diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.36643459142543056,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00018321729571271528,
+      "grad_norm": 7.481241625973833,
+      "learning_rate": 0.0,
+      "loss": 0.764,
+      "step": 1
+    },
+    {
+      "epoch": 0.00036643459142543056,
+      "grad_norm": 7.936839352385908,
+      "learning_rate": 1.8315018315018317e-08,
+      "loss": 0.7505,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005496518871381459,
+      "grad_norm": 8.381229660678587,
+      "learning_rate": 3.6630036630036635e-08,
+      "loss": 0.7767,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007328691828508611,
+      "grad_norm": 7.533434041856819,
+      "learning_rate": 5.494505494505495e-08,
+      "loss": 0.7545,
+      "step": 4
+    },
+    {
+      "epoch": 0.0009160864785635764,
+      "grad_norm": 6.4876435204860545,
+      "learning_rate": 7.326007326007327e-08,
+      "loss": 0.7254,
+      "step": 5
+    },
+    {
+      "epoch": 0.0010993037742762918,
+      "grad_norm": 6.984439562746247,
+      "learning_rate": 9.157509157509159e-08,
+      "loss": 0.7238,
+      "step": 6
+    },
+    {
+      "epoch": 0.001282521069989007,
+      "grad_norm": 8.432795226889532,
+      "learning_rate": 1.098901098901099e-07,
+      "loss": 0.7667,
+      "step": 7
+    },
+    {
+      "epoch": 0.0014657383657017222,
+      "grad_norm": 6.615389400475257,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 0.739,
+      "step": 8
+    },
+    {
+      "epoch": 0.0016489556614144375,
+      "grad_norm": 7.043438048491804,
+      "learning_rate": 1.4652014652014654e-07,
+      "loss": 0.7412,
+      "step": 9
+    },
+    {
+      "epoch": 0.001832172957127153,
+      "grad_norm": 7.763888342467084,
+      "learning_rate": 1.6483516483516484e-07,
+      "loss": 0.739,
+      "step": 10
+    },
+    {
+      "epoch": 0.002015390252839868,
+      "grad_norm": 7.115680896600135,
+      "learning_rate": 1.8315018315018317e-07,
+      "loss": 0.6814,
+      "step": 11
+    },
+    {
+      "epoch": 0.0021986075485525836,
+      "grad_norm": 9.472751553447468,
+      "learning_rate": 2.014652014652015e-07,
+      "loss": 0.7694,
+      "step": 12
+    },
+    {
+      "epoch": 0.0023818248442652986,
+      "grad_norm": 9.164365397122241,
+      "learning_rate": 2.197802197802198e-07,
+      "loss": 0.8101,
+      "step": 13
+    },
+    {
+      "epoch": 0.002565042139978014,
+      "grad_norm": 7.831934204465113,
+      "learning_rate": 2.3809523809523811e-07,
+      "loss": 0.7504,
+      "step": 14
+    },
+    {
+      "epoch": 0.002748259435690729,
+      "grad_norm": 6.7423320112288865,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 0.7442,
+      "step": 15
+    },
+    {
+      "epoch": 0.0029314767314034445,
+      "grad_norm": 7.8103949840893145,
+      "learning_rate": 2.7472527472527475e-07,
+      "loss": 0.7782,
+      "step": 16
+    },
+    {
+      "epoch": 0.00311469402711616,
+      "grad_norm": 6.1519268993583465,
+      "learning_rate": 2.930402930402931e-07,
+      "loss": 0.663,
+      "step": 17
+    },
+    {
+      "epoch": 0.003297911322828875,
+      "grad_norm": 8.049047835270914,
+      "learning_rate": 3.113553113553114e-07,
+      "loss": 0.7934,
+      "step": 18
+    },
+    {
+      "epoch": 0.0034811286185415903,
+      "grad_norm": 6.388709624292669,
+      "learning_rate": 3.296703296703297e-07,
+      "loss": 0.7241,
+      "step": 19
+    },
+    {
+      "epoch": 0.003664345914254306,
+      "grad_norm": 7.039170133500172,
+      "learning_rate": 3.47985347985348e-07,
+      "loss": 0.7133,
+      "step": 20
+    },
+    {
+      "epoch": 0.003847563209967021,
+      "grad_norm": 8.416142106876258,
+      "learning_rate": 3.6630036630036635e-07,
+      "loss": 0.7787,
+      "step": 21
+    },
+    {
+      "epoch": 0.004030780505679736,
+      "grad_norm": 5.942542945633484,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 0.692,
+      "step": 22
+    },
+    {
+      "epoch": 0.004213997801392451,
+      "grad_norm": 6.397620608740201,
+      "learning_rate": 4.02930402930403e-07,
+      "loss": 0.6885,
+      "step": 23
+    },
+    {
+      "epoch": 0.004397215097105167,
+      "grad_norm": 4.789506080162222,
+      "learning_rate": 4.212454212454213e-07,
+      "loss": 0.6818,
+      "step": 24
+    },
+    {
+      "epoch": 0.004580432392817882,
+      "grad_norm": 5.4323837663846275,
+      "learning_rate": 4.395604395604396e-07,
+      "loss": 0.7202,
+      "step": 25
+    },
+    {
+      "epoch": 0.004763649688530597,
+      "grad_norm": 6.001418955741802,
+      "learning_rate": 4.578754578754579e-07,
+      "loss": 0.6915,
+      "step": 26
+    },
+    {
+      "epoch": 0.004946866984243312,
+      "grad_norm": 6.931290264891086,
+      "learning_rate": 4.7619047619047623e-07,
+      "loss": 0.7214,
+      "step": 27
+    },
+    {
+      "epoch": 0.005130084279956028,
+      "grad_norm": 5.423177642103934,
+      "learning_rate": 4.945054945054946e-07,
+      "loss": 0.6627,
+      "step": 28
+    },
+    {
+      "epoch": 0.005313301575668743,
+      "grad_norm": 5.824082478920574,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 0.7128,
+      "step": 29
+    },
+    {
+      "epoch": 0.005496518871381458,
+      "grad_norm": 4.837515053542817,
+      "learning_rate": 5.311355311355311e-07,
+      "loss": 0.7063,
+      "step": 30
+    },
+    {
+      "epoch": 0.005679736167094174,
+      "grad_norm": 4.2178492878732,
+      "learning_rate": 5.494505494505495e-07,
+      "loss": 0.6667,
+      "step": 31
+    },
+    {
+      "epoch": 0.005862953462806889,
+      "grad_norm": 3.7310273500960283,
+      "learning_rate": 5.677655677655678e-07,
+      "loss": 0.6724,
+      "step": 32
+    },
+    {
+      "epoch": 0.006046170758519604,
+      "grad_norm": 3.4436166557785013,
+      "learning_rate": 5.860805860805862e-07,
+      "loss": 0.6158,
+      "step": 33
+    },
+    {
+      "epoch": 0.00622938805423232,
+      "grad_norm": 2.6914219359650264,
+      "learning_rate": 6.043956043956044e-07,
+      "loss": 0.6384,
+      "step": 34
+    },
+    {
+      "epoch": 0.006412605349945035,
+      "grad_norm": 2.8611154351289803,
+      "learning_rate": 6.227106227106228e-07,
+      "loss": 0.6564,
+      "step": 35
+    },
+    {
+      "epoch": 0.00659582264565775,
+      "grad_norm": 2.915608387835296,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 0.6556,
+      "step": 36
+    },
+    {
+      "epoch": 0.006779039941370466,
+      "grad_norm": 3.104210670394775,
+      "learning_rate": 6.593406593406594e-07,
+      "loss": 0.6571,
+      "step": 37
+    },
+    {
+      "epoch": 0.006962257237083181,
+      "grad_norm": 2.416024185643152,
+      "learning_rate": 6.776556776556777e-07,
+      "loss": 0.6118,
+      "step": 38
+    },
+    {
+      "epoch": 0.007145474532795896,
+      "grad_norm": 2.2840327527366187,
+      "learning_rate": 6.95970695970696e-07,
+      "loss": 0.5773,
+      "step": 39
+    },
+    {
+      "epoch": 0.007328691828508612,
+      "grad_norm": 2.994955066139323,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 0.6621,
