    },
+    {
+      "epoch": 0.6001687974742896,
+      "grad_norm": 0.1930224597454071,
+      "learning_rate": 0.00016022727272727274,
+      "loss": 0.4334,
+      "step": 600
+    },
+    {
+      "epoch": 0.6011690788034134,
+      "grad_norm": 0.1898859143257141,
+      "learning_rate": 0.00016016042780748663,
+      "loss": 0.4082,
+      "step": 601
+    },
+    {
+      "epoch": 0.6021693601325373,
+      "grad_norm": 0.18250437080860138,
+      "learning_rate": 0.00016009358288770052,
+      "loss": 0.4356,
+      "step": 602
+    },
+    {
+      "epoch": 0.6031696414616611,
+      "grad_norm": 0.16886284947395325,
+      "learning_rate": 0.00016002673796791444,
+      "loss": 0.4097,
+      "step": 603
+    },
+    {
+      "epoch": 0.604169922790785,
+      "grad_norm": 0.18321110308170319,
+      "learning_rate": 0.00015995989304812836,
+      "loss": 0.4082,
+      "step": 604
+    },
+    {
+      "epoch": 0.6051702041199087,
+      "grad_norm": 0.18324923515319824,
+      "learning_rate": 0.00015989304812834225,
+      "loss": 0.4378,
+      "step": 605
+    },
+    {
+      "epoch": 0.6061704854490325,
+      "grad_norm": 0.17818456888198853,
+      "learning_rate": 0.00015982620320855614,
+      "loss": 0.4212,
+      "step": 606
+    },
+    {
+      "epoch": 0.6071707667781564,
+      "grad_norm": 0.17111051082611084,
+      "learning_rate": 0.00015975935828877006,
+      "loss": 0.3938,
+      "step": 607
+    },
+    {
+      "epoch": 0.6081710481072802,
+      "grad_norm": 0.171490877866745,
+      "learning_rate": 0.00015969251336898398,
+      "loss": 0.4357,
+      "step": 608
+    },
+    {
+      "epoch": 0.609171329436404,
+      "grad_norm": 0.1783282458782196,
+      "learning_rate": 0.00015962566844919787,
+      "loss": 0.4241,
+      "step": 609
+    },
+    {
+      "epoch": 0.6101716107655278,
+      "grad_norm": 0.1665242314338684,
+      "learning_rate": 0.00015955882352941178,
+      "loss": 0.3968,
+      "step": 610
+    },
+    {
+      "epoch": 0.6111718920946516,
+      "grad_norm": 0.17004479467868805,
+      "learning_rate": 0.00015949197860962568,
+      "loss": 0.4215,
+      "step": 611
+    },
+    {
+      "epoch": 0.6121721734237754,
+      "grad_norm": 0.190244659781456,
+      "learning_rate": 0.00015942513368983957,
+      "loss": 0.4211,
+      "step": 612
+    },
+    {
+      "epoch": 0.6131724547528993,
+      "grad_norm": 0.17213775217533112,
+      "learning_rate": 0.00015935828877005348,
+      "loss": 0.4124,
+      "step": 613
+    },
+    {
+      "epoch": 0.614172736082023,
+      "grad_norm": 0.18273095786571503,
+      "learning_rate": 0.0001592914438502674,
+      "loss": 0.4075,
+      "step": 614
+    },
+    {
+      "epoch": 0.6151730174111469,
+      "grad_norm": 0.1769276261329651,
+      "learning_rate": 0.0001592245989304813,
+      "loss": 0.4242,
+      "step": 615
+    },
+    {
+      "epoch": 0.6161732987402707,
+      "grad_norm": 0.18576711416244507,
+      "learning_rate": 0.00015915775401069518,
+      "loss": 0.4247,
+      "step": 616
+    },
+    {
+      "epoch": 0.6171735800693945,
+      "grad_norm": 0.1755998581647873,
+      "learning_rate": 0.0001590909090909091,
+      "loss": 0.401,
+      "step": 617
+    },
+    {
+      "epoch": 0.6181738613985184,
+      "grad_norm": 0.17052628099918365,
+      "learning_rate": 0.00015902406417112302,
+      "loss": 0.4184,
+      "step": 618
+    },
+    {
+      "epoch": 0.6191741427276422,
+      "grad_norm": 0.16688649356365204,
+      "learning_rate": 0.0001589572192513369,
+      "loss": 0.4058,
+      "step": 619
+    },
+    {
+      "epoch": 0.6201744240567659,
+      "grad_norm": 0.15762940049171448,
+      "learning_rate": 0.0001588903743315508,
+      "loss": 0.4139,
+      "step": 620
+    },
+    {
+      "epoch": 0.6211747053858898,
+      "grad_norm": 0.1699974089860916,
+      "learning_rate": 0.0001588235294117647,
+      "loss": 0.4342,
+      "step": 621
+    },
+    {
+      "epoch": 0.6221749867150136,
+      "grad_norm": 0.1690402776002884,
+      "learning_rate": 0.00015875668449197864,
+      "loss": 0.4102,
+      "step": 622
+    },
+    {
+      "epoch": 0.6231752680441374,
+      "grad_norm": 0.17223213613033295,
+      "learning_rate": 0.00015868983957219253,
+      "loss": 0.417,
+      "step": 623
+    },
+    {
+      "epoch": 0.6241755493732613,
+      "grad_norm": 0.18006610870361328,
+      "learning_rate": 0.00015862299465240642,
+      "loss": 0.4183,
+      "step": 624
+    },
+    {
+      "epoch": 0.625175830702385,
+      "grad_norm": 0.16844414174556732,
+      "learning_rate": 0.0001585561497326203,
+      "loss": 0.4084,
+      "step": 625
+    },
+    {
+      "epoch": 0.6261761120315089,
+      "grad_norm": 0.17007195949554443,
+      "learning_rate": 0.00015848930481283423,
+      "loss": 0.3945,
+      "step": 626
+    },
+    {
+      "epoch": 0.6271763933606327,
+      "grad_norm": 0.17890439927577972,
+      "learning_rate": 0.00015842245989304815,
+      "loss": 0.3977,
+      "step": 627
+    },
+    {
+      "epoch": 0.6281766746897565,
+      "grad_norm": 0.17793171107769012,
+      "learning_rate": 0.00015835561497326204,
+      "loss": 0.397,
+      "step": 628
+    },
+    {
+      "epoch": 0.6291769560188804,
+      "grad_norm": 0.18395744264125824,
+      "learning_rate": 0.00015828877005347595,
+      "loss": 0.4228,
+      "step": 629
+    },
+    {
+      "epoch": 0.6301772373480041,
+      "grad_norm": 0.17771929502487183,
+      "learning_rate": 0.00015822192513368984,
+      "loss": 0.4093,
+      "step": 630
+    },
+    {
+      "epoch": 0.6311775186771279,
+      "grad_norm": 0.1839386522769928,
+      "learning_rate": 0.00015815508021390376,
+      "loss": 0.4028,
+      "step": 631
+    },
+    {
+      "epoch": 0.6321778000062518,
+      "grad_norm": 0.16454514861106873,
+      "learning_rate": 0.00015808823529411765,
+      "loss": 0.3936,
+      "step": 632
+    },
+    {
+      "epoch": 0.6331780813353756,
+      "grad_norm": 0.1736559122800827,
+      "learning_rate": 0.00015802139037433157,
+      "loss": 0.416,
+      "step": 633
+    },
+    {
+      "epoch": 0.6341783626644993,
+      "grad_norm": 0.1678173691034317,
+      "learning_rate": 0.00015795454545454546,
+      "loss": 0.4278,
+      "step": 634
+    },
+    {
+      "epoch": 0.6351786439936232,
+      "grad_norm": 0.17004351317882538,
+      "learning_rate": 0.00015788770053475935,
+      "loss": 0.4212,
+      "step": 635
+    },
+    {
+      "epoch": 0.636178925322747,
+      "grad_norm": 0.17825426161289215,
+      "learning_rate": 0.00015782085561497327,
+      "loss": 0.3898,
+      "step": 636
+    },
+    {
+      "epoch": 0.6371792066518709,
+      "grad_norm": 0.16482678055763245,
+      "learning_rate": 0.0001577540106951872,
+      "loss": 0.4055,
+      "step": 637
+    },
+    {
+      "epoch": 0.6381794879809947,
+      "grad_norm": 0.1691804975271225,
+      "learning_rate": 0.00015768716577540108,
+      "loss": 0.4137,
+      "step": 638
+    },
+    {
+      "epoch": 0.6391797693101184,
+      "grad_norm": 0.16740399599075317,
+      "learning_rate": 0.00015762032085561497,
+      "loss": 0.3891,
+      "step": 639
+    },
+    {
+      "epoch": 0.6401800506392423,
+      "grad_norm": 0.1827842891216278,
+      "learning_rate": 0.0001575534759358289,
+      "loss": 0.4156,
+      "step": 640
+    },
+    {
+      "epoch": 0.6411803319683661,
+      "grad_norm": 0.1753103882074356,
+      "learning_rate": 0.0001574866310160428,
+      "loss": 0.4046,
+      "step": 641
+    },
+    {
+      "epoch": 0.6421806132974899,
+      "grad_norm": 0.18446987867355347,
+      "learning_rate": 0.0001574197860962567,
+      "loss": 0.4195,
+      "step": 642
+    },
+    {
+      "epoch": 0.6431808946266138,
+      "grad_norm": 0.17731043696403503,
+      "learning_rate": 0.0001573529411764706,
+      "loss": 0.4227,
+      "step": 643
+    },
+    {
+      "epoch": 0.6441811759557375,
+      "grad_norm": 0.1867007315158844,
+      "learning_rate": 0.00015728609625668448,
+      "loss": 0.4386,
+      "step": 644
+    },
+    {
+      "epoch": 0.6451814572848613,
+      "grad_norm": 0.18152248859405518,
+      "learning_rate": 0.0001572192513368984,
+      "loss": 0.4237,
+      "step": 645
+    },
+    {
+      "epoch": 0.6461817386139852,
+      "grad_norm": 0.18429970741271973,
+      "learning_rate": 0.00015715240641711231,
+      "loss": 0.4346,
+      "step": 646
+    },
+    {
+      "epoch": 0.647182019943109,
+      "grad_norm": 0.16958941519260406,
+      "learning_rate": 0.0001570855614973262,
+      "loss": 0.4136,
+      "step": 647
+    },
+    {
+      "epoch": 0.6481823012722329,
+      "grad_norm": 0.17754653096199036,
+      "learning_rate": 0.0001570187165775401,
+      "loss": 0.4054,
+      "step": 648
+    },
+    {
+      "epoch": 0.6491825826013566,
+      "grad_norm": 0.1706933081150055,
+      "learning_rate": 0.00015695187165775401,
+      "loss": 0.3872,
+      "step": 649
+    },
+    {
+      "epoch": 0.6501828639304804,
+      "grad_norm": 0.1693519949913025,
+      "learning_rate": 0.00015688502673796793,
+      "loss": 0.4023,
+      "step": 650
+    },
+    {
+      "epoch": 0.6511831452596043,
+      "grad_norm": 0.18293803930282593,
+      "learning_rate": 0.00015681818181818182,
+      "loss": 0.4426,
+      "step": 651
+    },
+    {
+      "epoch": 0.6521834265887281,
+      "grad_norm": 0.18341802060604095,
+      "learning_rate": 0.00015675133689839574,
+      "loss": 0.4348,
+      "step": 652
+    },
+    {
+      "epoch": 0.6531837079178519,
+      "grad_norm": 0.17068350315093994,
+      "learning_rate": 0.00015668449197860963,
+      "loss": 0.3951,
+      "step": 653
+    },
+    {
+      "epoch": 0.6541839892469757,
+      "grad_norm": 0.1712416708469391,
+      "learning_rate": 0.00015661764705882352,
+      "loss": 0.4018,
+      "step": 654
+    },
+    {
+      "epoch": 0.6551842705760995,
+      "grad_norm": 0.17853021621704102,
+      "learning_rate": 0.00015655080213903744,
+      "loss": 0.4425,
+      "step": 655
+    },
+    {
+      "epoch": 0.6561845519052233,
+      "grad_norm": 0.17284280061721802,
+      "learning_rate": 0.00015648395721925136,
+      "loss": 0.4306,
+      "step": 656
+    },
+    {
+      "epoch": 0.6571848332343472,
+      "grad_norm": 0.17899803817272186,
+      "learning_rate": 0.00015641711229946525,
+      "loss": 0.4023,
+      "step": 657
+    },
+    {
+      "epoch": 0.658185114563471,
+      "grad_norm": 0.1919001191854477,
+      "learning_rate": 0.00015635026737967914,
+      "loss": 0.4415,
+      "step": 658
+    },
+    {
+      "epoch": 0.6591853958925948,
+      "grad_norm": 0.1722707599401474,
+      "learning_rate": 0.00015628342245989306,
+      "loss": 0.4176,
+      "step": 659
+    },
+    {
+      "epoch": 0.6601856772217186,
+      "grad_norm": 0.17643126845359802,
+      "learning_rate": 0.00015621657754010698,
+      "loss": 0.4014,
+      "step": 660
+    },
+    {
+      "epoch": 0.6611859585508424,
+      "grad_norm": 0.18269233405590057,
+      "learning_rate": 0.00015614973262032087,
+      "loss": 0.4113,
+      "step": 661
+    },
+    {
+      "epoch": 0.6621862398799663,
+      "grad_norm": 0.16375958919525146,
+      "learning_rate": 0.00015608288770053476,
+      "loss": 0.367,
+      "step": 662
+    },
+    {
+      "epoch": 0.66318652120909,
+      "grad_norm": 0.18737877905368805,
+      "learning_rate": 0.00015601604278074865,
+      "loss": 0.4145,
+      "step": 663
+    },
+    {
+      "epoch": 0.6641868025382138,
+      "grad_norm": 0.17648738622665405,
+      "learning_rate": 0.0001559491978609626,
+      "loss": 0.3852,
+      "step": 664
+    },
+    {
+      "epoch": 0.6651870838673377,
+      "grad_norm": 0.18149302899837494,
+      "learning_rate": 0.00015588235294117648,
+      "loss": 0.4187,
+      "step": 665
+    },
+    {
+      "epoch": 0.6661873651964615,
+      "grad_norm": 0.17750906944274902,
+      "learning_rate": 0.00015581550802139037,
+      "loss": 0.4126,
+      "step": 666
+    },
+    {
+      "epoch": 0.6671876465255854,
+      "grad_norm": 0.17197686433792114,
+      "learning_rate": 0.00015574866310160427,
+      "loss": 0.4182,
+      "step": 667
+    },
+    {
+      "epoch": 0.6681879278547092,
+      "grad_norm": 0.1686813086271286,
+      "learning_rate": 0.00015568181818181818,
+      "loss": 0.3887,
+      "step": 668
+    },
+    {
+      "epoch": 0.6691882091838329,
+      "grad_norm": 0.1689004749059677,
+      "learning_rate": 0.0001556149732620321,
+      "loss": 0.4121,
+      "step": 669
+    },
+    {
+      "epoch": 0.6701884905129568,
+      "grad_norm": 0.18247756361961365,
+      "learning_rate": 0.000155548128342246,
+      "loss": 0.4158,
+      "step": 670
+    },
+    {
+      "epoch": 0.6711887718420806,
+      "grad_norm": 0.18172405660152435,
+      "learning_rate": 0.0001554812834224599,
+      "loss": 0.4221,
+      "step": 671
+    },
+    {
+      "epoch": 0.6721890531712044,
+      "grad_norm": 0.18004614114761353,
+      "learning_rate": 0.0001554144385026738,
+      "loss": 0.4157,
+      "step": 672
+    },
+    {
+      "epoch": 0.6731893345003283,
+      "grad_norm": 0.1740560680627823,
+      "learning_rate": 0.00015534759358288772,
+      "loss": 0.3958,
+      "step": 673
+    },
+    {
+      "epoch": 0.674189615829452,
+      "grad_norm": 0.1756337434053421,
+      "learning_rate": 0.0001552807486631016,
+      "loss": 0.4217,
+      "step": 674
+    },
+    {
+      "epoch": 0.6751898971585758,
+      "grad_norm": 0.16236595809459686,
+      "learning_rate": 0.00015521390374331553,
+      "loss": 0.4009,
+      "step": 675
+    },
+    {
+      "epoch": 0.6761901784876997,
+      "grad_norm": 0.17603887617588043,
+      "learning_rate": 0.00015514705882352942,
+      "loss": 0.4351,
+      "step": 676
+    },
+    {
+      "epoch": 0.6771904598168235,
+      "grad_norm": 0.1763356328010559,
+      "learning_rate": 0.0001550802139037433,
+      "loss": 0.4334,
+      "step": 677
+    },
+    {
+      "epoch": 0.6781907411459474,
+      "grad_norm": 0.17445343732833862,
+      "learning_rate": 0.00015501336898395723,
+      "loss": 0.4199,
+      "step": 678
+    },
+    {
+      "epoch": 0.6791910224750711,
+      "grad_norm": 0.17688532173633575,
+      "learning_rate": 0.00015494652406417115,
+      "loss": 0.4314,
+      "step": 679
+    },
+    {
+      "epoch": 0.6801913038041949,
+      "grad_norm": 0.17520317435264587,
+      "learning_rate": 0.00015487967914438504,
+      "loss": 0.4128,
+      "step": 680
+    },
+    {
+      "epoch": 0.6811915851333188,
+      "grad_norm": 0.17253170907497406,
+      "learning_rate": 0.00015481283422459893,
+      "loss": 0.4095,
+      "step": 681
+    },
+    {
+      "epoch": 0.6821918664624426,
+      "grad_norm": 0.17683500051498413,
+      "learning_rate": 0.00015474598930481284,
+      "loss": 0.4037,
+      "step": 682
+    },
+    {
+      "epoch": 0.6831921477915663,
+      "grad_norm": 0.16928136348724365,
+      "learning_rate": 0.00015467914438502676,
+      "loss": 0.397,
+      "step": 683
+    },
+    {
+      "epoch": 0.6841924291206902,
+      "grad_norm": 0.18264882266521454,
+      "learning_rate": 0.00015461229946524065,
+      "loss": 0.4387,
+      "step": 684
+    },
+    {
+      "epoch": 0.685192710449814,
+      "grad_norm": 0.18944154679775238,
+      "learning_rate": 0.00015454545454545454,
+      "loss": 0.4114,
+      "step": 685
+    },
+    {
+      "epoch": 0.6861929917789378,
+      "grad_norm": 0.17461170256137848,
+      "learning_rate": 0.00015447860962566844,
+      "loss": 0.4068,
+      "step": 686
+    },
+    {
+      "epoch": 0.6871932731080617,
+      "grad_norm": 0.16725128889083862,
+      "learning_rate": 0.00015441176470588238,
+      "loss": 0.3937,
+      "step": 687
+    },
+    {
+      "epoch": 0.6881935544371854,
+      "grad_norm": 0.1661311835050583,
+      "learning_rate": 0.00015434491978609627,
+      "loss": 0.4102,
+      "step": 688
+    },
+    {
+      "epoch": 0.6891938357663093,
+      "grad_norm": 0.17577211558818817,
+      "learning_rate": 0.00015427807486631016,
+      "loss": 0.4033,
+      "step": 689
+    },
+    {
+      "epoch": 0.6901941170954331,
+      "grad_norm": 0.17130760848522186,
+      "learning_rate": 0.00015421122994652405,
+      "loss": 0.412,
+      "step": 690
+    },
+    {
+      "epoch": 0.6911943984245569,
+      "grad_norm": 0.1697661131620407,
+      "learning_rate": 0.00015414438502673797,
+      "loss": 0.3953,
+      "step": 691
+    },
+    {
+      "epoch": 0.6921946797536808,
+      "grad_norm": 0.17887946963310242,
+      "learning_rate": 0.0001540775401069519,
+      "loss": 0.431,
+      "step": 692
+    },
+    {
+      "epoch": 0.6931949610828045,
+      "grad_norm": 0.1887810230255127,
+      "learning_rate": 0.00015401069518716578,
+      "loss": 0.4294,
+      "step": 693
+    },
+    {
+      "epoch": 0.6941952424119283,
+      "grad_norm": 0.19596914947032928,
+      "learning_rate": 0.0001539438502673797,
+      "loss": 0.4292,
+      "step": 694
+    },
+    {
+      "epoch": 0.6951955237410522,
+      "grad_norm": 0.18208423256874084,
+      "learning_rate": 0.0001538770053475936,
+      "loss": 0.4102,
+      "step": 695
+    },
+    {
+      "epoch": 0.696195805070176,
+      "grad_norm": 0.1716201901435852,
+      "learning_rate": 0.0001538101604278075,
+      "loss": 0.4067,
+      "step": 696
+    },
+    {
+      "epoch": 0.6971960863992998,
+      "grad_norm": 0.18010923266410828,
+      "learning_rate": 0.0001537433155080214,
+      "loss": 0.4016,
+      "step": 697
+    },
+    {
+      "epoch": 0.6981963677284236,
+      "grad_norm": 0.1718294769525528,
+      "learning_rate": 0.00015367647058823531,
+      "loss": 0.3937,
+      "step": 698
+    },
+    {
+      "epoch": 0.6991966490575474,
+      "grad_norm": 0.16809529066085815,
+      "learning_rate": 0.0001536096256684492,
+      "loss": 0.3989,
+      "step": 699
+    },
+    {
+      "epoch": 0.7001969303866713,
+      "grad_norm": 0.17799147963523865,
+      "learning_rate": 0.0001535427807486631,
+      "loss": 0.4316,
+      "step": 700
+    },
+    {
+      "epoch": 0.7011972117157951,
+      "grad_norm": 0.16764914989471436,
+      "learning_rate": 0.00015347593582887701,
+      "loss": 0.3922,
+      "step": 701
+    },
+    {
+      "epoch": 0.7021974930449189,
+      "grad_norm": 0.1715439260005951,
+      "learning_rate": 0.00015340909090909093,
+      "loss": 0.3947,
+      "step": 702
+    },
+    {
+      "epoch": 0.7031977743740427,
+      "grad_norm": 0.17045070230960846,
+      "learning_rate": 0.00015334224598930482,
+      "loss": 0.3825,
+      "step": 703
+    },
+    {
+      "epoch": 0.7041980557031665,
+      "grad_norm": 0.16872522234916687,
+      "learning_rate": 0.00015327540106951871,
+      "loss": 0.4129,
+      "step": 704
+    },
+    {
+      "epoch": 0.7051983370322903,
+      "grad_norm": 0.18743546307086945,
+      "learning_rate": 0.00015320855614973263,
+      "loss": 0.42,
+      "step": 705
+    },
+    {
+      "epoch": 0.7061986183614142,
+      "grad_norm": 0.1783863753080368,
+      "learning_rate": 0.00015314171122994655,
+      "loss": 0.4162,
+      "step": 706
+    },
+    {
+      "epoch": 0.707198899690538,
+      "grad_norm": 0.17544718086719513,
+      "learning_rate": 0.00015307486631016044,
+      "loss": 0.4107,
+      "step": 707
+    },
+    {
+      "epoch": 0.7081991810196617,
+      "grad_norm": 0.16872859001159668,
+      "learning_rate": 0.00015300802139037433,
+      "loss": 0.4039,
+      "step": 708
+    },
+    {
+      "epoch": 0.7091994623487856,
+      "grad_norm": 0.16296930611133575,
+      "learning_rate": 0.00015294117647058822,
+      "loss": 0.3857,
+      "step": 709
+    },
+    {
+      "epoch": 0.7101997436779094,
+      "grad_norm": 0.16989687085151672,
+      "learning_rate": 0.00015287433155080214,
+      "loss": 0.4097,
+      "step": 710
+    },
+    {
+      "epoch": 0.7112000250070333,
+      "grad_norm": 0.17646653950214386,
+      "learning_rate": 0.00015280748663101606,
+      "loss": 0.4004,
+      "step": 711
+    },
+    {
+      "epoch": 0.712200306336157,
+      "grad_norm": 0.16280825436115265,
+      "learning_rate": 0.00015274064171122995,
+      "loss": 0.386,
+      "step": 712
+    },
+    {
+      "epoch": 0.7132005876652808,
+      "grad_norm": 0.18282656371593475,
+      "learning_rate": 0.00015267379679144387,
+      "loss": 0.4271,
+      "step": 713
+    },
+    {
+      "epoch": 0.7142008689944047,
+      "grad_norm": 0.17502710223197937,
+      "learning_rate": 0.00015260695187165776,
+      "loss": 0.4426,
+      "step": 714
+    },
+    {
+      "epoch": 0.7152011503235285,
+      "grad_norm": 0.16774967312812805,
+      "learning_rate": 0.00015254010695187168,
+      "loss": 0.3889,
+      "step": 715
+    },
+    {
+      "epoch": 0.7162014316526523,
+      "grad_norm": 0.17474064230918884,
+      "learning_rate": 0.00015247326203208557,
+      "loss": 0.4182,
+      "step": 716
+    },
+    {
+      "epoch": 0.7172017129817762,
+      "grad_norm": 0.1727539449930191,
+      "learning_rate": 0.00015240641711229948,
+      "loss": 0.3914,
+      "step": 717
+    },
+    {
+      "epoch": 0.7182019943108999,
+      "grad_norm": 0.17711150646209717,
+      "learning_rate": 0.00015233957219251337,
+      "loss": 0.3967,
+      "step": 718
+    },
+    {
+      "epoch": 0.7192022756400237,
+      "grad_norm": 0.17282328009605408,
+      "learning_rate": 0.00015227272727272727,
+      "loss": 0.4186,
+      "step": 719
+    },
+    {
+      "epoch": 0.7202025569691476,
+      "grad_norm": 0.18157802522182465,
+      "learning_rate": 0.00015220588235294118,
+      "loss": 0.4219,
+      "step": 720
+    },
+    {
+      "epoch": 0.7212028382982714,
+      "grad_norm": 0.1840929538011551,
+      "learning_rate": 0.0001521390374331551,
+      "loss": 0.4004,
+      "step": 721
+    },
+    {
+      "epoch": 0.7222031196273953,
+      "grad_norm": 0.17365121841430664,
+      "learning_rate": 0.000152072192513369,
+      "loss": 0.399,
+      "step": 722
+    },
+    {
+      "epoch": 0.723203400956519,
+      "grad_norm": 0.1816079169511795,
+      "learning_rate": 0.00015200534759358288,
+      "loss": 0.4073,
+      "step": 723
+    },
+    {
+      "epoch": 0.7242036822856428,
+      "grad_norm": 0.17313836514949799,
+      "learning_rate": 0.0001519385026737968,
+      "loss": 0.3875,
+      "step": 724
+    },
+    {
+      "epoch": 0.7252039636147667,
+      "grad_norm": 0.16790300607681274,
+      "learning_rate": 0.00015187165775401072,
+      "loss": 0.4075,
+      "step": 725
+    },
+    {
+      "epoch": 0.7262042449438905,
+      "grad_norm": 0.16900648176670074,
+      "learning_rate": 0.0001518048128342246,
+      "loss": 0.4126,
+      "step": 726
+    },
+    {
+      "epoch": 0.7272045262730142,
+      "grad_norm": 0.16767656803131104,
+      "learning_rate": 0.0001517379679144385,
+      "loss": 0.409,
+      "step": 727
+    },
+    {
+      "epoch": 0.7282048076021381,
+      "grad_norm": 0.17568565905094147,
+      "learning_rate": 0.0001516711229946524,
+      "loss": 0.4075,
+      "step": 728
+    },
+    {
+      "epoch": 0.7292050889312619,
+      "grad_norm": 0.17152975499629974,
+      "learning_rate": 0.00015160427807486634,
+      "loss": 0.4199,
+      "step": 729
+    },
+    {
+      "epoch": 0.7302053702603858,
+      "grad_norm": 0.17456263303756714,
+      "learning_rate": 0.00015153743315508023,
+      "loss": 0.4115,
+      "step": 730
+    },
+    {
+      "epoch": 0.7312056515895096,
+      "grad_norm": 0.1764313280582428,
+      "learning_rate": 0.00015147058823529412,
+      "loss": 0.4074,
+      "step": 731
+    },
+    {
+      "epoch": 0.7322059329186333,
+      "grad_norm": 0.17887412011623383,
+      "learning_rate": 0.000151403743315508,
+      "loss": 0.4105,
+      "step": 732
+    },
+    {
+      "epoch": 0.7332062142477572,
+      "grad_norm": 0.18053936958312988,
+      "learning_rate": 0.00015133689839572193,
+      "loss": 0.4224,
+      "step": 733
+    },
+    {
+      "epoch": 0.734206495576881,
+      "grad_norm": 0.1864321231842041,
+      "learning_rate": 0.00015127005347593584,
+      "loss": 0.3945,
+      "step": 734
+    },
+    {
+      "epoch": 0.7352067769060048,
+      "grad_norm": 0.17719848453998566,
+      "learning_rate": 0.00015120320855614974,
+      "loss": 0.4173,
+      "step": 735
+    },
+    {
+      "epoch": 0.7362070582351287,
+      "grad_norm": 0.17382051050662994,
+      "learning_rate": 0.00015113636363636365,
+      "loss": 0.412,
+      "step": 736
+    },
+    {
+      "epoch": 0.7372073395642524,
+      "grad_norm": 0.17831949889659882,
+      "learning_rate": 0.00015106951871657754,
+      "loss": 0.3926,
+      "step": 737
+    },
+    {
+      "epoch": 0.7382076208933762,
+      "grad_norm": 0.19203975796699524,
+      "learning_rate": 0.00015100267379679146,
+      "loss": 0.4061,
+      "step": 738
+    },
+    {
+      "epoch": 0.7392079022225001,
+      "grad_norm": 0.1851518303155899,
+      "learning_rate": 0.00015093582887700535,
+      "loss": 0.4254,
+      "step": 739
+    },
+    {
+      "epoch": 0.7402081835516239,
+      "grad_norm": 0.17518197000026703,
+      "learning_rate": 0.00015086898395721927,
+      "loss": 0.4094,
+      "step": 740
+    },
+    {
+      "epoch": 0.7412084648807478,
+      "grad_norm": 0.19005051255226135,
+      "learning_rate": 0.00015080213903743316,
+      "loss": 0.435,
+      "step": 741
+    },
+    {
+      "epoch": 0.7422087462098715,
+      "grad_norm": 0.17512501776218414,
+      "learning_rate": 0.00015073529411764705,
+      "loss": 0.3818,
+      "step": 742
+    },
+    {
+      "epoch": 0.7432090275389953,
+      "grad_norm": 0.17218168079853058,
+      "learning_rate": 0.00015066844919786097,
+      "loss": 0.3938,
+      "step": 743
+    },
+    {
+      "epoch": 0.7442093088681192,
+      "grad_norm": 0.16856470704078674,
+      "learning_rate": 0.0001506016042780749,
+      "loss": 0.4186,
+      "step": 744
+    },
+    {
+      "epoch": 0.745209590197243,
+      "grad_norm": 0.185384601354599,
+      "learning_rate": 0.00015053475935828878,
+      "loss": 0.3965,
+      "step": 745
+    },
+    {
+      "epoch": 0.7462098715263668,
+      "grad_norm": 0.18854860961437225,
+      "learning_rate": 0.00015046791443850267,
+      "loss": 0.4281,
+      "step": 746
+    },
+    {
+      "epoch": 0.7472101528554906,
+      "grad_norm": 0.17106805741786957,
+      "learning_rate": 0.0001504010695187166,
+      "loss": 0.403,
+      "step": 747
+    },
+    {
+      "epoch": 0.7482104341846144,
+      "grad_norm": 0.17635445296764374,
+      "learning_rate": 0.0001503342245989305,
+      "loss": 0.4045,
+      "step": 748
+    },
+    {
+      "epoch": 0.7492107155137382,
+      "grad_norm": 0.17554466426372528,
+      "learning_rate": 0.0001502673796791444,
+      "loss": 0.4216,
+      "step": 749
+    },
+    {
+      "epoch": 0.7502109968428621,
+      "grad_norm": 0.16902974247932434,
+      "learning_rate": 0.0001502005347593583,
+      "loss": 0.3842,
+      "step": 750
+    },
+    {
+      "epoch": 0.7512112781719859,
+      "grad_norm": 0.17811493575572968,
+      "learning_rate": 0.00015013368983957218,
+      "loss": 0.4085,
+      "step": 751
+    },
+    {
+      "epoch": 0.7522115595011097,
+      "grad_norm": 0.17670631408691406,
+      "learning_rate": 0.0001500668449197861,
+      "loss": 0.4142,
+      "step": 752
+    },
+    {
+      "epoch": 0.7532118408302335,
+      "grad_norm": 0.1737818568944931,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.4171,
+      "step": 753
+    },
+    {
+      "epoch": 0.7542121221593573,
+      "grad_norm": 0.17644794285297394,
+      "learning_rate": 0.0001499331550802139,
+      "loss": 0.41,
+      "step": 754
+    },
+    {
+      "epoch": 0.7552124034884812,
+      "grad_norm": 0.17215976119041443,
+      "learning_rate": 0.00014986631016042782,
+      "loss": 0.397,
+      "step": 755
+    },
+    {
+      "epoch": 0.756212684817605,
+      "grad_norm": 0.17125418782234192,
+      "learning_rate": 0.0001497994652406417,
+      "loss": 0.42,
+      "step": 756
+    },
+    {
+      "epoch": 0.7572129661467287,
+      "grad_norm": 0.19707784056663513,
+      "learning_rate": 0.00014973262032085563,
+      "loss": 0.4085,
+      "step": 757
+    },
+    {
+      "epoch": 0.7582132474758526,
+      "grad_norm": 0.17742498219013214,
+      "learning_rate": 0.00014966577540106952,
+      "loss": 0.4297,
+      "step": 758
+    },
+    {
+      "epoch": 0.7592135288049764,
+      "grad_norm": 0.17373313009738922,
+      "learning_rate": 0.00014959893048128344,
+      "loss": 0.4203,
+      "step": 759
+    },
+    {
+      "epoch": 0.7602138101341002,
+      "grad_norm": 0.17742851376533508,
+      "learning_rate": 0.00014953208556149733,
+      "loss": 0.4225,
+      "step": 760
+    },
+    {
+      "epoch": 0.761214091463224,
+      "grad_norm": 0.18568547070026398,
+      "learning_rate": 0.00014946524064171122,
+      "loss": 0.4175,
+      "step": 761
+    },
+    {
+      "epoch": 0.7622143727923478,
+      "grad_norm": 0.16926662623882294,
+      "learning_rate": 0.00014939839572192514,
+      "loss": 0.4182,
+      "step": 762
+    },
+    {
+      "epoch": 0.7632146541214717,
+      "grad_norm": 0.1670301854610443,
+      "learning_rate": 0.00014933155080213906,
+      "loss": 0.4032,
+      "step": 763
+    },
+    {
+      "epoch": 0.7642149354505955,
+      "grad_norm": 0.17339031398296356,
+      "learning_rate": 0.00014926470588235295,
+      "loss": 0.43,
+      "step": 764
+    },
+    {
+      "epoch": 0.7652152167797193,
+      "grad_norm": 0.17211312055587769,
+      "learning_rate": 0.00014919786096256684,
+      "loss": 0.4086,
+      "step": 765
+    },
+    {
+      "epoch": 0.7662154981088432,
+      "grad_norm": 0.17297977209091187,
+      "learning_rate": 0.00014913101604278076,
+      "loss": 0.3909,
+      "step": 766
+    },
+    {
+      "epoch": 0.7672157794379669,
+      "grad_norm": 0.18003138899803162,
+      "learning_rate": 0.00014906417112299468,
+      "loss": 0.4073,
+      "step": 767
+    },
+    {
+      "epoch": 0.7682160607670907,
+      "grad_norm": 0.18356819450855255,
+      "learning_rate": 0.00014899732620320857,
+      "loss": 0.4105,
+      "step": 768
+    },
+    {
+      "epoch": 0.7692163420962146,
+      "grad_norm": 0.1751221865415573,
+      "learning_rate": 0.00014893048128342246,
+      "loss": 0.4033,
+      "step": 769
+    },
+    {
+      "epoch": 0.7702166234253384,
+      "grad_norm": 0.1809505969285965,
+      "learning_rate": 0.00014886363636363635,
+      "loss": 0.4033,
+      "step": 770
+    },
+    {
+      "epoch": 0.7712169047544621,
+      "grad_norm": 0.17458176612854004,
+      "learning_rate": 0.0001487967914438503,
+      "loss": 0.415,
+      "step": 771
+    },
+    {
+      "epoch": 0.772217186083586,
+      "grad_norm": 0.17972098290920258,
+      "learning_rate": 0.00014872994652406418,
+      "loss": 0.4304,
+      "step": 772
+    },
+    {
+      "epoch": 0.7732174674127098,
+      "grad_norm": 0.17451392114162445,
+      "learning_rate": 0.00014866310160427807,
+      "loss": 0.3945,
+      "step": 773
+    },
+    {
+      "epoch": 0.7742177487418337,
+      "grad_norm": 0.18923582136631012,
+      "learning_rate": 0.000148596256684492,
+      "loss": 0.4039,
+      "step": 774
+    },
+    {
+      "epoch": 0.7752180300709575,
+      "grad_norm": 0.17579270899295807,