+      "step": 40
+    },
+    {
+      "epoch": 0.007511909124221327,
+      "grad_norm": 2.4977146662969956,
+      "learning_rate": 7.326007326007327e-07,
+      "loss": 0.5878,
+      "step": 41
+    },
+    {
+      "epoch": 0.007695126419934042,
+      "grad_norm": 2.2697538551105443,
+      "learning_rate": 7.50915750915751e-07,
+      "loss": 0.6161,
+      "step": 42
+    },
+    {
+      "epoch": 0.007878343715646757,
+      "grad_norm": 2.040084870467292,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 0.625,
+      "step": 43
+    },
+    {
+      "epoch": 0.008061561011359472,
+      "grad_norm": 1.4264410515550054,
+      "learning_rate": 7.875457875457876e-07,
+      "loss": 0.5991,
+      "step": 44
+    },
+    {
+      "epoch": 0.008244778307072188,
+      "grad_norm": 1.3307329745484617,
+      "learning_rate": 8.05860805860806e-07,
+      "loss": 0.5802,
+      "step": 45
+    },
+    {
+      "epoch": 0.008427995602784902,
+      "grad_norm": 1.3186002922814006,
+      "learning_rate": 8.241758241758242e-07,
+      "loss": 0.5971,
+      "step": 46
+    },
+    {
+      "epoch": 0.008611212898497618,
+      "grad_norm": 1.4454419288227918,
+      "learning_rate": 8.424908424908426e-07,
+      "loss": 0.6152,
+      "step": 47
+    },
+    {
+      "epoch": 0.008794430194210334,
+      "grad_norm": 1.2728865809977838,
+      "learning_rate": 8.608058608058609e-07,
+      "loss": 0.6176,
+      "step": 48
+    },
+    {
+      "epoch": 0.008977647489923048,
+      "grad_norm": 1.2387163683388445,
+      "learning_rate": 8.791208791208792e-07,
+      "loss": 0.6233,
+      "step": 49
+    },
+    {
+      "epoch": 0.009160864785635764,
+      "grad_norm": 1.1465107255795282,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 0.5795,
+      "step": 50
+    },
+    {
+      "epoch": 0.00934408208134848,
+      "grad_norm": 1.264265250343769,
+      "learning_rate": 9.157509157509158e-07,
+      "loss": 0.6262,
+      "step": 51
+    },
+    {
+      "epoch": 0.009527299377061194,
+      "grad_norm": 0.919312835052228,
+      "learning_rate": 9.340659340659342e-07,
+      "loss": 0.5535,
+      "step": 52
+    },
+    {
+      "epoch": 0.00971051667277391,
+      "grad_norm": 0.8672735287012353,
+      "learning_rate": 9.523809523809525e-07,
+      "loss": 0.5327,
+      "step": 53
+    },
+    {
+      "epoch": 0.009893733968486624,
+      "grad_norm": 0.8288169490471791,
+      "learning_rate": 9.706959706959708e-07,
+      "loss": 0.5673,
+      "step": 54
+    },
+    {
+      "epoch": 0.01007695126419934,
+      "grad_norm": 0.8603030677007734,
+      "learning_rate": 9.890109890109891e-07,
+      "loss": 0.5547,
+      "step": 55
+    },
+    {
+      "epoch": 0.010260168559912056,
+      "grad_norm": 0.7482006424960487,
+      "learning_rate": 1.0073260073260074e-06,
+      "loss": 0.6012,
+      "step": 56
+    },
+    {
+      "epoch": 0.01044338585562477,
+      "grad_norm": 0.6144742944660406,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 0.5231,
+      "step": 57
+    },
+    {
+      "epoch": 0.010626603151337486,
+      "grad_norm": 0.634897676727318,
+      "learning_rate": 1.0439560439560442e-06,
+      "loss": 0.5726,
+      "step": 58
+    },
+    {
+      "epoch": 0.010809820447050202,
+      "grad_norm": 0.6410545094813134,
+      "learning_rate": 1.0622710622710622e-06,
+      "loss": 0.5695,
+      "step": 59
+    },
+    {
+      "epoch": 0.010993037742762916,
+      "grad_norm": 0.6609456014192859,
+      "learning_rate": 1.0805860805860807e-06,
+      "loss": 0.5805,
+      "step": 60
+    },
+    {
+      "epoch": 0.011176255038475632,
+      "grad_norm": 0.7416419125176478,
+      "learning_rate": 1.098901098901099e-06,
+      "loss": 0.634,
+      "step": 61
+    },
+    {
+      "epoch": 0.011359472334188348,
+      "grad_norm": 0.6851293138177442,
+      "learning_rate": 1.1172161172161173e-06,
+      "loss": 0.5544,
+      "step": 62
+    },
+    {
+      "epoch": 0.011542689629901062,
+      "grad_norm": 0.6383351643137376,
+      "learning_rate": 1.1355311355311355e-06,
+      "loss": 0.557,
+      "step": 63
+    },
+    {
+      "epoch": 0.011725906925613778,
+      "grad_norm": 0.6348085314283121,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 0.5632,
+      "step": 64
+    },
+    {
+      "epoch": 0.011909124221326494,
+      "grad_norm": 0.7024363928163538,
+      "learning_rate": 1.1721611721611723e-06,
+      "loss": 0.5839,
+      "step": 65
+    },
+    {
+      "epoch": 0.012092341517039208,
+      "grad_norm": 0.5900911656210056,
+      "learning_rate": 1.1904761904761906e-06,
+      "loss": 0.5344,
+      "step": 66
+    },
+    {
+      "epoch": 0.012275558812751924,
+      "grad_norm": 0.6275963867777459,
+      "learning_rate": 1.2087912087912089e-06,
+      "loss": 0.5584,
+      "step": 67
+    },
+    {
+      "epoch": 0.01245877610846464,
+      "grad_norm": 0.5865171487667508,
+      "learning_rate": 1.2271062271062271e-06,
+      "loss": 0.575,
+      "step": 68
+    },
+    {
+      "epoch": 0.012641993404177354,
+      "grad_norm": 0.674046870317235,
+      "learning_rate": 1.2454212454212456e-06,
+      "loss": 0.5744,
+      "step": 69
+    },
+    {
+      "epoch": 0.01282521069989007,
+      "grad_norm": 0.60918911566989,
+      "learning_rate": 1.2637362637362637e-06,
+      "loss": 0.5423,
+      "step": 70
+    },
+    {
+      "epoch": 0.013008427995602785,
+      "grad_norm": 0.6221860528344966,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 0.5652,
+      "step": 71
+    },
+    {
+      "epoch": 0.0131916452913155,
+      "grad_norm": 0.606665778835579,
+      "learning_rate": 1.3003663003663005e-06,
+      "loss": 0.6123,
+      "step": 72
+    },
+    {
+      "epoch": 0.013374862587028215,
+      "grad_norm": 0.5884723188789052,
+      "learning_rate": 1.3186813186813187e-06,
+      "loss": 0.5863,
+      "step": 73
+    },
+    {
+      "epoch": 0.013558079882740931,
+      "grad_norm": 0.5824062487652404,
+      "learning_rate": 1.336996336996337e-06,
+      "loss": 0.5918,
+      "step": 74
+    },
+    {
+      "epoch": 0.013741297178453645,
+      "grad_norm": 0.5460196146561194,
+      "learning_rate": 1.3553113553113553e-06,
+      "loss": 0.5912,
+      "step": 75
+    },
+    {
+      "epoch": 0.013924514474166361,
+      "grad_norm": 0.49213553293102813,
+      "learning_rate": 1.3736263736263738e-06,
+      "loss": 0.5421,
+      "step": 76
+    },
+    {
+      "epoch": 0.014107731769879077,
+      "grad_norm": 0.5594168254817149,
+      "learning_rate": 1.391941391941392e-06,
+      "loss": 0.5122,
+      "step": 77
+    },
+    {
+      "epoch": 0.014290949065591791,
+      "grad_norm": 0.5277088034821339,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 0.5678,
+      "step": 78
+    },
+    {
+      "epoch": 0.014474166361304507,
+      "grad_norm": 0.49574136870511754,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.5498,
+      "step": 79
+    },
+    {
+      "epoch": 0.014657383657017223,
+      "grad_norm": 0.5944104688680958,
+      "learning_rate": 1.4468864468864471e-06,
+      "loss": 0.5469,
+      "step": 80
+    },
+    {
+      "epoch": 0.014840600952729937,
+      "grad_norm": 0.4732970613815555,
+      "learning_rate": 1.4652014652014654e-06,
+      "loss": 0.5233,
+      "step": 81
+    },
+    {
+      "epoch": 0.015023818248442653,
+      "grad_norm": 0.49591541086638596,
+      "learning_rate": 1.4835164835164837e-06,
+      "loss": 0.5367,
+      "step": 82
+    },
+    {
+      "epoch": 0.015207035544155369,
+      "grad_norm": 0.4883648219378977,
+      "learning_rate": 1.501831501831502e-06,
+      "loss": 0.5782,
+      "step": 83
+    },
+    {
+      "epoch": 0.015390252839868083,
+      "grad_norm": 0.48559293967760114,
+      "learning_rate": 1.5201465201465202e-06,
+      "loss": 0.5578,
+      "step": 84
+    },
+    {
+      "epoch": 0.015573470135580799,
+      "grad_norm": 0.5802435990379926,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 0.5562,
+      "step": 85
+    },
+    {
+      "epoch": 0.015756687431293513,
+      "grad_norm": 0.6655689083082568,
+      "learning_rate": 1.556776556776557e-06,
+      "loss": 0.5916,
+      "step": 86
+    },
+    {
+      "epoch": 0.01593990472700623,
+      "grad_norm": 0.5272919501805459,
+      "learning_rate": 1.5750915750915753e-06,
+      "loss": 0.5462,
+      "step": 87
+    },
+    {
+      "epoch": 0.016123122022718945,
+      "grad_norm": 0.5025768009972991,
+      "learning_rate": 1.5934065934065933e-06,
+      "loss": 0.564,
+      "step": 88
+    },
+    {
+      "epoch": 0.01630633931843166,
+      "grad_norm": 0.4868371012830415,
+      "learning_rate": 1.611721611721612e-06,
+      "loss": 0.557,
+      "step": 89
+    },
+    {
+      "epoch": 0.016489556614144377,
+      "grad_norm": 0.4724914851279723,
+      "learning_rate": 1.6300366300366301e-06,
+      "loss": 0.534,
+      "step": 90
+    },
+    {
+      "epoch": 0.01667277390985709,
+      "grad_norm": 0.4933114328584066,
+      "learning_rate": 1.6483516483516484e-06,
+      "loss": 0.5628,
+      "step": 91
+    },
+    {
+      "epoch": 0.016855991205569805,
+      "grad_norm": 0.5429724547645147,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.5693,
+      "step": 92
+    },
+    {
+      "epoch": 0.017039208501282523,
+      "grad_norm": 0.46970029586030615,
+      "learning_rate": 1.6849816849816852e-06,
+      "loss": 0.5761,
+      "step": 93
+    },
+    {
+      "epoch": 0.017222425796995237,
+      "grad_norm": 0.5466204682716642,
+      "learning_rate": 1.7032967032967034e-06,
+      "loss": 0.5407,
+      "step": 94
+    },
+    {
+      "epoch": 0.01740564309270795,
+      "grad_norm": 0.46448768993410167,
+      "learning_rate": 1.7216117216117217e-06,
+      "loss": 0.5784,
+      "step": 95
+    },
+    {
+      "epoch": 0.01758886038842067,
+      "grad_norm": 0.46769210989227256,
+      "learning_rate": 1.73992673992674e-06,
+      "loss": 0.5262,
+      "step": 96
+    },
+    {
+      "epoch": 0.017772077684133383,
+      "grad_norm": 0.4412027593954725,
+      "learning_rate": 1.7582417582417585e-06,
+      "loss": 0.5608,
+      "step": 97
+    },
+    {
+      "epoch": 0.017955294979846097,
+      "grad_norm": 0.5975733130220022,
+      "learning_rate": 1.7765567765567768e-06,
+      "loss": 0.5633,
+      "step": 98
+    },
+    {
+      "epoch": 0.018138512275558814,
+      "grad_norm": 0.46156861116006753,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 0.5576,
+      "step": 99
+    },
+    {
+      "epoch": 0.01832172957127153,
+      "grad_norm": 0.5193644526534718,
+      "learning_rate": 1.8131868131868133e-06,
+      "loss": 0.5533,
+      "step": 100
+    },
+    {
+      "epoch": 0.018504946866984243,
+      "grad_norm": 0.479596247036775,
+      "learning_rate": 1.8315018315018316e-06,
+      "loss": 0.5337,
+      "step": 101
+    },
+    {
+      "epoch": 0.01868816416269696,
+      "grad_norm": 0.4610500867184236,
+      "learning_rate": 1.84981684981685e-06,
+      "loss": 0.5314,
+      "step": 102
+    },
+    {
+      "epoch": 0.018871381458409674,
+      "grad_norm": 0.481950984787821,
+      "learning_rate": 1.8681318681318684e-06,
+      "loss": 0.5865,
+      "step": 103
+    },
+    {
+      "epoch": 0.01905459875412239,
+      "grad_norm": 0.5225151739123198,
+      "learning_rate": 1.8864468864468866e-06,
+      "loss": 0.5694,
+      "step": 104
+    },
+    {
+      "epoch": 0.019237816049835106,
+      "grad_norm": 0.46381468108353424,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 0.557,
+      "step": 105
+    },
+    {
+      "epoch": 0.01942103334554782,
+      "grad_norm": 0.45411743556679485,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 0.5648,
+      "step": 106
+    },
+    {
+      "epoch": 0.019604250641260534,
+      "grad_norm": 0.4859210831259208,
+      "learning_rate": 1.9413919413919417e-06,
+      "loss": 0.5685,
+      "step": 107
+    },
+    {
+      "epoch": 0.01978746793697325,
+      "grad_norm": 0.4205701080898951,
+      "learning_rate": 1.95970695970696e-06,
+      "loss": 0.5426,
+      "step": 108
+    },
+    {
+      "epoch": 0.019970685232685966,
+      "grad_norm": 0.5044757429436201,
+      "learning_rate": 1.9780219780219782e-06,
+      "loss": 0.5509,
+      "step": 109
+    },
+    {
+      "epoch": 0.02015390252839868,
+      "grad_norm": 0.47571345015861244,
+      "learning_rate": 1.9963369963369965e-06,
+      "loss": 0.5281,
+      "step": 110
+    },
+    {
+      "epoch": 0.020337119824111394,
+      "grad_norm": 0.5681283820477757,
+      "learning_rate": 2.0146520146520148e-06,
+      "loss": 0.5398,
+      "step": 111
+    },
+    {
+      "epoch": 0.020520337119824112,
+      "grad_norm": 0.48199138528425167,
+      "learning_rate": 2.032967032967033e-06,
+      "loss": 0.56,
+      "step": 112
+    },
+    {
+      "epoch": 0.020703554415536826,
+      "grad_norm": 0.4841461198775233,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 0.5454,
+      "step": 113
+    },
+    {
+      "epoch": 0.02088677171124954,
+      "grad_norm": 0.4680411378560794,
+      "learning_rate": 2.0695970695970696e-06,
+      "loss": 0.516,
+      "step": 114
+    },
+    {
+      "epoch": 0.021069989006962258,
+      "grad_norm": 0.4426302566203345,
+      "learning_rate": 2.0879120879120883e-06,
+      "loss": 0.5493,
+      "step": 115
+    },
+    {
+      "epoch": 0.021253206302674972,
+      "grad_norm": 0.5379521772056074,
+      "learning_rate": 2.1062271062271066e-06,
+      "loss": 0.5642,
+      "step": 116
+    },
+    {
+      "epoch": 0.021436423598387686,
+      "grad_norm": 0.42644965401865687,
+      "learning_rate": 2.1245421245421245e-06,
+      "loss": 0.564,
+      "step": 117
+    },
+    {
+      "epoch": 0.021619640894100404,
+      "grad_norm": 1.3341683924694292,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 0.5831,
+      "step": 118
+    },
+    {
+      "epoch": 0.021802858189813118,
+      "grad_norm": 0.5613781478585862,
+      "learning_rate": 2.1611721611721614e-06,
+      "loss": 0.5598,
+      "step": 119
+    },
+    {
+      "epoch": 0.021986075485525832,
+      "grad_norm": 0.4389899980262906,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 0.5498,