+      "learning_rate": 0.00014852941176470588,
+      "loss": 0.4115,
+      "step": 775
+    },
+    {
+      "epoch": 0.7762183114000812,
+      "grad_norm": 0.18641281127929688,
+      "learning_rate": 0.0001484625668449198,
+      "loss": 0.4028,
+      "step": 776
+    },
+    {
+      "epoch": 0.7772185927292051,
+      "grad_norm": 0.1804158091545105,
+      "learning_rate": 0.0001483957219251337,
+      "loss": 0.363,
+      "step": 777
+    },
+    {
+      "epoch": 0.7782188740583289,
+      "grad_norm": 0.1881331503391266,
+      "learning_rate": 0.0001483288770053476,
+      "loss": 0.3875,
+      "step": 778
+    },
+    {
+      "epoch": 0.7792191553874527,
+      "grad_norm": 0.17641088366508484,
+      "learning_rate": 0.0001482620320855615,
+      "loss": 0.4124,
+      "step": 779
+    },
+    {
+      "epoch": 0.7802194367165766,
+      "grad_norm": 0.18274515867233276,
+      "learning_rate": 0.00014819518716577542,
+      "loss": 0.4048,
+      "step": 780
+    },
+    {
+      "epoch": 0.7812197180457003,
+      "grad_norm": 0.17852523922920227,
+      "learning_rate": 0.0001481283422459893,
+      "loss": 0.4094,
+      "step": 781
+    },
+    {
+      "epoch": 0.7822199993748241,
+      "grad_norm": 0.1843167096376419,
+      "learning_rate": 0.00014806149732620323,
+      "loss": 0.4222,
+      "step": 782
+    },
+    {
+      "epoch": 0.783220280703948,
+      "grad_norm": 0.19953066110610962,
+      "learning_rate": 0.00014799465240641712,
+      "loss": 0.3951,
+      "step": 783
+    },
+    {
+      "epoch": 0.7842205620330718,
+      "grad_norm": 0.1708746999502182,
+      "learning_rate": 0.000147927807486631,
+      "loss": 0.4331,
+      "step": 784
+    },
+    {
+      "epoch": 0.7852208433621957,
+      "grad_norm": 0.16868069767951965,
+      "learning_rate": 0.00014786096256684493,
+      "loss": 0.3986,
+      "step": 785
+    },
+    {
+      "epoch": 0.7862211246913194,
+      "grad_norm": 0.17892426252365112,
+      "learning_rate": 0.00014779411764705884,
+      "loss": 0.4108,
+      "step": 786
+    },
+    {
+      "epoch": 0.7872214060204432,
+      "grad_norm": 0.16507907211780548,
+      "learning_rate": 0.00014772727272727274,
+      "loss": 0.3735,
+      "step": 787
+    },
+    {
+      "epoch": 0.7882216873495671,
+      "grad_norm": 0.17571504414081573,
+      "learning_rate": 0.00014766042780748663,
+      "loss": 0.4135,
+      "step": 788
+    },
+    {
+      "epoch": 0.7892219686786909,
+      "grad_norm": 0.16615907847881317,
+      "learning_rate": 0.00014759358288770054,
+      "loss": 0.4048,
+      "step": 789
+    },
+    {
+      "epoch": 0.7902222500078147,
+      "grad_norm": 0.18569040298461914,
+      "learning_rate": 0.00014752673796791446,
+      "loss": 0.427,
+      "step": 790
+    },
+    {
+      "epoch": 0.7912225313369385,
+      "grad_norm": 0.17103785276412964,
+      "learning_rate": 0.00014745989304812835,
+      "loss": 0.3877,
+      "step": 791
+    },
+    {
+      "epoch": 0.7922228126660623,
+      "grad_norm": 0.1752656102180481,
+      "learning_rate": 0.00014739304812834224,
+      "loss": 0.393,
+      "step": 792
+    },
+    {
+      "epoch": 0.7932230939951862,
+      "grad_norm": 0.17465591430664062,
+      "learning_rate": 0.00014732620320855613,
+      "loss": 0.4149,
+      "step": 793
+    },
+    {
+      "epoch": 0.79422337532431,
+      "grad_norm": 0.18063925206661224,
+      "learning_rate": 0.00014725935828877008,
+      "loss": 0.3961,
+      "step": 794
+    },
+    {
+      "epoch": 0.7952236566534338,
+      "grad_norm": 0.1805940717458725,
+      "learning_rate": 0.00014719251336898397,
+      "loss": 0.4112,
+      "step": 795
+    },
+    {
+      "epoch": 0.7962239379825576,
+      "grad_norm": 0.17591600120067596,
+      "learning_rate": 0.00014712566844919786,
+      "loss": 0.3919,
+      "step": 796
+    },
+    {
+      "epoch": 0.7972242193116814,
+      "grad_norm": 0.17145437002182007,
+      "learning_rate": 0.00014705882352941178,
+      "loss": 0.4263,
+      "step": 797
+    },
+    {
+      "epoch": 0.7982245006408052,
+      "grad_norm": 0.17273344099521637,
+      "learning_rate": 0.00014699197860962567,
+      "loss": 0.3945,
+      "step": 798
+    },
+    {
+      "epoch": 0.7992247819699291,
+      "grad_norm": 0.16053451597690582,
+      "learning_rate": 0.0001469251336898396,
+      "loss": 0.3728,
+      "step": 799
+    },
+    {
+      "epoch": 0.8002250632990529,
+      "grad_norm": 0.1720341145992279,
+      "learning_rate": 0.00014685828877005348,
+      "loss": 0.4029,
+      "step": 800
+    },
+    {
+      "epoch": 0.8012253446281766,
+      "grad_norm": 0.1686064451932907,
+      "learning_rate": 0.0001467914438502674,
+      "loss": 0.3996,
+      "step": 801
+    },
+    {
+      "epoch": 0.8022256259573005,
+      "grad_norm": 0.17165598273277283,
+      "learning_rate": 0.0001467245989304813,
+      "loss": 0.3785,
+      "step": 802
+    },
+    {
+      "epoch": 0.8032259072864243,
+      "grad_norm": 0.17315958440303802,
+      "learning_rate": 0.0001466577540106952,
+      "loss": 0.3989,
+      "step": 803
+    },
+    {
+      "epoch": 0.8042261886155482,
+      "grad_norm": 0.17488616704940796,
+      "learning_rate": 0.0001465909090909091,
+      "loss": 0.4087,
+      "step": 804
+    },
+    {
+      "epoch": 0.805226469944672,
+      "grad_norm": 0.17673259973526,
+      "learning_rate": 0.00014652406417112301,
+      "loss": 0.4126,
+      "step": 805
+    },
+    {
+      "epoch": 0.8062267512737957,
+      "grad_norm": 0.17650727927684784,
+      "learning_rate": 0.0001464572192513369,
+      "loss": 0.4115,
+      "step": 806
+    },
+    {
+      "epoch": 0.8072270326029196,
+      "grad_norm": 0.17311464250087738,
+      "learning_rate": 0.0001463903743315508,
+      "loss": 0.4175,
+      "step": 807
+    },
+    {
+      "epoch": 0.8082273139320434,
+      "grad_norm": 0.18814238905906677,
+      "learning_rate": 0.0001463235294117647,
+      "loss": 0.4082,
+      "step": 808
+    },
+    {
+      "epoch": 0.8092275952611672,
+      "grad_norm": 0.16839633882045746,
+      "learning_rate": 0.00014625668449197863,
+      "loss": 0.3909,
+      "step": 809
+    },
+    {
+      "epoch": 0.810227876590291,
+      "grad_norm": 0.1769394725561142,
+      "learning_rate": 0.00014618983957219252,
+      "loss": 0.3913,
+      "step": 810
+    },
+    {
+      "epoch": 0.8112281579194148,
+      "grad_norm": 0.17125903069972992,
+      "learning_rate": 0.0001461229946524064,
+      "loss": 0.3914,
+      "step": 811
+    },
+    {
+      "epoch": 0.8122284392485386,
+      "grad_norm": 0.17443816363811493,
+      "learning_rate": 0.00014605614973262033,
+      "loss": 0.4183,
+      "step": 812
+    },
+    {
+      "epoch": 0.8132287205776625,
+      "grad_norm": 0.18099425733089447,
+      "learning_rate": 0.00014598930481283425,
+      "loss": 0.4043,
+      "step": 813
+    },
+    {
+      "epoch": 0.8142290019067863,
+      "grad_norm": 0.17355240881443024,
+      "learning_rate": 0.00014592245989304814,
+      "loss": 0.4082,
+      "step": 814
+    },
+    {
+      "epoch": 0.8152292832359102,
+      "grad_norm": 0.16909727454185486,
+      "learning_rate": 0.00014585561497326203,
+      "loss": 0.4056,
+      "step": 815
+    },
+    {
+      "epoch": 0.8162295645650339,
+      "grad_norm": 0.1898905634880066,
+      "learning_rate": 0.00014578877005347595,
+      "loss": 0.4352,
+      "step": 816
+    },
+    {
+      "epoch": 0.8172298458941577,
+      "grad_norm": 0.17900870740413666,
+      "learning_rate": 0.00014572192513368984,
+      "loss": 0.4205,
+      "step": 817
+    },
+    {
+      "epoch": 0.8182301272232816,
+      "grad_norm": 0.17821773886680603,
+      "learning_rate": 0.00014565508021390376,
+      "loss": 0.4248,
+      "step": 818
+    },
+    {
+      "epoch": 0.8192304085524054,
+      "grad_norm": 0.19676046073436737,
+      "learning_rate": 0.00014558823529411765,
+      "loss": 0.4096,
+      "step": 819
+    },
+    {
+      "epoch": 0.8202306898815291,
+      "grad_norm": 0.17447529733181,
+      "learning_rate": 0.00014552139037433157,
+      "loss": 0.4089,
+      "step": 820
+    },
+    {
+      "epoch": 0.821230971210653,
+      "grad_norm": 0.1961289793252945,
+      "learning_rate": 0.00014545454545454546,
+      "loss": 0.4046,
+      "step": 821
+    },
+    {
+      "epoch": 0.8222312525397768,
+      "grad_norm": 0.18811632692813873,
+      "learning_rate": 0.00014538770053475937,
+      "loss": 0.4265,
+      "step": 822
+    },
+    {
+      "epoch": 0.8232315338689006,
+      "grad_norm": 0.16100798547267914,
+      "learning_rate": 0.00014532085561497327,
+      "loss": 0.3986,
+      "step": 823
+    },
+    {
+      "epoch": 0.8242318151980245,
+      "grad_norm": 0.1786595582962036,
+      "learning_rate": 0.00014525401069518718,
+      "loss": 0.4095,
+      "step": 824
+    },
+    {
+      "epoch": 0.8252320965271482,
+      "grad_norm": 0.17071866989135742,
+      "learning_rate": 0.00014518716577540107,
+      "loss": 0.3784,
+      "step": 825
+    },
+    {
+      "epoch": 0.8262323778562721,
+      "grad_norm": 0.18947453796863556,
+      "learning_rate": 0.00014512032085561496,
+      "loss": 0.4101,
+      "step": 826
+    },
+    {
+      "epoch": 0.8272326591853959,
+      "grad_norm": 0.17812004685401917,
+      "learning_rate": 0.00014505347593582888,
+      "loss": 0.4102,
+      "step": 827
+    },
+    {
+      "epoch": 0.8282329405145197,
+      "grad_norm": 0.17082251608371735,
+      "learning_rate": 0.0001449866310160428,
+      "loss": 0.3902,
+      "step": 828
+    },
+    {
+      "epoch": 0.8292332218436436,
+      "grad_norm": 0.19824115931987762,
+      "learning_rate": 0.0001449197860962567,
+      "loss": 0.4108,
+      "step": 829
+    },
+    {
+      "epoch": 0.8302335031727673,
+      "grad_norm": 0.19035981595516205,
+      "learning_rate": 0.00014485294117647058,
+      "loss": 0.4299,
+      "step": 830
+    },
+    {
+      "epoch": 0.8312337845018911,
+      "grad_norm": 0.17072008550167084,
+      "learning_rate": 0.0001447860962566845,
+      "loss": 0.3892,
+      "step": 831
+    },
+    {
+      "epoch": 0.832234065831015,
+      "grad_norm": 0.16947698593139648,
+      "learning_rate": 0.00014471925133689842,
+      "loss": 0.3799,
+      "step": 832
+    },
+    {
+      "epoch": 0.8332343471601388,
+      "grad_norm": 0.173639714717865,
+      "learning_rate": 0.0001446524064171123,
+      "loss": 0.3889,
+      "step": 833
+    },
+    {
+      "epoch": 0.8342346284892626,
+      "grad_norm": 0.17889994382858276,
+      "learning_rate": 0.0001445855614973262,
+      "loss": 0.4097,
+      "step": 834
+    },
+    {
+      "epoch": 0.8352349098183864,
+      "grad_norm": 0.17331556975841522,
+      "learning_rate": 0.0001445187165775401,
+      "loss": 0.4041,
+      "step": 835
+    },
+    {
+      "epoch": 0.8362351911475102,
+      "grad_norm": 0.17438872158527374,
+      "learning_rate": 0.00014445187165775404,
+      "loss": 0.3966,
+      "step": 836
+    },
+    {
+      "epoch": 0.8372354724766341,
+      "grad_norm": 0.1759490966796875,
+      "learning_rate": 0.00014438502673796793,
+      "loss": 0.4043,
+      "step": 837
+    },
+    {
+      "epoch": 0.8382357538057579,
+      "grad_norm": 0.16946400701999664,
+      "learning_rate": 0.00014431818181818182,
+      "loss": 0.4282,
+      "step": 838
+    },
+    {
+      "epoch": 0.8392360351348817,
+      "grad_norm": 0.16447113454341888,
+      "learning_rate": 0.00014425133689839574,
+      "loss": 0.3731,
+      "step": 839
+    },
+    {
+      "epoch": 0.8402363164640055,
+      "grad_norm": 0.1674252301454544,
+      "learning_rate": 0.00014418449197860963,
+      "loss": 0.3879,
+      "step": 840
+    },
+    {
+      "epoch": 0.8412365977931293,
+      "grad_norm": 0.1837848722934723,
+      "learning_rate": 0.00014411764705882354,
+      "loss": 0.4217,
+      "step": 841
+    },
+    {
+      "epoch": 0.8422368791222531,
+      "grad_norm": 0.1785300076007843,
+      "learning_rate": 0.00014405080213903743,
+      "loss": 0.4125,
+      "step": 842
+    },
+    {
+      "epoch": 0.843237160451377,
+      "grad_norm": 0.20810818672180176,
+      "learning_rate": 0.00014398395721925135,
+      "loss": 0.409,
+      "step": 843
+    },
+    {
+      "epoch": 0.8442374417805008,
+      "grad_norm": 0.17327488958835602,
+      "learning_rate": 0.00014391711229946524,
+      "loss": 0.3981,
+      "step": 844
+    },
+    {
+      "epoch": 0.8452377231096245,
+      "grad_norm": 0.2608954608440399,
+      "learning_rate": 0.00014385026737967916,
+      "loss": 0.3927,
+      "step": 845
+    },
+    {
+      "epoch": 0.8462380044387484,
+      "grad_norm": 0.18573161959648132,
+      "learning_rate": 0.00014378342245989305,
+      "loss": 0.4287,
+      "step": 846
+    },
+    {
+      "epoch": 0.8472382857678722,
+      "grad_norm": 0.1860964000225067,
+      "learning_rate": 0.00014371657754010697,
+      "loss": 0.3982,
+      "step": 847
+    },
+    {
+      "epoch": 0.8482385670969961,
+      "grad_norm": 0.17765875160694122,
+      "learning_rate": 0.00014364973262032086,
+      "loss": 0.3898,
+      "step": 848
+    },
+    {
+      "epoch": 0.8492388484261199,
+      "grad_norm": 0.17560985684394836,
+      "learning_rate": 0.00014358288770053475,
+      "loss": 0.4192,
+      "step": 849
+    },
+    {
+      "epoch": 0.8502391297552436,
+      "grad_norm": 0.17350338399410248,
+      "learning_rate": 0.00014351604278074867,
+      "loss": 0.4087,
+      "step": 850
+    },
+    {
+      "epoch": 0.8512394110843675,
+      "grad_norm": 0.17332197725772858,
+      "learning_rate": 0.0001434491978609626,
+      "loss": 0.4037,
+      "step": 851
+    },
+    {
+      "epoch": 0.8522396924134913,
+      "grad_norm": 0.17164768278598785,
+      "learning_rate": 0.00014338235294117648,
+      "loss": 0.4132,
+      "step": 852
+    },
+    {
+      "epoch": 0.8532399737426151,
+      "grad_norm": 0.17493771016597748,
+      "learning_rate": 0.00014331550802139037,
+      "loss": 0.3898,
+      "step": 853
+    },
+    {
+      "epoch": 0.854240255071739,
+      "grad_norm": 0.1740700900554657,
+      "learning_rate": 0.0001432486631016043,
+      "loss": 0.3857,
+      "step": 854
+    },
+    {
+      "epoch": 0.8552405364008627,
+      "grad_norm": 0.17457234859466553,
+      "learning_rate": 0.0001431818181818182,
+      "loss": 0.401,
+      "step": 855
+    },
+    {
+      "epoch": 0.8562408177299866,
+      "grad_norm": 0.1648370772600174,
+      "learning_rate": 0.0001431149732620321,
+      "loss": 0.3838,
+      "step": 856
+    },
+    {
+      "epoch": 0.8572410990591104,
+      "grad_norm": 0.1766262650489807,
+      "learning_rate": 0.000143048128342246,
+      "loss": 0.4083,
+      "step": 857
+    },
+    {
+      "epoch": 0.8582413803882342,
+      "grad_norm": 0.16761666536331177,
+      "learning_rate": 0.0001429812834224599,
+      "loss": 0.3771,
+      "step": 858
+    },
+    {
+      "epoch": 0.859241661717358,
+      "grad_norm": 0.1663917452096939,
+      "learning_rate": 0.0001429144385026738,
+      "loss": 0.4198,
+      "step": 859
+    },
+    {
+      "epoch": 0.8602419430464818,
+      "grad_norm": 0.17130149900913239,
+      "learning_rate": 0.0001428475935828877,
+      "loss": 0.3755,
+      "step": 860
+    },
+    {
+      "epoch": 0.8612422243756056,
+      "grad_norm": 0.16928842663764954,
+      "learning_rate": 0.0001427807486631016,
+      "loss": 0.3922,
+      "step": 861
+    },
+    {
+      "epoch": 0.8622425057047295,
+      "grad_norm": 0.1808835119009018,
+      "learning_rate": 0.00014271390374331552,
+      "loss": 0.3831,
+      "step": 862
+    },
+    {
+      "epoch": 0.8632427870338533,
+      "grad_norm": 0.17441269755363464,
+      "learning_rate": 0.0001426470588235294,
+      "loss": 0.4159,
+      "step": 863
+    },
+    {
+      "epoch": 0.864243068362977,
+      "grad_norm": 0.17643436789512634,
+      "learning_rate": 0.00014258021390374333,
+      "loss": 0.4119,
+      "step": 864
+    },
+    {
+      "epoch": 0.8652433496921009,
+      "grad_norm": 0.1707470417022705,
+      "learning_rate": 0.00014251336898395722,
+      "loss": 0.3818,
+      "step": 865
+    },
+    {
+      "epoch": 0.8662436310212247,
+      "grad_norm": 0.16326646506786346,
+      "learning_rate": 0.00014244652406417114,
+      "loss": 0.3755,
+      "step": 866
+    },
+    {
+      "epoch": 0.8672439123503486,
+      "grad_norm": 0.17867451906204224,
+      "learning_rate": 0.00014237967914438503,
+      "loss": 0.4158,
+      "step": 867
+    },
+    {
+      "epoch": 0.8682441936794724,
+      "grad_norm": 0.17721618711948395,
+      "learning_rate": 0.00014231283422459892,
+      "loss": 0.4329,
+      "step": 868
+    },
+    {
+      "epoch": 0.8692444750085961,
+      "grad_norm": 0.16564089059829712,
+      "learning_rate": 0.00014224598930481284,
+      "loss": 0.4011,
+      "step": 869
+    },
+    {
+      "epoch": 0.87024475633772,
+      "grad_norm": 0.174989715218544,
+      "learning_rate": 0.00014217914438502676,
+      "loss": 0.389,
+      "step": 870
+    },
+    {
+      "epoch": 0.8712450376668438,
+      "grad_norm": 0.17899653315544128,
+      "learning_rate": 0.00014211229946524065,
+      "loss": 0.4123,
+      "step": 871
+    },
+    {
+      "epoch": 0.8722453189959676,
+      "grad_norm": 0.17878752946853638,
+      "learning_rate": 0.00014204545454545454,
+      "loss": 0.4248,
+      "step": 872
+    },
+    {
+      "epoch": 0.8732456003250915,
+      "grad_norm": 0.17918603122234344,
+      "learning_rate": 0.00014197860962566846,
+      "loss": 0.4108,
+      "step": 873
+    },
+    {
+      "epoch": 0.8742458816542152,
+      "grad_norm": 0.16828817129135132,
+      "learning_rate": 0.00014191176470588237,
+      "loss": 0.4016,
+      "step": 874
+    },
+    {
+      "epoch": 0.875246162983339,
+      "grad_norm": 0.17155155539512634,
+      "learning_rate": 0.00014184491978609627,
+      "loss": 0.4061,
+      "step": 875
+    },
+    {
+      "epoch": 0.8762464443124629,
+      "grad_norm": 0.1909843236207962,
+      "learning_rate": 0.00014177807486631016,
+      "loss": 0.4019,
+      "step": 876
+    },
+    {
+      "epoch": 0.8772467256415867,
+      "grad_norm": 0.17694242298603058,
+      "learning_rate": 0.00014171122994652405,
+      "loss": 0.4005,
+      "step": 877
+    },
+    {
+      "epoch": 0.8782470069707106,
+      "grad_norm": 0.18169617652893066,
+      "learning_rate": 0.000141644385026738,
+      "loss": 0.395,
+      "step": 878
+    },
+    {
+      "epoch": 0.8792472882998343,
+      "grad_norm": 0.18381339311599731,
+      "learning_rate": 0.00014157754010695188,
+      "loss": 0.3825,
+      "step": 879
+    },
+    {
+      "epoch": 0.8802475696289581,
+      "grad_norm": 0.16577623784542084,
+      "learning_rate": 0.00014151069518716577,
+      "loss": 0.3815,
+      "step": 880
+    },
+    {
+      "epoch": 0.881247850958082,
+      "grad_norm": 0.19959154725074768,
+      "learning_rate": 0.0001414438502673797,
+      "loss": 0.3994,
+      "step": 881
+    },
+    {
+      "epoch": 0.8822481322872058,
+      "grad_norm": 0.19829948246479034,
+      "learning_rate": 0.00014137700534759358,
+      "loss": 0.412,
+      "step": 882
+    },
+    {
+      "epoch": 0.8832484136163296,
+      "grad_norm": 0.17270344495773315,
+      "learning_rate": 0.0001413101604278075,
+      "loss": 0.408,
+      "step": 883
+    },
+    {
+      "epoch": 0.8842486949454534,
+      "grad_norm": 0.17711827158927917,
+      "learning_rate": 0.0001412433155080214,
+      "loss": 0.3942,
+      "step": 884
+    },
+    {
+      "epoch": 0.8852489762745772,
+      "grad_norm": 0.20845437049865723,
+      "learning_rate": 0.0001411764705882353,
+      "loss": 0.4221,
+      "step": 885
+    },
+    {
+      "epoch": 0.886249257603701,
+      "grad_norm": 0.1848880648612976,
+      "learning_rate": 0.0001411096256684492,
+      "loss": 0.4029,
+      "step": 886
+    },
+    {
+      "epoch": 0.8872495389328249,
+      "grad_norm": 0.16786153614521027,
+      "learning_rate": 0.00014104278074866312,
+      "loss": 0.3785,
+      "step": 887
+    },
+    {
+      "epoch": 0.8882498202619487,
+      "grad_norm": 0.16691294312477112,
+      "learning_rate": 0.000140975935828877,
+      "loss": 0.4067,
+      "step": 888
+    },
+    {
+      "epoch": 0.8892501015910725,
+      "grad_norm": 0.17728787660598755,
+      "learning_rate": 0.00014090909090909093,
+      "loss": 0.397,
+      "step": 889
+    },
+    {
+      "epoch": 0.8902503829201963,
+      "grad_norm": 0.1761750876903534,
+      "learning_rate": 0.00014084224598930482,
+      "loss": 0.3979,
+      "step": 890
+    },
+    {
+      "epoch": 0.8912506642493201,
+      "grad_norm": 0.1840093433856964,
+      "learning_rate": 0.0001407754010695187,
+      "loss": 0.4029,
+      "step": 891
+    },
+    {
+      "epoch": 0.892250945578444,
+      "grad_norm": 0.17632125318050385,
+      "learning_rate": 0.00014070855614973263,
+      "loss": 0.3997,
+      "step": 892
+    },
+    {
+      "epoch": 0.8932512269075678,
+      "grad_norm": 0.17613720893859863,
+      "learning_rate": 0.00014064171122994654,
+      "loss": 0.3784,
+      "step": 893
+    },
+    {
+      "epoch": 0.8942515082366915,
+      "grad_norm": 0.18408696353435516,
+      "learning_rate": 0.00014057486631016043,
+      "loss": 0.4104,
+      "step": 894
+    },
+    {
+      "epoch": 0.8952517895658154,
+      "grad_norm": 0.18052643537521362,
+      "learning_rate": 0.00014050802139037433,
+      "loss": 0.4014,
+      "step": 895
+    },
+    {
+      "epoch": 0.8962520708949392,
+      "grad_norm": 0.18135356903076172,
+      "learning_rate": 0.00014044117647058824,
+      "loss": 0.4188,
+      "step": 896
+    },
+    {
+      "epoch": 0.897252352224063,
+      "grad_norm": 0.1768425852060318,
+      "learning_rate": 0.00014037433155080216,
+      "loss": 0.4026,
+      "step": 897
+    },
+    {
+      "epoch": 0.8982526335531869,
+      "grad_norm": 0.17669373750686646,
+      "learning_rate": 0.00014030748663101605,
+      "loss": 0.4249,
+      "step": 898
+    },
+    {
+      "epoch": 0.8992529148823106,
+      "grad_norm": 0.1866510659456253,
+      "learning_rate": 0.00014024064171122994,
+      "loss": 0.447,
+      "step": 899
+    },
+    {
+      "epoch": 0.9002531962114345,
+      "grad_norm": 0.18631310760974884,
+      "learning_rate": 0.00014017379679144386,
+      "loss": 0.4072,
+      "step": 900
+    },
+    {
+      "epoch": 0.9012534775405583,
+      "grad_norm": 0.1782829463481903,
+      "learning_rate": 0.00014010695187165778,
+      "loss": 0.4176,
+      "step": 901
+    },
+    {
+      "epoch": 0.9022537588696821,
+      "grad_norm": 0.16321073472499847,
+      "learning_rate": 0.00014004010695187167,
+      "loss": 0.3762,
+      "step": 902
+    },
+    {
+      "epoch": 0.903254040198806,
+      "grad_norm": 0.1676628589630127,
+      "learning_rate": 0.00013997326203208556,
+      "loss": 0.3868,
+      "step": 903
+    },
+    {
+      "epoch": 0.9042543215279297,
+      "grad_norm": 0.18179528415203094,
+      "learning_rate": 0.00013990641711229948,
+      "loss": 0.4031,
+      "step": 904
+    },
+    {
+      "epoch": 0.9052546028570535,
+      "grad_norm": 0.17398111522197723,
+      "learning_rate": 0.00013983957219251337,
+      "loss": 0.3926,
+      "step": 905
+    },
+    {
+      "epoch": 0.9062548841861774,
+      "grad_norm": 0.17181190848350525,
+      "learning_rate": 0.0001397727272727273,
+      "loss": 0.4142,
+      "step": 906
+    },
+    {
+      "epoch": 0.9072551655153012,
+      "grad_norm": 0.17833548784255981,
+      "learning_rate": 0.00013970588235294118,
+      "loss": 0.4167,
+      "step": 907
+    },
+    {
+      "epoch": 0.9082554468444249,
+      "grad_norm": 0.17353063821792603,
+      "learning_rate": 0.0001396390374331551,
+      "loss": 0.4221,
+      "step": 908
+    },
+    {
+      "epoch": 0.9092557281735488,
+      "grad_norm": 0.17322175204753876,
+      "learning_rate": 0.000139572192513369,
+      "loss": 0.4298,
+      "step": 909
+    },
+    {
+      "epoch": 0.9102560095026726,
+      "grad_norm": 0.17294059693813324,
+      "learning_rate": 0.0001395053475935829,
+      "loss": 0.3897,
+      "step": 910
+    },
+    {
+      "epoch": 0.9112562908317965,
+      "grad_norm": 0.17554591596126556,
+      "learning_rate": 0.0001394385026737968,
+      "loss": 0.3742,
+      "step": 911
+    },
+    {
+      "epoch": 0.9122565721609203,
+      "grad_norm": 0.1788209229707718,
+      "learning_rate": 0.0001393716577540107,
+      "loss": 0.4048,
+      "step": 912
+    },
+    {
+      "epoch": 0.913256853490044,
+      "grad_norm": 0.1844097226858139,
+      "learning_rate": 0.0001393048128342246,
+      "loss": 0.4148,
+      "step": 913
+    },
+    {
+      "epoch": 0.9142571348191679,
+      "grad_norm": 0.18024982511997223,
+      "learning_rate": 0.0001392379679144385,
+      "loss": 0.4042,
+      "step": 914
+    },
+    {
+      "epoch": 0.9152574161482917,
+      "grad_norm": 0.18731427192687988,
+      "learning_rate": 0.0001391711229946524,
+      "loss": 0.3902,
+      "step": 915
+    },
+    {
+      "epoch": 0.9162576974774155,
+      "grad_norm": 0.17596861720085144,
+      "learning_rate": 0.00013910427807486633,
+      "loss": 0.3782,
+      "step": 916
+    },
+    {
+      "epoch": 0.9172579788065394,
+      "grad_norm": 0.18986783921718597,
+      "learning_rate": 0.00013903743315508022,
+      "loss": 0.4147,
+      "step": 917
+    },
+    {
+      "epoch": 0.9182582601356631,
+      "grad_norm": 0.16781824827194214,
+      "learning_rate": 0.0001389705882352941,
+      "loss": 0.4042,
+      "step": 918
+    },
+    {
+      "epoch": 0.919258541464787,
+      "grad_norm": 0.17034654319286346,
+      "learning_rate": 0.00013890374331550803,
+      "loss": 0.3906,
+      "step": 919
+    },
+    {
+      "epoch": 0.9202588227939108,
+      "grad_norm": 0.16829004883766174,
+      "learning_rate": 0.00013883689839572195,
+      "loss": 0.4141,
+      "step": 920
+    },
+    {
+      "epoch": 0.9212591041230346,
+      "grad_norm": 0.19006861746311188,
+      "learning_rate": 0.00013877005347593584,
+      "loss": 0.4561,
+      "step": 921
+    },
+    {
+      "epoch": 0.9222593854521585,
+      "grad_norm": 0.16582091152668,
+      "learning_rate": 0.00013870320855614973,
+      "loss": 0.4091,
+      "step": 922
+    },
+    {
+      "epoch": 0.9232596667812822,
+      "grad_norm": 0.17240959405899048,
+      "learning_rate": 0.00013863636363636365,
+      "loss": 0.4028,
+      "step": 923
+    },
+    {
+      "epoch": 0.924259948110406,
+      "grad_norm": 0.169050931930542,
+      "learning_rate": 0.00013856951871657754,
+      "loss": 0.3787,
+      "step": 924
+    },
+    {
+      "epoch": 0.9252602294395299,
+      "grad_norm": 0.18316373229026794,
+      "learning_rate": 0.00013850267379679146,
+      "loss": 0.4219,
+      "step": 925
+    },
+    {
+      "epoch": 0.9262605107686537,
+      "grad_norm": 0.1770108938217163,
+      "learning_rate": 0.00013843582887700535,
+      "loss": 0.3965,
+      "step": 926
+    },
+    {
+      "epoch": 0.9272607920977775,
+      "grad_norm": 0.17968682944774628,
+      "learning_rate": 0.00013836898395721927,
+      "loss": 0.3796,
+      "step": 927
+    },
+    {
+      "epoch": 0.9282610734269013,
+      "grad_norm": 0.17519252002239227,
+      "learning_rate": 0.00013830213903743316,
+      "loss": 0.3973,
+      "step": 928
+    },
+    {
+      "epoch": 0.9292613547560251,
+      "grad_norm": 0.17505863308906555,
+      "learning_rate": 0.00013823529411764707,
+      "loss": 0.3923,
+      "step": 929
+    },
+    {
+      "epoch": 0.930261636085149,
+      "grad_norm": 0.1817988157272339,
+      "learning_rate": 0.00013816844919786096,
+      "loss": 0.4159,
+      "step": 930
+    },
+    {
+      "epoch": 0.9312619174142728,
+      "grad_norm": 0.18939583003520966,
+      "learning_rate": 0.00013810160427807488,
+      "loss": 0.3866,
+      "step": 931
+    },
+    {
+      "epoch": 0.9322621987433966,
+      "grad_norm": 0.17609402537345886,
+      "learning_rate": 0.00013803475935828877,
+      "loss": 0.3984,
+      "step": 932
+    },
+    {
+      "epoch": 0.9332624800725204,
+      "grad_norm": 0.17766056954860687,
+      "learning_rate": 0.00013796791443850266,
+      "loss": 0.4082,
+      "step": 933
+    },
+    {
+      "epoch": 0.9342627614016442,
+      "grad_norm": 0.17946885526180267,
+      "learning_rate": 0.00013790106951871658,
+      "loss": 0.3907,
+      "step": 934
+    },
+    {
+      "epoch": 0.935263042730768,
+      "grad_norm": 0.17798246443271637,
+      "learning_rate": 0.0001378342245989305,
+      "loss": 0.4033,
+      "step": 935
+    },
+    {
+      "epoch": 0.9362633240598919,
+      "grad_norm": 0.17503587901592255,
+      "learning_rate": 0.0001377673796791444,
+      "loss": 0.3984,
+      "step": 936
+    },
+    {
+      "epoch": 0.9372636053890157,
+      "grad_norm": 0.18128138780593872,
+      "learning_rate": 0.00013770053475935828,
+      "loss": 0.3902,
+      "step": 937
+    },
+    {
+      "epoch": 0.9382638867181394,
+      "grad_norm": 0.17769305408000946,
+      "learning_rate": 0.0001376336898395722,
+      "loss": 0.4268,
+      "step": 938
+    },
+    {
+      "epoch": 0.9392641680472633,
+      "grad_norm": 0.178322896361351,
+      "learning_rate": 0.00013756684491978612,
+      "loss": 0.4037,
+      "step": 939
+    },
+    {
+      "epoch": 0.9402644493763871,
+      "grad_norm": 0.17999599874019623,
+      "learning_rate": 0.0001375,
+      "loss": 0.4092,
+      "step": 940
+    },
+    {
+      "epoch": 0.941264730705511,
+      "grad_norm": 0.1779446005821228,
+      "learning_rate": 0.0001374331550802139,
+      "loss": 0.3998,
+      "step": 941
+    },
+    {
+      "epoch": 0.9422650120346348,
+      "grad_norm": 0.17861850559711456,
+      "learning_rate": 0.00013736631016042782,
+      "loss": 0.4,
+      "step": 942
+    },
+    {
+      "epoch": 0.9432652933637585,
+      "grad_norm": 0.1744145005941391,
+      "learning_rate": 0.00013729946524064173,
+      "loss": 0.384,
+      "step": 943
+    },
+    {
+      "epoch": 0.9442655746928824,
+      "grad_norm": 0.17079755663871765,
+      "learning_rate": 0.00013723262032085563,
+      "loss": 0.4012,
+      "step": 944
+    },
+    {
+      "epoch": 0.9452658560220062,
+      "grad_norm": 0.17308571934700012,
+      "learning_rate": 0.00013716577540106952,
+      "loss": 0.4116,
+      "step": 945
+    },
+    {
+      "epoch": 0.94626613735113,
+      "grad_norm": 0.1663939356803894,
+      "learning_rate": 0.00013709893048128343,
+      "loss": 0.3813,
+      "step": 946
+    },
+    {
+      "epoch": 0.9472664186802539,
+      "grad_norm": 0.1654575616121292,
+      "learning_rate": 0.00013703208556149733,
+      "loss": 0.3903,
+      "step": 947
+    },
+    {
+      "epoch": 0.9482667000093776,
+      "grad_norm": 0.17540469765663147,
+      "learning_rate": 0.00013696524064171124,
+      "loss": 0.401,
+      "step": 948
+    },
+    {
+      "epoch": 0.9492669813385014,
+      "grad_norm": 0.17629101872444153,
+      "learning_rate": 0.00013689839572192513,
+      "loss": 0.3975,
+      "step": 949
+    },
+    {
+      "epoch": 0.9502672626676253,
+      "grad_norm": 0.170530304312706,
+      "learning_rate": 0.00013683155080213905,
+      "loss": 0.3889,
+      "step": 950