+      "step": 120
+    },
+    {
+      "epoch": 0.02216929278123855,
+      "grad_norm": 0.5433793107490897,
+      "learning_rate": 2.197802197802198e-06,
+      "loss": 0.5838,
+      "step": 121
+    },
+    {
+      "epoch": 0.022352510076951264,
+      "grad_norm": 0.49992126724035435,
+      "learning_rate": 2.2161172161172163e-06,
+      "loss": 0.5581,
+      "step": 122
+    },
+    {
+      "epoch": 0.022535727372663978,
+      "grad_norm": 0.5192253683114394,
+      "learning_rate": 2.2344322344322345e-06,
+      "loss": 0.5963,
+      "step": 123
+    },
+    {
+      "epoch": 0.022718944668376696,
+      "grad_norm": 0.5218720529273816,
+      "learning_rate": 2.252747252747253e-06,
+      "loss": 0.5568,
+      "step": 124
+    },
+    {
+      "epoch": 0.02290216196408941,
+      "grad_norm": 0.41135006043138017,
+      "learning_rate": 2.271062271062271e-06,
+      "loss": 0.5256,
+      "step": 125
+    },
+    {
+      "epoch": 0.023085379259802124,
+      "grad_norm": 0.4960132808732071,
+      "learning_rate": 2.2893772893772894e-06,
+      "loss": 0.5636,
+      "step": 126
+    },
+    {
+      "epoch": 0.02326859655551484,
+      "grad_norm": 0.40294874528313,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 0.5373,
+      "step": 127
+    },
+    {
+      "epoch": 0.023451813851227556,
+      "grad_norm": 0.49129951716432685,
+      "learning_rate": 2.3260073260073264e-06,
+      "loss": 0.5571,
+      "step": 128
+    },
+    {
+      "epoch": 0.02363503114694027,
+      "grad_norm": 0.4541626724950761,
+      "learning_rate": 2.3443223443223446e-06,
+      "loss": 0.5407,
+      "step": 129
+    },
+    {
+      "epoch": 0.023818248442652987,
+      "grad_norm": 0.44854257290105054,
+      "learning_rate": 2.362637362637363e-06,
+      "loss": 0.5516,
+      "step": 130
+    },
+    {
+      "epoch": 0.0240014657383657,
+      "grad_norm": 0.4439039076319426,
+      "learning_rate": 2.380952380952381e-06,
+      "loss": 0.5192,
+      "step": 131
+    },
+    {
+      "epoch": 0.024184683034078416,
+      "grad_norm": 0.4410959524536652,
+      "learning_rate": 2.3992673992673995e-06,
+      "loss": 0.5678,
+      "step": 132
+    },
+    {
+      "epoch": 0.024367900329791133,
+      "grad_norm": 0.49361868947998544,
+      "learning_rate": 2.4175824175824177e-06,
+      "loss": 0.5591,
+      "step": 133
+    },
+    {
+      "epoch": 0.024551117625503847,
+      "grad_norm": 0.5023022664764156,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 0.5818,
+      "step": 134
+    },
+    {
+      "epoch": 0.02473433492121656,
+      "grad_norm": 0.4287249843742935,
+      "learning_rate": 2.4542124542124543e-06,
+      "loss": 0.5532,
+      "step": 135
+    },
+    {
+      "epoch": 0.02491755221692928,
+      "grad_norm": 0.43595844702277614,
+      "learning_rate": 2.472527472527473e-06,
+      "loss": 0.534,
+      "step": 136
+    },
+    {
+      "epoch": 0.025100769512641993,
+      "grad_norm": 0.5546618594341888,
+      "learning_rate": 2.4908424908424913e-06,
+      "loss": 0.5058,
+      "step": 137
+    },
+    {
+      "epoch": 0.025283986808354707,
+      "grad_norm": 0.5017549598118495,
+      "learning_rate": 2.509157509157509e-06,
+      "loss": 0.5416,
+      "step": 138
+    },
+    {
+      "epoch": 0.025467204104067425,
+      "grad_norm": 0.47047331776441903,
+      "learning_rate": 2.5274725274725274e-06,
+      "loss": 0.5557,
+      "step": 139
+    },
+    {
+      "epoch": 0.02565042139978014,
+      "grad_norm": 0.4881037742451912,
+      "learning_rate": 2.5457875457875457e-06,
+      "loss": 0.5494,
+      "step": 140
+    },
+    {
+      "epoch": 0.025833638695492853,
+      "grad_norm": 0.41933017046045795,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 0.5729,
+      "step": 141
+    },
+    {
+      "epoch": 0.02601685599120557,
+      "grad_norm": 0.39974209370594543,
+      "learning_rate": 2.582417582417583e-06,
+      "loss": 0.5381,
+      "step": 142
+    },
+    {
+      "epoch": 0.026200073286918285,
+      "grad_norm": 0.4473481461077892,
+      "learning_rate": 2.600732600732601e-06,
+      "loss": 0.5336,
+      "step": 143
+    },
+    {
+      "epoch": 0.026383290582631,
+      "grad_norm": 0.45197895778035446,
+      "learning_rate": 2.6190476190476192e-06,
+      "loss": 0.5271,
+      "step": 144
+    },
+    {
+      "epoch": 0.026566507878343717,
+      "grad_norm": 0.49529182062519755,
+      "learning_rate": 2.6373626373626375e-06,
+      "loss": 0.5468,
+      "step": 145
+    },
+    {
+      "epoch": 0.02674972517405643,
+      "grad_norm": 0.5413416851523152,
+      "learning_rate": 2.6556776556776558e-06,
+      "loss": 0.5807,
+      "step": 146
+    },
+    {
+      "epoch": 0.026932942469769145,
+      "grad_norm": 0.4703582853460863,
+      "learning_rate": 2.673992673992674e-06,
+      "loss": 0.538,
+      "step": 147
+    },
+    {
+      "epoch": 0.027116159765481863,
+      "grad_norm": 0.4332295907365602,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 0.5167,
+      "step": 148
+    },
+    {
+      "epoch": 0.027299377061194577,
+      "grad_norm": 0.458616583095158,
+      "learning_rate": 2.7106227106227106e-06,
+      "loss": 0.5226,
+      "step": 149
+    },
+    {
+      "epoch": 0.02748259435690729,
+      "grad_norm": 0.43740744109233864,
+      "learning_rate": 2.728937728937729e-06,
+      "loss": 0.541,
+      "step": 150
+    },
+    {
+      "epoch": 0.02766581165262001,
+      "grad_norm": 0.5735574364330706,
+      "learning_rate": 2.7472527472527476e-06,
+      "loss": 0.5835,
+      "step": 151
+    },
+    {
+      "epoch": 0.027849028948332723,
+      "grad_norm": 0.5055518766772779,
+      "learning_rate": 2.765567765567766e-06,
+      "loss": 0.5631,
+      "step": 152
+    },
+    {
+      "epoch": 0.028032246244045437,
+      "grad_norm": 0.43271616931082146,
+      "learning_rate": 2.783882783882784e-06,
+      "loss": 0.5426,
+      "step": 153
+    },
+    {
+      "epoch": 0.028215463539758154,
+      "grad_norm": 0.4971463373765023,
+      "learning_rate": 2.8021978021978024e-06,
+      "loss": 0.5389,
+      "step": 154
+    },
+    {
+      "epoch": 0.02839868083547087,
+      "grad_norm": 0.4591734918280654,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 0.5309,
+      "step": 155
+    },
+    {
+      "epoch": 0.028581898131183583,
+      "grad_norm": 0.45247087926212365,
+      "learning_rate": 2.838827838827839e-06,
+      "loss": 0.5536,
+      "step": 156
+    },
+    {
+      "epoch": 0.0287651154268963,
+      "grad_norm": 0.46626015341742694,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.559,
+      "step": 157
+    },
+    {
+      "epoch": 0.028948332722609015,
+      "grad_norm": 0.4834854871636071,
+      "learning_rate": 2.8754578754578755e-06,
+      "loss": 0.5476,
+      "step": 158
+    },
+    {
+      "epoch": 0.02913155001832173,
+      "grad_norm": 0.570546858013933,
+      "learning_rate": 2.8937728937728942e-06,
+      "loss": 0.5394,
+      "step": 159
+    },
+    {
+      "epoch": 0.029314767314034446,
+      "grad_norm": 0.45441480104130544,
+      "learning_rate": 2.9120879120879125e-06,
+      "loss": 0.5029,
+      "step": 160
+    },
+    {
+      "epoch": 0.02949798460974716,
+      "grad_norm": 0.5206871098196969,
+      "learning_rate": 2.930402930402931e-06,
+      "loss": 0.528,
+      "step": 161
+    },
+    {
+      "epoch": 0.029681201905459875,
+      "grad_norm": 0.43359542229542136,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 0.5502,
+      "step": 162
+    },
+    {
+      "epoch": 0.029864419201172592,
+      "grad_norm": 0.46302617065984364,
+      "learning_rate": 2.9670329670329673e-06,
+      "loss": 0.5447,
+      "step": 163
+    },
+    {
+      "epoch": 0.030047636496885306,
+      "grad_norm": 0.3915642271916536,
+      "learning_rate": 2.9853479853479856e-06,
+      "loss": 0.503,
+      "step": 164
+    },
+    {
+      "epoch": 0.03023085379259802,
+      "grad_norm": 0.42991246482419715,
+      "learning_rate": 3.003663003663004e-06,
+      "loss": 0.521,
+      "step": 165
+    },
+    {
+      "epoch": 0.030414071088310738,
+      "grad_norm": 0.468175424095518,
+      "learning_rate": 3.021978021978022e-06,
+      "loss": 0.5101,
+      "step": 166
+    },
+    {
+      "epoch": 0.030597288384023452,
+      "grad_norm": 0.6468735604295471,
+      "learning_rate": 3.0402930402930405e-06,
+      "loss": 0.5617,
+      "step": 167
+    },
+    {
+      "epoch": 0.030780505679736166,
+      "grad_norm": 0.5058923699848836,
+      "learning_rate": 3.058608058608059e-06,
+      "loss": 0.5154,
+      "step": 168
+    },
+    {
+      "epoch": 0.030963722975448884,
+      "grad_norm": 0.45437537978064513,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.4993,
+      "step": 169
+    },
+    {
+      "epoch": 0.031146940271161598,
+      "grad_norm": 0.467183819204265,
+      "learning_rate": 3.0952380952380957e-06,
+      "loss": 0.5701,
+      "step": 170
+    },
+    {
+      "epoch": 0.031330157566874316,
+      "grad_norm": 0.3835674459614267,
+      "learning_rate": 3.113553113553114e-06,
+      "loss": 0.4902,
+      "step": 171
+    },
+    {
+      "epoch": 0.031513374862587026,
+      "grad_norm": 0.4292795112150875,
+      "learning_rate": 3.1318681318681323e-06,
+      "loss": 0.5237,
+      "step": 172
+    },
+    {
+      "epoch": 0.031696592158299744,
+      "grad_norm": 0.6593051731116806,
+      "learning_rate": 3.1501831501831505e-06,
+      "loss": 0.5361,
+      "step": 173
+    },
+    {
+      "epoch": 0.03187980945401246,
+      "grad_norm": 0.4718436483558741,
+      "learning_rate": 3.1684981684981684e-06,
+      "loss": 0.5324,
+      "step": 174
+    },
+    {
+      "epoch": 0.03206302674972517,
+      "grad_norm": 0.48018416222395494,
+      "learning_rate": 3.1868131868131867e-06,
+      "loss": 0.5273,
+      "step": 175
+    },
+    {
+      "epoch": 0.03224624404543789,
+      "grad_norm": 0.6316028239985647,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 0.5346,
+      "step": 176
+    },
+    {
+      "epoch": 0.03242946134115061,
+      "grad_norm": 0.4596646339699305,
+      "learning_rate": 3.223443223443224e-06,
+      "loss": 0.5263,
+      "step": 177
+    },
+    {
+      "epoch": 0.03261267863686332,
+      "grad_norm": 0.6529252419894329,
+      "learning_rate": 3.2417582417582424e-06,
+      "loss": 0.5442,
+      "step": 178
+    },
+    {
+      "epoch": 0.032795895932576036,
+      "grad_norm": 0.5538108341969676,
+      "learning_rate": 3.2600732600732602e-06,
+      "loss": 0.543,
+      "step": 179
+    },
+    {
+      "epoch": 0.03297911322828875,
+      "grad_norm": 0.46058359751530825,
+      "learning_rate": 3.2783882783882785e-06,
+      "loss": 0.5101,
+      "step": 180
+    },
+    {
+      "epoch": 0.033162330524001464,
+      "grad_norm": 0.42238032667898895,
+      "learning_rate": 3.2967032967032968e-06,
+      "loss": 0.5487,
+      "step": 181
+    },
+    {
+      "epoch": 0.03334554781971418,
+      "grad_norm": 0.49973876221605035,
+      "learning_rate": 3.315018315018315e-06,
+      "loss": 0.5333,
+      "step": 182
+    },
+    {
+      "epoch": 0.0335287651154269,
+      "grad_norm": 0.49185079032879564,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.5733,
+      "step": 183
+    },
+    {
+      "epoch": 0.03371198241113961,
+      "grad_norm": 0.41357500203470565,
+      "learning_rate": 3.3516483516483516e-06,
+      "loss": 0.5108,
+      "step": 184
+    },
+    {
+      "epoch": 0.03389519970685233,
+      "grad_norm": 0.45597659979596383,
+      "learning_rate": 3.3699633699633703e-06,
+      "loss": 0.5316,
+      "step": 185
+    },
+    {
+      "epoch": 0.034078417002565045,
+      "grad_norm": 0.523873134306111,
+      "learning_rate": 3.3882783882783886e-06,
+      "loss": 0.5516,
+      "step": 186
+    },
+    {
+      "epoch": 0.034261634298277756,
+      "grad_norm": 0.4604705346503309,
+      "learning_rate": 3.406593406593407e-06,
+      "loss": 0.5463,
+      "step": 187
+    },
+    {
+      "epoch": 0.03444485159399047,
+      "grad_norm": 0.4494660179181277,
+      "learning_rate": 3.424908424908425e-06,
+      "loss": 0.5195,
+      "step": 188
+    },
+    {
+      "epoch": 0.03462806888970319,
+      "grad_norm": 0.47161535336220833,
+      "learning_rate": 3.4432234432234434e-06,
+      "loss": 0.5495,
+      "step": 189
+    },
+    {
+      "epoch": 0.0348112861854159,
+      "grad_norm": 0.41422931205807795,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 0.5165,
+      "step": 190
+    },
+    {
+      "epoch": 0.03499450348112862,
+      "grad_norm": 0.48552414748511474,
+      "learning_rate": 3.47985347985348e-06,
+      "loss": 0.5207,
+      "step": 191
+    },
+    {
+      "epoch": 0.03517772077684134,
+      "grad_norm": 0.49918231460984896,
+      "learning_rate": 3.4981684981684982e-06,
+      "loss": 0.5055,
+      "step": 192
+    },
+    {
+      "epoch": 0.03536093807255405,
+      "grad_norm": 0.5047784031000427,
+      "learning_rate": 3.516483516483517e-06,
+      "loss": 0.5541,
+      "step": 193
+    },
+    {
+      "epoch": 0.035544155368266765,
+      "grad_norm": 0.45515393167769386,
+      "learning_rate": 3.5347985347985352e-06,
+      "loss": 0.5324,
+      "step": 194
+    },
+    {
+      "epoch": 0.03572737266397948,
+      "grad_norm": 0.4697755766255993,
+      "learning_rate": 3.5531135531135535e-06,
+      "loss": 0.5317,
+      "step": 195
+    },
+    {
+      "epoch": 0.035910589959692193,
+      "grad_norm": 0.47670842684148323,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 0.5331,
+      "step": 196
+    },
+    {
+      "epoch": 0.03609380725540491,
+      "grad_norm": 0.46160363428859325,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 0.5645,
+      "step": 197
+    },
+    {
+      "epoch": 0.03627702455111763,
+      "grad_norm": 0.5456279444043086,
+      "learning_rate": 3.6080586080586083e-06,
+      "loss": 0.4923,
+      "step": 198
+    },
+    {
+      "epoch": 0.03646024184683034,
+      "grad_norm": 0.45679324633890633,
+      "learning_rate": 3.6263736263736266e-06,
+      "loss": 0.5408,
+      "step": 199
+    },
+    {
+      "epoch": 0.03664345914254306,