+    },
+    {
+      "epoch": 0.9512675439967491,
+      "grad_norm": 0.1654834896326065,
+      "learning_rate": 0.00013676470588235294,
+      "loss": 0.3861,
+      "step": 951
+    },
+    {
+      "epoch": 0.952267825325873,
+      "grad_norm": 0.17197169363498688,
+      "learning_rate": 0.00013669786096256686,
+      "loss": 0.4027,
+      "step": 952
+    },
+    {
+      "epoch": 0.9532681066549967,
+      "grad_norm": 0.1747899055480957,
+      "learning_rate": 0.00013663101604278075,
+      "loss": 0.3811,
+      "step": 953
+    },
+    {
+      "epoch": 0.9542683879841205,
+      "grad_norm": 0.17641963064670563,
+      "learning_rate": 0.00013656417112299467,
+      "loss": 0.4064,
+      "step": 954
+    },
+    {
+      "epoch": 0.9552686693132444,
+      "grad_norm": 0.18137142062187195,
+      "learning_rate": 0.00013649732620320856,
+      "loss": 0.3985,
+      "step": 955
+    },
+    {
+      "epoch": 0.9562689506423682,
+      "grad_norm": 0.1665244996547699,
+      "learning_rate": 0.00013643048128342245,
+      "loss": 0.4019,
+      "step": 956
+    },
+    {
+      "epoch": 0.9572692319714919,
+      "grad_norm": 0.17225103080272675,
+      "learning_rate": 0.00013636363636363637,
+      "loss": 0.401,
+      "step": 957
+    },
+    {
+      "epoch": 0.9582695133006158,
+      "grad_norm": 0.17319133877754211,
+      "learning_rate": 0.0001362967914438503,
+      "loss": 0.4115,
+      "step": 958
+    },
+    {
+      "epoch": 0.9592697946297396,
+      "grad_norm": 0.1801668256521225,
+      "learning_rate": 0.00013622994652406418,
+      "loss": 0.4055,
+      "step": 959
+    },
+    {
+      "epoch": 0.9602700759588634,
+      "grad_norm": 0.1737590879201889,
+      "learning_rate": 0.00013616310160427807,
+      "loss": 0.3978,
+      "step": 960
+    },
+    {
+      "epoch": 0.9612703572879873,
+      "grad_norm": 0.17786799371242523,
+      "learning_rate": 0.00013609625668449199,
+      "loss": 0.4066,
+      "step": 961
+    },
+    {
+      "epoch": 0.962270638617111,
+      "grad_norm": 0.17783451080322266,
+      "learning_rate": 0.0001360294117647059,
+      "loss": 0.423,
+      "step": 962
+    },
+    {
+      "epoch": 0.9632709199462349,
+      "grad_norm": 0.17744936048984528,
+      "learning_rate": 0.0001359625668449198,
+      "loss": 0.3976,
+      "step": 963
+    },
+    {
+      "epoch": 0.9642712012753587,
+      "grad_norm": 0.1820724457502365,
+      "learning_rate": 0.00013589572192513369,
+      "loss": 0.3925,
+      "step": 964
+    },
+    {
+      "epoch": 0.9652714826044825,
+      "grad_norm": 0.16475753486156464,
+      "learning_rate": 0.0001358288770053476,
+      "loss": 0.3804,
+      "step": 965
+    },
+    {
+      "epoch": 0.9662717639336064,
+      "grad_norm": 0.17936855554580688,
+      "learning_rate": 0.0001357620320855615,
+      "loss": 0.4008,
+      "step": 966
+    },
+    {
+      "epoch": 0.9672720452627301,
+      "grad_norm": 0.17178383469581604,
+      "learning_rate": 0.0001356951871657754,
+      "loss": 0.4043,
+      "step": 967
+    },
+    {
+      "epoch": 0.9682723265918539,
+      "grad_norm": 0.18698321282863617,
+      "learning_rate": 0.0001356283422459893,
+      "loss": 0.4209,
+      "step": 968
+    },
+    {
+      "epoch": 0.9692726079209778,
+      "grad_norm": 0.16066111624240875,
+      "learning_rate": 0.00013556149732620322,
+      "loss": 0.3693,
+      "step": 969
+    },
+    {
+      "epoch": 0.9702728892501016,
+      "grad_norm": 0.17552712559700012,
+      "learning_rate": 0.0001354946524064171,
+      "loss": 0.3985,
+      "step": 970
+    },
+    {
+      "epoch": 0.9712731705792254,
+      "grad_norm": 0.1760571002960205,
+      "learning_rate": 0.00013542780748663103,
+      "loss": 0.3913,
+      "step": 971
+    },
+    {
+      "epoch": 0.9722734519083492,
+      "grad_norm": 0.1769057661294937,
+      "learning_rate": 0.00013536096256684492,
+      "loss": 0.4154,
+      "step": 972
+    },
+    {
+      "epoch": 0.973273733237473,
+      "grad_norm": 0.1754721999168396,
+      "learning_rate": 0.00013529411764705884,
+      "loss": 0.3763,
+      "step": 973
+    },
+    {
+      "epoch": 0.9742740145665969,
+      "grad_norm": 0.1725761741399765,
+      "learning_rate": 0.00013522727272727273,
+      "loss": 0.3847,
+      "step": 974
+    },
+    {
+      "epoch": 0.9752742958957207,
+      "grad_norm": 0.18031996488571167,
+      "learning_rate": 0.00013516042780748662,
+      "loss": 0.3996,
+      "step": 975
+    },
+    {
+      "epoch": 0.9762745772248445,
+      "grad_norm": 0.17676669359207153,
+      "learning_rate": 0.00013509358288770054,
+      "loss": 0.4087,
+      "step": 976
+    },
+    {
+      "epoch": 0.9772748585539683,
+      "grad_norm": 0.17073461413383484,
+      "learning_rate": 0.00013502673796791446,
+      "loss": 0.4114,
+      "step": 977
+    },
+    {
+      "epoch": 0.9782751398830921,
+      "grad_norm": 0.17377017438411713,
+      "learning_rate": 0.00013495989304812835,
+      "loss": 0.406,
+      "step": 978
+    },
+    {
+      "epoch": 0.9792754212122159,
+      "grad_norm": 0.1675245761871338,
+      "learning_rate": 0.00013489304812834224,
+      "loss": 0.3855,
+      "step": 979
+    },
+    {
+      "epoch": 0.9802757025413398,
+      "grad_norm": 0.17281413078308105,
+      "learning_rate": 0.00013482620320855616,
+      "loss": 0.3893,
+      "step": 980
+    },
+    {
+      "epoch": 0.9812759838704636,
+      "grad_norm": 0.17831061780452728,
+      "learning_rate": 0.00013475935828877007,
+      "loss": 0.4151,
+      "step": 981
+    },
+    {
+      "epoch": 0.9822762651995874,
+      "grad_norm": 0.18597757816314697,
+      "learning_rate": 0.00013469251336898396,
+      "loss": 0.4072,
+      "step": 982
+    },
+    {
+      "epoch": 0.9832765465287112,
+      "grad_norm": 0.185447558760643,
+      "learning_rate": 0.00013462566844919786,
+      "loss": 0.4194,
+      "step": 983
+    },
+    {
+      "epoch": 0.984276827857835,
+      "grad_norm": 0.16914793848991394,
+      "learning_rate": 0.00013455882352941177,
+      "loss": 0.3941,
+      "step": 984
+    },
+    {
+      "epoch": 0.9852771091869589,
+      "grad_norm": 0.17165033519268036,
+      "learning_rate": 0.0001344919786096257,
+      "loss": 0.4108,
+      "step": 985
+    },
+    {
+      "epoch": 0.9862773905160827,
+      "grad_norm": 0.17159344255924225,
+      "learning_rate": 0.00013442513368983958,
+      "loss": 0.3675,
+      "step": 986
+    },
+    {
+      "epoch": 0.9872776718452064,
+      "grad_norm": 0.16202309727668762,
+      "learning_rate": 0.00013435828877005347,
+      "loss": 0.3808,
+      "step": 987
+    },
+    {
+      "epoch": 0.9882779531743303,
+      "grad_norm": 0.18378107249736786,
+      "learning_rate": 0.0001342914438502674,
+      "loss": 0.4001,
+      "step": 988
+    },
+    {
+      "epoch": 0.9892782345034541,
+      "grad_norm": 0.17740057408809662,
+      "learning_rate": 0.00013422459893048128,
+      "loss": 0.3983,
+      "step": 989
+    },
+    {
+      "epoch": 0.9902785158325779,
+      "grad_norm": 0.18121828138828278,
+      "learning_rate": 0.0001341577540106952,
+      "loss": 0.3996,
+      "step": 990
+    },
+    {
+      "epoch": 0.9912787971617018,
+      "grad_norm": 0.17484596371650696,
+      "learning_rate": 0.0001340909090909091,
+      "loss": 0.3984,
+      "step": 991
+    },
+    {
+      "epoch": 0.9922790784908255,
+      "grad_norm": 0.1730891317129135,
+      "learning_rate": 0.000134024064171123,
+      "loss": 0.392,
+      "step": 992
+    },
+    {
+      "epoch": 0.9932793598199494,
+      "grad_norm": 0.1800169050693512,
+      "learning_rate": 0.0001339572192513369,
+      "loss": 0.4135,
+      "step": 993
+    },
+    {
+      "epoch": 0.9942796411490732,
+      "grad_norm": 0.17655795812606812,
+      "learning_rate": 0.00013389037433155082,
+      "loss": 0.4061,
+      "step": 994
+    },
+    {
+      "epoch": 0.995279922478197,
+      "grad_norm": 0.17109893262386322,
+      "learning_rate": 0.0001338235294117647,
+      "loss": 0.394,
+      "step": 995
+    },
+    {
+      "epoch": 0.9962802038073209,
+      "grad_norm": 0.16945064067840576,
+      "learning_rate": 0.00013375668449197863,
+      "loss": 0.368,
+      "step": 996
+    },
+    {
+      "epoch": 0.9972804851364446,
+      "grad_norm": 0.17989976704120636,
+      "learning_rate": 0.00013368983957219252,
+      "loss": 0.382,
+      "step": 997
+    },
+    {
+      "epoch": 0.9982807664655684,
+      "grad_norm": 0.1649048626422882,
+      "learning_rate": 0.0001336229946524064,
+      "loss": 0.4012,
+      "step": 998
+    },
+    {
+      "epoch": 0.9992810477946923,
+      "grad_norm": 0.17571307718753815,
+      "learning_rate": 0.00013355614973262033,
+      "loss": 0.4168,
+      "step": 999
+    },
+    {
+      "epoch": 1.0002813291238162,
+      "grad_norm": 0.17617368698120117,
+      "learning_rate": 0.00013348930481283424,
+      "loss": 0.4042,
+      "step": 1000
+    },
+    {
+      "epoch": 1.0012816104529398,
+      "grad_norm": 0.19696933031082153,
+      "learning_rate": 0.00013342245989304813,
+      "loss": 0.3582,
+      "step": 1001
+    },
+    {
+      "epoch": 1.0022818917820637,
+      "grad_norm": 0.17075574398040771,
+      "learning_rate": 0.00013335561497326202,
+      "loss": 0.3637,
+      "step": 1002
+    },
+    {
+      "epoch": 1.0032821731111876,
+      "grad_norm": 0.16586962342262268,
+      "learning_rate": 0.00013328877005347594,
+      "loss": 0.3634,
+      "step": 1003
+    },
+    {
+      "epoch": 1.0042824544403113,
+      "grad_norm": 0.1692279875278473,
+      "learning_rate": 0.00013322192513368986,
+      "loss": 0.3626,
+      "step": 1004
+    },
+    {
+      "epoch": 1.0052827357694352,
+      "grad_norm": 0.1674485206604004,
+      "learning_rate": 0.00013315508021390375,
+      "loss": 0.3515,
+      "step": 1005
+    },
+    {
+      "epoch": 1.006283017098559,
+      "grad_norm": 0.19807380437850952,
+      "learning_rate": 0.00013308823529411764,
+      "loss": 0.3851,
+      "step": 1006
+    },
+    {
+      "epoch": 1.0072832984276827,
+      "grad_norm": 0.17747996747493744,
+      "learning_rate": 0.00013302139037433156,
+      "loss": 0.3357,
+      "step": 1007
+    },
+    {
+      "epoch": 1.0082835797568066,
+      "grad_norm": 0.20884305238723755,
+      "learning_rate": 0.00013295454545454548,
+      "loss": 0.3397,
+      "step": 1008
+    },
+    {
+      "epoch": 1.0092838610859305,
+      "grad_norm": 0.19564613699913025,
+      "learning_rate": 0.00013288770053475937,
+      "loss": 0.3659,
+      "step": 1009
+    },
+    {
+      "epoch": 1.0102841424150542,
+      "grad_norm": 0.19661381840705872,
+      "learning_rate": 0.00013282085561497326,
+      "loss": 0.3534,
+      "step": 1010
+    },
+    {
+      "epoch": 1.011284423744178,
+      "grad_norm": 0.18916454911231995,
+      "learning_rate": 0.00013275401069518718,
+      "loss": 0.3728,
+      "step": 1011
+    },
+    {
+      "epoch": 1.012284705073302,
+      "grad_norm": 0.17743521928787231,
+      "learning_rate": 0.00013268716577540107,
+      "loss": 0.3349,
+      "step": 1012
+    },
+    {
+      "epoch": 1.0132849864024256,
+      "grad_norm": 0.18179528415203094,
+      "learning_rate": 0.00013262032085561499,
+      "loss": 0.3513,
+      "step": 1013
+    },
+    {
+      "epoch": 1.0142852677315495,
+      "grad_norm": 0.18252994120121002,
+      "learning_rate": 0.00013255347593582888,
+      "loss": 0.3555,
+      "step": 1014
+    },
+    {
+      "epoch": 1.0152855490606734,
+      "grad_norm": 0.17366494238376617,
+      "learning_rate": 0.0001324866310160428,
+      "loss": 0.3489,
+      "step": 1015
+    },
+    {
+      "epoch": 1.016285830389797,
+      "grad_norm": 0.17742004990577698,
+      "learning_rate": 0.00013241978609625669,
+      "loss": 0.3678,
+      "step": 1016
+    },
+    {
+      "epoch": 1.017286111718921,
+      "grad_norm": 0.1723395437002182,
+      "learning_rate": 0.0001323529411764706,
+      "loss": 0.3359,
+      "step": 1017
+    },
+    {
+      "epoch": 1.0182863930480448,
+      "grad_norm": 0.17242787778377533,
+      "learning_rate": 0.0001322860962566845,
+      "loss": 0.3563,
+      "step": 1018
+    },
+    {
+      "epoch": 1.0192866743771685,
+      "grad_norm": 0.18579402565956116,
+      "learning_rate": 0.0001322192513368984,
+      "loss": 0.3597,
+      "step": 1019
+    },
+    {
+      "epoch": 1.0202869557062924,
+      "grad_norm": 0.17989873886108398,
+      "learning_rate": 0.0001321524064171123,
+      "loss": 0.3695,
+      "step": 1020
+    },
+    {
+      "epoch": 1.0212872370354162,
+      "grad_norm": 0.1743343472480774,
+      "learning_rate": 0.0001320855614973262,
+      "loss": 0.3621,
+      "step": 1021
+    },
+    {
+      "epoch": 1.0222875183645401,
+      "grad_norm": 0.1741182506084442,
+      "learning_rate": 0.0001320187165775401,
+      "loss": 0.3397,
+      "step": 1022
+    },
+    {
+      "epoch": 1.0232877996936638,
+      "grad_norm": 0.1851324737071991,
+      "learning_rate": 0.00013195187165775403,
+      "loss": 0.3758,
+      "step": 1023
+    },
+    {
+      "epoch": 1.0242880810227877,
+      "grad_norm": 0.18184123933315277,
+      "learning_rate": 0.00013188502673796792,
+      "loss": 0.3589,
+      "step": 1024
+    },
+    {
+      "epoch": 1.0252883623519116,
+      "grad_norm": 0.17170466482639313,
+      "learning_rate": 0.0001318181818181818,
+      "loss": 0.3479,
+      "step": 1025
+    },
+    {
+      "epoch": 1.0262886436810352,
+      "grad_norm": 0.17408423125743866,
+      "learning_rate": 0.00013175133689839573,
+      "loss": 0.3308,
+      "step": 1026
+    },
+    {
+      "epoch": 1.0272889250101591,
+      "grad_norm": 0.18442901968955994,
+      "learning_rate": 0.00013168449197860965,
+      "loss": 0.358,
+      "step": 1027
+    },
+    {
+      "epoch": 1.028289206339283,
+      "grad_norm": 0.19452296197414398,
+      "learning_rate": 0.00013161764705882354,
+      "loss": 0.3649,
+      "step": 1028
+    },
+    {
+      "epoch": 1.0292894876684067,
+      "grad_norm": 0.185035839676857,
+      "learning_rate": 0.00013155080213903743,
+      "loss": 0.3685,
+      "step": 1029
+    },
+    {
+      "epoch": 1.0302897689975306,
+      "grad_norm": 0.18745794892311096,
+      "learning_rate": 0.00013148395721925135,
+      "loss": 0.3494,
+      "step": 1030
+    },
+    {
+      "epoch": 1.0312900503266544,
+      "grad_norm": 0.18403327465057373,
+      "learning_rate": 0.00013141711229946524,
+      "loss": 0.3584,
+      "step": 1031
+    },
+    {
+      "epoch": 1.032290331655778,
+      "grad_norm": 0.1871235966682434,
+      "learning_rate": 0.00013135026737967916,
+      "loss": 0.3675,
+      "step": 1032
+    },
+    {
+      "epoch": 1.033290612984902,
+      "grad_norm": 0.17956537008285522,
+      "learning_rate": 0.00013128342245989305,
+      "loss": 0.3323,
+      "step": 1033
+    },
+    {
+      "epoch": 1.0342908943140259,
+      "grad_norm": 0.19827093183994293,
+      "learning_rate": 0.00013121657754010696,
+      "loss": 0.3988,
+      "step": 1034
+    },
+    {
+      "epoch": 1.0352911756431495,
+      "grad_norm": 0.17778335511684418,
+      "learning_rate": 0.00013114973262032086,
+      "loss": 0.33,
+      "step": 1035
+    },
+    {
+      "epoch": 1.0362914569722734,
+      "grad_norm": 0.1932620108127594,
+      "learning_rate": 0.00013108288770053477,
+      "loss": 0.3808,
+      "step": 1036
+    },
+    {
+      "epoch": 1.0372917383013973,
+      "grad_norm": 0.18136858940124512,
+      "learning_rate": 0.00013101604278074866,
+      "loss": 0.3591,
+      "step": 1037
+    },
+    {
+      "epoch": 1.038292019630521,
+      "grad_norm": 0.181020587682724,
+      "learning_rate": 0.00013094919786096258,
+      "loss": 0.3621,
+      "step": 1038
+    },
+    {
+      "epoch": 1.0392923009596449,
+      "grad_norm": 0.1801241934299469,
+      "learning_rate": 0.00013088235294117647,
+      "loss": 0.3464,
+      "step": 1039
+    },
+    {
+      "epoch": 1.0402925822887688,
+      "grad_norm": 0.18890109658241272,
+      "learning_rate": 0.00013081550802139036,
+      "loss": 0.3668,
+      "step": 1040
+    },
+    {
+      "epoch": 1.0412928636178926,
+      "grad_norm": 0.1839684247970581,
+      "learning_rate": 0.00013074866310160428,
+      "loss": 0.3621,
+      "step": 1041
+    },
+    {
+      "epoch": 1.0422931449470163,
+      "grad_norm": 0.18589378893375397,
+      "learning_rate": 0.0001306818181818182,
+      "loss": 0.3451,
+      "step": 1042
+    },
+    {
+      "epoch": 1.0432934262761402,
+      "grad_norm": 0.18868054449558258,
+      "learning_rate": 0.0001306149732620321,
+      "loss": 0.3517,
+      "step": 1043
+    },
+    {
+      "epoch": 1.044293707605264,
+      "grad_norm": 0.1799829751253128,
+      "learning_rate": 0.00013054812834224598,
+      "loss": 0.3316,
+      "step": 1044
+    },
+    {
+      "epoch": 1.0452939889343877,
+      "grad_norm": 0.195167675614357,
+      "learning_rate": 0.0001304812834224599,
+      "loss": 0.3556,
+      "step": 1045
+    },
+    {
+      "epoch": 1.0462942702635116,
+      "grad_norm": 0.17397986352443695,
+      "learning_rate": 0.00013041443850267382,
+      "loss": 0.3453,
+      "step": 1046
+    },
+    {
+      "epoch": 1.0472945515926355,
+      "grad_norm": 0.17894266545772552,
+      "learning_rate": 0.0001303475935828877,
+      "loss": 0.3296,
+      "step": 1047
+    },
+    {
+      "epoch": 1.0482948329217592,
+      "grad_norm": 0.19452843070030212,
+      "learning_rate": 0.0001302807486631016,
+      "loss": 0.3537,
+      "step": 1048
+    },
+    {
+      "epoch": 1.049295114250883,
+      "grad_norm": 0.18535330891609192,
+      "learning_rate": 0.00013021390374331552,
+      "loss": 0.3602,
+      "step": 1049
+    },
+    {
+      "epoch": 1.050295395580007,
+      "grad_norm": 0.18628297746181488,
+      "learning_rate": 0.00013014705882352943,
+      "loss": 0.3544,
+      "step": 1050
+    },
+    {
+      "epoch": 1.0512956769091306,
+      "grad_norm": 0.18558001518249512,
+      "learning_rate": 0.00013008021390374333,
+      "loss": 0.3579,
+      "step": 1051
+    },
+    {
+      "epoch": 1.0522959582382545,
+      "grad_norm": 0.19161471724510193,
+      "learning_rate": 0.00013001336898395722,
+      "loss": 0.3486,
+      "step": 1052
+    },
+    {
+      "epoch": 1.0532962395673784,
+      "grad_norm": 0.17937242984771729,
+      "learning_rate": 0.00012994652406417113,
+      "loss": 0.3336,
+      "step": 1053
+    },
+    {
+      "epoch": 1.054296520896502,
+      "grad_norm": 0.19983892142772675,
+      "learning_rate": 0.00012987967914438502,
+      "loss": 0.364,
+      "step": 1054
+    },
+    {
+      "epoch": 1.055296802225626,
+      "grad_norm": 0.19824886322021484,
+      "learning_rate": 0.00012981283422459894,
+      "loss": 0.3607,
+      "step": 1055
+    },
+    {
+      "epoch": 1.0562970835547498,
+      "grad_norm": 0.18773949146270752,
+      "learning_rate": 0.00012974598930481283,
+      "loss": 0.3488,
+      "step": 1056
+    },
+    {
+      "epoch": 1.0572973648838735,
+      "grad_norm": 0.18821211159229279,
+      "learning_rate": 0.00012967914438502675,
+      "loss": 0.3512,
+      "step": 1057
+    },
+    {
+      "epoch": 1.0582976462129974,
+      "grad_norm": 0.18271654844284058,
+      "learning_rate": 0.00012961229946524064,
+      "loss": 0.3552,
+      "step": 1058
+    },
+    {
+      "epoch": 1.0592979275421213,
+      "grad_norm": 0.18984054028987885,
+      "learning_rate": 0.00012954545454545456,
+      "loss": 0.3407,
+      "step": 1059
+    },
+    {
+      "epoch": 1.0602982088712452,
+      "grad_norm": 0.17650173604488373,
+      "learning_rate": 0.00012947860962566845,
+      "loss": 0.3423,
+      "step": 1060
+    },
+    {
+      "epoch": 1.0612984902003688,
+      "grad_norm": 0.18469476699829102,
+      "learning_rate": 0.00012941176470588237,
+      "loss": 0.3377,
+      "step": 1061
+    },
+    {
+      "epoch": 1.0622987715294927,
+      "grad_norm": 0.17992320656776428,
+      "learning_rate": 0.00012934491978609626,
+      "loss": 0.3464,
+      "step": 1062
+    },
+    {
+      "epoch": 1.0632990528586166,
+      "grad_norm": 0.18346770107746124,
+      "learning_rate": 0.00012927807486631015,
+      "loss": 0.3643,
+      "step": 1063
+    },
+    {
+      "epoch": 1.0642993341877403,
+      "grad_norm": 0.18397624790668488,
+      "learning_rate": 0.00012921122994652407,
+      "loss": 0.3585,
+      "step": 1064
+    },
+    {
+      "epoch": 1.0652996155168641,
+      "grad_norm": 0.17817001044750214,
+      "learning_rate": 0.00012914438502673799,
+      "loss": 0.3387,
+      "step": 1065
+    },
+    {
+      "epoch": 1.066299896845988,
+      "grad_norm": 0.1891738623380661,
+      "learning_rate": 0.00012907754010695188,
+      "loss": 0.3446,
+      "step": 1066
+    },
+    {
+      "epoch": 1.0673001781751117,
+      "grad_norm": 0.206379696726799,
+      "learning_rate": 0.00012901069518716577,
+      "loss": 0.3593,
+      "step": 1067
+    },
+    {
+      "epoch": 1.0683004595042356,
+      "grad_norm": 0.1817658245563507,
+      "learning_rate": 0.00012894385026737969,
+      "loss": 0.3481,
+      "step": 1068
+    },
+    {
+      "epoch": 1.0693007408333595,
+      "grad_norm": 0.17719882726669312,
+      "learning_rate": 0.0001288770053475936,
+      "loss": 0.3425,
+      "step": 1069
+    },
+    {
+      "epoch": 1.0703010221624831,
+      "grad_norm": 0.18691854178905487,
+      "learning_rate": 0.0001288101604278075,
+      "loss": 0.355,
+      "step": 1070
+    },
+    {
+      "epoch": 1.071301303491607,
+      "grad_norm": 0.1885235756635666,
+      "learning_rate": 0.00012874331550802139,
+      "loss": 0.3489,
+      "step": 1071
+    },
+    {
+      "epoch": 1.072301584820731,
+      "grad_norm": 0.19461680948734283,
+      "learning_rate": 0.0001286764705882353,
+      "loss": 0.3571,
+      "step": 1072
+    },
+    {
+      "epoch": 1.0733018661498546,
+      "grad_norm": 0.1823570430278778,
+      "learning_rate": 0.0001286096256684492,
+      "loss": 0.3209,
+      "step": 1073
+    },
+    {
+      "epoch": 1.0743021474789785,
+      "grad_norm": 0.18070855736732483,
+      "learning_rate": 0.0001285427807486631,
+      "loss": 0.3518,
+      "step": 1074
+    },
+    {
+      "epoch": 1.0753024288081023,
+      "grad_norm": 0.1850992739200592,
+      "learning_rate": 0.000128475935828877,
+      "loss": 0.3322,
+      "step": 1075
+    },
+    {
+      "epoch": 1.076302710137226,
+      "grad_norm": 0.18650782108306885,
+      "learning_rate": 0.00012840909090909092,
+      "loss": 0.3638,
+      "step": 1076
+    },
+    {
+      "epoch": 1.07730299146635,
+      "grad_norm": 0.18977458775043488,
+      "learning_rate": 0.0001283422459893048,
+      "loss": 0.357,
+      "step": 1077
+    },
+    {
+      "epoch": 1.0783032727954738,
+      "grad_norm": 0.18222202360630035,
+      "learning_rate": 0.00012827540106951873,
+      "loss": 0.339,
+      "step": 1078
+    },
+    {
+      "epoch": 1.0793035541245974,
+      "grad_norm": 0.1832960993051529,
+      "learning_rate": 0.00012820855614973262,
+      "loss": 0.3424,
+      "step": 1079
+    },
+    {
+      "epoch": 1.0803038354537213,
+      "grad_norm": 0.18659837543964386,
+      "learning_rate": 0.00012814171122994654,
+      "loss": 0.3742,
+      "step": 1080
+    },
+    {
+      "epoch": 1.0813041167828452,
+      "grad_norm": 0.18844346702098846,
+      "learning_rate": 0.00012807486631016043,
+      "loss": 0.355,
+      "step": 1081
+    },
+    {
+      "epoch": 1.0823043981119689,
+      "grad_norm": 0.1926846206188202,
+      "learning_rate": 0.00012800802139037432,
+      "loss": 0.3648,
+      "step": 1082
+    },
+    {
+      "epoch": 1.0833046794410928,
+      "grad_norm": 0.18758529424667358,
+      "learning_rate": 0.00012794117647058824,
+      "loss": 0.3482,
+      "step": 1083
+    },
+    {
+      "epoch": 1.0843049607702167,
+      "grad_norm": 0.18731693923473358,
+      "learning_rate": 0.00012787433155080216,
+      "loss": 0.3441,
+      "step": 1084
+    },
+    {
+      "epoch": 1.0853052420993405,
+      "grad_norm": 0.18570996820926666,
+      "learning_rate": 0.00012780748663101605,
+      "loss": 0.3511,
+      "step": 1085
+    },
+    {
+      "epoch": 1.0863055234284642,
+      "grad_norm": 0.18744036555290222,
+      "learning_rate": 0.00012774064171122994,
+      "loss": 0.3302,
+      "step": 1086
+    },
+    {
+      "epoch": 1.087305804757588,
+      "grad_norm": 0.19644097983837128,
+      "learning_rate": 0.00012767379679144386,
+      "loss": 0.3734,
+      "step": 1087
+    },
+    {
+      "epoch": 1.088306086086712,
+      "grad_norm": 0.19182941317558289,
+      "learning_rate": 0.00012760695187165777,
+      "loss": 0.3313,
+      "step": 1088
+    },
+    {
+      "epoch": 1.0893063674158356,
+      "grad_norm": 0.20565946400165558,
+      "learning_rate": 0.00012754010695187166,
+      "loss": 0.3941,
+      "step": 1089
+    },
+    {
+      "epoch": 1.0903066487449595,
+      "grad_norm": 0.19885022938251495,
+      "learning_rate": 0.00012747326203208555,
+      "loss": 0.3617,
+      "step": 1090
+    },
+    {
+      "epoch": 1.0913069300740834,
+      "grad_norm": 0.19223029911518097,
+      "learning_rate": 0.00012740641711229947,
+      "loss": 0.3562,
+      "step": 1091
+    },
+    {
+      "epoch": 1.092307211403207,
+      "grad_norm": 0.2061058133840561,
+      "learning_rate": 0.0001273395721925134,
+      "loss": 0.3702,
+      "step": 1092
+    },
+    {
+      "epoch": 1.093307492732331,
+      "grad_norm": 0.1882254183292389,
+      "learning_rate": 0.00012727272727272728,
+      "loss": 0.3741,
+      "step": 1093
+    },
+    {
+      "epoch": 1.0943077740614549,
+      "grad_norm": 0.18210825324058533,
+      "learning_rate": 0.00012720588235294117,
+      "loss": 0.3415,
+      "step": 1094
+    },
+    {
+      "epoch": 1.0953080553905785,
+      "grad_norm": 0.1922857016324997,
+      "learning_rate": 0.0001271390374331551,
+      "loss": 0.3665,
+      "step": 1095
+    },
+    {
+      "epoch": 1.0963083367197024,
+      "grad_norm": 0.18596512079238892,
+      "learning_rate": 0.00012707219251336898,
+      "loss": 0.3526,
+      "step": 1096
+    },
+    {
+      "epoch": 1.0973086180488263,
+      "grad_norm": 0.17666245996952057,
+      "learning_rate": 0.0001270053475935829,
+      "loss": 0.3374,
+      "step": 1097
+    },
+    {
+      "epoch": 1.09830889937795,
+      "grad_norm": 0.18581458926200867,
+      "learning_rate": 0.0001269385026737968,
+      "loss": 0.3482,
+      "step": 1098
+    },
+    {
+      "epoch": 1.0993091807070738,
+      "grad_norm": 0.18548209965229034,
+      "learning_rate": 0.0001268716577540107,
+      "loss": 0.3364,
+      "step": 1099
+    },
+    {
+      "epoch": 1.1003094620361977,
+      "grad_norm": 0.1841896027326584,
+      "learning_rate": 0.0001268048128342246,
+      "loss": 0.3595,
+      "step": 1100
+    },
+    {
+      "epoch": 1.1013097433653214,
+      "grad_norm": 0.19389475882053375,
+      "learning_rate": 0.00012673796791443852,
+      "loss": 0.365,
+      "step": 1101
+    },
+    {
+      "epoch": 1.1023100246944453,
+      "grad_norm": 0.18105150759220123,
+      "learning_rate": 0.0001266711229946524,
+      "loss": 0.3456,
+      "step": 1102
+    },
+    {
+      "epoch": 1.1033103060235692,
+      "grad_norm": 0.1881769448518753,
+      "learning_rate": 0.00012660427807486632,
+      "loss": 0.3385,
+      "step": 1103
+    },
+    {
+      "epoch": 1.104310587352693,
+      "grad_norm": 0.1934771090745926,
+      "learning_rate": 0.00012653743315508022,
+      "loss": 0.3586,
+      "step": 1104
+    },
+    {
+      "epoch": 1.1053108686818167,
+      "grad_norm": 0.20060613751411438,
+      "learning_rate": 0.0001264705882352941,
+      "loss": 0.3736,
+      "step": 1105
+    },
+    {
+      "epoch": 1.1063111500109406,
+      "grad_norm": 0.1890273094177246,
+      "learning_rate": 0.00012640374331550802,
+      "loss": 0.3714,
+      "step": 1106
+    },
+    {
+      "epoch": 1.1073114313400645,
+      "grad_norm": 0.20768089592456818,
+      "learning_rate": 0.00012633689839572194,
+      "loss": 0.361,
+      "step": 1107
+    },
+    {
+      "epoch": 1.1083117126691882,
+      "grad_norm": 0.1920238435268402,
+      "learning_rate": 0.00012627005347593583,
+      "loss": 0.3746,
+      "step": 1108
+    },
+    {
+      "epoch": 1.109311993998312,
+      "grad_norm": 0.18602445721626282,
+      "learning_rate": 0.00012620320855614972,
+      "loss": 0.3463,
+      "step": 1109
+    },
+    {
+      "epoch": 1.110312275327436,
+      "grad_norm": 0.1848316341638565,
+      "learning_rate": 0.00012613636363636364,
+      "loss": 0.3673,
+      "step": 1110
+    },
+    {
+      "epoch": 1.1113125566565596,
+      "grad_norm": 0.18250003457069397,
+      "learning_rate": 0.00012606951871657756,
+      "loss": 0.3587,
+      "step": 1111
+    },
+    {
+      "epoch": 1.1123128379856835,
+      "grad_norm": 0.17966102063655853,
+      "learning_rate": 0.00012600267379679145,
+      "loss": 0.3471,
+      "step": 1112
+    },
+    {
+      "epoch": 1.1133131193148074,
+      "grad_norm": 0.19589273631572723,
+      "learning_rate": 0.00012593582887700534,
+      "loss": 0.3576,
+      "step": 1113
+    },
+    {
+      "epoch": 1.114313400643931,
+      "grad_norm": 0.19104252755641937,
+      "learning_rate": 0.00012586898395721926,
+      "loss": 0.3461,
+      "step": 1114
+    },
+    {
+      "epoch": 1.115313681973055,
+      "grad_norm": 0.20112526416778564,
+      "learning_rate": 0.00012580213903743318,
+      "loss": 0.3704,
+      "step": 1115
+    },
+    {
+      "epoch": 1.1163139633021788,
+      "grad_norm": 0.19293449819087982,
+      "learning_rate": 0.00012573529411764707,
+      "loss": 0.3469,
+      "step": 1116
+    },
+    {
+      "epoch": 1.1173142446313025,
+      "grad_norm": 0.201870396733284,
+      "learning_rate": 0.00012566844919786096,
+      "loss": 0.3686,
+      "step": 1117
+    },
+    {
+      "epoch": 1.1183145259604264,
+      "grad_norm": 0.19286701083183289,
+      "learning_rate": 0.00012560160427807488,
+      "loss": 0.3457,
+      "step": 1118
+    },
+    {
+      "epoch": 1.1193148072895502,
+      "grad_norm": 0.19193173944950104,
+      "learning_rate": 0.00012553475935828877,
+      "loss": 0.3569,
+      "step": 1119
+    },
+    {
+      "epoch": 1.120315088618674,
+      "grad_norm": 0.19239270687103271,
+      "learning_rate": 0.00012546791443850269,
+      "loss": 0.3457,
+      "step": 1120
+    },
+    {
+      "epoch": 1.1213153699477978,
+      "grad_norm": 0.18718737363815308,
+      "learning_rate": 0.00012540106951871658,
+      "loss": 0.365,
+      "step": 1121
+    },
+    {
+      "epoch": 1.1223156512769217,
+      "grad_norm": 0.19814792275428772,
+      "learning_rate": 0.0001253342245989305,
+      "loss": 0.3624,
+      "step": 1122
+    },
+    {
+      "epoch": 1.1233159326060456,
+      "grad_norm": 0.20927958190441132,
+      "learning_rate": 0.00012526737967914439,
+      "loss": 0.3676,
+      "step": 1123
+    },
+    {
+      "epoch": 1.1243162139351692,
+      "grad_norm": 0.18859633803367615,
+      "learning_rate": 0.0001252005347593583,
+      "loss": 0.3672,
+      "step": 1124
+    },
+    {
+      "epoch": 1.1253164952642931,
+      "grad_norm": 0.18651632964611053,
+      "learning_rate": 0.0001251336898395722,