+      "grad_norm": 0.4038620329916062,
+      "learning_rate": 3.644688644688645e-06,
+      "loss": 0.5542,
+      "step": 200
+    },
+    {
+      "epoch": 0.036826676438255775,
+      "grad_norm": 0.42338470275172285,
+      "learning_rate": 3.663003663003663e-06,
+      "loss": 0.5509,
+      "step": 201
+    },
+    {
+      "epoch": 0.037009893733968485,
+      "grad_norm": 0.48862631901451187,
+      "learning_rate": 3.681318681318682e-06,
+      "loss": 0.533,
+      "step": 202
+    },
+    {
+      "epoch": 0.0371931110296812,
+      "grad_norm": 0.4339739485667441,
+      "learning_rate": 3.6996336996337e-06,
+      "loss": 0.5331,
+      "step": 203
+    },
+    {
+      "epoch": 0.03737632832539392,
+      "grad_norm": 0.4040593227125272,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 0.568,
+      "step": 204
+    },
+    {
+      "epoch": 0.03755954562110663,
+      "grad_norm": 0.5025361443024049,
+      "learning_rate": 3.7362637362637367e-06,
+      "loss": 0.5616,
+      "step": 205
+    },
+    {
+      "epoch": 0.03774276291681935,
+      "grad_norm": 0.43671528407867277,
+      "learning_rate": 3.754578754578755e-06,
+      "loss": 0.5468,
+      "step": 206
+    },
+    {
+      "epoch": 0.037925980212532066,
+      "grad_norm": 0.5426409729355149,
+      "learning_rate": 3.7728937728937733e-06,
+      "loss": 0.5274,
+      "step": 207
+    },
+    {
+      "epoch": 0.03810919750824478,
+      "grad_norm": 0.4383066270524436,
+      "learning_rate": 3.7912087912087915e-06,
+      "loss": 0.5491,
+      "step": 208
+    },
+    {
+      "epoch": 0.038292414803957495,
+      "grad_norm": 0.4541571165503901,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 0.511,
+      "step": 209
+    },
+    {
+      "epoch": 0.03847563209967021,
+      "grad_norm": 0.42914694472434756,
+      "learning_rate": 3.827838827838828e-06,
+      "loss": 0.4891,
+      "step": 210
+    },
+    {
+      "epoch": 0.03865884939538292,
+      "grad_norm": 0.4583586099579229,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.5359,
+      "step": 211
+    },
+    {
+      "epoch": 0.03884206669109564,
+      "grad_norm": 0.46332307938408596,
+      "learning_rate": 3.864468864468865e-06,
+      "loss": 0.506,
+      "step": 212
+    },
+    {
+      "epoch": 0.03902528398680836,
+      "grad_norm": 0.4979093495563886,
+      "learning_rate": 3.882783882783883e-06,
+      "loss": 0.5523,
+      "step": 213
+    },
+    {
+      "epoch": 0.03920850128252107,
+      "grad_norm": 0.524225251704003,
+      "learning_rate": 3.901098901098901e-06,
+      "loss": 0.543,
+      "step": 214
+    },
+    {
+      "epoch": 0.039391718578233786,
+      "grad_norm": 0.5147235908704881,
+      "learning_rate": 3.91941391941392e-06,
+      "loss": 0.5465,
+      "step": 215
+    },
+    {
+      "epoch": 0.0395749358739465,
+      "grad_norm": 0.4170039710750516,
+      "learning_rate": 3.937728937728938e-06,
+      "loss": 0.557,
+      "step": 216
+    },
+    {
+      "epoch": 0.039758153169659215,
+      "grad_norm": 0.4459151957597281,
+      "learning_rate": 3.9560439560439565e-06,
+      "loss": 0.5176,
+      "step": 217
+    },
+    {
+      "epoch": 0.03994137046537193,
+      "grad_norm": 0.4387028825678643,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 0.5433,
+      "step": 218
+    },
+    {
+      "epoch": 0.04012458776108464,
+      "grad_norm": 0.4606220492260897,
+      "learning_rate": 3.992673992673993e-06,
+      "loss": 0.5558,
+      "step": 219
+    },
+    {
+      "epoch": 0.04030780505679736,
+      "grad_norm": 0.43876788524484817,
+      "learning_rate": 4.010989010989012e-06,
+      "loss": 0.496,
+      "step": 220
+    },
+    {
+      "epoch": 0.04049102235251008,
+      "grad_norm": 0.5354442317270937,
+      "learning_rate": 4.0293040293040296e-06,
+      "loss": 0.4883,
+      "step": 221
+    },
+    {
+      "epoch": 0.04067423964822279,
+      "grad_norm": 0.4471338769246311,
+      "learning_rate": 4.047619047619048e-06,
+      "loss": 0.5248,
+      "step": 222
+    },
+    {
+      "epoch": 0.040857456943935506,
+      "grad_norm": 0.4728450428012797,
+      "learning_rate": 4.065934065934066e-06,
+      "loss": 0.5412,
+      "step": 223
+    },
+    {
+      "epoch": 0.041040674239648224,
+      "grad_norm": 0.5021628515290991,
+      "learning_rate": 4.084249084249085e-06,
+      "loss": 0.5398,
+      "step": 224
+    },
+    {
+      "epoch": 0.041223891535360935,
+      "grad_norm": 0.379469323319607,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 0.5484,
+      "step": 225
+    },
+    {
+      "epoch": 0.04140710883107365,
+      "grad_norm": 0.4682923741442823,
+      "learning_rate": 4.120879120879121e-06,
+      "loss": 0.5221,
+      "step": 226
+    },
+    {
+      "epoch": 0.04159032612678637,
+      "grad_norm": 0.5298230539031403,
+      "learning_rate": 4.139194139194139e-06,
+      "loss": 0.5248,
+      "step": 227
+    },
+    {
+      "epoch": 0.04177354342249908,
+      "grad_norm": 0.448919135267925,
+      "learning_rate": 4.157509157509158e-06,
+      "loss": 0.5492,
+      "step": 228
+    },
+    {
+      "epoch": 0.0419567607182118,
+      "grad_norm": 0.4651550006247672,
+      "learning_rate": 4.175824175824177e-06,
+      "loss": 0.5244,
+      "step": 229
+    },
+    {
+      "epoch": 0.042139978013924516,
+      "grad_norm": 0.46885786974024124,
+      "learning_rate": 4.1941391941391945e-06,
+      "loss": 0.5145,
+      "step": 230
+    },
+    {
+      "epoch": 0.04232319530963723,
+      "grad_norm": 0.9939588576814024,
+      "learning_rate": 4.212454212454213e-06,
+      "loss": 0.507,
+      "step": 231
+    },
+    {
+      "epoch": 0.042506412605349944,
+      "grad_norm": 0.42445387506164906,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 0.528,
+      "step": 232
+    },
+    {
+      "epoch": 0.04268962990106266,
+      "grad_norm": 0.4386077325175301,
+      "learning_rate": 4.249084249084249e-06,
+      "loss": 0.5235,
+      "step": 233
+    },
+    {
+      "epoch": 0.04287284719677537,
+      "grad_norm": 0.4275787850644743,
+      "learning_rate": 4.267399267399268e-06,
+      "loss": 0.5268,
+      "step": 234
+    },
+    {
+      "epoch": 0.04305606449248809,
+      "grad_norm": 0.3819534900324145,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.5108,
+      "step": 235
+    },
+    {
+      "epoch": 0.04323928178820081,
+      "grad_norm": 0.5069196499403993,
+      "learning_rate": 4.304029304029305e-06,
+      "loss": 0.4914,
+      "step": 236
+    },
+    {
+      "epoch": 0.04342249908391352,
+      "grad_norm": 0.4224671659155117,
+      "learning_rate": 4.322344322344323e-06,
+      "loss": 0.531,
+      "step": 237
+    },
+    {
+      "epoch": 0.043605716379626236,
+      "grad_norm": 0.49892435842001814,
+      "learning_rate": 4.340659340659341e-06,
+      "loss": 0.5309,
+      "step": 238
+    },
+    {
+      "epoch": 0.043788933675338954,
+      "grad_norm": 0.4435906661481072,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 0.5386,
+      "step": 239
+    },
+    {
+      "epoch": 0.043972150971051664,
+      "grad_norm": 0.4646701721197805,
+      "learning_rate": 4.377289377289377e-06,
+      "loss": 0.5445,
+      "step": 240
+    },
+    {
+      "epoch": 0.04415536826676438,
+      "grad_norm": 0.42345412346700445,
+      "learning_rate": 4.395604395604396e-06,
+      "loss": 0.5171,
+      "step": 241
+    },
+    {
+      "epoch": 0.0443385855624771,
+      "grad_norm": 0.4664513677280782,
+      "learning_rate": 4.413919413919414e-06,
+      "loss": 0.5471,
+      "step": 242
+    },
+    {
+      "epoch": 0.04452180285818981,
+      "grad_norm": 0.5007403600133091,
+      "learning_rate": 4.4322344322344325e-06,
+      "loss": 0.5224,
+      "step": 243
+    },
+    {
+      "epoch": 0.04470502015390253,
+      "grad_norm": 0.4202775330369337,
+      "learning_rate": 4.45054945054945e-06,
+      "loss": 0.5283,
+      "step": 244
+    },
+    {
+      "epoch": 0.044888237449615245,
+      "grad_norm": 0.485312343325994,
+      "learning_rate": 4.468864468864469e-06,
+      "loss": 0.5365,
+      "step": 245
+    },
+    {
+      "epoch": 0.045071454745327956,
+      "grad_norm": 0.4796744192562041,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 0.5481,
+      "step": 246
+    },
+    {
+      "epoch": 0.045254672041040674,
+      "grad_norm": 0.4699525411145978,
+      "learning_rate": 4.505494505494506e-06,
+      "loss": 0.526,
+      "step": 247
+    },
+    {
+      "epoch": 0.04543788933675339,
+      "grad_norm": 0.37235134296143163,
+      "learning_rate": 4.523809523809524e-06,
+      "loss": 0.4929,
+      "step": 248
+    },
+    {
+      "epoch": 0.0456211066324661,
+      "grad_norm": 0.4594065464998237,
+      "learning_rate": 4.542124542124542e-06,
+      "loss": 0.5435,
+      "step": 249
+    },
+    {
+      "epoch": 0.04580432392817882,
+      "grad_norm": 0.4812140084395847,
+      "learning_rate": 4.560439560439561e-06,
+      "loss": 0.4714,
+      "step": 250
+    },
+    {
+      "epoch": 0.04598754122389154,
+      "grad_norm": 0.44275937617791644,
+      "learning_rate": 4.578754578754579e-06,
+      "loss": 0.5289,
+      "step": 251
+    },
+    {
+      "epoch": 0.04617075851960425,
+      "grad_norm": 0.543477861772032,
+      "learning_rate": 4.5970695970695975e-06,
+      "loss": 0.5176,
+      "step": 252
+    },
+    {
+      "epoch": 0.046353975815316965,
+      "grad_norm": 0.9195336625704912,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.5315,
+      "step": 253
+    },
+    {
+      "epoch": 0.04653719311102968,
+      "grad_norm": 0.4922315670719196,
+      "learning_rate": 4.633699633699634e-06,
+      "loss": 0.5364,
+      "step": 254
+    },
+    {
+      "epoch": 0.046720410406742394,
+      "grad_norm": 0.47361690353516367,
+      "learning_rate": 4.652014652014653e-06,
+      "loss": 0.5591,
+      "step": 255
+    },
+    {
+      "epoch": 0.04690362770245511,
+      "grad_norm": 0.5164779414047217,
+      "learning_rate": 4.6703296703296706e-06,
+      "loss": 0.5538,
+      "step": 256
+    },
+    {
+      "epoch": 0.04708684499816783,
+      "grad_norm": 0.421397036777767,
+      "learning_rate": 4.688644688644689e-06,
+      "loss": 0.5338,
+      "step": 257
+    },
+    {
+      "epoch": 0.04727006229388054,
+      "grad_norm": 0.5053871259325204,
+      "learning_rate": 4.706959706959707e-06,
+      "loss": 0.544,
+      "step": 258
+    },
+    {
+      "epoch": 0.04745327958959326,
+      "grad_norm": 0.4607447877406368,
+      "learning_rate": 4.725274725274726e-06,
+      "loss": 0.4812,
+      "step": 259
+    },
+    {
+      "epoch": 0.047636496885305975,
+      "grad_norm": 0.4875422302168998,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 0.5614,
+      "step": 260
+    },
+    {
+      "epoch": 0.047819714181018685,
+      "grad_norm": 0.41579068200919733,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 0.4729,
+      "step": 261
+    },
+    {
+      "epoch": 0.0480029314767314,
+      "grad_norm": 0.46282242693556186,
+      "learning_rate": 4.780219780219781e-06,
+      "loss": 0.5224,
+      "step": 262
+    },
+    {
+      "epoch": 0.04818614877244412,
+      "grad_norm": 0.4482830307148575,
+      "learning_rate": 4.798534798534799e-06,
+      "loss": 0.5326,
+      "step": 263
+    },
+    {
+      "epoch": 0.04836936606815683,
+      "grad_norm": 0.4342637761169385,
+      "learning_rate": 4.816849816849818e-06,
+      "loss": 0.5056,
+      "step": 264
+    },
+    {
+      "epoch": 0.04855258336386955,
+      "grad_norm": 0.42342307710917526,
+      "learning_rate": 4.8351648351648355e-06,
+      "loss": 0.5049,
+      "step": 265
+    },
+    {
+      "epoch": 0.04873580065958227,
+      "grad_norm": 0.41567729548709964,
+      "learning_rate": 4.853479853479854e-06,
+      "loss": 0.5313,
+      "step": 266
+    },
+    {
+      "epoch": 0.04891901795529498,
+      "grad_norm": 0.44877730848158315,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 0.5027,
+      "step": 267
+    },
+    {
+      "epoch": 0.049102235251007695,
+      "grad_norm": 0.46298601715996757,
+      "learning_rate": 4.890109890109891e-06,
+      "loss": 0.5418,
+      "step": 268
+    },
+    {
+      "epoch": 0.04928545254672041,
+      "grad_norm": 0.4233152082129357,
+      "learning_rate": 4.908424908424909e-06,
+      "loss": 0.534,
+      "step": 269
+    },
+    {
+      "epoch": 0.04946866984243312,
+      "grad_norm": 0.42327618076780654,
+      "learning_rate": 4.926739926739927e-06,
+      "loss": 0.5073,
+      "step": 270
+    },
+    {
+      "epoch": 0.04965188713814584,
+      "grad_norm": 0.47132160003804374,
+      "learning_rate": 4.945054945054946e-06,
+      "loss": 0.564,
+      "step": 271
+    },
+    {
+      "epoch": 0.04983510443385856,
+      "grad_norm": 0.467395758671848,
+      "learning_rate": 4.963369963369964e-06,
+      "loss": 0.5194,
+      "step": 272
+    },
+    {
+      "epoch": 0.05001832172957127,
+      "grad_norm": 0.4377241688268797,
+      "learning_rate": 4.9816849816849826e-06,
+      "loss": 0.5473,
+      "step": 273
+    },
+    {
+      "epoch": 0.05020153902528399,
+      "grad_norm": 0.4344310773487788,
+      "learning_rate": 5e-06,
+      "loss": 0.5198,
+      "step": 274
+    },
+    {
+      "epoch": 0.050384756320996704,
+      "grad_norm": 0.505691771505538,
+      "learning_rate": 5.018315018315018e-06,
+      "loss": 0.5001,
+      "step": 275
+    },
+    {
+      "epoch": 0.050567973616709415,
+      "grad_norm": 0.4741702696479342,
+      "learning_rate": 5.036630036630037e-06,
+      "loss": 0.5269,
+      "step": 276
+    },
+    {
+      "epoch": 0.05075119091242213,
+      "grad_norm": 0.4123760853657366,
+      "learning_rate": 5.054945054945055e-06,
+      "loss": 0.5366,
+      "step": 277
+    },
+    {
+      "epoch": 0.05093440820813485,
+      "grad_norm": 0.5165952852785715,
+      "learning_rate": 5.0732600732600735e-06,
+      "loss": 0.5629,
+      "step": 278
+    },
+    {
+      "epoch": 0.05111762550384756,
+      "grad_norm": 0.6018743338441076,
+      "learning_rate": 5.091575091575091e-06,
+      "loss": 0.5268,
+      "step": 279
+    },
+    {
+      "epoch": 0.05130084279956028,