+      "loss": 0.3248,
+      "step": 1125
+    },
+    {
+      "epoch": 1.1263167765934168,
+      "grad_norm": 0.19732235372066498,
+      "learning_rate": 0.0001250668449197861,
+      "loss": 0.3734,
+      "step": 1126
+    },
+    {
+      "epoch": 1.1273170579225407,
+      "grad_norm": 0.2037302851676941,
+      "learning_rate": 0.000125,
+      "loss": 0.371,
+      "step": 1127
+    },
+    {
+      "epoch": 1.1283173392516646,
+      "grad_norm": 0.19439762830734253,
+      "learning_rate": 0.0001249331550802139,
+      "loss": 0.3552,
+      "step": 1128
+    },
+    {
+      "epoch": 1.1293176205807884,
+      "grad_norm": 0.21107234060764313,
+      "learning_rate": 0.0001248663101604278,
+      "loss": 0.4027,
+      "step": 1129
+    },
+    {
+      "epoch": 1.130317901909912,
+      "grad_norm": 0.19808197021484375,
+      "learning_rate": 0.00012479946524064173,
+      "loss": 0.3568,
+      "step": 1130
+    },
+    {
+      "epoch": 1.131318183239036,
+      "grad_norm": 0.18348826467990875,
+      "learning_rate": 0.00012473262032085562,
+      "loss": 0.3504,
+      "step": 1131
+    },
+    {
+      "epoch": 1.1323184645681599,
+      "grad_norm": 0.19102126359939575,
+      "learning_rate": 0.0001246657754010695,
+      "loss": 0.3604,
+      "step": 1132
+    },
+    {
+      "epoch": 1.1333187458972835,
+      "grad_norm": 0.19289638102054596,
+      "learning_rate": 0.00012459893048128343,
+      "loss": 0.3441,
+      "step": 1133
+    },
+    {
+      "epoch": 1.1343190272264074,
+      "grad_norm": 0.1917727291584015,
+      "learning_rate": 0.00012453208556149735,
+      "loss": 0.3566,
+      "step": 1134
+    },
+    {
+      "epoch": 1.1353193085555313,
+      "grad_norm": 0.19141171872615814,
+      "learning_rate": 0.00012446524064171124,
+      "loss": 0.3524,
+      "step": 1135
+    },
+    {
+      "epoch": 1.136319589884655,
+      "grad_norm": 0.19459858536720276,
+      "learning_rate": 0.00012439839572192513,
+      "loss": 0.3542,
+      "step": 1136
+    },
+    {
+      "epoch": 1.1373198712137789,
+      "grad_norm": 0.1860855668783188,
+      "learning_rate": 0.00012433155080213905,
+      "loss": 0.3402,
+      "step": 1137
+    },
+    {
+      "epoch": 1.1383201525429028,
+      "grad_norm": 0.1922546923160553,
+      "learning_rate": 0.00012426470588235294,
+      "loss": 0.3671,
+      "step": 1138
+    },
+    {
+      "epoch": 1.1393204338720264,
+      "grad_norm": 0.1903626173734665,
+      "learning_rate": 0.00012419786096256685,
+      "loss": 0.3738,
+      "step": 1139
+    },
+    {
+      "epoch": 1.1403207152011503,
+      "grad_norm": 0.2029544562101364,
+      "learning_rate": 0.00012413101604278075,
+      "loss": 0.3636,
+      "step": 1140
+    },
+    {
+      "epoch": 1.1413209965302742,
+      "grad_norm": 0.19865334033966064,
+      "learning_rate": 0.00012406417112299466,
+      "loss": 0.3589,
+      "step": 1141
+    },
+    {
+      "epoch": 1.142321277859398,
+      "grad_norm": 0.18906521797180176,
+      "learning_rate": 0.00012399732620320855,
+      "loss": 0.377,
+      "step": 1142
+    },
+    {
+      "epoch": 1.1433215591885217,
+      "grad_norm": 0.18444448709487915,
+      "learning_rate": 0.00012393048128342247,
+      "loss": 0.3463,
+      "step": 1143
+    },
+    {
+      "epoch": 1.1443218405176456,
+      "grad_norm": 0.18665602803230286,
+      "learning_rate": 0.00012386363636363636,
+      "loss": 0.3489,
+      "step": 1144
+    },
+    {
+      "epoch": 1.1453221218467693,
+      "grad_norm": 0.17723019421100616,
+      "learning_rate": 0.00012379679144385028,
+      "loss": 0.3333,
+      "step": 1145
+    },
+    {
+      "epoch": 1.1463224031758932,
+      "grad_norm": 0.19831480085849762,
+      "learning_rate": 0.00012372994652406417,
+      "loss": 0.3578,
+      "step": 1146
+    },
+    {
+      "epoch": 1.147322684505017,
+      "grad_norm": 0.18058457970619202,
+      "learning_rate": 0.00012366310160427806,
+      "loss": 0.3404,
+      "step": 1147
+    },
+    {
+      "epoch": 1.148322965834141,
+      "grad_norm": 0.18793006241321564,
+      "learning_rate": 0.000123596256684492,
+      "loss": 0.3481,
+      "step": 1148
+    },
+    {
+      "epoch": 1.1493232471632646,
+      "grad_norm": 0.20226594805717468,
+      "learning_rate": 0.0001235294117647059,
+      "loss": 0.3552,
+      "step": 1149
+    },
+    {
+      "epoch": 1.1503235284923885,
+      "grad_norm": 0.19378265738487244,
+      "learning_rate": 0.0001234625668449198,
+      "loss": 0.344,
+      "step": 1150
+    },
+    {
+      "epoch": 1.1513238098215124,
+      "grad_norm": 0.18742580711841583,
+      "learning_rate": 0.00012339572192513368,
+      "loss": 0.3338,
+      "step": 1151
+    },
+    {
+      "epoch": 1.152324091150636,
+      "grad_norm": 0.20786020159721375,
+      "learning_rate": 0.0001233288770053476,
+      "loss": 0.3582,
+      "step": 1152
+    },
+    {
+      "epoch": 1.15332437247976,
+      "grad_norm": 0.19193722307682037,
+      "learning_rate": 0.00012326203208556152,
+      "loss": 0.3432,
+      "step": 1153
+    },
+    {
+      "epoch": 1.1543246538088838,
+      "grad_norm": 0.1848604828119278,
+      "learning_rate": 0.0001231951871657754,
+      "loss": 0.3649,
+      "step": 1154
+    },
+    {
+      "epoch": 1.1553249351380075,
+      "grad_norm": 0.19047203660011292,
+      "learning_rate": 0.0001231283422459893,
+      "loss": 0.3547,
+      "step": 1155
+    },
+    {
+      "epoch": 1.1563252164671314,
+      "grad_norm": 0.19730713963508606,
+      "learning_rate": 0.00012306149732620322,
+      "loss": 0.3727,
+      "step": 1156
+    },
+    {
+      "epoch": 1.1573254977962553,
+      "grad_norm": 0.18616151809692383,
+      "learning_rate": 0.00012299465240641713,
+      "loss": 0.3304,
+      "step": 1157
+    },
+    {
+      "epoch": 1.158325779125379,
+      "grad_norm": 0.18702992796897888,
+      "learning_rate": 0.00012292780748663102,
+      "loss": 0.3378,
+      "step": 1158
+    },
+    {
+      "epoch": 1.1593260604545028,
+      "grad_norm": 0.1955466866493225,
+      "learning_rate": 0.00012286096256684492,
+      "loss": 0.3663,
+      "step": 1159
+    },
+    {
+      "epoch": 1.1603263417836267,
+      "grad_norm": 0.1863570213317871,
+      "learning_rate": 0.00012279411764705883,
+      "loss": 0.3528,
+      "step": 1160
+    },
+    {
+      "epoch": 1.1613266231127504,
+      "grad_norm": 0.19376443326473236,
+      "learning_rate": 0.00012272727272727272,
+      "loss": 0.3689,
+      "step": 1161
+    },
+    {
+      "epoch": 1.1623269044418743,
+      "grad_norm": 0.1831565499305725,
+      "learning_rate": 0.00012266042780748664,
+      "loss": 0.3462,
+      "step": 1162
+    },
+    {
+      "epoch": 1.1633271857709981,
+      "grad_norm": 0.19377927482128143,
+      "learning_rate": 0.00012259358288770053,
+      "loss": 0.3809,
+      "step": 1163
+    },
+    {
+      "epoch": 1.1643274671001218,
+      "grad_norm": 0.1892099231481552,
+      "learning_rate": 0.00012252673796791445,
+      "loss": 0.3466,
+      "step": 1164
+    },
+    {
+      "epoch": 1.1653277484292457,
+      "grad_norm": 0.1814454048871994,
+      "learning_rate": 0.00012245989304812834,
+      "loss": 0.3445,
+      "step": 1165
+    },
+    {
+      "epoch": 1.1663280297583696,
+      "grad_norm": 0.2038552314043045,
+      "learning_rate": 0.00012239304812834226,
+      "loss": 0.3705,
+      "step": 1166
+    },
+    {
+      "epoch": 1.1673283110874935,
+      "grad_norm": 0.18704229593276978,
+      "learning_rate": 0.00012232620320855615,
+      "loss": 0.3606,
+      "step": 1167
+    },
+    {
+      "epoch": 1.1683285924166171,
+      "grad_norm": 0.17329853773117065,
+      "learning_rate": 0.00012225935828877007,
+      "loss": 0.3161,
+      "step": 1168
+    },
+    {
+      "epoch": 1.169328873745741,
+      "grad_norm": 0.22162747383117676,
+      "learning_rate": 0.00012219251336898396,
+      "loss": 0.344,
+      "step": 1169
+    },
+    {
+      "epoch": 1.170329155074865,
+      "grad_norm": 0.204198956489563,
+      "learning_rate": 0.00012212566844919785,
+      "loss": 0.3685,
+      "step": 1170
+    },
+    {
+      "epoch": 1.1713294364039886,
+      "grad_norm": 0.18648314476013184,
+      "learning_rate": 0.00012205882352941178,
+      "loss": 0.3511,
+      "step": 1171
+    },
+    {
+      "epoch": 1.1723297177331125,
+      "grad_norm": 0.21162071824073792,
+      "learning_rate": 0.00012199197860962569,
+      "loss": 0.377,
+      "step": 1172
+    },
+    {
+      "epoch": 1.1733299990622363,
+      "grad_norm": 0.21572354435920715,
+      "learning_rate": 0.00012192513368983958,
+      "loss": 0.3521,
+      "step": 1173
+    },
+    {
+      "epoch": 1.17433028039136,
+      "grad_norm": 0.1933080404996872,
+      "learning_rate": 0.00012185828877005348,
+      "loss": 0.3507,
+      "step": 1174
+    },
+    {
+      "epoch": 1.175330561720484,
+      "grad_norm": 0.19826601445674896,
+      "learning_rate": 0.0001217914438502674,
+      "loss": 0.36,
+      "step": 1175
+    },
+    {
+      "epoch": 1.1763308430496078,
+      "grad_norm": 0.18432265520095825,
+      "learning_rate": 0.00012172459893048129,
+      "loss": 0.3461,
+      "step": 1176
+    },
+    {
+      "epoch": 1.1773311243787314,
+      "grad_norm": 0.19286255538463593,
+      "learning_rate": 0.0001216577540106952,
+      "loss": 0.3693,
+      "step": 1177
+    },
+    {
+      "epoch": 1.1783314057078553,
+      "grad_norm": 0.18874181807041168,
+      "learning_rate": 0.00012159090909090908,
+      "loss": 0.3363,
+      "step": 1178
+    },
+    {
+      "epoch": 1.1793316870369792,
+      "grad_norm": 0.1872003823518753,
+      "learning_rate": 0.000121524064171123,
+      "loss": 0.3433,
+      "step": 1179
+    },
+    {
+      "epoch": 1.1803319683661029,
+      "grad_norm": 0.19510194659233093,
+      "learning_rate": 0.00012145721925133691,
+      "loss": 0.3636,
+      "step": 1180
+    },
+    {
+      "epoch": 1.1813322496952268,
+      "grad_norm": 0.18251201510429382,
+      "learning_rate": 0.00012139037433155081,
+      "loss": 0.3367,
+      "step": 1181
+    },
+    {
+      "epoch": 1.1823325310243507,
+      "grad_norm": 0.19301854074001312,
+      "learning_rate": 0.0001213235294117647,
+      "loss": 0.3647,
+      "step": 1182
+    },
+    {
+      "epoch": 1.1833328123534743,
+      "grad_norm": 0.19702407717704773,
+      "learning_rate": 0.00012125668449197862,
+      "loss": 0.3368,
+      "step": 1183
+    },
+    {
+      "epoch": 1.1843330936825982,
+      "grad_norm": 0.186981201171875,
+      "learning_rate": 0.00012118983957219252,
+      "loss": 0.3326,
+      "step": 1184
+    },
+    {
+      "epoch": 1.185333375011722,
+      "grad_norm": 0.20133353769779205,
+      "learning_rate": 0.00012112299465240642,
+      "loss": 0.3592,
+      "step": 1185
+    },
+    {
+      "epoch": 1.186333656340846,
+      "grad_norm": 0.1925647258758545,
+      "learning_rate": 0.00012105614973262032,
+      "loss": 0.3418,
+      "step": 1186
+    },
+    {
+      "epoch": 1.1873339376699696,
+      "grad_norm": 0.19109131395816803,
+      "learning_rate": 0.00012098930481283424,
+      "loss": 0.3618,
+      "step": 1187
+    },
+    {
+      "epoch": 1.1883342189990935,
+      "grad_norm": 0.201302170753479,
+      "learning_rate": 0.00012092245989304813,
+      "loss": 0.352,
+      "step": 1188
+    },
+    {
+      "epoch": 1.1893345003282172,
+      "grad_norm": 0.18659594655036926,
+      "learning_rate": 0.00012085561497326203,
+      "loss": 0.3472,
+      "step": 1189
+    },
+    {
+      "epoch": 1.190334781657341,
+      "grad_norm": 0.19188572466373444,
+      "learning_rate": 0.00012078877005347595,
+      "loss": 0.35,
+      "step": 1190
+    },
+    {
+      "epoch": 1.191335062986465,
+      "grad_norm": 0.19020341336727142,
+      "learning_rate": 0.00012072192513368985,
+      "loss": 0.3455,
+      "step": 1191
+    },
+    {
+      "epoch": 1.1923353443155889,
+      "grad_norm": 0.1902928203344345,
+      "learning_rate": 0.00012065508021390375,
+      "loss": 0.342,
+      "step": 1192
+    },
+    {
+      "epoch": 1.1933356256447125,
+      "grad_norm": 0.1958119124174118,
+      "learning_rate": 0.00012058823529411765,
+      "loss": 0.3592,
+      "step": 1193
+    },
+    {
+      "epoch": 1.1943359069738364,
+      "grad_norm": 0.18990500271320343,
+      "learning_rate": 0.00012052139037433157,
+      "loss": 0.3433,
+      "step": 1194
+    },
+    {
+      "epoch": 1.1953361883029603,
+      "grad_norm": 0.19543765485286713,
+      "learning_rate": 0.00012045454545454546,
+      "loss": 0.3683,
+      "step": 1195
+    },
+    {
+      "epoch": 1.196336469632084,
+      "grad_norm": 0.18834958970546722,
+      "learning_rate": 0.00012038770053475936,
+      "loss": 0.3421,
+      "step": 1196
+    },
+    {
+      "epoch": 1.1973367509612078,
+      "grad_norm": 0.19068677723407745,
+      "learning_rate": 0.00012032085561497325,
+      "loss": 0.3525,
+      "step": 1197
+    },
+    {
+      "epoch": 1.1983370322903317,
+      "grad_norm": 0.19191397726535797,
+      "learning_rate": 0.00012025401069518719,
+      "loss": 0.3547,
+      "step": 1198
+    },
+    {
+      "epoch": 1.1993373136194554,
+      "grad_norm": 0.1843283474445343,
+      "learning_rate": 0.00012018716577540108,
+      "loss": 0.3385,
+      "step": 1199
+    },
+    {
+      "epoch": 1.2003375949485793,
+      "grad_norm": 0.18444311618804932,
+      "learning_rate": 0.00012012032085561498,
+      "loss": 0.3493,
+      "step": 1200
+    },
+    {
+      "epoch": 1.2013378762777032,
+      "grad_norm": 0.20181988179683685,
+      "learning_rate": 0.00012005347593582887,
+      "loss": 0.3684,
+      "step": 1201
+    },
+    {
+      "epoch": 1.2023381576068268,
+      "grad_norm": 0.19645917415618896,
+      "learning_rate": 0.00011998663101604279,
+      "loss": 0.3556,
+      "step": 1202
+    },
+    {
+      "epoch": 1.2033384389359507,
+      "grad_norm": 0.19454020261764526,
+      "learning_rate": 0.0001199197860962567,
+      "loss": 0.3377,
+      "step": 1203
+    },
+    {
+      "epoch": 1.2043387202650746,
+      "grad_norm": 0.1820986121892929,
+      "learning_rate": 0.00011985294117647058,
+      "loss": 0.3538,
+      "step": 1204
+    },
+    {
+      "epoch": 1.2053390015941985,
+      "grad_norm": 0.19814157485961914,
+      "learning_rate": 0.00011978609625668449,
+      "loss": 0.3536,
+      "step": 1205
+    },
+    {
+      "epoch": 1.2063392829233222,
+      "grad_norm": 0.19701911509037018,
+      "learning_rate": 0.00011971925133689841,
+      "loss": 0.3592,
+      "step": 1206
+    },
+    {
+      "epoch": 1.207339564252446,
+      "grad_norm": 0.18923602998256683,
+      "learning_rate": 0.00011965240641711231,
+      "loss": 0.3526,
+      "step": 1207
+    },
+    {
+      "epoch": 1.2083398455815697,
+      "grad_norm": 0.206210657954216,
+      "learning_rate": 0.0001195855614973262,
+      "loss": 0.3712,
+      "step": 1208
+    },
+    {
+      "epoch": 1.2093401269106936,
+      "grad_norm": 0.21036550402641296,
+      "learning_rate": 0.0001195187165775401,
+      "loss": 0.375,
+      "step": 1209
+    },
+    {
+      "epoch": 1.2103404082398175,
+      "grad_norm": 0.19063423573970795,
+      "learning_rate": 0.00011945187165775402,
+      "loss": 0.3605,
+      "step": 1210
+    },
+    {
+      "epoch": 1.2113406895689414,
+      "grad_norm": 0.18491782248020172,
+      "learning_rate": 0.00011938502673796792,
+      "loss": 0.3435,
+      "step": 1211
+    },
+    {
+      "epoch": 1.212340970898065,
+      "grad_norm": 0.18909581005573273,
+      "learning_rate": 0.00011931818181818182,
+      "loss": 0.3572,
+      "step": 1212
+    },
+    {
+      "epoch": 1.213341252227189,
+      "grad_norm": 0.19811712205410004,
+      "learning_rate": 0.00011925133689839574,
+      "loss": 0.3798,
+      "step": 1213
+    },
+    {
+      "epoch": 1.2143415335563128,
+      "grad_norm": 0.20707352459430695,
+      "learning_rate": 0.00011918449197860964,
+      "loss": 0.3925,
+      "step": 1214
+    },
+    {
+      "epoch": 1.2153418148854365,
+      "grad_norm": 0.18582026660442352,
+      "learning_rate": 0.00011911764705882353,
+      "loss": 0.3467,
+      "step": 1215
+    },
+    {
+      "epoch": 1.2163420962145604,
+      "grad_norm": 0.19096308946609497,
+      "learning_rate": 0.00011905080213903744,
+      "loss": 0.3563,
+      "step": 1216
+    },
+    {
+      "epoch": 1.2173423775436842,
+      "grad_norm": 0.18377093970775604,
+      "learning_rate": 0.00011898395721925135,
+      "loss": 0.3561,
+      "step": 1217
+    },
+    {
+      "epoch": 1.218342658872808,
+      "grad_norm": 0.19126996397972107,
+      "learning_rate": 0.00011891711229946525,
+      "loss": 0.3379,
+      "step": 1218
+    },
+    {
+      "epoch": 1.2193429402019318,
+      "grad_norm": 0.19350090622901917,
+      "learning_rate": 0.00011885026737967915,
+      "loss": 0.3541,
+      "step": 1219
+    },
+    {
+      "epoch": 1.2203432215310557,
+      "grad_norm": 0.19457250833511353,
+      "learning_rate": 0.00011878342245989304,
+      "loss": 0.3496,
+      "step": 1220
+    },
+    {
+      "epoch": 1.2213435028601793,
+      "grad_norm": 0.20100454986095428,
+      "learning_rate": 0.00011871657754010697,
+      "loss": 0.3614,
+      "step": 1221
+    },
+    {
+      "epoch": 1.2223437841893032,
+      "grad_norm": 0.18846745789051056,
+      "learning_rate": 0.00011864973262032086,
+      "loss": 0.3586,
+      "step": 1222
+    },
+    {
+      "epoch": 1.2233440655184271,
+      "grad_norm": 0.18870976567268372,
+      "learning_rate": 0.00011858288770053477,
+      "loss": 0.3424,
+      "step": 1223
+    },
+    {
+      "epoch": 1.2243443468475508,
+      "grad_norm": 0.20238159596920013,
+      "learning_rate": 0.00011851604278074866,
+      "loss": 0.3716,
+      "step": 1224
+    },
+    {
+      "epoch": 1.2253446281766747,
+      "grad_norm": 0.18229271471500397,
+      "learning_rate": 0.00011844919786096258,
+      "loss": 0.3271,
+      "step": 1225
+    },
+    {
+      "epoch": 1.2263449095057986,
+      "grad_norm": 0.1907934993505478,
+      "learning_rate": 0.00011838235294117648,
+      "loss": 0.3573,
+      "step": 1226
+    },
+    {
+      "epoch": 1.2273451908349222,
+      "grad_norm": 0.1861458122730255,
+      "learning_rate": 0.00011831550802139037,
+      "loss": 0.3555,
+      "step": 1227
+    },
+    {
+      "epoch": 1.228345472164046,
+      "grad_norm": 0.198177307844162,
+      "learning_rate": 0.00011824866310160428,
+      "loss": 0.365,
+      "step": 1228
+    },
+    {
+      "epoch": 1.22934575349317,
+      "grad_norm": 0.18466392159461975,
+      "learning_rate": 0.0001181818181818182,
+      "loss": 0.3362,
+      "step": 1229
+    },
+    {
+      "epoch": 1.2303460348222939,
+      "grad_norm": 0.19253143668174744,
+      "learning_rate": 0.0001181149732620321,
+      "loss": 0.3335,
+      "step": 1230
+    },
+    {
+      "epoch": 1.2313463161514175,
+      "grad_norm": 0.18905387818813324,
+      "learning_rate": 0.00011804812834224599,
+      "loss": 0.3354,
+      "step": 1231
+    },
+    {
+      "epoch": 1.2323465974805414,
+      "grad_norm": 0.18816527724266052,
+      "learning_rate": 0.0001179812834224599,
+      "loss": 0.3532,
+      "step": 1232
+    },
+    {
+      "epoch": 1.2333468788096653,
+      "grad_norm": 0.19693224132061005,
+      "learning_rate": 0.00011791443850267381,
+      "loss": 0.3675,
+      "step": 1233
+    },
+    {
+      "epoch": 1.234347160138789,
+      "grad_norm": 0.20398585498332977,
+      "learning_rate": 0.0001178475935828877,
+      "loss": 0.3551,
+      "step": 1234
+    },
+    {
+      "epoch": 1.2353474414679129,
+      "grad_norm": 0.19057269394397736,
+      "learning_rate": 0.0001177807486631016,
+      "loss": 0.3578,
+      "step": 1235
+    },
+    {
+      "epoch": 1.2363477227970368,
+      "grad_norm": 0.18603938817977905,
+      "learning_rate": 0.00011771390374331552,
+      "loss": 0.3462,
+      "step": 1236
+    },
+    {
+      "epoch": 1.2373480041261604,
+      "grad_norm": 0.19099555909633636,
+      "learning_rate": 0.00011764705882352942,
+      "loss": 0.3699,
+      "step": 1237
+    },
+    {
+      "epoch": 1.2383482854552843,
+      "grad_norm": 0.19030873477458954,
+      "learning_rate": 0.00011758021390374332,
+      "loss": 0.3306,
+      "step": 1238
+    },
+    {
+      "epoch": 1.2393485667844082,
+      "grad_norm": 0.1933780461549759,
+      "learning_rate": 0.00011751336898395722,
+      "loss": 0.3579,
+      "step": 1239
+    },
+    {
+      "epoch": 1.2403488481135319,
+      "grad_norm": 0.1877300888299942,
+      "learning_rate": 0.00011744652406417114,
+      "loss": 0.3511,
+      "step": 1240
+    },
+    {
+      "epoch": 1.2413491294426557,
+      "grad_norm": 0.21098686754703522,
+      "learning_rate": 0.00011737967914438503,
+      "loss": 0.3637,
+      "step": 1241
+    },
+    {
+      "epoch": 1.2423494107717796,
+      "grad_norm": 0.18884582817554474,
+      "learning_rate": 0.00011731283422459894,
+      "loss": 0.354,
+      "step": 1242
+    },
+    {
+      "epoch": 1.2433496921009033,
+      "grad_norm": 0.1900736540555954,
+      "learning_rate": 0.00011724598930481283,
+      "loss": 0.3439,
+      "step": 1243
+    },
+    {
+      "epoch": 1.2443499734300272,
+      "grad_norm": 0.1971302628517151,
+      "learning_rate": 0.00011717914438502675,
+      "loss": 0.3744,
+      "step": 1244
+    },
+    {
+      "epoch": 1.245350254759151,
+      "grad_norm": 0.18847867846488953,
+      "learning_rate": 0.00011711229946524065,
+      "loss": 0.3627,
+      "step": 1245
+    },
+    {
+      "epoch": 1.2463505360882747,
+      "grad_norm": 0.19386990368366241,
+      "learning_rate": 0.00011704545454545454,
+      "loss": 0.349,
+      "step": 1246
+    },
+    {
+      "epoch": 1.2473508174173986,
+      "grad_norm": 0.19323894381523132,
+      "learning_rate": 0.00011697860962566845,
+      "loss": 0.3709,
+      "step": 1247
+    },
+    {
+      "epoch": 1.2483510987465225,
+      "grad_norm": 0.18393279612064362,
+      "learning_rate": 0.00011691176470588236,
+      "loss": 0.3523,
+      "step": 1248
+    },
+    {
+      "epoch": 1.2493513800756464,
+      "grad_norm": 0.18940933048725128,
+      "learning_rate": 0.00011684491978609627,
+      "loss": 0.3657,
+      "step": 1249
+    },
+    {
+      "epoch": 1.25035166140477,
+      "grad_norm": 0.1846853494644165,
+      "learning_rate": 0.00011677807486631016,
+      "loss": 0.3258,
+      "step": 1250
+    },
+    {
+      "epoch": 1.251351942733894,
+      "grad_norm": 0.19235248863697052,
+      "learning_rate": 0.00011671122994652406,
+      "loss": 0.3552,
+      "step": 1251
+    },
+    {
+      "epoch": 1.2523522240630176,
+      "grad_norm": 0.18935024738311768,
+      "learning_rate": 0.00011664438502673798,
+      "loss": 0.3638,
+      "step": 1252
+    },
+    {
+      "epoch": 1.2533525053921415,
+      "grad_norm": 0.18913301825523376,
+      "learning_rate": 0.00011657754010695187,
+      "loss": 0.343,
+      "step": 1253
+    },
+    {
+      "epoch": 1.2543527867212654,
+      "grad_norm": 0.18621708452701569,
+      "learning_rate": 0.00011651069518716578,
+      "loss": 0.3401,
+      "step": 1254
+    },
+    {
+      "epoch": 1.2553530680503893,
+      "grad_norm": 0.19752287864685059,
+      "learning_rate": 0.0001164438502673797,
+      "loss": 0.349,
+      "step": 1255
+    },
+    {
+      "epoch": 1.256353349379513,
+      "grad_norm": 0.18767908215522766,
+      "learning_rate": 0.0001163770053475936,
+      "loss": 0.3311,
+      "step": 1256
+    },
+    {
+      "epoch": 1.2573536307086368,
+      "grad_norm": 0.19681666791439056,
+      "learning_rate": 0.00011631016042780749,
+      "loss": 0.3312,
+      "step": 1257
+    },
+    {
+      "epoch": 1.2583539120377607,
+      "grad_norm": 0.1895011067390442,
+      "learning_rate": 0.00011624331550802139,
+      "loss": 0.3359,
+      "step": 1258
+    },
+    {
+      "epoch": 1.2593541933668844,
+      "grad_norm": 0.20513348281383514,
+      "learning_rate": 0.00011617647058823531,
+      "loss": 0.3495,
+      "step": 1259
+    },
+    {
+      "epoch": 1.2603544746960083,
+      "grad_norm": 0.2113916128873825,
+      "learning_rate": 0.0001161096256684492,
+      "loss": 0.3742,
+      "step": 1260
+    },
+    {
+      "epoch": 1.2613547560251321,
+      "grad_norm": 0.19102343916893005,
+      "learning_rate": 0.0001160427807486631,
+      "loss": 0.3327,
+      "step": 1261
+    },
+    {
+      "epoch": 1.2623550373542558,
+      "grad_norm": 0.20323969423770905,
+      "learning_rate": 0.000115975935828877,
+      "loss": 0.356,
+      "step": 1262
+    },
+    {
+      "epoch": 1.2633553186833797,
+      "grad_norm": 0.2006496638059616,
+      "learning_rate": 0.00011590909090909093,
+      "loss": 0.3498,
+      "step": 1263
+    },
+    {
+      "epoch": 1.2643556000125036,
+      "grad_norm": 0.18970783054828644,
+      "learning_rate": 0.00011584224598930482,
+      "loss": 0.3254,
+      "step": 1264
+    },
+    {
+      "epoch": 1.2653558813416272,
+      "grad_norm": 0.19461022317409515,
+      "learning_rate": 0.00011577540106951872,
+      "loss": 0.3672,
+      "step": 1265
+    },
+    {
+      "epoch": 1.2663561626707511,
+      "grad_norm": 0.1997997760772705,
+      "learning_rate": 0.00011570855614973261,
+      "loss": 0.368,
+      "step": 1266
+    },
+    {
+      "epoch": 1.267356443999875,
+      "grad_norm": 0.19266700744628906,
+      "learning_rate": 0.00011564171122994653,
+      "loss": 0.3506,
+      "step": 1267
+    },
+    {
+      "epoch": 1.268356725328999,
+      "grad_norm": 0.19196540117263794,
+      "learning_rate": 0.00011557486631016044,
+      "loss": 0.3466,
+      "step": 1268
+    },
+    {
+      "epoch": 1.2693570066581226,
+      "grad_norm": 0.19035640358924866,
+      "learning_rate": 0.00011550802139037433,
+      "loss": 0.3544,
+      "step": 1269
+    },
+    {
+      "epoch": 1.2703572879872465,
+      "grad_norm": 0.20412512123584747,
+      "learning_rate": 0.00011544117647058823,
+      "loss": 0.3691,
+      "step": 1270
+    },
+    {
+      "epoch": 1.2713575693163701,
+      "grad_norm": 0.20337072014808655,
+      "learning_rate": 0.00011537433155080215,
+      "loss": 0.4002,
+      "step": 1271
+    },
+    {
+      "epoch": 1.272357850645494,
+      "grad_norm": 0.19159816205501556,
+      "learning_rate": 0.00011530748663101605,
+      "loss": 0.3502,
+      "step": 1272
+    },
+    {
+      "epoch": 1.273358131974618,
+      "grad_norm": 0.19147491455078125,
+      "learning_rate": 0.00011524064171122995,
+      "loss": 0.3651,
+      "step": 1273
+    },
+    {
+      "epoch": 1.2743584133037418,
+      "grad_norm": 0.19759878516197205,
+      "learning_rate": 0.00011517379679144386,
+      "loss": 0.3756,
+      "step": 1274
+    },
+    {
+      "epoch": 1.2753586946328654,
+      "grad_norm": 0.19178122282028198,
+      "learning_rate": 0.00011510695187165777,
+      "loss": 0.3445,
+      "step": 1275
+    },
+    {
+      "epoch": 1.2763589759619893,
+      "grad_norm": 0.20684480667114258,
+      "learning_rate": 0.00011504010695187166,
+      "loss": 0.3405,
+      "step": 1276
+    },
+    {
+      "epoch": 1.277359257291113,
+      "grad_norm": 0.202129065990448,
+      "learning_rate": 0.00011497326203208556,
+      "loss": 0.3619,
+      "step": 1277
+    },
+    {
+      "epoch": 1.2783595386202369,
+      "grad_norm": 0.20815640687942505,
+      "learning_rate": 0.00011490641711229948,
+      "loss": 0.3489,
+      "step": 1278
+    },
+    {
+      "epoch": 1.2793598199493608,
+      "grad_norm": 0.19774824380874634,
+      "learning_rate": 0.00011483957219251338,
+      "loss": 0.371,
+      "step": 1279
+    },
+    {
+      "epoch": 1.2803601012784847,
+      "grad_norm": 0.19183914363384247,
+      "learning_rate": 0.00011477272727272728,
+      "loss": 0.3411,
+      "step": 1280
+    },
+    {
+      "epoch": 1.2813603826076083,
+      "grad_norm": 0.20658662915229797,
+      "learning_rate": 0.00011470588235294118,
+      "loss": 0.3612,
+      "step": 1281
+    },
+    {
+      "epoch": 1.2823606639367322,
+      "grad_norm": 0.19902697205543518,
+      "learning_rate": 0.0001146390374331551,
+      "loss": 0.3465,
+      "step": 1282
+    },
+    {
+      "epoch": 1.283360945265856,
+      "grad_norm": 0.18690262734889984,
+      "learning_rate": 0.00011457219251336899,
+      "loss": 0.3493,
+      "step": 1283
+    },
+    {
+      "epoch": 1.2843612265949798,
+      "grad_norm": 0.1825837343931198,
+      "learning_rate": 0.00011450534759358289,
+      "loss": 0.3336,
+      "step": 1284
+    },
+    {
+      "epoch": 1.2853615079241036,
+      "grad_norm": 0.19882430136203766,
+      "learning_rate": 0.00011443850267379678,
+      "loss": 0.3641,
+      "step": 1285
+    },
+    {
+      "epoch": 1.2863617892532275,
+      "grad_norm": 0.1894419640302658,
+      "learning_rate": 0.0001143716577540107,
+      "loss": 0.3409,
+      "step": 1286
+    },
+    {
+      "epoch": 1.2873620705823514,
+      "grad_norm": 0.19247683882713318,
+      "learning_rate": 0.0001143048128342246,
+      "loss": 0.328,
+      "step": 1287
+    },
+    {
+      "epoch": 1.288362351911475,
+      "grad_norm": 0.19267721474170685,
+      "learning_rate": 0.00011423796791443851,
+      "loss": 0.3534,
+      "step": 1288
+    },
+    {
+      "epoch": 1.289362633240599,
+      "grad_norm": 0.19307327270507812,
+      "learning_rate": 0.0001141711229946524,
+      "loss": 0.3594,
+      "step": 1289
+    },
+    {
+      "epoch": 1.2903629145697226,
+      "grad_norm": 0.19241534173488617,
+      "learning_rate": 0.00011410427807486632,
+      "loss": 0.3475,
+      "step": 1290
+    },
+    {
+      "epoch": 1.2913631958988465,
+      "grad_norm": 0.19015994668006897,
+      "learning_rate": 0.00011403743315508022,
+      "loss": 0.3492,
+      "step": 1291
+    },
+    {
+      "epoch": 1.2923634772279704,
+      "grad_norm": 0.1863250881433487,
+      "learning_rate": 0.00011397058823529411,
+      "loss": 0.3367,
+      "step": 1292
+    },
+    {
+      "epoch": 1.2933637585570943,
+      "grad_norm": 0.19604456424713135,
+      "learning_rate": 0.00011390374331550802,
+      "loss": 0.3623,
+      "step": 1293
+    },
+    {
+      "epoch": 1.294364039886218,
+      "grad_norm": 0.19332492351531982,
+      "learning_rate": 0.00011383689839572194,
+      "loss": 0.3606,
+      "step": 1294
+    },
+    {
+      "epoch": 1.2953643212153418,
+      "grad_norm": 0.19089557230472565,
+      "learning_rate": 0.00011377005347593583,
+      "loss": 0.3507,
+      "step": 1295
+    },
+    {
+      "epoch": 1.2963646025444655,
+      "grad_norm": 0.1952594518661499,
+      "learning_rate": 0.00011370320855614973,
+      "loss": 0.3446,
+      "step": 1296
+    },
+    {
+      "epoch": 1.2973648838735894,
+      "grad_norm": 0.20317956805229187,
+      "learning_rate": 0.00011363636363636365,
+      "loss": 0.3773,
+      "step": 1297
+    },
+    {
+      "epoch": 1.2983651652027133,
+      "grad_norm": 0.19343924522399902,
+      "learning_rate": 0.00011356951871657755,
+      "loss": 0.3581,
+      "step": 1298
+    },
+    {
+      "epoch": 1.2993654465318372,
+      "grad_norm": 0.19085943698883057,
+      "learning_rate": 0.00011350267379679144,
+      "loss": 0.3422,
+      "step": 1299
+    },
+    {
+      "epoch": 1.3003657278609608,
+      "grad_norm": 0.19981223344802856,
+      "learning_rate": 0.00011343582887700535,
+      "loss": 0.3576,
+      "step": 1300
+    },
+    {
+      "epoch": 1.3013660091900847,
+      "grad_norm": 0.18372730910778046,
+      "learning_rate": 0.00011336898395721927,
+      "loss": 0.353,
+      "step": 1301
+    },
+    {
+      "epoch": 1.3023662905192086,
+      "grad_norm": 0.18826976418495178,
+      "learning_rate": 0.00011330213903743316,
+      "loss": 0.337,
+      "step": 1302
+    },
+    {
+      "epoch": 1.3033665718483323,
+      "grad_norm": 0.19734327495098114,
+      "learning_rate": 0.00011323529411764706,
+      "loss": 0.3362,
+      "step": 1303
+    },
+    {
+      "epoch": 1.3043668531774562,
+      "grad_norm": 0.19954827427864075,
+      "learning_rate": 0.00011316844919786095,
+      "loss": 0.3478,
+      "step": 1304
+    },
+    {
+      "epoch": 1.30536713450658,
+      "grad_norm": 0.2039194405078888,
+      "learning_rate": 0.00011310160427807488,
+      "loss": 0.3665,
+      "step": 1305
+    },
+    {
+      "epoch": 1.306367415835704,
+      "grad_norm": 0.22923427820205688,
+      "learning_rate": 0.00011303475935828878,
+      "loss": 0.3693,
+      "step": 1306
+    },
+    {
+      "epoch": 1.3073676971648276,
+      "grad_norm": 0.19368557631969452,
+      "learning_rate": 0.00011296791443850268,
+      "loss": 0.356,
+      "step": 1307
+    },
+    {
+      "epoch": 1.3083679784939515,
+      "grad_norm": 0.20185476541519165,
+      "learning_rate": 0.00011290106951871657,
+      "loss": 0.3448,
+      "step": 1308
+    },
+    {
+      "epoch": 1.3093682598230751,
+      "grad_norm": 0.18888550996780396,
+      "learning_rate": 0.00011283422459893049,
+      "loss": 0.3223,
+      "step": 1309
+    },
+    {
+      "epoch": 1.310368541152199,
+      "grad_norm": 0.19785448908805847,
+      "learning_rate": 0.00011276737967914439,
+      "loss": 0.3622,
+      "step": 1310
+    },
+    {
+      "epoch": 1.311368822481323,
+      "grad_norm": 0.20938999950885773,
+      "learning_rate": 0.00011270053475935828,
+      "loss": 0.3465,
+      "step": 1311
+    },
+    {
+      "epoch": 1.3123691038104468,
+      "grad_norm": 0.18876370787620544,
+      "learning_rate": 0.00011263368983957219,
+      "loss": 0.3474,
+      "step": 1312
+    },
+    {
+      "epoch": 1.3133693851395705,
+      "grad_norm": 0.19659797847270966,
+      "learning_rate": 0.0001125668449197861,
+      "loss": 0.346,
+      "step": 1313
+    },
+    {
+      "epoch": 1.3143696664686944,
+      "grad_norm": 0.19584700465202332,
+      "learning_rate": 0.00011250000000000001,
+      "loss": 0.3347,
+      "step": 1314
+    },
+    {
+      "epoch": 1.315369947797818,
+      "grad_norm": 0.18816378712654114,
+      "learning_rate": 0.0001124331550802139,
+      "loss": 0.3363,
+      "step": 1315
+    },
+    {
+      "epoch": 1.316370229126942,
+      "grad_norm": 0.19282221794128418,
+      "learning_rate": 0.00011236631016042782,
+      "loss": 0.3578,
+      "step": 1316
+    },
+    {
+      "epoch": 1.3173705104560658,
+      "grad_norm": 0.1960417926311493,
+      "learning_rate": 0.00011229946524064172,
+      "loss": 0.3594,
+      "step": 1317
+    },
+    {
+      "epoch": 1.3183707917851897,
+      "grad_norm": 0.19245748221874237,
+      "learning_rate": 0.00011223262032085561,
+      "loss": 0.3507,
+      "step": 1318
+    },
+    {
+      "epoch": 1.3193710731143133,
+      "grad_norm": 0.18782652914524078,
+      "learning_rate": 0.00011216577540106952,
+      "loss": 0.3395,
+      "step": 1319
+    },
+    {
+      "epoch": 1.3203713544434372,
+      "grad_norm": 0.19783708453178406,
+      "learning_rate": 0.00011209893048128344,
+      "loss": 0.3875,
+      "step": 1320
+    },
+    {
+      "epoch": 1.3213716357725611,
+      "grad_norm": 0.18575870990753174,
+      "learning_rate": 0.00011203208556149734,
+      "loss": 0.3545,
+      "step": 1321
+    },
+    {
+      "epoch": 1.3223719171016848,
+      "grad_norm": 0.1973956972360611,
+      "learning_rate": 0.00011196524064171123,
+      "loss": 0.3332,
+      "step": 1322
+    },
+    {
+      "epoch": 1.3233721984308087,
+      "grad_norm": 0.1955372542142868,
+      "learning_rate": 0.00011189839572192514,
+      "loss": 0.3628,
+      "step": 1323
+    },
+    {
+      "epoch": 1.3243724797599326,
+      "grad_norm": 0.18606936931610107,
+      "learning_rate": 0.00011183155080213905,
+      "loss": 0.3394,
+      "step": 1324
+    },
+    {
+      "epoch": 1.3253727610890562,
+      "grad_norm": 0.20660845935344696,
+      "learning_rate": 0.00011176470588235294,
+      "loss": 0.351,
+      "step": 1325
+    },
+    {
+      "epoch": 1.32637304241818,
+      "grad_norm": 0.19645744562149048,
+      "learning_rate": 0.00011169786096256685,
+      "loss": 0.3383,
+      "step": 1326
+    },
+    {
+      "epoch": 1.327373323747304,
+      "grad_norm": 0.1907406896352768,
+      "learning_rate": 0.00011163101604278074,
+      "loss": 0.3386,
+      "step": 1327
+    },
+    {
+      "epoch": 1.3283736050764277,
+      "grad_norm": 0.1935541331768036,
+      "learning_rate": 0.00011156417112299467,
+      "loss": 0.3706,
+      "step": 1328
+    },
+    {
+      "epoch": 1.3293738864055515,
+      "grad_norm": 0.20446737110614777,
+      "learning_rate": 0.00011149732620320856,
+      "loss": 0.3695,
+      "step": 1329
+    },
+    {
+      "epoch": 1.3303741677346754,
+      "grad_norm": 0.19279111921787262,
+      "learning_rate": 0.00011143048128342247,
+      "loss": 0.3537,
+      "step": 1330
+    },
+    {
+      "epoch": 1.3313744490637993,
+      "grad_norm": 0.1932227909564972,
+      "learning_rate": 0.00011136363636363636,
+      "loss": 0.3315,
+      "step": 1331
+    },
+    {
+      "epoch": 1.332374730392923,
+      "grad_norm": 0.19266147911548615,
+      "learning_rate": 0.00011129679144385028,
+      "loss": 0.3588,
+      "step": 1332
+    },
+    {
+      "epoch": 1.3333750117220469,
+      "grad_norm": 0.19882942736148834,
+      "learning_rate": 0.00011122994652406418,
+      "loss": 0.3685,
+      "step": 1333
+    },
+    {
+      "epoch": 1.3343752930511705,
+      "grad_norm": 0.19788067042827606,
+      "learning_rate": 0.00011116310160427807,
+      "loss": 0.3612,
+      "step": 1334
+    },
+    {
+      "epoch": 1.3353755743802944,
+      "grad_norm": 0.19140426814556122,
+      "learning_rate": 0.00011109625668449199,
+      "loss": 0.3454,
+      "step": 1335
+    },
+    {
+      "epoch": 1.3363758557094183,
+      "grad_norm": 0.19551077485084534,
+      "learning_rate": 0.00011102941176470589,
+      "loss": 0.3613,
+      "step": 1336
+    },
+    {
+      "epoch": 1.3373761370385422,
+      "grad_norm": 0.1953306943178177,
+      "learning_rate": 0.0001109625668449198,
+      "loss": 0.3591,
+      "step": 1337
+    },
+    {
+      "epoch": 1.3383764183676659,
+      "grad_norm": 0.19102926552295685,
+      "learning_rate": 0.00011089572192513369,
+      "loss": 0.3638,
+      "step": 1338
+    },
+    {
+      "epoch": 1.3393766996967897,
+      "grad_norm": 0.195556178689003,
+      "learning_rate": 0.0001108288770053476,
+      "loss": 0.3648,
+      "step": 1339
+    },
+    {
+      "epoch": 1.3403769810259134,
+      "grad_norm": 0.20362123847007751,
+      "learning_rate": 0.00011076203208556151,
+      "loss": 0.3614,
+      "step": 1340
+    },
+    {
+      "epoch": 1.3413772623550373,
+      "grad_norm": 0.226017564535141,
+      "learning_rate": 0.0001106951871657754,
+      "loss": 0.3472,
+      "step": 1341
+    },
+    {
+      "epoch": 1.3423775436841612,
+      "grad_norm": 0.20594573020935059,
+      "learning_rate": 0.0001106283422459893,
+      "loss": 0.3835,
+      "step": 1342
+    },
+    {
+      "epoch": 1.343377825013285,
+      "grad_norm": 0.20177976787090302,
+      "learning_rate": 0.00011056149732620322,
+      "loss": 0.3511,
+      "step": 1343
+    },
+    {
+      "epoch": 1.3443781063424087,
+      "grad_norm": 0.1984052062034607,
+      "learning_rate": 0.00011049465240641711,
+      "loss": 0.3739,
+      "step": 1344
+    },
+    {
+      "epoch": 1.3453783876715326,
+      "grad_norm": 0.20187319815158844,
+      "learning_rate": 0.00011042780748663102,
+      "loss": 0.3711,
+      "step": 1345
+    },
+    {
+      "epoch": 1.3463786690006565,
+      "grad_norm": 0.19345791637897491,
+      "learning_rate": 0.00011036096256684492,
+      "loss": 0.3497,
+      "step": 1346
+    },
+    {
+      "epoch": 1.3473789503297802,
+      "grad_norm": 0.19715888798236847,
+      "learning_rate": 0.00011029411764705884,
+      "loss": 0.3614,
+      "step": 1347
+    },
+    {
+      "epoch": 1.348379231658904,
+      "grad_norm": 0.19497311115264893,
+      "learning_rate": 0.00011022727272727273,
+      "loss": 0.34,
+      "step": 1348
+    },
+    {
+      "epoch": 1.349379512988028,
+      "grad_norm": 0.1859533041715622,
+      "learning_rate": 0.00011016042780748664,
+      "loss": 0.3604,
+      "step": 1349
+    },
+    {
+      "epoch": 1.3503797943171518,
+      "grad_norm": 0.1890004277229309,
+      "learning_rate": 0.00011009358288770053,
+      "loss": 0.3381,
+      "step": 1350
+    },
+    {
+      "epoch": 1.3513800756462755,
+      "grad_norm": 0.20360521972179413,
+      "learning_rate": 0.00011002673796791444,
+      "loss": 0.3624,
+      "step": 1351
+    },
+    {
+      "epoch": 1.3523803569753994,
+      "grad_norm": 0.2020767629146576,
+      "learning_rate": 0.00010995989304812835,
+      "loss": 0.3409,
+      "step": 1352
+    },
+    {
+      "epoch": 1.353380638304523,
+      "grad_norm": 0.19499342143535614,
+      "learning_rate": 0.00010989304812834224,
+      "loss": 0.351,
+      "step": 1353
+    },
+    {
+      "epoch": 1.354380919633647,
+      "grad_norm": 0.19958187639713287,
+      "learning_rate": 0.00010982620320855614,
+      "loss": 0.344,
+      "step": 1354
+    },
+    {
+      "epoch": 1.3553812009627708,
+      "grad_norm": 0.21096356213092804,
+      "learning_rate": 0.00010975935828877006,
+      "loss": 0.3443,
+      "step": 1355
+    },
+    {
+      "epoch": 1.3563814822918947,
+      "grad_norm": 0.18848828971385956,
+      "learning_rate": 0.00010969251336898397,
+      "loss": 0.3437,
+      "step": 1356
+    },
+    {
+      "epoch": 1.3573817636210184,
+      "grad_norm": 0.19669117033481598,
+      "learning_rate": 0.00010962566844919786,
+      "loss": 0.3419,
+      "step": 1357
+    },
+    {
+      "epoch": 1.3583820449501423,
+      "grad_norm": 0.18661530315876007,
+      "learning_rate": 0.00010955882352941178,
+      "loss": 0.3362,
+      "step": 1358
+    },
+    {
+      "epoch": 1.359382326279266,
+      "grad_norm": 0.2140861451625824,
+      "learning_rate": 0.00010949197860962568,
+      "loss": 0.3618,
+      "step": 1359
+    },
+    {
+      "epoch": 1.3603826076083898,
+      "grad_norm": 0.20248393714427948,
+      "learning_rate": 0.00010942513368983957,
+      "loss": 0.3508,
+      "step": 1360
+    },
+    {
+      "epoch": 1.3613828889375137,
+      "grad_norm": 0.206039160490036,
+      "learning_rate": 0.00010935828877005347,
+      "loss": 0.3585,
+      "step": 1361
+    },
+    {
+      "epoch": 1.3623831702666376,
+      "grad_norm": 0.2040003091096878,
+      "learning_rate": 0.00010929144385026739,
+      "loss": 0.3438,
+      "step": 1362
+    },
+    {
+      "epoch": 1.3633834515957612,
+      "grad_norm": 0.19484709203243256,
+      "learning_rate": 0.0001092245989304813,
+      "loss": 0.3315,
+      "step": 1363
+    },
+    {
+      "epoch": 1.3643837329248851,
+      "grad_norm": 0.19427084922790527,
+      "learning_rate": 0.00010915775401069519,
+      "loss": 0.3308,
+      "step": 1364
+    },
+    {
+      "epoch": 1.365384014254009,
+      "grad_norm": 0.199051171541214,
+      "learning_rate": 0.00010909090909090909,
+      "loss": 0.3397,
+      "step": 1365
+    },
+    {
+      "epoch": 1.3663842955831327,
+      "grad_norm": 0.18874703347682953,
+      "learning_rate": 0.00010902406417112301,
+      "loss": 0.3543,
+      "step": 1366
+    },
+    {
+      "epoch": 1.3673845769122566,
+      "grad_norm": 0.1966288536787033,
+      "learning_rate": 0.0001089572192513369,
+      "loss": 0.3484,
+      "step": 1367
+    },
+    {
+      "epoch": 1.3683848582413805,
+      "grad_norm": 0.20508556067943573,
+      "learning_rate": 0.0001088903743315508,
+      "loss": 0.3725,
+      "step": 1368
+    },
+    {
+      "epoch": 1.3693851395705043,
+      "grad_norm": 0.18867287039756775,
+      "learning_rate": 0.0001088235294117647,
+      "loss": 0.3376,
+      "step": 1369
+    },
+    {
+      "epoch": 1.370385420899628,
+      "grad_norm": 0.19050075113773346,
+      "learning_rate": 0.00010875668449197863,
+      "loss": 0.3531,
+      "step": 1370
+    },
+    {
+      "epoch": 1.371385702228752,
+      "grad_norm": 0.19328930974006653,
+      "learning_rate": 0.00010868983957219252,
+      "loss": 0.3485,
+      "step": 1371
+    },
+    {
+      "epoch": 1.3723859835578756,
+      "grad_norm": 0.19735842943191528,
+      "learning_rate": 0.00010862299465240642,
+      "loss": 0.3583,
+      "step": 1372
+    },
+    {
+      "epoch": 1.3733862648869994,
+      "grad_norm": 0.19491377472877502,
+      "learning_rate": 0.00010855614973262031,
+      "loss": 0.3616,
+      "step": 1373
+    },
+    {
+      "epoch": 1.3743865462161233,
+      "grad_norm": 0.20703378319740295,
+      "learning_rate": 0.00010848930481283423,
+      "loss": 0.3683,
+      "step": 1374
+    },
+    {
+      "epoch": 1.3753868275452472,
+      "grad_norm": 0.19440965354442596,
+      "learning_rate": 0.00010842245989304814,
+      "loss": 0.3598,
+      "step": 1375
+    },
+    {
+      "epoch": 1.3763871088743709,
+      "grad_norm": 0.1931820511817932,
+      "learning_rate": 0.00010835561497326203,
+      "loss": 0.3683,
+      "step": 1376
+    },
+    {
+      "epoch": 1.3773873902034948,
+      "grad_norm": 0.1991053968667984,
+      "learning_rate": 0.00010828877005347596,
+      "loss": 0.3536,
+      "step": 1377
+    },
+    {
+      "epoch": 1.3783876715326184,
+      "grad_norm": 0.19572791457176208,
+      "learning_rate": 0.00010822192513368985,
+      "loss": 0.3613,
+      "step": 1378
+    },
+    {
+      "epoch": 1.3793879528617423,
+      "grad_norm": 0.19673775136470795,
+      "learning_rate": 0.00010815508021390375,
+      "loss": 0.357,
+      "step": 1379
+    },
+    {
+      "epoch": 1.3803882341908662,
+      "grad_norm": 0.20348121225833893,
+      "learning_rate": 0.00010808823529411764,
+      "loss": 0.3584,
+      "step": 1380
+    },
+    {
+      "epoch": 1.38138851551999,
+      "grad_norm": 0.18706169724464417,
+      "learning_rate": 0.00010802139037433156,
+      "loss": 0.3633,
+      "step": 1381
+    },
+    {
+      "epoch": 1.3823887968491138,
+      "grad_norm": 0.1849328875541687,
+      "learning_rate": 0.00010795454545454547,
+      "loss": 0.3262,
+      "step": 1382
+    },
+    {
+      "epoch": 1.3833890781782376,
+      "grad_norm": 0.1902342140674591,
+      "learning_rate": 0.00010788770053475936,
+      "loss": 0.3273,
+      "step": 1383
+    },
+    {
+      "epoch": 1.3843893595073615,
+      "grad_norm": 0.194757342338562,
+      "learning_rate": 0.00010782085561497326,
+      "loss": 0.3541,
+      "step": 1384
+    },
+    {
+      "epoch": 1.3853896408364852,
+      "grad_norm": 0.2075491100549698,
+      "learning_rate": 0.00010775401069518718,
+      "loss": 0.3717,
+      "step": 1385
+    },
+    {
+      "epoch": 1.386389922165609,
+      "grad_norm": 0.19040700793266296,
+      "learning_rate": 0.00010768716577540108,
+      "loss": 0.3397,
+      "step": 1386
+    },
+    {
+      "epoch": 1.387390203494733,
+      "grad_norm": 0.19135387241840363,
+      "learning_rate": 0.00010762032085561497,
+      "loss": 0.3366,
+      "step": 1387
+    },
+    {
+      "epoch": 1.3883904848238566,
+      "grad_norm": 0.1972009539604187,
+      "learning_rate": 0.00010755347593582888,
+      "loss": 0.3517,
+      "step": 1388
+    },
+    {
+      "epoch": 1.3893907661529805,
+      "grad_norm": 0.2090437114238739,
+      "learning_rate": 0.0001074866310160428,
+      "loss": 0.3422,
+      "step": 1389
+    },
+    {
+      "epoch": 1.3903910474821044,
+      "grad_norm": 0.19525732100009918,
+      "learning_rate": 0.00010741978609625669,
+      "loss": 0.3354,
+      "step": 1390
+    },
+    {
+      "epoch": 1.391391328811228,
+      "grad_norm": 0.1889064460992813,
+      "learning_rate": 0.00010735294117647059,
+      "loss": 0.3561,
+      "step": 1391
+    },
+    {
+      "epoch": 1.392391610140352,
+      "grad_norm": 0.1869310587644577,
+      "learning_rate": 0.00010728609625668448,
+      "loss": 0.3506,
+      "step": 1392
+    },
+    {
+      "epoch": 1.3933918914694758,
+      "grad_norm": 0.19872039556503296,
+      "learning_rate": 0.0001072192513368984,
+      "loss": 0.3511,
+      "step": 1393
+    },
+    {
+      "epoch": 1.3943921727985997,
+      "grad_norm": 0.18522930145263672,
+      "learning_rate": 0.0001071524064171123,
+      "loss": 0.3395,
+      "step": 1394
+    },
+    {
+      "epoch": 1.3953924541277234,
+      "grad_norm": 0.20646034181118011,
+      "learning_rate": 0.00010708556149732621,
+      "loss": 0.3604,
+      "step": 1395
+    },
+    {
+      "epoch": 1.3963927354568473,
+      "grad_norm": 0.1998060792684555,
+      "learning_rate": 0.0001070187165775401,
+      "loss": 0.3603,
+      "step": 1396
+    },
+    {
+      "epoch": 1.397393016785971,
+      "grad_norm": 0.1944441944360733,
+      "learning_rate": 0.00010695187165775402,
+      "loss": 0.3465,
+      "step": 1397
+    },
+    {
+      "epoch": 1.3983932981150948,
+      "grad_norm": 0.19120155274868011,
+      "learning_rate": 0.00010688502673796792,
+      "loss": 0.355,
+      "step": 1398
+    },
+    {
+      "epoch": 1.3993935794442187,
+      "grad_norm": 0.1935286819934845,
+      "learning_rate": 0.00010681818181818181,
+      "loss": 0.3521,
+      "step": 1399
+    },
+    {
+      "epoch": 1.4003938607733426,
+      "grad_norm": 0.19738948345184326,
+      "learning_rate": 0.00010675133689839573,
+      "loss": 0.3639,
+      "step": 1400
+    },
+    {
+      "epoch": 1.4013941421024663,
+      "grad_norm": 0.19463558495044708,
+      "learning_rate": 0.00010668449197860964,
+      "loss": 0.361,
+      "step": 1401
+    },
+    {
+      "epoch": 1.4023944234315902,
+      "grad_norm": 0.19433772563934326,
+      "learning_rate": 0.00010661764705882353,
+      "loss": 0.3783,
+      "step": 1402
+    },
+    {
+      "epoch": 1.4033947047607138,
+      "grad_norm": 0.20147088170051575,
+      "learning_rate": 0.00010655080213903743,
+      "loss": 0.3768,
+      "step": 1403
+    },
+    {
+      "epoch": 1.4043949860898377,
+      "grad_norm": 0.1883205771446228,
+      "learning_rate": 0.00010648395721925135,
+      "loss": 0.3349,
+      "step": 1404
+    },
+    {
+      "epoch": 1.4053952674189616,
+      "grad_norm": 0.19207331538200378,
+      "learning_rate": 0.00010641711229946525,
+      "loss": 0.3485,
+      "step": 1405
+    },
+    {
+      "epoch": 1.4063955487480855,
+      "grad_norm": 0.18761774897575378,
+      "learning_rate": 0.00010635026737967914,
+      "loss": 0.3212,
+      "step": 1406
+    },
+    {
+      "epoch": 1.4073958300772091,
+      "grad_norm": 0.19692522287368774,
+      "learning_rate": 0.00010628342245989305,
+      "loss": 0.352,
+      "step": 1407
+    },
+    {
+      "epoch": 1.408396111406333,
+      "grad_norm": 0.19802354276180267,
+      "learning_rate": 0.00010621657754010697,
+      "loss": 0.3447,
+      "step": 1408
+    },
+    {
+      "epoch": 1.409396392735457,
+      "grad_norm": 0.18945840001106262,
+      "learning_rate": 0.00010614973262032086,
+      "loss": 0.344,
+      "step": 1409
+    },
+    {
+      "epoch": 1.4103966740645806,
+      "grad_norm": 0.1912890374660492,
+      "learning_rate": 0.00010608288770053476,
+      "loss": 0.3458,
+      "step": 1410
+    },
+    {
+      "epoch": 1.4113969553937045,
+      "grad_norm": 0.2135806530714035,
+      "learning_rate": 0.00010601604278074865,
+      "loss": 0.3685,
+      "step": 1411
+    },
+    {
+      "epoch": 1.4123972367228284,
+      "grad_norm": 0.18862971663475037,
+      "learning_rate": 0.00010594919786096258,
+      "loss": 0.3341,
+      "step": 1412
+    },
+    {
+      "epoch": 1.4133975180519522,
+      "grad_norm": 0.19299282133579254,
+      "learning_rate": 0.00010588235294117647,
+      "loss": 0.3337,
+      "step": 1413
+    },
+    {
+      "epoch": 1.414397799381076,
+      "grad_norm": 0.19964806735515594,
+      "learning_rate": 0.00010581550802139038,
+      "loss": 0.3599,
+      "step": 1414
+    },
+    {
+      "epoch": 1.4153980807101998,
+      "grad_norm": 0.19632697105407715,
+      "learning_rate": 0.00010574866310160427,
+      "loss": 0.352,
+      "step": 1415
+    },
+    {
+      "epoch": 1.4163983620393235,
+      "grad_norm": 0.20215977728366852,
+      "learning_rate": 0.00010568181818181819,
+      "loss": 0.352,
+      "step": 1416
+    },
+    {
+      "epoch": 1.4173986433684473,
+      "grad_norm": 0.20333050191402435,
+      "learning_rate": 0.00010561497326203209,
+      "loss": 0.3602,
+      "step": 1417
+    },
+    {
+      "epoch": 1.4183989246975712,
+      "grad_norm": 0.1925247758626938,
+      "learning_rate": 0.00010554812834224598,
+      "loss": 0.3475,
+      "step": 1418
+    },
+    {
+      "epoch": 1.4193992060266951,
+      "grad_norm": 0.19947625696659088,
+      "learning_rate": 0.00010548128342245991,
+      "loss": 0.3648,
+      "step": 1419
+    },
+    {
+      "epoch": 1.4203994873558188,
+      "grad_norm": 0.20450519025325775,
+      "learning_rate": 0.0001054144385026738,
+      "loss": 0.3414,
+      "step": 1420
+    },
+    {
+      "epoch": 1.4213997686849427,
+      "grad_norm": 0.1868099719285965,
+      "learning_rate": 0.00010534759358288771,
+      "loss": 0.3223,
+      "step": 1421
+    },
+    {
+      "epoch": 1.4224000500140663,
+      "grad_norm": 0.1968180239200592,
+      "learning_rate": 0.0001052807486631016,
+      "loss": 0.35,
+      "step": 1422
+    },
+    {
+      "epoch": 1.4234003313431902,
+      "grad_norm": 0.1937379240989685,
+      "learning_rate": 0.00010521390374331552,
+      "loss": 0.3467,
+      "step": 1423
+    },
+    {
+      "epoch": 1.424400612672314,
+      "grad_norm": 0.1855948269367218,
+      "learning_rate": 0.00010514705882352942,
+      "loss": 0.3554,
+      "step": 1424
+    },
+    {
+      "epoch": 1.425400894001438,
+      "grad_norm": 0.20068863034248352,
+      "learning_rate": 0.00010508021390374331,
+      "loss": 0.3579,
+      "step": 1425
+    },
+    {
+      "epoch": 1.4264011753305617,
+      "grad_norm": 0.20534196496009827,
+      "learning_rate": 0.00010501336898395722,
+      "loss": 0.3724,
+      "step": 1426
+    },
+    {
+      "epoch": 1.4274014566596855,
+      "grad_norm": 0.195881187915802,
+      "learning_rate": 0.00010494652406417114,
+      "loss": 0.3455,
+      "step": 1427
+    },
+    {
+      "epoch": 1.4284017379888094,
+      "grad_norm": 0.1928311437368393,
+      "learning_rate": 0.00010487967914438504,
+      "loss": 0.3374,
+      "step": 1428
+    },
+    {
+      "epoch": 1.429402019317933,
+      "grad_norm": 0.19880704581737518,
+      "learning_rate": 0.00010481283422459893,
+      "loss": 0.3531,
+      "step": 1429
+    },
+    {
+      "epoch": 1.430402300647057,
+      "grad_norm": 0.19416001439094543,
+      "learning_rate": 0.00010474598930481284,
+      "loss": 0.343,
+      "step": 1430
+    },
+    {
+      "epoch": 1.4314025819761809,
+      "grad_norm": 0.20308318734169006,
+      "learning_rate": 0.00010467914438502675,
+      "loss": 0.3581,
+      "step": 1431
+    },
+    {
+      "epoch": 1.4324028633053048,
+      "grad_norm": 0.19536352157592773,
+      "learning_rate": 0.00010461229946524064,
+      "loss": 0.3612,
+      "step": 1432
+    },
+    {
+      "epoch": 1.4334031446344284,
+      "grad_norm": 0.19380682706832886,
+      "learning_rate": 0.00010454545454545455,
+      "loss": 0.3332,
+      "step": 1433
+    },
+    {
+      "epoch": 1.4344034259635523,
+      "grad_norm": 0.19428761303424835,
+      "learning_rate": 0.00010447860962566844,
+      "loss": 0.3658,
+      "step": 1434
+    },
+    {
+      "epoch": 1.435403707292676,
+      "grad_norm": 0.1936434507369995,
+      "learning_rate": 0.00010441176470588237,
+      "loss": 0.3531,
+      "step": 1435
+    },
+    {
+      "epoch": 1.4364039886217999,
+      "grad_norm": 0.1886702924966812,
+      "learning_rate": 0.00010434491978609626,
+      "loss": 0.3547,
+      "step": 1436
+    },
+    {
+      "epoch": 1.4374042699509237,
+      "grad_norm": 0.1851184368133545,
+      "learning_rate": 0.00010427807486631017,
+      "loss": 0.3447,
+      "step": 1437
+    },
+    {
+      "epoch": 1.4384045512800476,
+      "grad_norm": 0.19021812081336975,
+      "learning_rate": 0.00010421122994652406,
+      "loss": 0.3369,
+      "step": 1438
+    },
+    {
+      "epoch": 1.4394048326091713,
+      "grad_norm": 0.1922377347946167,
+      "learning_rate": 0.00010414438502673797,
+      "loss": 0.3552,
+      "step": 1439
+    },
+    {
+      "epoch": 1.4404051139382952,
+      "grad_norm": 0.19438506662845612,
+      "learning_rate": 0.00010407754010695188,
+      "loss": 0.3576,
+      "step": 1440
+    },
+    {
+      "epoch": 1.4414053952674188,
+      "grad_norm": 0.20057320594787598,
+      "learning_rate": 0.00010401069518716577,
+      "loss": 0.3674,
+      "step": 1441
+    },
+    {
+      "epoch": 1.4424056765965427,
+      "grad_norm": 0.20331524312496185,
+      "learning_rate": 0.00010394385026737969,
+      "loss": 0.3574,
+      "step": 1442
+    },
+    {
+      "epoch": 1.4434059579256666,
+      "grad_norm": 0.20183008909225464,
+      "learning_rate": 0.00010387700534759359,
+      "loss": 0.3509,
+      "step": 1443
+    },
+    {
+      "epoch": 1.4444062392547905,
+      "grad_norm": 0.19722557067871094,
+      "learning_rate": 0.0001038101604278075,
+      "loss": 0.3645,
+      "step": 1444
+    },
+    {
+      "epoch": 1.4454065205839142,
+      "grad_norm": 0.1835191249847412,
+      "learning_rate": 0.00010374331550802139,
+      "loss": 0.3369,
+      "step": 1445
+    },
+    {
+      "epoch": 1.446406801913038,
+      "grad_norm": 0.1879657804965973,
+      "learning_rate": 0.0001036764705882353,
+      "loss": 0.335,
+      "step": 1446
+    },
+    {
+      "epoch": 1.447407083242162,
+      "grad_norm": 0.19872035086154938,
+      "learning_rate": 0.00010360962566844921,
+      "loss": 0.351,
+      "step": 1447
+    },
+    {
+      "epoch": 1.4484073645712856,
+      "grad_norm": 0.19605733454227448,
+      "learning_rate": 0.0001035427807486631,
+      "loss": 0.3417,
+      "step": 1448
+    },
+    {
+      "epoch": 1.4494076459004095,
+      "grad_norm": 0.22014260292053223,
+      "learning_rate": 0.000103475935828877,
+      "loss": 0.365,
+      "step": 1449
+    },
+    {
+      "epoch": 1.4504079272295334,
+      "grad_norm": 0.2031819373369217,
+      "learning_rate": 0.00010340909090909092,
+      "loss": 0.3541,
+      "step": 1450
+    },
+    {
+      "epoch": 1.451408208558657,
+      "grad_norm": 0.19159074127674103,
+      "learning_rate": 0.00010334224598930481,
+      "loss": 0.3626,
+      "step": 1451
+    },
+    {
+      "epoch": 1.452408489887781,
+      "grad_norm": 0.1996089071035385,
+      "learning_rate": 0.00010327540106951872,
+      "loss": 0.3473,
+      "step": 1452
+    },
+    {
+      "epoch": 1.4534087712169048,
+      "grad_norm": 0.19533640146255493,
+      "learning_rate": 0.00010320855614973262,
+      "loss": 0.3441,
+      "step": 1453
+    },
+    {
+      "epoch": 1.4544090525460285,
+      "grad_norm": 0.19409118592739105,
+      "learning_rate": 0.00010314171122994654,
+      "loss": 0.3426,
+      "step": 1454
+    },
+    {
+      "epoch": 1.4554093338751524,
+      "grad_norm": 0.19071871042251587,
+      "learning_rate": 0.00010307486631016043,
+      "loss": 0.3333,
+      "step": 1455
+    },
+    {
+      "epoch": 1.4564096152042763,
+      "grad_norm": 0.18473486602306366,
+      "learning_rate": 0.00010300802139037434,
+      "loss": 0.3495,
+      "step": 1456
+    },
+    {
+      "epoch": 1.4574098965334001,
+      "grad_norm": 0.1941157430410385,
+      "learning_rate": 0.00010294117647058823,
+      "loss": 0.3596,
+      "step": 1457
+    },
+    {
+      "epoch": 1.4584101778625238,
+      "grad_norm": 0.1890484243631363,
+      "learning_rate": 0.00010287433155080214,
+      "loss": 0.3526,
+      "step": 1458
+    },
+    {
+      "epoch": 1.4594104591916477,
+      "grad_norm": 0.1993981897830963,
+      "learning_rate": 0.00010280748663101605,
+      "loss": 0.3348,
+      "step": 1459
+    },
+    {
+      "epoch": 1.4604107405207714,
+      "grad_norm": 0.19111202657222748,
+      "learning_rate": 0.00010274064171122994,
+      "loss": 0.3464,
+      "step": 1460
+    },
+    {
+      "epoch": 1.4614110218498952,
+      "grad_norm": 0.19217447936534882,
+      "learning_rate": 0.00010267379679144387,
+      "loss": 0.3453,
+      "step": 1461
+    },
+    {
+      "epoch": 1.4624113031790191,
+      "grad_norm": 0.18713563680648804,
+      "learning_rate": 0.00010260695187165776,
+      "loss": 0.3465,
+      "step": 1462
+    },
+    {
+      "epoch": 1.463411584508143,
+      "grad_norm": 0.19021406769752502,
+      "learning_rate": 0.00010254010695187167,
+      "loss": 0.3397,
+      "step": 1463
+    },
+    {
+      "epoch": 1.4644118658372667,
+      "grad_norm": 0.19233457744121552,
+      "learning_rate": 0.00010247326203208556,
+      "loss": 0.3457,
+      "step": 1464
+    },
+    {
+      "epoch": 1.4654121471663906,
+      "grad_norm": 0.19287459552288055,
+      "learning_rate": 0.00010240641711229947,
+      "loss": 0.3493,
+      "step": 1465
+    },
+    {
+      "epoch": 1.4664124284955142,
+      "grad_norm": 0.19790810346603394,
+      "learning_rate": 0.00010233957219251338,
+      "loss": 0.3497,
+      "step": 1466
+    },
+    {
+      "epoch": 1.4674127098246381,
+      "grad_norm": 0.20381097495555878,
+      "learning_rate": 0.00010227272727272727,
+      "loss": 0.353,
+      "step": 1467
+    },
+    {
+      "epoch": 1.468412991153762,
+      "grad_norm": 0.21507777273654938,
+      "learning_rate": 0.00010220588235294117,
+      "loss": 0.3752,
+      "step": 1468
+    },
+    {
+      "epoch": 1.469413272482886,
+      "grad_norm": 0.20002609491348267,
+      "learning_rate": 0.00010213903743315509,
+      "loss": 0.3499,
+      "step": 1469
+    },
+    {
+      "epoch": 1.4704135538120096,
+      "grad_norm": 0.19354285299777985,
+      "learning_rate": 0.000102072192513369,
+      "loss": 0.3615,
+      "step": 1470
+    },
+    {
+      "epoch": 1.4714138351411334,
+      "grad_norm": 0.20071446895599365,
+      "learning_rate": 0.00010200534759358289,
+      "loss": 0.3621,
+      "step": 1471
+    },
+    {
+      "epoch": 1.4724141164702573,
+      "grad_norm": 0.20678704977035522,
+      "learning_rate": 0.00010193850267379679,
+      "loss": 0.3496,
+      "step": 1472
+    },
+    {
+      "epoch": 1.473414397799381,
+      "grad_norm": 0.20452506840229034,
+      "learning_rate": 0.00010187165775401071,
+      "loss": 0.3609,
+      "step": 1473
+    },
+    {
+      "epoch": 1.4744146791285049,
+      "grad_norm": 0.1930905282497406,
+      "learning_rate": 0.0001018048128342246,
+      "loss": 0.3363,
+      "step": 1474
+    },
+    {
+      "epoch": 1.4754149604576288,
+      "grad_norm": 0.19852490723133087,
+      "learning_rate": 0.0001017379679144385,
+      "loss": 0.365,
+      "step": 1475
+    },
+    {
+      "epoch": 1.4764152417867527,
+      "grad_norm": 0.1999901980161667,
+      "learning_rate": 0.0001016711229946524,
+      "loss": 0.3582,
+      "step": 1476
+    },
+    {
+      "epoch": 1.4774155231158763,
+      "grad_norm": 0.20496493577957153,
+      "learning_rate": 0.00010160427807486633,
+      "loss": 0.3705,
+      "step": 1477
+    },
+    {
+      "epoch": 1.4784158044450002,
+      "grad_norm": 0.2017916440963745,
+      "learning_rate": 0.00010153743315508022,
+      "loss": 0.3613,
+      "step": 1478
+    },
+    {
+      "epoch": 1.4794160857741239,
+      "grad_norm": 0.19047458469867706,
+      "learning_rate": 0.00010147058823529412,
+      "loss": 0.3358,
+      "step": 1479
+    },
+    {
+      "epoch": 1.4804163671032478,
+      "grad_norm": 0.2044849395751953,
+      "learning_rate": 0.00010140374331550801,
+      "loss": 0.375,
+      "step": 1480
+    },
+    {
+      "epoch": 1.4814166484323716,
+      "grad_norm": 0.19748015701770782,
+      "learning_rate": 0.00010133689839572193,
+      "loss": 0.3428,
+      "step": 1481
+    },
+    {
+      "epoch": 1.4824169297614955,
+      "grad_norm": 0.1886800080537796,
+      "learning_rate": 0.00010127005347593584,
+      "loss": 0.3338,
+      "step": 1482
+    },
+    {
+      "epoch": 1.4834172110906192,
+      "grad_norm": 0.18916280567646027,
+      "learning_rate": 0.00010120320855614973,
+      "loss": 0.3264,
+      "step": 1483
+    },
+    {
+      "epoch": 1.484417492419743,
+      "grad_norm": 0.19210568070411682,
+      "learning_rate": 0.00010113636363636366,
+      "loss": 0.3563,
+      "step": 1484
+    },
+    {
+      "epoch": 1.4854177737488667,
+      "grad_norm": 0.18612031638622284,
+      "learning_rate": 0.00010106951871657755,
+      "loss": 0.3464,
+      "step": 1485
+    },
+    {
+      "epoch": 1.4864180550779906,
+      "grad_norm": 0.20126496255397797,
+      "learning_rate": 0.00010100267379679145,
+      "loss": 0.3616,
+      "step": 1486
+    },
+    {
+      "epoch": 1.4874183364071145,
+      "grad_norm": 0.20939086377620697,
+      "learning_rate": 0.00010093582887700534,
+      "loss": 0.3604,
+      "step": 1487
+    },
+    {
+      "epoch": 1.4884186177362384,
+      "grad_norm": 0.20146135985851288,
+      "learning_rate": 0.00010086898395721926,
+      "loss": 0.3564,
+      "step": 1488
+    },
+    {
+      "epoch": 1.489418899065362,
+      "grad_norm": 0.19922378659248352,
+      "learning_rate": 0.00010080213903743317,
+      "loss": 0.3527,
+      "step": 1489
+    },
+    {
+      "epoch": 1.490419180394486,
+      "grad_norm": 0.19193412363529205,
+      "learning_rate": 0.00010073529411764706,
+      "loss": 0.3467,
+      "step": 1490
+    },
+    {
+      "epoch": 1.4914194617236098,
+      "grad_norm": 0.1944057047367096,
+      "learning_rate": 0.00010066844919786096,
+      "loss": 0.3523,
+      "step": 1491
+    },
+    {
+      "epoch": 1.4924197430527335,
+      "grad_norm": 0.19009791314601898,
+      "learning_rate": 0.00010060160427807488,
+      "loss": 0.3367,
+      "step": 1492
+    },
+    {
+      "epoch": 1.4934200243818574,
+      "grad_norm": 0.1931905895471573,
+      "learning_rate": 0.00010053475935828878,
+      "loss": 0.3206,
+      "step": 1493
+    },
+    {
+      "epoch": 1.4944203057109813,
+      "grad_norm": 0.20467343926429749,
+      "learning_rate": 0.00010046791443850267,
+      "loss": 0.3606,
+      "step": 1494
+    },
+    {
+      "epoch": 1.4954205870401052,
+      "grad_norm": 0.19655448198318481,
+      "learning_rate": 0.00010040106951871658,
+      "loss": 0.3513,
+      "step": 1495
+    },
+    {
+      "epoch": 1.4964208683692288,
+      "grad_norm": 0.2011963427066803,
+      "learning_rate": 0.0001003342245989305,
+      "loss": 0.3631,
+      "step": 1496
+    },
+    {
+      "epoch": 1.4974211496983527,
+      "grad_norm": 0.20077554881572723,
+      "learning_rate": 0.00010026737967914439,
+      "loss": 0.3358,
+      "step": 1497
+    },
+    {
+      "epoch": 1.4984214310274764,
+      "grad_norm": 0.18944434821605682,
+      "learning_rate": 0.00010020053475935829,
+      "loss": 0.3316,
+      "step": 1498
+    },
+    {
+      "epoch": 1.4994217123566003,
+      "grad_norm": 0.19132313132286072,
+      "learning_rate": 0.00010013368983957218,
+      "loss": 0.3472,
+      "step": 1499
+    },
+    {
+      "epoch": 1.5004219936857242,
+      "grad_norm": 0.20129220187664032,
+      "learning_rate": 0.0001000668449197861,
+      "loss": 0.3516,
+      "step": 1500
+    },
+    {
+      "epoch": 1.501422275014848,
+      "grad_norm": 0.20574413239955902,
+      "learning_rate": 0.0001,
+      "loss": 0.3553,
+      "step": 1501
+    },
+    {
+      "epoch": 1.5024225563439717,
+      "grad_norm": 0.19799011945724487,
+      "learning_rate": 9.993315508021391e-05,
+      "loss": 0.3506,
+      "step": 1502
+    },
+    {
+      "epoch": 1.5034228376730956,
+      "grad_norm": 0.20432278513908386,
+      "learning_rate": 9.986631016042781e-05,
+      "loss": 0.3549,
+      "step": 1503
+    },
+    {
+      "epoch": 1.5044231190022193,
+      "grad_norm": 0.20404933393001556,
+      "learning_rate": 9.979946524064172e-05,
+      "loss": 0.3345,
+      "step": 1504
+    },
+    {
+      "epoch": 1.5054234003313431,
+      "grad_norm": 0.19238418340682983,
+      "learning_rate": 9.973262032085562e-05,
+      "loss": 0.3367,
+      "step": 1505
+    },
+    {
+      "epoch": 1.506423681660467,
+      "grad_norm": 0.195791557431221,
+      "learning_rate": 9.966577540106953e-05,
+      "loss": 0.3581,
+      "step": 1506
+    },
+    {
+      "epoch": 1.507423962989591,
+      "grad_norm": 0.19825923442840576,
+      "learning_rate": 9.959893048128343e-05,
+      "loss": 0.3536,
+      "step": 1507
+    },
+    {
+      "epoch": 1.5084242443187148,
+      "grad_norm": 0.20793968439102173,
+      "learning_rate": 9.953208556149734e-05,
+      "loss": 0.369,
+      "step": 1508
+    },
+    {
+      "epoch": 1.5094245256478385,
+      "grad_norm": 0.1966966837644577,
+      "learning_rate": 9.946524064171123e-05,
+      "loss": 0.3454,
+      "step": 1509
+    },
+    {
+      "epoch": 1.5104248069769621,
+      "grad_norm": 0.19071269035339355,
+      "learning_rate": 9.939839572192514e-05,
+      "loss": 0.37,
+      "step": 1510
+    },
+    {
+      "epoch": 1.511425088306086,
+      "grad_norm": 0.19878540933132172,
+      "learning_rate": 9.933155080213903e-05,
+      "loss": 0.3466,
+      "step": 1511
+    },
+    {
+      "epoch": 1.51242536963521,
+      "grad_norm": 0.18890559673309326,
+      "learning_rate": 9.926470588235295e-05,
+      "loss": 0.3199,
+      "step": 1512
+    },
+    {
+      "epoch": 1.5134256509643338,
+      "grad_norm": 0.19429443776607513,
+      "learning_rate": 9.919786096256684e-05,
+      "loss": 0.3457,
+      "step": 1513
+    },
+    {
+      "epoch": 1.5144259322934577,
+      "grad_norm": 0.1940256804227829,
+      "learning_rate": 9.913101604278076e-05,
+      "loss": 0.3377,
+      "step": 1514
+    },
+    {
+      "epoch": 1.5154262136225813,
+      "grad_norm": 0.19646622240543365,
+      "learning_rate": 9.906417112299465e-05,
+      "loss": 0.3569,
+      "step": 1515
+    },
+    {
+      "epoch": 1.516426494951705,
+      "grad_norm": 0.19872450828552246,
+      "learning_rate": 9.899732620320856e-05,
+      "loss": 0.3532,
+      "step": 1516
+    },
+    {
+      "epoch": 1.517426776280829,
+      "grad_norm": 0.19802603125572205,
+      "learning_rate": 9.893048128342246e-05,
+      "loss": 0.3453,
+      "step": 1517
+    },
+    {
+      "epoch": 1.5184270576099528,
+      "grad_norm": 0.19828654825687408,
+      "learning_rate": 9.886363636363637e-05,
+      "loss": 0.3391,
+      "step": 1518
+    },
+    {
+      "epoch": 1.5194273389390767,
+      "grad_norm": 0.21418526768684387,
+      "learning_rate": 9.879679144385027e-05,
+      "loss": 0.3615,
+      "step": 1519
+    },
+    {
+      "epoch": 1.5204276202682006,
+      "grad_norm": 0.2027418315410614,
+      "learning_rate": 9.872994652406417e-05,
+      "loss": 0.3505,
+      "step": 1520
+    },
+    {
+      "epoch": 1.5214279015973242,
+      "grad_norm": 0.19818469882011414,
+      "learning_rate": 9.866310160427808e-05,
+      "loss": 0.3539,
+      "step": 1521
+    },
+    {
+      "epoch": 1.522428182926448,
+      "grad_norm": 0.20705671608448029,
+      "learning_rate": 9.859625668449198e-05,
+      "loss": 0.3574,
+      "step": 1522
+    },
+    {
+      "epoch": 1.5234284642555718,
+      "grad_norm": 0.19522565603256226,
+      "learning_rate": 9.852941176470589e-05,
+      "loss": 0.3387,
+      "step": 1523
+    },
+    {
+      "epoch": 1.5244287455846957,
+      "grad_norm": 0.19582891464233398,
+      "learning_rate": 9.846256684491979e-05,
+      "loss": 0.3383,
+      "step": 1524
+    },
+    {
+      "epoch": 1.5254290269138195,
+      "grad_norm": 0.19448718428611755,
+      "learning_rate": 9.83957219251337e-05,
+      "loss": 0.3483,
+      "step": 1525
+    },
+    {
+      "epoch": 1.5264293082429434,
+      "grad_norm": 0.20851725339889526,
+      "learning_rate": 9.83288770053476e-05,
+      "loss": 0.3667,
+      "step": 1526
+    },
+    {
+      "epoch": 1.527429589572067,
+      "grad_norm": 0.20255693793296814,
+      "learning_rate": 9.82620320855615e-05,
+      "loss": 0.366,
+      "step": 1527
+    },
+    {
+      "epoch": 1.528429870901191,
+      "grad_norm": 0.19125273823738098,
+      "learning_rate": 9.819518716577541e-05,
+      "loss": 0.3539,
+      "step": 1528
+    },
+    {
+      "epoch": 1.5294301522303146,
+      "grad_norm": 0.1900012344121933,
+      "learning_rate": 9.812834224598931e-05,
+      "loss": 0.3338,
+      "step": 1529
+    },
+    {
+      "epoch": 1.5304304335594385,
+      "grad_norm": 0.1967354714870453,
+      "learning_rate": 9.806149732620322e-05,
+      "loss": 0.3605,
+      "step": 1530
+    },
+    {
+      "epoch": 1.5314307148885624,
+      "grad_norm": 0.187136709690094,
+      "learning_rate": 9.799465240641712e-05,
+      "loss": 0.3349,
+      "step": 1531
+    },
+    {
+      "epoch": 1.5324309962176863,
+      "grad_norm": 0.19811485707759857,
+      "learning_rate": 9.792780748663101e-05,
+      "loss": 0.3373,
+      "step": 1532
+    },
+    {
+      "epoch": 1.5334312775468102,
+      "grad_norm": 0.19568456709384918,
+      "learning_rate": 9.786096256684493e-05,
+      "loss": 0.359,
+      "step": 1533
+    },
+    {
+      "epoch": 1.5344315588759339,
+      "grad_norm": 0.19473132491111755,
+      "learning_rate": 9.779411764705882e-05,
+      "loss": 0.3301,
+      "step": 1534
+    },
+    {
+      "epoch": 1.5354318402050575,
+      "grad_norm": 0.20854467153549194,
+      "learning_rate": 9.772727272727274e-05,
+      "loss": 0.3617,
+      "step": 1535
+    },
+    {
+      "epoch": 1.5364321215341814,
+      "grad_norm": 0.20137538015842438,
+      "learning_rate": 9.766042780748663e-05,
+      "loss": 0.3466,
+      "step": 1536
+    },
+    {
+      "epoch": 1.5374324028633053,
+      "grad_norm": 0.19850364327430725,
+      "learning_rate": 9.759358288770053e-05,
+      "loss": 0.3553,
+      "step": 1537
+    },
+    {
+      "epoch": 1.5384326841924292,
+      "grad_norm": 0.19727839529514313,
+      "learning_rate": 9.752673796791444e-05,
+      "loss": 0.3573,
+      "step": 1538
+    },
+    {
+      "epoch": 1.539432965521553,
+      "grad_norm": 0.19953922927379608,
+      "learning_rate": 9.745989304812834e-05,
+      "loss": 0.3431,
+      "step": 1539
+    },
+    {
+      "epoch": 1.5404332468506767,
+      "grad_norm": 0.19373491406440735,
+      "learning_rate": 9.739304812834225e-05,
+      "loss": 0.3384,
+      "step": 1540
+    },
+    {
+      "epoch": 1.5414335281798006,
+      "grad_norm": 0.20158186554908752,
+      "learning_rate": 9.732620320855615e-05,
+      "loss": 0.3325,
+      "step": 1541
+    },
+    {
+      "epoch": 1.5424338095089243,
+      "grad_norm": 0.20183441042900085,
+      "learning_rate": 9.725935828877006e-05,
+      "loss": 0.3619,
+      "step": 1542
+    },
+    {
+      "epoch": 1.5434340908380482,
+      "grad_norm": 0.19980648159980774,
+      "learning_rate": 9.719251336898396e-05,
+      "loss": 0.3461,
+      "step": 1543
+    },
+    {
+      "epoch": 1.544434372167172,
+      "grad_norm": 0.19913893938064575,
+      "learning_rate": 9.712566844919787e-05,
+      "loss": 0.3478,
+      "step": 1544
+    },
+    {
+      "epoch": 1.545434653496296,
+      "grad_norm": 0.19230477511882782,
+      "learning_rate": 9.705882352941177e-05,
+      "loss": 0.3426,
+      "step": 1545
+    },
+    {
+      "epoch": 1.5464349348254196,
+      "grad_norm": 0.19588300585746765,
+      "learning_rate": 9.699197860962567e-05,
+      "loss": 0.3443,
+      "step": 1546
+    },
+    {
+      "epoch": 1.5474352161545435,
+      "grad_norm": 0.1868908703327179,
+      "learning_rate": 9.692513368983958e-05,
+      "loss": 0.345,
+      "step": 1547
+    },
+    {
+      "epoch": 1.5484354974836672,
+      "grad_norm": 0.19833886623382568,
+      "learning_rate": 9.685828877005348e-05,
+      "loss": 0.3511,
+      "step": 1548
+    },
+    {
+      "epoch": 1.549435778812791,
+      "grad_norm": 0.1904633641242981,
+      "learning_rate": 9.679144385026739e-05,
+      "loss": 0.3419,
+      "step": 1549
+    },
+    {
+      "epoch": 1.550436060141915,
+      "grad_norm": 0.2014801949262619,
+      "learning_rate": 9.672459893048129e-05,
+      "loss": 0.3469,
+      "step": 1550
+    },
+    {
+      "epoch": 1.5514363414710388,
+      "grad_norm": 0.18893307447433472,
+      "learning_rate": 9.66577540106952e-05,
+      "loss": 0.3338,
+      "step": 1551
+    },
+    {
+      "epoch": 1.5524366228001627,
+      "grad_norm": 0.19834621250629425,
+      "learning_rate": 9.65909090909091e-05,
+      "loss": 0.3332,
+      "step": 1552
+    },
+    {
+      "epoch": 1.5534369041292864,
+      "grad_norm": 0.2053605020046234,
+      "learning_rate": 9.652406417112299e-05,
+      "loss": 0.3611,
+      "step": 1553
+    },
+    {
+      "epoch": 1.55443718545841,
+      "grad_norm": 0.19630397856235504,
+      "learning_rate": 9.645721925133691e-05,
+      "loss": 0.3309,
+      "step": 1554
+    },
+    {
+      "epoch": 1.555437466787534,
+      "grad_norm": 0.1943737119436264,
+      "learning_rate": 9.63903743315508e-05,
+      "loss": 0.3349,
+      "step": 1555
+    },
+    {
+      "epoch": 1.5564377481166578,
+      "grad_norm": 0.2033100724220276,
+      "learning_rate": 9.632352941176472e-05,
+      "loss": 0.3552,
+      "step": 1556
+    },
+    {
+      "epoch": 1.5574380294457817,
+      "grad_norm": 0.1950339674949646,
+      "learning_rate": 9.625668449197861e-05,
+      "loss": 0.3402,
+      "step": 1557
+    },
+    {
+      "epoch": 1.5584383107749056,
+      "grad_norm": 0.2025507688522339,
+      "learning_rate": 9.618983957219251e-05,
+      "loss": 0.3575,
+      "step": 1558
+    },
+    {
+      "epoch": 1.5594385921040292,
+      "grad_norm": 0.19221533834934235,
+      "learning_rate": 9.612299465240642e-05,
+      "loss": 0.354,
+      "step": 1559
+    },
+    {
+      "epoch": 1.5604388734331531,
+      "grad_norm": 0.19286495447158813,
+      "learning_rate": 9.605614973262032e-05,
+      "loss": 0.3547,
+      "step": 1560
+    },
+    {
+      "epoch": 1.5614391547622768,
+      "grad_norm": 0.21111372113227844,
+      "learning_rate": 9.598930481283423e-05,
+      "loss": 0.3605,
+      "step": 1561
+    },
+    {
+      "epoch": 1.5624394360914007,
+      "grad_norm": 0.1878976970911026,
+      "learning_rate": 9.592245989304813e-05,
+      "loss": 0.3515,
+      "step": 1562
+    },
+    {
+      "epoch": 1.5634397174205246,
+      "grad_norm": 0.1857670396566391,
+      "learning_rate": 9.585561497326203e-05,
+      "loss": 0.3373,
+      "step": 1563
+    },
+    {
+      "epoch": 1.5644399987496485,
+      "grad_norm": 0.19104637205600739,
+      "learning_rate": 9.578877005347594e-05,
+      "loss": 0.3311,
+      "step": 1564
+    },
+    {
+      "epoch": 1.5654402800787721,
+      "grad_norm": 0.19491977989673615,
+      "learning_rate": 9.572192513368984e-05,
+      "loss": 0.3508,
+      "step": 1565
+    },
+    {
+      "epoch": 1.566440561407896,
+      "grad_norm": 0.19101645052433014,
+      "learning_rate": 9.565508021390375e-05,
+      "loss": 0.3361,
+      "step": 1566
+    },
+    {
+      "epoch": 1.5674408427370197,
+      "grad_norm": 0.19752353429794312,
+      "learning_rate": 9.558823529411765e-05,
+      "loss": 0.3413,
+      "step": 1567
+    },
+    {
+      "epoch": 1.5684411240661436,
+      "grad_norm": 0.18833643198013306,
+      "learning_rate": 9.552139037433156e-05,
+      "loss": 0.3559,
+      "step": 1568
+    },
+    {
+      "epoch": 1.5694414053952674,
+      "grad_norm": 0.19119861721992493,
+      "learning_rate": 9.545454545454546e-05,
+      "loss": 0.3488,
+      "step": 1569
+    },
+    {
+      "epoch": 1.5704416867243913,
+      "grad_norm": 0.1994052678346634,
+      "learning_rate": 9.538770053475937e-05,
+      "loss": 0.346,
+      "step": 1570
+    },
+    {
+      "epoch": 1.5714419680535152,
+      "grad_norm": 0.18726104497909546,
+      "learning_rate": 9.532085561497327e-05,
+      "loss": 0.3416,
+      "step": 1571
+    },
+    {
+      "epoch": 1.5724422493826389,
+      "grad_norm": 0.1924692988395691,
+      "learning_rate": 9.525401069518717e-05,
+      "loss": 0.3316,
+      "step": 1572
+    },
+    {
+      "epoch": 1.5734425307117625,
+      "grad_norm": 0.19445118308067322,
+      "learning_rate": 9.518716577540108e-05,
+      "loss": 0.3433,
+      "step": 1573
+    },
+    {
+      "epoch": 1.5744428120408864,
+      "grad_norm": 0.1873147189617157,
+      "learning_rate": 9.512032085561497e-05,
+      "loss": 0.3285,
+      "step": 1574
+    },
+    {
+      "epoch": 1.5754430933700103,
+      "grad_norm": 0.1980922967195511,
+      "learning_rate": 9.505347593582889e-05,
+      "loss": 0.3683,
+      "step": 1575
+    },
+    {
+      "epoch": 1.5764433746991342,
+      "grad_norm": 0.19233527779579163,
+      "learning_rate": 9.498663101604278e-05,
+      "loss": 0.3724,
+      "step": 1576
+    },
+    {
+      "epoch": 1.577443656028258,
+      "grad_norm": 0.19352510571479797,
+      "learning_rate": 9.49197860962567e-05,
+      "loss": 0.344,
+      "step": 1577
+    },
+    {
+      "epoch": 1.5784439373573818,
+      "grad_norm": 0.2041119635105133,
+      "learning_rate": 9.485294117647059e-05,
+      "loss": 0.3449,
+      "step": 1578
+    },
+    {
+      "epoch": 1.5794442186865054,
+      "grad_norm": 0.2055397629737854,
+      "learning_rate": 9.47860962566845e-05,
+      "loss": 0.3493,
+      "step": 1579
+    },
+    {
+      "epoch": 1.5804445000156293,
+      "grad_norm": 0.19196373224258423,
+      "learning_rate": 9.47192513368984e-05,
+      "loss": 0.3375,
+      "step": 1580
+    },
+    {
+      "epoch": 1.5814447813447532,
+      "grad_norm": 0.19856508076190948,
+      "learning_rate": 9.46524064171123e-05,
+      "loss": 0.3409,
+      "step": 1581
+    },
+    {
+      "epoch": 1.582445062673877,
+      "grad_norm": 0.19488084316253662,
+      "learning_rate": 9.45855614973262e-05,
+      "loss": 0.3444,
+      "step": 1582
+    },
+    {
+      "epoch": 1.583445344003001,
+      "grad_norm": 0.1891886442899704,
+      "learning_rate": 9.451871657754011e-05,
+      "loss": 0.3425,
+      "step": 1583
+    },
+    {
+      "epoch": 1.5844456253321246,
+      "grad_norm": 0.2060234248638153,
+      "learning_rate": 9.445187165775401e-05,
+      "loss": 0.3491,
+      "step": 1584
+    },
+    {
+      "epoch": 1.5854459066612485,
+      "grad_norm": 0.20644551515579224,
+      "learning_rate": 9.438502673796792e-05,
+      "loss": 0.3545,
+      "step": 1585
+    },
+    {
+      "epoch": 1.5864461879903722,
+      "grad_norm": 0.20658767223358154,
+      "learning_rate": 9.431818181818182e-05,
+      "loss": 0.356,
+      "step": 1586
+    },
+    {
+      "epoch": 1.587446469319496,
+      "grad_norm": 0.19904020428657532,
+      "learning_rate": 9.425133689839573e-05,
+      "loss": 0.3496,
+      "step": 1587
+    },
+    {
+      "epoch": 1.58844675064862,
+      "grad_norm": 0.1928834766149521,
+      "learning_rate": 9.418449197860963e-05,
+      "loss": 0.3469,
+      "step": 1588
+    },
+    {
+      "epoch": 1.5894470319777438,
+      "grad_norm": 0.19535012543201447,
+      "learning_rate": 9.411764705882353e-05,
+      "loss": 0.3508,
+      "step": 1589
+    },
+    {
+      "epoch": 1.5904473133068675,
+      "grad_norm": 0.20613259077072144,
+      "learning_rate": 9.405080213903744e-05,
+      "loss": 0.3496,
+      "step": 1590
+    },
+    {
+      "epoch": 1.5914475946359914,
+      "grad_norm": 0.20351542532444,
+      "learning_rate": 9.398395721925134e-05,
+      "loss": 0.3642,
+      "step": 1591
+    },
+    {
+      "epoch": 1.592447875965115,
+      "grad_norm": 0.19753794372081757,
+      "learning_rate": 9.391711229946525e-05,
+      "loss": 0.3511,
+      "step": 1592
+    },
+    {
+      "epoch": 1.593448157294239,
+      "grad_norm": 0.20048683881759644,
+      "learning_rate": 9.385026737967915e-05,
+      "loss": 0.3295,
+      "step": 1593
+    },
+    {
+      "epoch": 1.5944484386233628,
+      "grad_norm": 0.20266291499137878,
+      "learning_rate": 9.378342245989306e-05,
+      "loss": 0.3375,
+      "step": 1594
+    },
+    {
+      "epoch": 1.5954487199524867,
+      "grad_norm": 0.19047027826309204,
+      "learning_rate": 9.371657754010695e-05,
+      "loss": 0.3365,
+      "step": 1595
+    },
+    {
+      "epoch": 1.5964490012816106,
+      "grad_norm": 0.19788208603858948,
+      "learning_rate": 9.364973262032087e-05,
+      "loss": 0.3481,
+      "step": 1596
+    },
+    {
+      "epoch": 1.5974492826107343,
+      "grad_norm": 0.20020200312137604,
+      "learning_rate": 9.358288770053476e-05,
+      "loss": 0.3403,
+      "step": 1597
+    },
+    {
+      "epoch": 1.598449563939858,
+      "grad_norm": 0.20887427031993866,
+      "learning_rate": 9.351604278074867e-05,
+      "loss": 0.3495,
+      "step": 1598
+    },
+    {
+      "epoch": 1.5994498452689818,
+      "grad_norm": 0.19780832529067993,
+      "learning_rate": 9.344919786096256e-05,
+      "loss": 0.3558,
+      "step": 1599
+    },
+    {
+      "epoch": 1.6004501265981057,
+      "grad_norm": 0.1918174922466278,
+      "learning_rate": 9.338235294117648e-05,
+      "loss": 0.3278,
+      "step": 1600
+    },
+    {
+      "epoch": 1.6014504079272296,
+      "grad_norm": 0.19131650030612946,
+      "learning_rate": 9.331550802139037e-05,
+      "loss": 0.3289,
+      "step": 1601
+    },
+    {
+      "epoch": 1.6024506892563535,
+      "grad_norm": 0.20947818458080292,
+      "learning_rate": 9.324866310160428e-05,
+      "loss": 0.3763,
+      "step": 1602
+    },
+    {
+      "epoch": 1.6034509705854771,
+      "grad_norm": 0.20184044539928436,
+      "learning_rate": 9.318181818181818e-05,
+      "loss": 0.3492,
+      "step": 1603
+    },
+    {
+      "epoch": 1.604451251914601,
+      "grad_norm": 0.19801858067512512,
+      "learning_rate": 9.311497326203209e-05,
+      "loss": 0.3518,
+      "step": 1604
+    },
+    {
+      "epoch": 1.6054515332437247,
+      "grad_norm": 0.19429895281791687,
+      "learning_rate": 9.3048128342246e-05,
+      "loss": 0.3483,
+      "step": 1605
+    },
+    {
+      "epoch": 1.6064518145728486,
+      "grad_norm": 0.20766685903072357,
+      "learning_rate": 9.29812834224599e-05,
+      "loss": 0.3693,
+      "step": 1606
+    },
+    {
+      "epoch": 1.6074520959019725,
+      "grad_norm": 0.20628705620765686,
+      "learning_rate": 9.29144385026738e-05,
+      "loss": 0.3399,
+      "step": 1607
+    },
+    {
+      "epoch": 1.6084523772310964,
+      "grad_norm": 0.19459855556488037,
+      "learning_rate": 9.28475935828877e-05,
+      "loss": 0.3585,
+      "step": 1608
+    },
+    {
+      "epoch": 1.60945265856022,
+      "grad_norm": 0.19495713710784912,
+      "learning_rate": 9.278074866310161e-05,
+      "loss": 0.3313,
+      "step": 1609
+    },
+    {
+      "epoch": 1.610452939889344,
+      "grad_norm": 0.19682952761650085,
+      "learning_rate": 9.271390374331551e-05,
+      "loss": 0.3468,
+      "step": 1610
+    },
+    {
+      "epoch": 1.6114532212184676,
+      "grad_norm": 0.19663280248641968,
+      "learning_rate": 9.264705882352942e-05,
+      "loss": 0.3342,
+      "step": 1611
+    },
+    {
+      "epoch": 1.6124535025475915,
+      "grad_norm": 0.19158923625946045,
+      "learning_rate": 9.258021390374332e-05,
+      "loss": 0.328,
+      "step": 1612
+    },
+    {
+      "epoch": 1.6134537838767153,
+      "grad_norm": 0.20232926309108734,
+      "learning_rate": 9.251336898395723e-05,
+      "loss": 0.3591,
+      "step": 1613
+    },
+    {
+      "epoch": 1.6144540652058392,
+      "grad_norm": 0.1910116970539093,
+      "learning_rate": 9.244652406417113e-05,
+      "loss": 0.3381,
+      "step": 1614
+    },
+    {
+      "epoch": 1.6154543465349631,
+      "grad_norm": 0.20306135714054108,
+      "learning_rate": 9.237967914438503e-05,
+      "loss": 0.3573,
+      "step": 1615
+    },
+    {
+      "epoch": 1.6164546278640868,
+      "grad_norm": 0.20349003374576569,
+      "learning_rate": 9.231283422459893e-05,
+      "loss": 0.3617,
+      "step": 1616
+    },
+    {
+      "epoch": 1.6174549091932104,
+      "grad_norm": 0.1951102316379547,
+      "learning_rate": 9.224598930481284e-05,
+      "loss": 0.3468,
+      "step": 1617
+    },
+    {
+      "epoch": 1.6184551905223343,
+      "grad_norm": 0.19183678925037384,
+      "learning_rate": 9.217914438502673e-05,
+      "loss": 0.3335,
+      "step": 1618
+    },
+    {
+      "epoch": 1.6194554718514582,
+      "grad_norm": 0.21201933920383453,
+      "learning_rate": 9.211229946524065e-05,
+      "loss": 0.3663,
+      "step": 1619
+    },
+    {
+      "epoch": 1.620455753180582,
+      "grad_norm": 0.1917801946401596,
+      "learning_rate": 9.204545454545454e-05,
+      "loss": 0.3407,
+      "step": 1620
+    },
+    {
+      "epoch": 1.621456034509706,
+      "grad_norm": 0.19059982895851135,
+      "learning_rate": 9.197860962566846e-05,
+      "loss": 0.3374,
+      "step": 1621
+    },
+    {
+      "epoch": 1.6224563158388297,
+      "grad_norm": 0.19140656292438507,
+      "learning_rate": 9.191176470588235e-05,
+      "loss": 0.3439,
+      "step": 1622
+    },
+    {
+      "epoch": 1.6234565971679535,
+      "grad_norm": 0.20249876379966736,
+      "learning_rate": 9.184491978609626e-05,
+      "loss": 0.3711,
+      "step": 1623
+    },
+    {
+      "epoch": 1.6244568784970772,
+      "grad_norm": 0.18776743113994598,
+      "learning_rate": 9.177807486631016e-05,
+      "loss": 0.3284,
+      "step": 1624
+    },
+    {
+      "epoch": 1.625457159826201,
+      "grad_norm": 0.19860869646072388,
+      "learning_rate": 9.171122994652406e-05,
+      "loss": 0.356,
+      "step": 1625
+    },
+    {
+      "epoch": 1.626457441155325,
+      "grad_norm": 0.2006106972694397,
+      "learning_rate": 9.164438502673798e-05,
+      "loss": 0.3684,
+      "step": 1626
+    },
+    {
+      "epoch": 1.6274577224844489,
+      "grad_norm": 0.19273580610752106,
+      "learning_rate": 9.157754010695187e-05,
+      "loss": 0.3467,
+      "step": 1627
+    },
+    {
+      "epoch": 1.6284580038135725,
+      "grad_norm": 0.1979825496673584,
+      "learning_rate": 9.151069518716579e-05,
+      "loss": 0.3468,
+      "step": 1628
+    },
+    {
+      "epoch": 1.6294582851426964,
+      "grad_norm": 0.19896800816059113,
+      "learning_rate": 9.144385026737968e-05,
+      "loss": 0.3594,
+      "step": 1629
+    },
+    {
+      "epoch": 1.63045856647182,
+      "grad_norm": 0.19518133997917175,
+      "learning_rate": 9.137700534759359e-05,
+      "loss": 0.3588,
+      "step": 1630
+    },
+    {
+      "epoch": 1.631458847800944,
+      "grad_norm": 0.19391058385372162,
+      "learning_rate": 9.131016042780749e-05,
+      "loss": 0.3305,
+      "step": 1631
+    },
+    {
+      "epoch": 1.6324591291300679,
+      "grad_norm": 0.2022496610879898,
+      "learning_rate": 9.12433155080214e-05,
+      "loss": 0.3332,
+      "step": 1632
+    },
+    {
+      "epoch": 1.6334594104591917,
+      "grad_norm": 0.2037343978881836,
+      "learning_rate": 9.11764705882353e-05,
+      "loss": 0.3512,
+      "step": 1633
+    },
+    {
+      "epoch": 1.6344596917883156,
+      "grad_norm": 0.2004886269569397,
+      "learning_rate": 9.11096256684492e-05,
+      "loss": 0.348,
+      "step": 1634
+    },
+    {
+      "epoch": 1.6354599731174393,
+      "grad_norm": 0.19458657503128052,
+      "learning_rate": 9.104278074866311e-05,
+      "loss": 0.3477,
+      "step": 1635
+    },
+    {
+      "epoch": 1.636460254446563,
+      "grad_norm": 0.20311668515205383,
+      "learning_rate": 9.097593582887701e-05,
+      "loss": 0.3501,
+      "step": 1636
+    },
+    {
+      "epoch": 1.6374605357756868,
+      "grad_norm": 0.19749753177165985,
+      "learning_rate": 9.090909090909092e-05,
+      "loss": 0.348,
+      "step": 1637
+    },
+    {
+      "epoch": 1.6384608171048107,
+      "grad_norm": 0.1951054185628891,
+      "learning_rate": 9.084224598930482e-05,
+      "loss": 0.3401,
+      "step": 1638
+    },
+    {
+      "epoch": 1.6394610984339346,
+      "grad_norm": 0.1985030621290207,
+      "learning_rate": 9.077540106951871e-05,
+      "loss": 0.3476,
+      "step": 1639
+    },
+    {
+      "epoch": 1.6404613797630585,
+      "grad_norm": 0.19220323860645294,
+      "learning_rate": 9.070855614973263e-05,
+      "loss": 0.3398,
+      "step": 1640
+    },
+    {
+      "epoch": 1.6414616610921822,
+      "grad_norm": 0.20534905791282654,
+      "learning_rate": 9.064171122994652e-05,
+      "loss": 0.3791,
+      "step": 1641
+    },
+    {
+      "epoch": 1.6424619424213058,
+      "grad_norm": 0.20361755788326263,
+      "learning_rate": 9.057486631016044e-05,
+      "loss": 0.3668,
+      "step": 1642
+    },
+    {
+      "epoch": 1.6434622237504297,
+      "grad_norm": 0.19728124141693115,
+      "learning_rate": 9.050802139037433e-05,
+      "loss": 0.3559,
+      "step": 1643
+    },
+    {
+      "epoch": 1.6444625050795536,
+      "grad_norm": 0.197921484708786,
+      "learning_rate": 9.044117647058823e-05,
+      "loss": 0.35,
+      "step": 1644
+    },
+    {
+      "epoch": 1.6454627864086775,
+      "grad_norm": 0.19405105710029602,
+      "learning_rate": 9.037433155080214e-05,
+      "loss": 0.3444,
+      "step": 1645
+    },
+    {
+      "epoch": 1.6464630677378014,
+      "grad_norm": 0.201245978474617,
+      "learning_rate": 9.030748663101604e-05,
+      "loss": 0.3524,
+      "step": 1646
+    },
+    {
+      "epoch": 1.647463349066925,
+      "grad_norm": 0.18698742985725403,
+      "learning_rate": 9.024064171122996e-05,
+      "loss": 0.3306,
+      "step": 1647
+    },
+    {
+      "epoch": 1.648463630396049,
+      "grad_norm": 0.20364025235176086,
+      "learning_rate": 9.017379679144385e-05,
+      "loss": 0.3331,
+      "step": 1648
+    },
+    {
+      "epoch": 1.6494639117251726,
+      "grad_norm": 0.19230453670024872,
+      "learning_rate": 9.010695187165777e-05,
+      "loss": 0.3575,
+      "step": 1649
+    },
+    {
+      "epoch": 1.6504641930542965,
+      "grad_norm": 0.1860368847846985,
+      "learning_rate": 9.004010695187166e-05,
+      "loss": 0.3376,
+      "step": 1650
+    },
+    {
+      "epoch": 1.6514644743834204,
+      "grad_norm": 0.1872228980064392,
+      "learning_rate": 8.997326203208556e-05,
+      "loss": 0.3229,
+      "step": 1651
+    },
+    {
+      "epoch": 1.6524647557125443,
+      "grad_norm": 0.21056225895881653,
+      "learning_rate": 8.990641711229947e-05,
+      "loss": 0.3538,
+      "step": 1652
+    },
+    {
+      "epoch": 1.653465037041668,
+      "grad_norm": 0.21094004809856415,
+      "learning_rate": 8.983957219251337e-05,
+      "loss": 0.3333,
+      "step": 1653
+    },
+    {
+      "epoch": 1.6544653183707918,
+      "grad_norm": 0.1969030201435089,
+      "learning_rate": 8.977272727272728e-05,
+      "loss": 0.3302,
+      "step": 1654
+    },
+    {
+      "epoch": 1.6554655996999155,
+      "grad_norm": 0.19789241254329681,
+      "learning_rate": 8.970588235294118e-05,
+      "loss": 0.3405,
+      "step": 1655
+    },
+    {
+      "epoch": 1.6564658810290394,
+      "grad_norm": 0.1954168975353241,
+      "learning_rate": 8.963903743315509e-05,
+      "loss": 0.3374,
+      "step": 1656
+    },
+    {
+      "epoch": 1.6574661623581632,
+      "grad_norm": 0.20916850864887238,
+      "learning_rate": 8.957219251336899e-05,
+      "loss": 0.37,
+      "step": 1657
+    },
+    {
+      "epoch": 1.6584664436872871,
+      "grad_norm": 0.21837671101093292,
+      "learning_rate": 8.95053475935829e-05,
+      "loss": 0.3278,
+      "step": 1658
+    },
+    {
+      "epoch": 1.659466725016411,
+      "grad_norm": 0.21735529601573944,
+      "learning_rate": 8.94385026737968e-05,
+      "loss": 0.3691,
+      "step": 1659
+    },
+    {
+      "epoch": 1.6604670063455347,
+      "grad_norm": 0.21532346308231354,
+      "learning_rate": 8.937165775401069e-05,
+      "loss": 0.3505,
+      "step": 1660
+    },
+    {
+      "epoch": 1.6614672876746583,
+      "grad_norm": 0.20343716442584991,
+      "learning_rate": 8.930481283422461e-05,
+      "loss": 0.3502,
+      "step": 1661
+    },
+    {
+      "epoch": 1.6624675690037822,
+      "grad_norm": 0.1884985864162445,
+      "learning_rate": 8.92379679144385e-05,
+      "loss": 0.3365,
+      "step": 1662
+    },
+    {
+      "epoch": 1.6634678503329061,
+      "grad_norm": 0.20315489172935486,
+      "learning_rate": 8.917112299465242e-05,
+      "loss": 0.3457,
+      "step": 1663
+    },
+    {
+      "epoch": 1.66446813166203,
+      "grad_norm": 0.18640293180942535,
+      "learning_rate": 8.910427807486631e-05,
+      "loss": 0.3329,
+      "step": 1664
+    },
+    {
+      "epoch": 1.665468412991154,
+      "grad_norm": 0.18697793781757355,
+      "learning_rate": 8.903743315508021e-05,
+      "loss": 0.3141,
+      "step": 1665
+    },
+    {
+      "epoch": 1.6664686943202776,
+      "grad_norm": 0.19344624876976013,
+      "learning_rate": 8.897058823529412e-05,
+      "loss": 0.3467,
+      "step": 1666
+    },
+    {
+      "epoch": 1.6674689756494014,
+      "grad_norm": 0.19100771844387054,
+      "learning_rate": 8.890374331550802e-05,
+      "loss": 0.3406,
+      "step": 1667
+    },
+    {
+      "epoch": 1.668469256978525,
+      "grad_norm": 0.19917811453342438,
+      "learning_rate": 8.883689839572194e-05,
+      "loss": 0.3643,
+      "step": 1668
+    },
+    {
+      "epoch": 1.669469538307649,
+      "grad_norm": 0.20756728947162628,
+      "learning_rate": 8.877005347593583e-05,
+      "loss": 0.3409,
+      "step": 1669
+    },
+    {
+      "epoch": 1.6704698196367729,
+      "grad_norm": 0.1978703886270523,
+      "learning_rate": 8.870320855614975e-05,
+      "loss": 0.3425,
+      "step": 1670
+    },
+    {
+      "epoch": 1.6714701009658968,
+      "grad_norm": 0.1950322687625885,
+      "learning_rate": 8.863636363636364e-05,
+      "loss": 0.3447,
+      "step": 1671
+    },
+    {
+      "epoch": 1.6724703822950204,
+      "grad_norm": 0.1997329741716385,
+      "learning_rate": 8.856951871657754e-05,
+      "loss": 0.3511,
+      "step": 1672
+    },
+    {
+      "epoch": 1.6734706636241443,
+      "grad_norm": 0.20322859287261963,
+      "learning_rate": 8.850267379679145e-05,
+      "loss": 0.3686,
+      "step": 1673
+    },
+    {
+      "epoch": 1.674470944953268,
+      "grad_norm": 0.19842784106731415,
+      "learning_rate": 8.843582887700535e-05,
+      "loss": 0.36,
+      "step": 1674
+    },
+    {
+      "epoch": 1.6754712262823919,
+      "grad_norm": 0.1973329335451126,
+      "learning_rate": 8.836898395721926e-05,
+      "loss": 0.3745,
+      "step": 1675
+    },
+    {
+      "epoch": 1.6764715076115158,
+      "grad_norm": 0.19963189959526062,
+      "learning_rate": 8.830213903743316e-05,
+      "loss": 0.3439,
+      "step": 1676
+    },
+    {
+      "epoch": 1.6774717889406396,
+      "grad_norm": 0.19335278868675232,
+      "learning_rate": 8.823529411764706e-05,
+      "loss": 0.3127,
+      "step": 1677
+    },
+    {
+      "epoch": 1.6784720702697635,
+      "grad_norm": 0.19424577057361603,
+      "learning_rate": 8.816844919786097e-05,
+      "loss": 0.354,
+      "step": 1678
+    },
+    {
+      "epoch": 1.6794723515988872,
+      "grad_norm": 0.20377704501152039,
+      "learning_rate": 8.810160427807487e-05,
+      "loss": 0.3409,
+      "step": 1679
+    },
+    {
+      "epoch": 1.6804726329280109,
+      "grad_norm": 0.1898677796125412,
+      "learning_rate": 8.803475935828878e-05,
+      "loss": 0.3304,
+      "step": 1680
+    },
+    {
+      "epoch": 1.6814729142571347,
+      "grad_norm": 0.19297891855239868,
+      "learning_rate": 8.796791443850267e-05,
+      "loss": 0.3372,
+      "step": 1681
+    },
+    {
+      "epoch": 1.6824731955862586,
+      "grad_norm": 0.20704831182956696,
+      "learning_rate": 8.790106951871659e-05,
+      "loss": 0.3542,
+      "step": 1682
+    },
+    {
+      "epoch": 1.6834734769153825,
+      "grad_norm": 0.19447298347949982,
+      "learning_rate": 8.783422459893048e-05,
+      "loss": 0.3307,
+      "step": 1683
+    },
+    {
+      "epoch": 1.6844737582445064,
+      "grad_norm": 0.2159268707036972,
+      "learning_rate": 8.77673796791444e-05,
+      "loss": 0.3476,
+      "step": 1684
+    },
+    {
+      "epoch": 1.68547403957363,
+      "grad_norm": 0.20109276473522186,
+      "learning_rate": 8.770053475935829e-05,
+      "loss": 0.3463,
+      "step": 1685
+    },
+    {
+      "epoch": 1.686474320902754,
+      "grad_norm": 0.20048797130584717,
+      "learning_rate": 8.76336898395722e-05,
+      "loss": 0.3447,
+      "step": 1686
+    },
+    {
+      "epoch": 1.6874746022318776,
+      "grad_norm": 0.19835472106933594,
+      "learning_rate": 8.75668449197861e-05,
+      "loss": 0.3407,
+      "step": 1687
+    },
+    {
+      "epoch": 1.6884748835610015,
+      "grad_norm": 0.19807566702365875,
+      "learning_rate": 8.75e-05,
+      "loss": 0.3464,
+      "step": 1688
+    },
+    {
+      "epoch": 1.6894751648901254,
+      "grad_norm": 0.1917795091867447,
+      "learning_rate": 8.743315508021392e-05,
+      "loss": 0.3459,
+      "step": 1689
+    },
+    {
+      "epoch": 1.6904754462192493,
+      "grad_norm": 0.21322079002857208,
+      "learning_rate": 8.736631016042781e-05,
+      "loss": 0.3659,
+      "step": 1690
+    },
+    {
+      "epoch": 1.691475727548373,
+      "grad_norm": 0.2130478471517563,
+      "learning_rate": 8.729946524064173e-05,
+      "loss": 0.3362,
+      "step": 1691
+    },
+    {
+      "epoch": 1.6924760088774968,
+      "grad_norm": 0.2068314254283905,
+      "learning_rate": 8.723262032085562e-05,
+      "loss": 0.3389,
+      "step": 1692
+    },
+    {
+      "epoch": 1.6934762902066205,
+      "grad_norm": 0.20057711005210876,
+      "learning_rate": 8.716577540106952e-05,
+      "loss": 0.3615,
+      "step": 1693
+    },
+    {
+      "epoch": 1.6944765715357444,
+      "grad_norm": 0.19378772377967834,
+      "learning_rate": 8.709893048128343e-05,
+      "loss": 0.3585,
+      "step": 1694
+    },
+    {
+      "epoch": 1.6954768528648683,
+      "grad_norm": 0.19208310544490814,
+      "learning_rate": 8.703208556149733e-05,
+      "loss": 0.3484,
+      "step": 1695
+    },
+    {
+      "epoch": 1.6964771341939922,
+      "grad_norm": 0.20359662175178528,
+      "learning_rate": 8.696524064171123e-05,
+      "loss": 0.3554,
+      "step": 1696
+    },
+    {
+      "epoch": 1.697477415523116,
+      "grad_norm": 0.1945527344942093,
+      "learning_rate": 8.689839572192514e-05,
+      "loss": 0.3513,
+      "step": 1697
+    },
+    {
+      "epoch": 1.6984776968522397,
+      "grad_norm": 0.1934496909379959,
+      "learning_rate": 8.683155080213904e-05,
+      "loss": 0.3431,
+      "step": 1698
+    },
+    {
+      "epoch": 1.6994779781813634,
+      "grad_norm": 0.19848768413066864,
+      "learning_rate": 8.676470588235295e-05,
+      "loss": 0.3524,
+      "step": 1699
+    },
+    {
+      "epoch": 1.7004782595104873,
+      "grad_norm": 0.19682340323925018,
+      "learning_rate": 8.669786096256685e-05,
+      "loss": 0.3488,
+      "step": 1700
+    },
+    {
+      "epoch": 1.7014785408396111,
+      "grad_norm": 0.20419567823410034,
+      "learning_rate": 8.663101604278076e-05,
+      "loss": 0.3314,
+      "step": 1701
+    },
+    {
+      "epoch": 1.702478822168735,
+      "grad_norm": 0.1952189952135086,
+      "learning_rate": 8.656417112299465e-05,
+      "loss": 0.3572,
+      "step": 1702
+    },
+    {
+      "epoch": 1.703479103497859,
+      "grad_norm": 0.20256690680980682,
+      "learning_rate": 8.649732620320856e-05,
+      "loss": 0.3496,
+      "step": 1703
+    },
+    {
+      "epoch": 1.7044793848269826,
+      "grad_norm": 0.1902756243944168,
+      "learning_rate": 8.643048128342246e-05,
+      "loss": 0.3258,
+      "step": 1704
+    },
+    {
+      "epoch": 1.7054796661561062,
+      "grad_norm": 0.20101502537727356,
+      "learning_rate": 8.636363636363637e-05,
+      "loss": 0.3355,
+      "step": 1705
+    },
+    {
+      "epoch": 1.7064799474852301,
+      "grad_norm": 0.20885342359542847,
+      "learning_rate": 8.629679144385026e-05,
+      "loss": 0.3543,
+      "step": 1706
+    },
+    {
+      "epoch": 1.707480228814354,
+      "grad_norm": 0.2001396119594574,
+      "learning_rate": 8.622994652406418e-05,
+      "loss": 0.3427,
+      "step": 1707
+    },
+    {
+      "epoch": 1.708480510143478,
+      "grad_norm": 0.20777541399002075,
+      "learning_rate": 8.616310160427807e-05,
+      "loss": 0.357,
+      "step": 1708
+    },
+    {
+      "epoch": 1.7094807914726018,
+      "grad_norm": 0.20500263571739197,
+      "learning_rate": 8.609625668449198e-05,
+      "loss": 0.3481,
+      "step": 1709
+    },
+    {
+      "epoch": 1.7104810728017255,
+      "grad_norm": 0.21295663714408875,
+      "learning_rate": 8.60294117647059e-05,
+      "loss": 0.3637,
+      "step": 1710
+    },
+    {
+      "epoch": 1.7114813541308493,
+      "grad_norm": 0.20312169194221497,
+      "learning_rate": 8.596256684491979e-05,
+      "loss": 0.3625,
+      "step": 1711
+    },
+    {
+      "epoch": 1.712481635459973,
+      "grad_norm": 0.21271884441375732,
+      "learning_rate": 8.58957219251337e-05,
+      "loss": 0.3488,
+      "step": 1712
+    },
+    {
+      "epoch": 1.713481916789097,
+      "grad_norm": 0.20425869524478912,
+      "learning_rate": 8.58288770053476e-05,
+      "loss": 0.3687,
+      "step": 1713
+    },
+    {
+      "epoch": 1.7144821981182208,
+      "grad_norm": 0.19300135970115662,
+      "learning_rate": 8.57620320855615e-05,
+      "loss": 0.3465,
+      "step": 1714
+    },
+    {
+      "epoch": 1.7154824794473447,
+      "grad_norm": 0.21252281963825226,
+      "learning_rate": 8.56951871657754e-05,
+      "loss": 0.3565,
+      "step": 1715
+    },
+    {
+      "epoch": 1.7164827607764683,
+      "grad_norm": 0.19000127911567688,
+      "learning_rate": 8.562834224598931e-05,
+      "loss": 0.3462,
+      "step": 1716
+    },
+    {
+      "epoch": 1.7174830421055922,
+      "grad_norm": 0.20023488998413086,
+      "learning_rate": 8.556149732620321e-05,
+      "loss": 0.3479,
+      "step": 1717
+    },
+    {
+      "epoch": 1.7184833234347159,
+      "grad_norm": 0.2015649974346161,
+      "learning_rate": 8.549465240641712e-05,
+      "loss": 0.349,
+      "step": 1718
+    },
+    {
+      "epoch": 1.7194836047638398,
+      "grad_norm": 0.19616490602493286,
+      "learning_rate": 8.542780748663102e-05,
+      "loss": 0.3319,
+      "step": 1719
+    },
+    {
+      "epoch": 1.7204838860929637,
+      "grad_norm": 0.20526890456676483,
+      "learning_rate": 8.536096256684493e-05,
+      "loss": 0.3594,
+      "step": 1720
+    },
+    {
+      "epoch": 1.7214841674220875,
+      "grad_norm": 0.2075759917497635,
+      "learning_rate": 8.529411764705883e-05,
+      "loss": 0.3515,
+      "step": 1721
+    },
+    {
+      "epoch": 1.7224844487512114,
+      "grad_norm": 0.19730225205421448,
+      "learning_rate": 8.522727272727273e-05,
+      "loss": 0.3482,
+      "step": 1722
+    },
+    {
+      "epoch": 1.723484730080335,
+      "grad_norm": 0.20098266005516052,
+      "learning_rate": 8.516042780748662e-05,
+      "loss": 0.3371,
+      "step": 1723
+    },
+    {
+      "epoch": 1.7244850114094588,
+      "grad_norm": 0.20571312308311462,
+      "learning_rate": 8.509358288770054e-05,
+      "loss": 0.3493,
+      "step": 1724
+    },
+    {
+      "epoch": 1.7254852927385826,
+      "grad_norm": 0.1959822177886963,
+      "learning_rate": 8.502673796791443e-05,
+      "loss": 0.328,
+      "step": 1725
+    },
+    {
+      "epoch": 1.7264855740677065,
+      "grad_norm": 0.19571168720722198,
+      "learning_rate": 8.495989304812835e-05,
+      "loss": 0.3266,
+      "step": 1726
+    },
+    {
+      "epoch": 1.7274858553968304,
+      "grad_norm": 0.19124802947044373,
+      "learning_rate": 8.489304812834224e-05,
+      "loss": 0.3415,
+      "step": 1727
+    },
+    {
+      "epoch": 1.7284861367259543,
+      "grad_norm": 0.2007458508014679,
+      "learning_rate": 8.482620320855616e-05,
+      "loss": 0.3478,
+      "step": 1728
+    },
+    {
+      "epoch": 1.729486418055078,
+      "grad_norm": 0.19514642655849457,
+      "learning_rate": 8.475935828877005e-05,
+      "loss": 0.3328,
+      "step": 1729
+    },
+    {
+      "epoch": 1.7304866993842019,
+      "grad_norm": 0.20902828872203827,
+      "learning_rate": 8.469251336898396e-05,
+      "loss": 0.3444,
+      "step": 1730
+    },
+    {
+      "epoch": 1.7314869807133255,
+      "grad_norm": 0.20852535963058472,
+      "learning_rate": 8.462566844919787e-05,
+      "loss": 0.3471,
+      "step": 1731
+    },
+    {
+      "epoch": 1.7324872620424494,
+      "grad_norm": 0.20414194464683533,
+      "learning_rate": 8.455882352941176e-05,
+      "loss": 0.365,
+      "step": 1732
+    },
+    {
+      "epoch": 1.7334875433715733,
+      "grad_norm": 0.20225827395915985,
+      "learning_rate": 8.449197860962568e-05,
+      "loss": 0.353,
+      "step": 1733
+    },
+    {
+      "epoch": 1.7344878247006972,
+      "grad_norm": 0.197005033493042,
+      "learning_rate": 8.442513368983957e-05,
+      "loss": 0.354,
+      "step": 1734
+    },
+    {
+      "epoch": 1.7354881060298208,
+      "grad_norm": 0.19046489894390106,
+      "learning_rate": 8.435828877005349e-05,
+      "loss": 0.3269,
+      "step": 1735
+    },
+    {
+      "epoch": 1.7364883873589447,
+      "grad_norm": 0.2090376615524292,
+      "learning_rate": 8.429144385026738e-05,
+      "loss": 0.3626,
+      "step": 1736
+    },
+    {
+      "epoch": 1.7374886686880684,
+      "grad_norm": 0.1934148222208023,
+      "learning_rate": 8.422459893048129e-05,
+      "loss": 0.3721,
+      "step": 1737
+    },
+    {
+      "epoch": 1.7384889500171923,
+      "grad_norm": 0.20189803838729858,
+      "learning_rate": 8.415775401069519e-05,
+      "loss": 0.3567,
+      "step": 1738
+    },
+    {
+      "epoch": 1.7394892313463162,
+      "grad_norm": 0.19359588623046875,
+      "learning_rate": 8.40909090909091e-05,
+      "loss": 0.3456,
+      "step": 1739
+    },
+    {
+      "epoch": 1.74048951267544,
+      "grad_norm": 0.1905854046344757,
+      "learning_rate": 8.4024064171123e-05,
+      "loss": 0.3346,
+      "step": 1740
+    },
+    {
+      "epoch": 1.741489794004564,
+      "grad_norm": 0.18599115312099457,
+      "learning_rate": 8.39572192513369e-05,
+      "loss": 0.3505,
+      "step": 1741
+    },
+    {
+      "epoch": 1.7424900753336876,
+      "grad_norm": 0.20366983115673065,
+      "learning_rate": 8.389037433155081e-05,
+      "loss": 0.3615,
+      "step": 1742
+    },
+    {
+      "epoch": 1.7434903566628113,
+      "grad_norm": 0.1938333362340927,
+      "learning_rate": 8.382352941176471e-05,
+      "loss": 0.3454,
+      "step": 1743
+    },
+    {
+      "epoch": 1.7444906379919352,
+      "grad_norm": 0.19507014751434326,
+      "learning_rate": 8.375668449197862e-05,
+      "loss": 0.3387,
+      "step": 1744
+    },
+    {
+      "epoch": 1.745490919321059,
+      "grad_norm": 0.18158507347106934,
+      "learning_rate": 8.368983957219252e-05,
+      "loss": 0.3237,
+      "step": 1745
+    },
+    {
+      "epoch": 1.746491200650183,
+      "grad_norm": 0.2008907049894333,
+      "learning_rate": 8.362299465240641e-05,
+      "loss": 0.321,
+      "step": 1746
+    },
+    {
+      "epoch": 1.7474914819793068,
+      "grad_norm": 0.19543159008026123,
+      "learning_rate": 8.355614973262033e-05,
+      "loss": 0.3364,
+      "step": 1747
+    },
+    {
+      "epoch": 1.7484917633084305,
+      "grad_norm": 0.201371431350708,
+      "learning_rate": 8.348930481283422e-05,
+      "loss": 0.3442,
+      "step": 1748
+    },
+    {
+      "epoch": 1.7494920446375544,
+      "grad_norm": 0.20404581725597382,
+      "learning_rate": 8.342245989304814e-05,
+      "loss": 0.355,
+      "step": 1749
+    },
+    {
+      "epoch": 1.750492325966678,
+      "grad_norm": 0.20166908204555511,
+      "learning_rate": 8.335561497326203e-05,
+      "loss": 0.3582,
+      "step": 1750
+    },
+    {
+      "epoch": 1.751492607295802,
+      "grad_norm": 0.19458959996700287,
+      "learning_rate": 8.328877005347593e-05,
+      "loss": 0.3417,
+      "step": 1751
+    },
+    {
+      "epoch": 1.7524928886249258,
+      "grad_norm": 0.19952069222927094,
+      "learning_rate": 8.322192513368985e-05,
+      "loss": 0.3281,
+      "step": 1752
+    },
+    {
+      "epoch": 1.7534931699540497,
+      "grad_norm": 0.19865602254867554,
+      "learning_rate": 8.315508021390374e-05,
+      "loss": 0.355,
+      "step": 1753
+    },
+    {
+      "epoch": 1.7544934512831734,
+      "grad_norm": 0.19536755979061127,
+      "learning_rate": 8.308823529411766e-05,
+      "loss": 0.3336,
+      "step": 1754
+    },
+    {
+      "epoch": 1.7554937326122972,
+      "grad_norm": 0.20494486391544342,
+      "learning_rate": 8.302139037433155e-05,
+      "loss": 0.3467,
+      "step": 1755
+    },
+    {
+      "epoch": 1.756494013941421,
+      "grad_norm": 0.20910057425498962,
+      "learning_rate": 8.295454545454547e-05,
+      "loss": 0.3567,
+      "step": 1756
+    },
+    {
+      "epoch": 1.7574942952705448,
+      "grad_norm": 0.20393873751163483,
+      "learning_rate": 8.288770053475936e-05,
+      "loss": 0.3561,
+      "step": 1757
+    },
+    {
+      "epoch": 1.7584945765996687,
+      "grad_norm": 0.20237115025520325,
+      "learning_rate": 8.282085561497326e-05,
+      "loss": 0.3521,
+      "step": 1758
+    },
+    {
+      "epoch": 1.7594948579287926,
+      "grad_norm": 0.2113528996706009,
+      "learning_rate": 8.275401069518717e-05,
+      "loss": 0.3379,
+      "step": 1759
+    },
+    {
+      "epoch": 1.7604951392579165,
+      "grad_norm": 0.2003299593925476,
+      "learning_rate": 8.268716577540107e-05,
+      "loss": 0.3512,
+      "step": 1760
+    },
+    {
+      "epoch": 1.7614954205870401,
+      "grad_norm": 0.19545969367027283,
+      "learning_rate": 8.262032085561498e-05,
+      "loss": 0.3242,
+      "step": 1761
+    },
+    {
+      "epoch": 1.7624957019161638,
+      "grad_norm": 0.19287890195846558,
+      "learning_rate": 8.255347593582888e-05,
+      "loss": 0.3313,
+      "step": 1762
+    },
+    {
+      "epoch": 1.7634959832452877,
+      "grad_norm": 0.20761357247829437,
+      "learning_rate": 8.248663101604279e-05,
+      "loss": 0.3621,
+      "step": 1763
+    },
+    {
+      "epoch": 1.7644962645744116,
+      "grad_norm": 0.19602590799331665,
+      "learning_rate": 8.241978609625669e-05,
+      "loss": 0.3554,
+      "step": 1764
+    },
+    {
+      "epoch": 1.7654965459035354,
+      "grad_norm": 0.1940908581018448,
+      "learning_rate": 8.23529411764706e-05,
+      "loss": 0.3415,
+      "step": 1765
+    },
+    {
+      "epoch": 1.7664968272326593,
+      "grad_norm": 0.19870306551456451,
+      "learning_rate": 8.22860962566845e-05,
+      "loss": 0.3585,
+      "step": 1766
+    },
+    {
+      "epoch": 1.767497108561783,
+      "grad_norm": 0.18804779648780823,
+      "learning_rate": 8.221925133689839e-05,
+      "loss": 0.3405,
+      "step": 1767
+    },
+    {
+      "epoch": 1.7684973898909067,
+      "grad_norm": 0.19874359667301178,
+      "learning_rate": 8.215240641711231e-05,
+      "loss": 0.35,
+      "step": 1768
+    },
+    {
+      "epoch": 1.7694976712200305,
+      "grad_norm": 0.19613303244113922,
+      "learning_rate": 8.20855614973262e-05,
+      "loss": 0.3282,
+      "step": 1769
+    },
+    {
+      "epoch": 1.7704979525491544,
+      "grad_norm": 0.20094047486782074,
+      "learning_rate": 8.201871657754012e-05,
+      "loss": 0.3517,
+      "step": 1770
+    },
+    {
+      "epoch": 1.7714982338782783,
+      "grad_norm": 0.2059866040945053,
+      "learning_rate": 8.195187165775401e-05,
+      "loss": 0.3305,
+      "step": 1771
+    },
+    {
+      "epoch": 1.7724985152074022,
+      "grad_norm": 0.213811993598938,
+      "learning_rate": 8.188502673796791e-05,
+      "loss": 0.358,
+      "step": 1772
+    },
+    {
+      "epoch": 1.7734987965365259,
+      "grad_norm": 0.19075533747673035,
+      "learning_rate": 8.181818181818183e-05,
+      "loss": 0.3559,
+      "step": 1773
+    },
+    {
+      "epoch": 1.7744990778656498,
+      "grad_norm": 0.19604365527629852,
+      "learning_rate": 8.175133689839572e-05,
+      "loss": 0.3328,
+      "step": 1774
+    },
+    {
+      "epoch": 1.7754993591947734,
+      "grad_norm": 0.21061527729034424,
+      "learning_rate": 8.168449197860964e-05,
+      "loss": 0.381,
+      "step": 1775
+    },
+    {
+      "epoch": 1.7764996405238973,
+      "grad_norm": 0.19863006472587585,
+      "learning_rate": 8.161764705882353e-05,
+      "loss": 0.3277,
+      "step": 1776
+    },
+    {
+      "epoch": 1.7774999218530212,
+      "grad_norm": 0.19438400864601135,
+      "learning_rate": 8.155080213903745e-05,
+      "loss": 0.3414,
+      "step": 1777
+    },
+    {
+      "epoch": 1.778500203182145,
+      "grad_norm": 0.2047220915555954,
+      "learning_rate": 8.148395721925134e-05,
+      "loss": 0.3575,
+      "step": 1778
+    },
+    {
+      "epoch": 1.7795004845112687,
+      "grad_norm": 0.1970013976097107,
+      "learning_rate": 8.141711229946524e-05,
+      "loss": 0.3366,
+      "step": 1779
+    },
+    {
+      "epoch": 1.7805007658403926,
+      "grad_norm": 0.2031356394290924,
+      "learning_rate": 8.135026737967915e-05,
+      "loss": 0.3562,
+      "step": 1780
+    },
+    {
+      "epoch": 1.7815010471695163,
+      "grad_norm": 0.20029020309448242,
+      "learning_rate": 8.128342245989305e-05,
+      "loss": 0.3445,
+      "step": 1781
+    },
+    {
+      "epoch": 1.7825013284986402,
+      "grad_norm": 0.19670608639717102,
+      "learning_rate": 8.121657754010696e-05,
+      "loss": 0.3418,
+      "step": 1782
+    },
+    {
+      "epoch": 1.783501609827764,
+      "grad_norm": 0.18874157965183258,
+      "learning_rate": 8.114973262032086e-05,
+      "loss": 0.3181,
+      "step": 1783
+    },
+    {
+      "epoch": 1.784501891156888,
+      "grad_norm": 0.1998082548379898,
+      "learning_rate": 8.108288770053476e-05,
+      "loss": 0.3191,
+      "step": 1784
+    },
+    {
+      "epoch": 1.7855021724860118,
+      "grad_norm": 0.1978779286146164,
+      "learning_rate": 8.101604278074867e-05,
+      "loss": 0.3427,
+      "step": 1785
+    },
+    {
+      "epoch": 1.7865024538151355,
+      "grad_norm": 0.19890382885932922,
+      "learning_rate": 8.094919786096257e-05,
+      "loss": 0.3493,
+      "step": 1786
+    },
+    {
+      "epoch": 1.7875027351442592,
+      "grad_norm": 0.20784509181976318,
+      "learning_rate": 8.088235294117648e-05,
+      "loss": 0.3554,
+      "step": 1787
+    },
+    {
+      "epoch": 1.788503016473383,
+      "grad_norm": 0.20013681054115295,
+      "learning_rate": 8.081550802139037e-05,
+      "loss": 0.3472,
+      "step": 1788
+    },
+    {
+      "epoch": 1.789503297802507,
+      "grad_norm": 0.19317620992660522,
+      "learning_rate": 8.074866310160429e-05,
+      "loss": 0.3357,
+      "step": 1789
+    },
+    {
+      "epoch": 1.7905035791316308,
+      "grad_norm": 0.2011358141899109,
+      "learning_rate": 8.068181818181818e-05,
+      "loss": 0.3468,
+      "step": 1790
+    },
+    {
+      "epoch": 1.7915038604607547,
+      "grad_norm": 0.20211952924728394,
+      "learning_rate": 8.06149732620321e-05,
+      "loss": 0.3534,
+      "step": 1791
+    },
+    {
+      "epoch": 1.7925041417898784,
+      "grad_norm": 0.20216429233551025,
+      "learning_rate": 8.0548128342246e-05,
+      "loss": 0.3395,
+      "step": 1792
+    },
+    {
+      "epoch": 1.7935044231190023,
+      "grad_norm": 0.1996137797832489,
+      "learning_rate": 8.04812834224599e-05,
+      "loss": 0.3226,
+      "step": 1793
+    },
+    {
+      "epoch": 1.794504704448126,
+      "grad_norm": 0.20004905760288239,
+      "learning_rate": 8.041443850267381e-05,
+      "loss": 0.3223,
+      "step": 1794
+    },
+    {
+      "epoch": 1.7955049857772498,
+      "grad_norm": 0.19642871618270874,
+      "learning_rate": 8.03475935828877e-05,
+      "loss": 0.3425,
+      "step": 1795
+    },
+    {
+      "epoch": 1.7965052671063737,
+      "grad_norm": 0.1992112547159195,
+      "learning_rate": 8.028074866310162e-05,
+      "loss": 0.3354,
+      "step": 1796
+    },
+    {
+      "epoch": 1.7975055484354976,
+      "grad_norm": 0.19024769961833954,
+      "learning_rate": 8.021390374331551e-05,
+      "loss": 0.3308,
+      "step": 1797
+    },
+    {
+      "epoch": 1.7985058297646213,
+      "grad_norm": 0.20694081485271454,
+      "learning_rate": 8.014705882352943e-05,
+      "loss": 0.3331,
+      "step": 1798
+    },
+    {
+      "epoch": 1.7995061110937451,
+      "grad_norm": 0.20062243938446045,
+      "learning_rate": 8.008021390374332e-05,
+      "loss": 0.3277,
+      "step": 1799
+    },
+    {
+      "epoch": 1.8005063924228688,
+      "grad_norm": 0.1996837854385376,
+      "learning_rate": 8.001336898395722e-05,
+      "loss": 0.3356,
+      "step": 1800
+    },
+    {
+      "epoch": 1.8015066737519927,
+      "grad_norm": 0.202518492937088,
+      "learning_rate": 7.994652406417112e-05,
+      "loss": 0.3439,
+      "step": 1801
+    },
+    {
+      "epoch": 1.8025069550811166,
+      "grad_norm": 0.19633659720420837,
+      "learning_rate": 7.987967914438503e-05,
+      "loss": 0.334,
+      "step": 1802
+    },
+    {
+      "epoch": 1.8035072364102405,
+      "grad_norm": 0.20383204519748688,
+      "learning_rate": 7.981283422459893e-05,
+      "loss": 0.3493,
+      "step": 1803
+    },
+    {
+      "epoch": 1.8045075177393644,
+      "grad_norm": 0.20518562197685242,
+      "learning_rate": 7.974598930481284e-05,
+      "loss": 0.344,
+      "step": 1804
+    },
+    {
+      "epoch": 1.805507799068488,
+      "grad_norm": 0.20150910317897797,
+      "learning_rate": 7.967914438502674e-05,
+      "loss": 0.3317,
+      "step": 1805
+    },
+    {
+      "epoch": 1.8065080803976117,
+      "grad_norm": 0.19550201296806335,
+      "learning_rate": 7.961229946524065e-05,
+      "loss": 0.326,
+      "step": 1806
+    },
+    {
+      "epoch": 1.8075083617267356,
+      "grad_norm": 0.20320627093315125,
+      "learning_rate": 7.954545454545455e-05,
+      "loss": 0.339,
+      "step": 1807
+    },
+    {
+      "epoch": 1.8085086430558595,
+      "grad_norm": 0.2062893807888031,
+      "learning_rate": 7.947860962566846e-05,
+      "loss": 0.3226,
+      "step": 1808
+    },
+    {
+      "epoch": 1.8095089243849833,
+      "grad_norm": 0.20152324438095093,
+      "learning_rate": 7.941176470588235e-05,
+      "loss": 0.3443,
+      "step": 1809
+    },
+    {
+      "epoch": 1.8105092057141072,
+      "grad_norm": 0.19772818684577942,
+      "learning_rate": 7.934491978609626e-05,
+      "loss": 0.3595,
+      "step": 1810
+    },
+    {
+      "epoch": 1.811509487043231,
+      "grad_norm": 0.20267978310585022,
+      "learning_rate": 7.927807486631015e-05,
+      "loss": 0.35,
+      "step": 1811
+    },
+    {
+      "epoch": 1.8125097683723548,
+      "grad_norm": 0.2131519615650177,
+      "learning_rate": 7.921122994652407e-05,
+      "loss": 0.3683,
+      "step": 1812
+    },
+    {
+      "epoch": 1.8135100497014784,
+      "grad_norm": 0.1985757052898407,
+      "learning_rate": 7.914438502673798e-05,
+      "loss": 0.3457,
+      "step": 1813
+    },
+    {
+      "epoch": 1.8145103310306023,
+      "grad_norm": 0.1999262124300003,
+      "learning_rate": 7.907754010695188e-05,
+      "loss": 0.3275,
+      "step": 1814
+    },
+    {
+      "epoch": 1.8155106123597262,
+      "grad_norm": 0.20453926920890808,
+      "learning_rate": 7.901069518716579e-05,
+      "loss": 0.3666,
+      "step": 1815
+    },
+    {
+      "epoch": 1.81651089368885,
+      "grad_norm": 0.19602292776107788,
+      "learning_rate": 7.894385026737968e-05,
+      "loss": 0.348,
+      "step": 1816
+    },
+    {
+      "epoch": 1.8175111750179738,
+      "grad_norm": 0.19922220706939697,
+      "learning_rate": 7.88770053475936e-05,
+      "loss": 0.3454,
+      "step": 1817
+    },
+    {
+      "epoch": 1.8185114563470977,
+      "grad_norm": 0.2007213830947876,
+      "learning_rate": 7.881016042780749e-05,
+      "loss": 0.3479,
+      "step": 1818
+    },
+    {
+      "epoch": 1.8195117376762213,
+      "grad_norm": 0.19183677434921265,
+      "learning_rate": 7.87433155080214e-05,
+      "loss": 0.3375,
+      "step": 1819
+    },
+    {
+      "epoch": 1.8205120190053452,
+      "grad_norm": 0.19790297746658325,
+      "learning_rate": 7.86764705882353e-05,
+      "loss": 0.3488,
+      "step": 1820
+    },
+    {
+      "epoch": 1.821512300334469,
+      "grad_norm": 0.19029243290424347,
+      "learning_rate": 7.86096256684492e-05,
+      "loss": 0.3401,
+      "step": 1821
+    },
+    {
+      "epoch": 1.822512581663593,
+      "grad_norm": 0.2124333381652832,
+      "learning_rate": 7.85427807486631e-05,
+      "loss": 0.3468,
+      "step": 1822
+    },
+    {
+      "epoch": 1.8235128629927169,
+      "grad_norm": 0.20635844767093658,
+      "learning_rate": 7.847593582887701e-05,
+      "loss": 0.3523,
+      "step": 1823
+    },
+    {
+      "epoch": 1.8245131443218405,
+      "grad_norm": 0.20127947628498077,
+      "learning_rate": 7.840909090909091e-05,
+      "loss": 0.3417,
+      "step": 1824
+    },
+    {
+      "epoch": 1.8255134256509642,
+      "grad_norm": 0.20063266158103943,
+      "learning_rate": 7.834224598930482e-05,
+      "loss": 0.3518,
+      "step": 1825
+    },
+    {
+      "epoch": 1.826513706980088,
+      "grad_norm": 0.1868637204170227,
+      "learning_rate": 7.827540106951872e-05,
+      "loss": 0.3202,
+      "step": 1826
+    },
+    {
+      "epoch": 1.827513988309212,
+      "grad_norm": 0.2002502679824829,
+      "learning_rate": 7.820855614973262e-05,
+      "loss": 0.3383,
+      "step": 1827
+    },
+    {
+      "epoch": 1.8285142696383359,
+      "grad_norm": 0.19885621964931488,
+      "learning_rate": 7.814171122994653e-05,
+      "loss": 0.3466,
+      "step": 1828
+    },
+    {
+      "epoch": 1.8295145509674597,
+      "grad_norm": 0.2023029327392578,
+      "learning_rate": 7.807486631016043e-05,
+      "loss": 0.3466,
+      "step": 1829
+    },
+    {
+      "epoch": 1.8305148322965834,
+      "grad_norm": 0.2129087746143341,
+      "learning_rate": 7.800802139037432e-05,
+      "loss": 0.3577,
+      "step": 1830
+    },
+    {
+      "epoch": 1.831515113625707,
+      "grad_norm": 0.2082679718732834,
+      "learning_rate": 7.794117647058824e-05,
+      "loss": 0.3503,
+      "step": 1831
+    },
+    {
+      "epoch": 1.832515394954831,
+      "grad_norm": 0.21037320792675018,
+      "learning_rate": 7.787433155080213e-05,
+      "loss": 0.3728,
+      "step": 1832
+    },
+    {
+      "epoch": 1.8335156762839548,
+      "grad_norm": 0.20348478853702545,
+      "learning_rate": 7.780748663101605e-05,
+      "loss": 0.3312,
+      "step": 1833
+    },
+    {
+      "epoch": 1.8345159576130787,
+      "grad_norm": 0.1988702267408371,
+      "learning_rate": 7.774064171122996e-05,
+      "loss": 0.3234,
+      "step": 1834
+    },
+    {
+      "epoch": 1.8355162389422026,
+      "grad_norm": 0.21632571518421173,
+      "learning_rate": 7.767379679144386e-05,
+      "loss": 0.3618,
+      "step": 1835
+    },
+    {
+      "epoch": 1.8365165202713263,
+      "grad_norm": 0.19366499781608582,
+      "learning_rate": 7.760695187165776e-05,
+      "loss": 0.3267,
+      "step": 1836
+    },
+    {
+      "epoch": 1.8375168016004502,
+      "grad_norm": 0.1899450123310089,
+      "learning_rate": 7.754010695187165e-05,
+      "loss": 0.3173,
+      "step": 1837
+    },
+    {
+      "epoch": 1.8385170829295738,
+      "grad_norm": 0.20794934034347534,
+      "learning_rate": 7.747326203208557e-05,
+      "loss": 0.3628,
+      "step": 1838
+    },
+    {
+      "epoch": 1.8395173642586977,
+      "grad_norm": 0.19454708695411682,
+      "learning_rate": 7.740641711229946e-05,
+      "loss": 0.3186,
+      "step": 1839
+    },
+    {
+      "epoch": 1.8405176455878216,
+      "grad_norm": 0.202149897813797,
+      "learning_rate": 7.733957219251338e-05,
+      "loss": 0.3617,
+      "step": 1840
+    },
+    {
+      "epoch": 1.8415179269169455,
+      "grad_norm": 0.20746676623821259,
+      "learning_rate": 7.727272727272727e-05,
+      "loss": 0.3502,
+      "step": 1841
+    },
+    {
+      "epoch": 1.8425182082460692,
+      "grad_norm": 0.1954769343137741,
+      "learning_rate": 7.720588235294119e-05,
+      "loss": 0.3434,
+      "step": 1842
+    },
+    {
+      "epoch": 1.843518489575193,
+      "grad_norm": 0.19882524013519287,
+      "learning_rate": 7.713903743315508e-05,
+      "loss": 0.3504,
+      "step": 1843
+    },
+    {
+      "epoch": 1.8445187709043167,
+      "grad_norm": 0.19531351327896118,
+      "learning_rate": 7.707219251336899e-05,
+      "loss": 0.3509,
+      "step": 1844
+    },
+    {
+      "epoch": 1.8455190522334406,
+      "grad_norm": 0.20430710911750793,
+      "learning_rate": 7.700534759358289e-05,
+      "loss": 0.3315,
+      "step": 1845
+    },
+    {
+      "epoch": 1.8465193335625645,
+      "grad_norm": 0.20103976130485535,
+      "learning_rate": 7.69385026737968e-05,
+      "loss": 0.3358,
+      "step": 1846
+    },
+    {
+      "epoch": 1.8475196148916884,
+      "grad_norm": 0.19687388837337494,
+      "learning_rate": 7.68716577540107e-05,
+      "loss": 0.3337,
+      "step": 1847
+    },
+    {
+      "epoch": 1.8485198962208123,
+      "grad_norm": 0.19997209310531616,
+      "learning_rate": 7.68048128342246e-05,
+      "loss": 0.3519,
+      "step": 1848
+    },
+    {
+      "epoch": 1.849520177549936,
+      "grad_norm": 0.19581446051597595,
+      "learning_rate": 7.673796791443851e-05,
+      "loss": 0.3374,
+      "step": 1849
+    },
+    {
+      "epoch": 1.8505204588790596,
+      "grad_norm": 0.19563953578472137,
+      "learning_rate": 7.667112299465241e-05,
+      "loss": 0.3563,
+      "step": 1850
+    },
+    {
+      "epoch": 1.8515207402081835,
+      "grad_norm": 0.19860684871673584,
+      "learning_rate": 7.660427807486632e-05,
+      "loss": 0.3336,
+      "step": 1851
+    },
+    {
+      "epoch": 1.8525210215373074,
+      "grad_norm": 0.2035166323184967,
+      "learning_rate": 7.653743315508022e-05,
+      "loss": 0.3622,
+      "step": 1852
+    },
+    {
+      "epoch": 1.8535213028664312,
+      "grad_norm": 0.19381564855575562,
+      "learning_rate": 7.647058823529411e-05,
+      "loss": 0.3418,
+      "step": 1853
+    },
+    {
+      "epoch": 1.8545215841955551,
+      "grad_norm": 0.1977507472038269,
+      "learning_rate": 7.640374331550803e-05,
+      "loss": 0.3332,
+      "step": 1854
+    },
+    {
+      "epoch": 1.8555218655246788,
+      "grad_norm": 0.19295284152030945,
+      "learning_rate": 7.633689839572193e-05,
+      "loss": 0.3319,
+      "step": 1855
+    },
+    {
+      "epoch": 1.8565221468538027,
+      "grad_norm": 0.18499962985515594,
+      "learning_rate": 7.627005347593584e-05,
+      "loss": 0.3273,
+      "step": 1856
+    },
+    {
+      "epoch": 1.8575224281829263,
+      "grad_norm": 0.1941608190536499,
+      "learning_rate": 7.620320855614974e-05,
+      "loss": 0.3262,
+      "step": 1857
+    },
+    {
+      "epoch": 1.8585227095120502,
+      "grad_norm": 0.2033688873052597,
+      "learning_rate": 7.613636363636363e-05,
+      "loss": 0.3765,
+      "step": 1858
+    },
+    {
+      "epoch": 1.8595229908411741,
+      "grad_norm": 0.19875052571296692,
+      "learning_rate": 7.606951871657755e-05,
+      "loss": 0.3507,
+      "step": 1859
+    },
+    {
+      "epoch": 1.860523272170298,
+      "grad_norm": 0.21089790761470795,
+      "learning_rate": 7.600267379679144e-05,
+      "loss": 0.3558,
+      "step": 1860
+    },
+    {
+      "epoch": 1.8615235534994217,
+      "grad_norm": 0.19856560230255127,
+      "learning_rate": 7.593582887700536e-05,
+      "loss": 0.3257,
+      "step": 1861
+    },
+    {
+      "epoch": 1.8625238348285456,
+      "grad_norm": 0.2095073163509369,
+      "learning_rate": 7.586898395721925e-05,
+      "loss": 0.3523,
+      "step": 1862
+    },
+    {
+      "epoch": 1.8635241161576692,
+      "grad_norm": 0.2076517939567566,
+      "learning_rate": 7.580213903743317e-05,
+      "loss": 0.3466,
+      "step": 1863
+    },
+    {
+      "epoch": 1.864524397486793,
+      "grad_norm": 0.20204126834869385,
+      "learning_rate": 7.573529411764706e-05,
+      "loss": 0.3367,
+      "step": 1864
+    },
+    {
+      "epoch": 1.865524678815917,
+      "grad_norm": 0.19041043519973755,
+      "learning_rate": 7.566844919786096e-05,
+      "loss": 0.3244,
+      "step": 1865
+    },
+    {
+      "epoch": 1.8665249601450409,
+      "grad_norm": 0.196553573012352,
+      "learning_rate": 7.560160427807487e-05,
+      "loss": 0.3345,
+      "step": 1866
+    },
+    {
+      "epoch": 1.8675252414741648,
+      "grad_norm": 0.1995212733745575,
+      "learning_rate": 7.553475935828877e-05,
+      "loss": 0.3455,
+      "step": 1867
+    },
+    {
+      "epoch": 1.8685255228032884,
+      "grad_norm": 0.19919808208942413,
+      "learning_rate": 7.546791443850268e-05,
+      "loss": 0.3535,
+      "step": 1868
+    },
+    {
+      "epoch": 1.869525804132412,
+      "grad_norm": 0.20622293651103973,
+      "learning_rate": 7.540106951871658e-05,
+      "loss": 0.3466,
+      "step": 1869
+    },
+    {
+      "epoch": 1.870526085461536,
+      "grad_norm": 0.20773158967494965,
+      "learning_rate": 7.533422459893049e-05,
+      "loss": 0.3507,
+      "step": 1870
+    },
+    {
+      "epoch": 1.8715263667906599,
+      "grad_norm": 0.2090151607990265,
+      "learning_rate": 7.526737967914439e-05,
+      "loss": 0.3634,
+      "step": 1871
+    },
+    {
+      "epoch": 1.8725266481197838,
+      "grad_norm": 0.2056730091571808,
+      "learning_rate": 7.52005347593583e-05,
+      "loss": 0.347,
+      "step": 1872
+    },
+    {
+      "epoch": 1.8735269294489076,
+      "grad_norm": 0.1937808096408844,
+      "learning_rate": 7.51336898395722e-05,
+      "loss": 0.3478,
+      "step": 1873
+    },
+    {
+      "epoch": 1.8745272107780313,
+      "grad_norm": 0.20472519099712372,
+      "learning_rate": 7.506684491978609e-05,
+      "loss": 0.3804,
+      "step": 1874
+    },
+    {
+      "epoch": 1.8755274921071552,
+      "grad_norm": 0.19628198444843292,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.3401,
+      "step": 1875
+    },
+    {
+      "epoch": 1.8765277734362789,
+      "grad_norm": 0.19590778648853302,
+      "learning_rate": 7.493315508021391e-05,
+      "loss": 0.3424,
+      "step": 1876
+    },
+    {
+      "epoch": 1.8775280547654027,
+      "grad_norm": 0.21412725746631622,
+      "learning_rate": 7.486631016042782e-05,
+      "loss": 0.3881,
+      "step": 1877
+    },
+    {
+      "epoch": 1.8785283360945266,
+      "grad_norm": 0.18656885623931885,
+      "learning_rate": 7.479946524064172e-05,
+      "loss": 0.3255,
+      "step": 1878
+    },
+    {
+      "epoch": 1.8795286174236505,
+      "grad_norm": 0.1904265135526657,
+      "learning_rate": 7.473262032085561e-05,
+      "loss": 0.3196,
+      "step": 1879
+    },
+    {
+      "epoch": 1.8805288987527742,
+      "grad_norm": 0.18961571156978607,
+      "learning_rate": 7.466577540106953e-05,
+      "loss": 0.3333,
+      "step": 1880
+    },
+    {
+      "epoch": 1.881529180081898,
+      "grad_norm": 0.19472570717334747,
+      "learning_rate": 7.459893048128342e-05,
+      "loss": 0.3303,
+      "step": 1881
+    },
+    {
+      "epoch": 1.8825294614110217,
+      "grad_norm": 0.19604532420635223,
+      "learning_rate": 7.453208556149734e-05,
+      "loss": 0.3427,
+      "step": 1882
+    },
+    {
+      "epoch": 1.8835297427401456,
+      "grad_norm": 0.1914471685886383,
+      "learning_rate": 7.446524064171123e-05,
+      "loss": 0.3286,
+      "step": 1883
+    },
+    {
+      "epoch": 1.8845300240692695,
+      "grad_norm": 0.19296447932720184,
+      "learning_rate": 7.439839572192515e-05,
+      "loss": 0.3347,
+      "step": 1884
+    },
+    {
+      "epoch": 1.8855303053983934,
+      "grad_norm": 0.19976140558719635,
+      "learning_rate": 7.433155080213904e-05,
+      "loss": 0.3674,
+      "step": 1885
+    },
+    {
+      "epoch": 1.8865305867275173,
+      "grad_norm": 0.1885506510734558,
+      "learning_rate": 7.426470588235294e-05,
+      "loss": 0.3242,
+      "step": 1886
+    },
+    {
+      "epoch": 1.887530868056641,
+      "grad_norm": 0.21544072031974792,
+      "learning_rate": 7.419786096256685e-05,
+      "loss": 0.3673,
+      "step": 1887
+    },
+    {
+      "epoch": 1.8885311493857646,
+      "grad_norm": 0.20368769764900208,
+      "learning_rate": 7.413101604278075e-05,
+      "loss": 0.3461,
+      "step": 1888
+    },
+    {
+      "epoch": 1.8895314307148885,
+      "grad_norm": 0.20054277777671814,
+      "learning_rate": 7.406417112299465e-05,
+      "loss": 0.3397,
+      "step": 1889
+    },
+    {
+      "epoch": 1.8905317120440124,
+      "grad_norm": 0.207710862159729,
+      "learning_rate": 7.399732620320856e-05,
+      "loss": 0.3616,
+      "step": 1890
+    },
+    {
+      "epoch": 1.8915319933731363,
+      "grad_norm": 0.19878432154655457,
+      "learning_rate": 7.393048128342246e-05,
+      "loss": 0.3287,
+      "step": 1891
+    },
+    {
+      "epoch": 1.8925322747022602,
+      "grad_norm": 0.20300355553627014,
+      "learning_rate": 7.386363636363637e-05,
+      "loss": 0.3546,
+      "step": 1892
+    },
+    {
+      "epoch": 1.8935325560313838,
+      "grad_norm": 0.20177844166755676,
+      "learning_rate": 7.379679144385027e-05,
+      "loss": 0.3288,
+      "step": 1893
+    },
+    {
+      "epoch": 1.8945328373605075,
+      "grad_norm": 0.19801439344882965,
+      "learning_rate": 7.372994652406418e-05,
+      "loss": 0.3497,
+      "step": 1894
+    },
+    {
+      "epoch": 1.8955331186896314,
+      "grad_norm": 0.20265112817287445,
+      "learning_rate": 7.366310160427807e-05,
+      "loss": 0.3697,
+      "step": 1895
+    },
+    {
+      "epoch": 1.8965334000187553,
+      "grad_norm": 0.20634900033473969,
+      "learning_rate": 7.359625668449199e-05,
+      "loss": 0.3531,
+      "step": 1896
+    },
+    {
+      "epoch": 1.8975336813478791,
+      "grad_norm": 0.19173398613929749,
+      "learning_rate": 7.352941176470589e-05,
+      "loss": 0.3245,
+      "step": 1897
+    },
+    {
+      "epoch": 1.898533962677003,
+      "grad_norm": 0.20165519416332245,
+      "learning_rate": 7.34625668449198e-05,
+      "loss": 0.3575,
+      "step": 1898
+    },
+    {
+      "epoch": 1.8995342440061267,
+      "grad_norm": 0.19723007082939148,
+      "learning_rate": 7.33957219251337e-05,
+      "loss": 0.3513,
+      "step": 1899
+    },
+    {
+      "epoch": 1.9005345253352506,
+      "grad_norm": 0.19607777893543243,
+      "learning_rate": 7.33288770053476e-05,
+      "loss": 0.3515,
+      "step": 1900
+    },
+    {
+      "epoch": 1.9015348066643742,
+      "grad_norm": 0.19856400787830353,
+      "learning_rate": 7.326203208556151e-05,
+      "loss": 0.3361,
+      "step": 1901
+    },
+    {
+      "epoch": 1.9025350879934981,
+      "grad_norm": 0.19188867509365082,
+      "learning_rate": 7.31951871657754e-05,
+      "loss": 0.3113,
+      "step": 1902
+    },
+    {
+      "epoch": 1.903535369322622,
+      "grad_norm": 0.1967494636774063,
+      "learning_rate": 7.312834224598932e-05,
+      "loss": 0.3424,
+      "step": 1903
+    },
+    {
+      "epoch": 1.904535650651746,
+      "grad_norm": 0.2088395059108734,
+      "learning_rate": 7.30614973262032e-05,
+      "loss": 0.3685,
+      "step": 1904
+    },
+    {
+      "epoch": 1.9055359319808696,
+      "grad_norm": 0.19484959542751312,
+      "learning_rate": 7.299465240641712e-05,
+      "loss": 0.3361,
+      "step": 1905
+    },
+    {
+      "epoch": 1.9065362133099935,
+      "grad_norm": 0.19527581334114075,
+      "learning_rate": 7.292780748663102e-05,
+      "loss": 0.3359,
+      "step": 1906
+    },
+    {
+      "epoch": 1.9075364946391171,
+      "grad_norm": 0.19926346838474274,
+      "learning_rate": 7.286096256684492e-05,
+      "loss": 0.3463,
+      "step": 1907
+    },
+    {
+      "epoch": 1.908536775968241,
+      "grad_norm": 0.1828133910894394,
+      "learning_rate": 7.279411764705882e-05,
+      "loss": 0.3385,
+      "step": 1908
+    },
+    {
+      "epoch": 1.909537057297365,
+      "grad_norm": 0.19586998224258423,
+      "learning_rate": 7.272727272727273e-05,
+      "loss": 0.3501,
+      "step": 1909
+    },
+    {
+      "epoch": 1.9105373386264888,
+      "grad_norm": 0.20466144382953644,
+      "learning_rate": 7.266042780748663e-05,
+      "loss": 0.3486,
+      "step": 1910
+    },
+    {
+      "epoch": 1.9115376199556127,
+      "grad_norm": 0.1957923173904419,
+      "learning_rate": 7.259358288770054e-05,
+      "loss": 0.3221,
+      "step": 1911
+    },
+    {
+      "epoch": 1.9125379012847363,
+      "grad_norm": 0.2018577754497528,
+      "learning_rate": 7.252673796791444e-05,
+      "loss": 0.3488,
+      "step": 1912
+    },
+    {
+      "epoch": 1.91353818261386,
+      "grad_norm": 0.1903039664030075,
+      "learning_rate": 7.245989304812835e-05,
+      "loss": 0.3174,
+      "step": 1913
+    },
+    {
+      "epoch": 1.9145384639429839,
+      "grad_norm": 0.2060737907886505,
+      "learning_rate": 7.239304812834225e-05,
+      "loss": 0.3437,
+      "step": 1914
+    },
+    {
+      "epoch": 1.9155387452721078,
+      "grad_norm": 0.1924235075712204,
+      "learning_rate": 7.232620320855615e-05,
+      "loss": 0.3184,
+      "step": 1915
+    },
+    {
+      "epoch": 1.9165390266012317,
+      "grad_norm": 0.20357847213745117,
+      "learning_rate": 7.225935828877005e-05,
+      "loss": 0.3385,
+      "step": 1916
+    },
+    {
+      "epoch": 1.9175393079303555,
+      "grad_norm": 0.19940611720085144,
+      "learning_rate": 7.219251336898396e-05,
+      "loss": 0.3298,
+      "step": 1917
+    },
+    {
+      "epoch": 1.9185395892594792,
+      "grad_norm": 0.20188336074352264,
+      "learning_rate": 7.212566844919787e-05,
+      "loss": 0.333,
+      "step": 1918
+    },
+    {
+      "epoch": 1.919539870588603,
+      "grad_norm": 0.19728946685791016,
+      "learning_rate": 7.205882352941177e-05,
+      "loss": 0.3396,
+      "step": 1919
+    },
+    {
+      "epoch": 1.9205401519177268,
+      "grad_norm": 0.20807284116744995,
+      "learning_rate": 7.199197860962568e-05,
+      "loss": 0.3479,
+      "step": 1920
+    },
+    {
+      "epoch": 1.9215404332468506,
+      "grad_norm": 0.1932482272386551,
+      "learning_rate": 7.192513368983958e-05,
+      "loss": 0.3404,
+      "step": 1921
+    },
+    {
+      "epoch": 1.9225407145759745,
+      "grad_norm": 0.19407489895820618,
+      "learning_rate": 7.185828877005349e-05,
+      "loss": 0.3542,
+      "step": 1922
+    },
+    {
+      "epoch": 1.9235409959050984,
+      "grad_norm": 0.19956664741039276,
+      "learning_rate": 7.179144385026738e-05,
+      "loss": 0.3256,
+      "step": 1923
+    },
+    {
+      "epoch": 1.924541277234222,
+      "grad_norm": 0.1943155974149704,
+      "learning_rate": 7.17245989304813e-05,
+      "loss": 0.3306,
+      "step": 1924
+    },
+    {
+      "epoch": 1.925541558563346,
+      "grad_norm": 0.1975592076778412,
+      "learning_rate": 7.165775401069518e-05,
+      "loss": 0.333,
+      "step": 1925
+    },
+    {
+      "epoch": 1.9265418398924696,
+      "grad_norm": 0.1958656758069992,
+      "learning_rate": 7.15909090909091e-05,
+      "loss": 0.3284,
+      "step": 1926
+    },
+    {
+      "epoch": 1.9275421212215935,
+      "grad_norm": 0.20319488644599915,
+      "learning_rate": 7.1524064171123e-05,
+      "loss": 0.3473,
+      "step": 1927
+    },
+    {
+      "epoch": 1.9285424025507174,
+      "grad_norm": 0.20498576760292053,
+      "learning_rate": 7.14572192513369e-05,
+      "loss": 0.3626,
+      "step": 1928
+    },
+    {
+      "epoch": 1.9295426838798413,
+      "grad_norm": 0.19313092529773712,
+      "learning_rate": 7.13903743315508e-05,
+      "loss": 0.3393,
+      "step": 1929
+    },
+    {
+      "epoch": 1.9305429652089652,
+      "grad_norm": 0.19795788824558258,
+      "learning_rate": 7.13235294117647e-05,
+      "loss": 0.3491,
+      "step": 1930
+    },
+    {
+      "epoch": 1.9315432465380888,
+      "grad_norm": 0.2025424838066101,
+      "learning_rate": 7.125668449197861e-05,
+      "loss": 0.333,
+      "step": 1931
+    },
+    {
+      "epoch": 1.9325435278672125,
+      "grad_norm": 0.20232167840003967,
+      "learning_rate": 7.118983957219252e-05,
+      "loss": 0.3425,
+      "step": 1932
+    },
+    {
+      "epoch": 1.9335438091963364,
+      "grad_norm": 0.20693567395210266,
+      "learning_rate": 7.112299465240642e-05,
+      "loss": 0.3478,
+      "step": 1933
+    },
+    {
+      "epoch": 1.9345440905254603,
+      "grad_norm": 0.19980236887931824,
+      "learning_rate": 7.105614973262032e-05,
+      "loss": 0.3511,
+      "step": 1934
+    },
+    {
+      "epoch": 1.9355443718545842,
+      "grad_norm": 0.19819217920303345,
+      "learning_rate": 7.098930481283423e-05,
+      "loss": 0.3381,
+      "step": 1935
+    },
+    {
+      "epoch": 1.936544653183708,
+      "grad_norm": 0.1935405284166336,
+      "learning_rate": 7.092245989304813e-05,
+      "loss": 0.3269,
+      "step": 1936
+    },
+    {
+      "epoch": 1.9375449345128317,
+      "grad_norm": 0.20116673409938812,
+      "learning_rate": 7.085561497326202e-05,
+      "loss": 0.3392,
+      "step": 1937
+    },
+    {
+      "epoch": 1.9385452158419556,
+      "grad_norm": 0.19773957133293152,
+      "learning_rate": 7.078877005347594e-05,
+      "loss": 0.3287,
+      "step": 1938
+    },
+    {
+      "epoch": 1.9395454971710793,
+      "grad_norm": 0.2101377248764038,
+      "learning_rate": 7.072192513368985e-05,
+      "loss": 0.3423,
+      "step": 1939
+    },
+    {
+      "epoch": 1.9405457785002032,
+      "grad_norm": 0.2109636813402176,
+      "learning_rate": 7.065508021390375e-05,
+      "loss": 0.3279,
+      "step": 1940
+    },
+    {
+      "epoch": 1.941546059829327,
+      "grad_norm": 0.19801610708236694,
+      "learning_rate": 7.058823529411765e-05,
+      "loss": 0.327,
+      "step": 1941
+    },
+    {
+      "epoch": 1.942546341158451,
+      "grad_norm": 0.20899434387683868,
+      "learning_rate": 7.052139037433156e-05,
+      "loss": 0.3252,
+      "step": 1942
+    },
+    {
+      "epoch": 1.9435466224875746,
+      "grad_norm": 0.20892994105815887,
+      "learning_rate": 7.045454545454546e-05,
+      "loss": 0.3331,
+      "step": 1943
+    },
+    {
+      "epoch": 1.9445469038166985,
+      "grad_norm": 0.20394472777843475,
+      "learning_rate": 7.038770053475935e-05,
+      "loss": 0.3268,
+      "step": 1944
+    },
+    {
+      "epoch": 1.9455471851458221,
+      "grad_norm": 0.19372175633907318,
+      "learning_rate": 7.032085561497327e-05,
+      "loss": 0.3263,
+      "step": 1945
+    },
+    {
+      "epoch": 1.946547466474946,
+      "grad_norm": 0.20710428059101105,
+      "learning_rate": 7.025401069518716e-05,
+      "loss": 0.3506,
+      "step": 1946
+    },
+    {
+      "epoch": 1.94754774780407,
+      "grad_norm": 0.20341874659061432,
+      "learning_rate": 7.018716577540108e-05,
+      "loss": 0.3282,
+      "step": 1947
+    },
+    {
+      "epoch": 1.9485480291331938,
+      "grad_norm": 0.19373811781406403,
+      "learning_rate": 7.012032085561497e-05,
+      "loss": 0.319,
+      "step": 1948
+    },
+    {
+      "epoch": 1.9495483104623177,
+      "grad_norm": 0.19935563206672668,
+      "learning_rate": 7.005347593582889e-05,
+      "loss": 0.3532,
+      "step": 1949
+    },
+    {
+      "epoch": 1.9505485917914414,
+      "grad_norm": 0.19019944965839386,
+      "learning_rate": 6.998663101604278e-05,
+      "loss": 0.3422,
+      "step": 1950
+    },
+    {
+      "epoch": 1.951548873120565,
+      "grad_norm": 0.19666914641857147,
+      "learning_rate": 6.991978609625668e-05,
+      "loss": 0.3349,
+      "step": 1951
+    },
+    {
+      "epoch": 1.952549154449689,
+      "grad_norm": 0.2144993245601654,
+      "learning_rate": 6.985294117647059e-05,
+      "loss": 0.343,
+      "step": 1952
+    },
+    {
+      "epoch": 1.9535494357788128,
+      "grad_norm": 0.2118874043226242,
+      "learning_rate": 6.97860962566845e-05,
+      "loss": 0.3386,
+      "step": 1953
+    },
+    {
+      "epoch": 1.9545497171079367,
+      "grad_norm": 0.19923652708530426,
+      "learning_rate": 6.97192513368984e-05,
+      "loss": 0.3332,
+      "step": 1954
+    },
+    {
+      "epoch": 1.9555499984370606,
+      "grad_norm": 0.1976926177740097,
+      "learning_rate": 6.96524064171123e-05,
+      "loss": 0.3333,
+      "step": 1955
+    },
+    {
+      "epoch": 1.9565502797661842,
+      "grad_norm": 0.2049078792333603,
+      "learning_rate": 6.95855614973262e-05,
+      "loss": 0.3683,
+      "step": 1956
+    },
+    {
+      "epoch": 1.957550561095308,
+      "grad_norm": 0.20219020545482635,
+      "learning_rate": 6.951871657754011e-05,
+      "loss": 0.3461,
+      "step": 1957
+    },
+    {
+      "epoch": 1.9585508424244318,
+      "grad_norm": 0.19515986740589142,
+      "learning_rate": 6.945187165775402e-05,
+      "loss": 0.3341,
+      "step": 1958
+    },
+    {
+      "epoch": 1.9595511237535557,
+      "grad_norm": 0.19154398143291473,
+      "learning_rate": 6.938502673796792e-05,
+      "loss": 0.3268,
+      "step": 1959
+    },
+    {
+      "epoch": 1.9605514050826796,
+      "grad_norm": 0.19598530232906342,
+      "learning_rate": 6.931818181818182e-05,
+      "loss": 0.3462,
+      "step": 1960
+    },
+    {
+      "epoch": 1.9615516864118034,
+      "grad_norm": 0.20135371387004852,
+      "learning_rate": 6.925133689839573e-05,
+      "loss": 0.3495,
+      "step": 1961
+    },
+    {
+      "epoch": 1.962551967740927,
+      "grad_norm": 0.19717907905578613,
+      "learning_rate": 6.918449197860963e-05,
+      "loss": 0.3235,
+      "step": 1962
+    },
+    {
+      "epoch": 1.963552249070051,
+      "grad_norm": 0.20029330253601074,
+      "learning_rate": 6.911764705882354e-05,
+      "loss": 0.3315,
+      "step": 1963
+    },
+    {
+      "epoch": 1.9645525303991747,
+      "grad_norm": 0.19339440762996674,
+      "learning_rate": 6.905080213903744e-05,
+      "loss": 0.3409,
+      "step": 1964
+    },
+    {
+      "epoch": 1.9655528117282985,
+      "grad_norm": 0.19925855100154877,
+      "learning_rate": 6.898395721925133e-05,
+      "loss": 0.3461,
+      "step": 1965
+    },
+    {
+      "epoch": 1.9665530930574224,
+      "grad_norm": 0.19967736303806305,
+      "learning_rate": 6.891711229946525e-05,
+      "loss": 0.3424,
+      "step": 1966
+    },
+    {
+      "epoch": 1.9675533743865463,
+      "grad_norm": 0.2010488361120224,
+      "learning_rate": 6.885026737967914e-05,
+      "loss": 0.3597,
+      "step": 1967
+    },
+    {
+      "epoch": 1.96855365571567,
+      "grad_norm": 0.1985686719417572,
+      "learning_rate": 6.878342245989306e-05,
+      "loss": 0.3379,
+      "step": 1968
+    },
+    {
+      "epoch": 1.9695539370447939,
+      "grad_norm": 0.2022276222705841,
+      "learning_rate": 6.871657754010695e-05,
+      "loss": 0.3468,
+      "step": 1969
+    },
+    {
+      "epoch": 1.9705542183739175,
+      "grad_norm": 0.2181425839662552,
+      "learning_rate": 6.864973262032087e-05,
+      "loss": 0.3587,
+      "step": 1970
+    },
+    {
+      "epoch": 1.9715544997030414,
+      "grad_norm": 0.20604324340820312,
+      "learning_rate": 6.858288770053476e-05,
+      "loss": 0.3405,
+      "step": 1971
+    },
+    {
+      "epoch": 1.9725547810321653,
+      "grad_norm": 0.1980798840522766,
+      "learning_rate": 6.851604278074866e-05,
+      "loss": 0.3342,
+      "step": 1972
+    },
+    {
+      "epoch": 1.9735550623612892,
+      "grad_norm": 0.19435887038707733,
+      "learning_rate": 6.844919786096257e-05,
+      "loss": 0.3159,
+      "step": 1973
+    },
+    {
+      "epoch": 1.974555343690413,
+      "grad_norm": 0.19549942016601562,
+      "learning_rate": 6.838235294117647e-05,
+      "loss": 0.3173,
+      "step": 1974
+    },
+    {
+      "epoch": 1.9755556250195367,
+      "grad_norm": 0.2067781537771225,
+      "learning_rate": 6.831550802139038e-05,
+      "loss": 0.3399,
+      "step": 1975
+    },
+    {
+      "epoch": 1.9765559063486604,
+      "grad_norm": 0.1954166293144226,
+      "learning_rate": 6.824866310160428e-05,
+      "loss": 0.3256,
+      "step": 1976
+    },
+    {
+      "epoch": 1.9775561876777843,
+      "grad_norm": 0.20686157047748566,
+      "learning_rate": 6.818181818181818e-05,
+      "loss": 0.3402,
+      "step": 1977
+    },
+    {
+      "epoch": 1.9785564690069082,
+      "grad_norm": 0.1978681981563568,
+      "learning_rate": 6.811497326203209e-05,
+      "loss": 0.3563,
+      "step": 1978
+    },
+    {
+      "epoch": 1.979556750336032,
+      "grad_norm": 0.20511318743228912,
+      "learning_rate": 6.804812834224599e-05,
+      "loss": 0.3492,
+      "step": 1979
+    },
+    {
+      "epoch": 1.980557031665156,
+      "grad_norm": 0.19985926151275635,
+      "learning_rate": 6.79812834224599e-05,
+      "loss": 0.3265,
+      "step": 1980
+    },
+    {
+      "epoch": 1.9815573129942796,
+      "grad_norm": 0.19825492799282074,
+      "learning_rate": 6.79144385026738e-05,
+      "loss": 0.3382,
+      "step": 1981
+    },
+    {
+      "epoch": 1.9825575943234035,
+      "grad_norm": 0.19896507263183594,
+      "learning_rate": 6.78475935828877e-05,
+      "loss": 0.3325,
+      "step": 1982
+    },
+    {
+      "epoch": 1.9835578756525272,
+      "grad_norm": 0.20204363763332367,
+      "learning_rate": 6.778074866310161e-05,
+      "loss": 0.3422,
+      "step": 1983
+    },
+    {
+      "epoch": 1.984558156981651,
+      "grad_norm": 0.20977722108364105,
+      "learning_rate": 6.771390374331552e-05,
+      "loss": 0.3467,
+      "step": 1984
+    },
+    {
+      "epoch": 1.985558438310775,
+      "grad_norm": 0.19912801682949066,
+      "learning_rate": 6.764705882352942e-05,
+      "loss": 0.3435,
+      "step": 1985
+    },
+    {
+      "epoch": 1.9865587196398988,
+      "grad_norm": 0.2084273397922516,
+      "learning_rate": 6.758021390374331e-05,
+      "loss": 0.3572,
+      "step": 1986
+    },
+    {
+      "epoch": 1.9875590009690225,
+      "grad_norm": 0.20003779232501984,
+      "learning_rate": 6.751336898395723e-05,
+      "loss": 0.3376,
+      "step": 1987
+    },
+    {
+      "epoch": 1.9885592822981464,
+      "grad_norm": 0.20492292940616608,
+      "learning_rate": 6.744652406417112e-05,
+      "loss": 0.3537,
+      "step": 1988
+    },
+    {
+      "epoch": 1.98955956362727,
+      "grad_norm": 0.20428413152694702,
+      "learning_rate": 6.737967914438504e-05,
+      "loss": 0.3659,
+      "step": 1989
+    },
+    {
+      "epoch": 1.990559844956394,
+      "grad_norm": 0.19491609930992126,
+      "learning_rate": 6.731283422459893e-05,
+      "loss": 0.3352,
+      "step": 1990
+    },
+    {
+      "epoch": 1.9915601262855178,
+      "grad_norm": 0.20149634778499603,
+      "learning_rate": 6.724598930481285e-05,
+      "loss": 0.3652,
+      "step": 1991
+    },
+    {
+      "epoch": 1.9925604076146417,
+      "grad_norm": 0.2055903524160385,
+      "learning_rate": 6.717914438502674e-05,
+      "loss": 0.3358,
+      "step": 1992
+    },
+    {
+      "epoch": 1.9935606889437656,
+      "grad_norm": 0.20021459460258484,
+      "learning_rate": 6.711229946524064e-05,
+      "loss": 0.3636,
+      "step": 1993
+    },
+    {
+      "epoch": 1.9945609702728893,
+      "grad_norm": 0.21420694887638092,
+      "learning_rate": 6.704545454545455e-05,
+      "loss": 0.3397,
+      "step": 1994
+    },
+    {
+      "epoch": 1.995561251602013,
+      "grad_norm": 0.2025606632232666,
+      "learning_rate": 6.697860962566845e-05,
+      "loss": 0.3303,
+      "step": 1995
+    },
+    {
+      "epoch": 1.9965615329311368,
+      "grad_norm": 0.20822446048259735,
+      "learning_rate": 6.691176470588235e-05,
+      "loss": 0.3642,
+      "step": 1996
+    },
+    {
+      "epoch": 1.9975618142602607,
+      "grad_norm": 0.1889631152153015,
+      "learning_rate": 6.684491978609626e-05,
+      "loss": 0.3199,
+      "step": 1997
+    },
+    {
+      "epoch": 1.9985620955893846,
+      "grad_norm": 0.20107050240039825,
+      "learning_rate": 6.677807486631016e-05,
+      "loss": 0.3495,
+      "step": 1998
+    },
+    {
+      "epoch": 1.9995623769185085,
+      "grad_norm": 0.19668109714984894,
+      "learning_rate": 6.671122994652407e-05,
+      "loss": 0.3316,
+      "step": 1999
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2997,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.386426946941862e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}