+      "grad_norm": 0.4647905088113548,
+      "learning_rate": 5.10989010989011e-06,
+      "loss": 0.5353,
+      "step": 280
+    },
+    {
+      "epoch": 0.051484060095272996,
+      "grad_norm": 0.4666862939383661,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.5582,
+      "step": 281
+    },
+    {
+      "epoch": 0.05166727739098571,
+      "grad_norm": 0.42295571485014016,
+      "learning_rate": 5.146520146520147e-06,
+      "loss": 0.5389,
+      "step": 282
+    },
+    {
+      "epoch": 0.051850494686698424,
+      "grad_norm": 0.5088982589681916,
+      "learning_rate": 5.164835164835166e-06,
+      "loss": 0.5114,
+      "step": 283
+    },
+    {
+      "epoch": 0.05203371198241114,
+      "grad_norm": 0.39252264391052066,
+      "learning_rate": 5.183150183150184e-06,
+      "loss": 0.5078,
+      "step": 284
+    },
+    {
+      "epoch": 0.05221692927812385,
+      "grad_norm": 0.47464881831711925,
+      "learning_rate": 5.201465201465202e-06,
+      "loss": 0.5719,
+      "step": 285
+    },
+    {
+      "epoch": 0.05240014657383657,
+      "grad_norm": 0.42811181510690394,
+      "learning_rate": 5.219780219780221e-06,
+      "loss": 0.536,
+      "step": 286
+    },
+    {
+      "epoch": 0.05258336386954929,
+      "grad_norm": 0.4983761768332983,
+      "learning_rate": 5.2380952380952384e-06,
+      "loss": 0.5454,
+      "step": 287
+    },
+    {
+      "epoch": 0.052766581165262,
+      "grad_norm": 0.3871752405078846,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 0.5172,
+      "step": 288
+    },
+    {
+      "epoch": 0.052949798460974716,
+      "grad_norm": 0.44878294094458826,
+      "learning_rate": 5.274725274725275e-06,
+      "loss": 0.4965,
+      "step": 289
+    },
+    {
+      "epoch": 0.053133015756687434,
+      "grad_norm": 0.4193937449864018,
+      "learning_rate": 5.293040293040294e-06,
+      "loss": 0.5418,
+      "step": 290
+    },
+    {
+      "epoch": 0.053316233052400144,
+      "grad_norm": 0.47326184468203625,
+      "learning_rate": 5.3113553113553116e-06,
+      "loss": 0.5465,
+      "step": 291
+    },
+    {
+      "epoch": 0.05349945034811286,
+      "grad_norm": 0.46890596965933473,
+      "learning_rate": 5.32967032967033e-06,
+      "loss": 0.5427,
+      "step": 292
+    },
+    {
+      "epoch": 0.05368266764382558,
+      "grad_norm": 0.4718404244115825,
+      "learning_rate": 5.347985347985348e-06,
+      "loss": 0.5315,
+      "step": 293
+    },
+    {
+      "epoch": 0.05386588493953829,
+      "grad_norm": 0.5313479853203268,
+      "learning_rate": 5.366300366300367e-06,
+      "loss": 0.5261,
+      "step": 294
+    },
+    {
+      "epoch": 0.05404910223525101,
+      "grad_norm": 0.4619862699623299,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.5224,
+      "step": 295
+    },
+    {
+      "epoch": 0.054232319530963725,
+      "grad_norm": 0.4235637361399875,
+      "learning_rate": 5.402930402930403e-06,
+      "loss": 0.5484,
+      "step": 296
+    },
+    {
+      "epoch": 0.054415536826676436,
+      "grad_norm": 0.4826898937600368,
+      "learning_rate": 5.421245421245421e-06,
+      "loss": 0.524,
+      "step": 297
+    },
+    {
+      "epoch": 0.054598754122389154,
+      "grad_norm": 0.43904512964940123,
+      "learning_rate": 5.43956043956044e-06,
+      "loss": 0.5486,
+      "step": 298
+    },
+    {
+      "epoch": 0.05478197141810187,
+      "grad_norm": 0.5045757484045217,
+      "learning_rate": 5.457875457875458e-06,
+      "loss": 0.5407,
+      "step": 299
+    },
+    {
+      "epoch": 0.05496518871381458,
+      "grad_norm": 0.47829971819207484,
+      "learning_rate": 5.476190476190477e-06,
+      "loss": 0.5344,
+      "step": 300
+    },
+    {
+      "epoch": 0.0551484060095273,
+      "grad_norm": 0.416644246441645,
+      "learning_rate": 5.494505494505495e-06,
+      "loss": 0.5111,
+      "step": 301
+    },
+    {
+      "epoch": 0.05533162330524002,
+      "grad_norm": 0.488275746902462,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 0.54,
+      "step": 302
+    },
+    {
+      "epoch": 0.05551484060095273,
+      "grad_norm": 0.43082352297647686,
+      "learning_rate": 5.531135531135532e-06,
+      "loss": 0.5219,
+      "step": 303
+    },
+    {
+      "epoch": 0.055698057896665445,
+      "grad_norm": 0.41708996725660685,
+      "learning_rate": 5.5494505494505504e-06,
+      "loss": 0.5272,
+      "step": 304
+    },
+    {
+      "epoch": 0.05588127519237816,
+      "grad_norm": 0.4748217492221608,
+      "learning_rate": 5.567765567765568e-06,
+      "loss": 0.5439,
+      "step": 305
+    },
+    {
+      "epoch": 0.056064492488090874,
+      "grad_norm": 0.5257169187612324,
+      "learning_rate": 5.586080586080587e-06,
+      "loss": 0.5671,
+      "step": 306
+    },
+    {
+      "epoch": 0.05624770978380359,
+      "grad_norm": 0.4243472668028098,
+      "learning_rate": 5.604395604395605e-06,
+      "loss": 0.5044,
+      "step": 307
+    },
+    {
+      "epoch": 0.05643092707951631,
+      "grad_norm": 0.43876355547814727,
+      "learning_rate": 5.6227106227106235e-06,
+      "loss": 0.4815,
+      "step": 308
+    },
+    {
+      "epoch": 0.05661414437522902,
+      "grad_norm": 0.43342398130791976,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 0.5308,
+      "step": 309
+    },
+    {
+      "epoch": 0.05679736167094174,
+      "grad_norm": 0.4660705177035744,
+      "learning_rate": 5.65934065934066e-06,
+      "loss": 0.5686,
+      "step": 310
+    },
+    {
+      "epoch": 0.056980578966654455,
+      "grad_norm": 0.4170384834874546,
+      "learning_rate": 5.677655677655678e-06,
+      "loss": 0.4965,
+      "step": 311
+    },
+    {
+      "epoch": 0.057163796262367166,
+      "grad_norm": 0.4568771189697619,
+      "learning_rate": 5.695970695970696e-06,
+      "loss": 0.5074,
+      "step": 312
+    },
+    {
+      "epoch": 0.05734701355807988,
+      "grad_norm": 0.4666988354433752,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.5225,
+      "step": 313
+    },
+    {
+      "epoch": 0.0575302308537926,
+      "grad_norm": 0.38305953336818827,
+      "learning_rate": 5.732600732600732e-06,
+      "loss": 0.5057,
+      "step": 314
+    },
+    {
+      "epoch": 0.05771344814950531,
+      "grad_norm": 0.8971752127635111,
+      "learning_rate": 5.750915750915751e-06,
+      "loss": 0.5556,
+      "step": 315
+    },
+    {
+      "epoch": 0.05789666544521803,
+      "grad_norm": 0.4580428245754137,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 0.5344,
+      "step": 316
+    },
+    {
+      "epoch": 0.05807988274093075,
+      "grad_norm": 0.414647309534276,
+      "learning_rate": 5.7875457875457885e-06,
+      "loss": 0.5177,
+      "step": 317
+    },
+    {
+      "epoch": 0.05826310003664346,
+      "grad_norm": 0.45495854796733415,
+      "learning_rate": 5.805860805860807e-06,
+      "loss": 0.5214,
+      "step": 318
+    },
+    {
+      "epoch": 0.058446317332356175,
+      "grad_norm": 0.47098300998769715,
+      "learning_rate": 5.824175824175825e-06,
+      "loss": 0.5519,
+      "step": 319
+    },
+    {
+      "epoch": 0.05862953462806889,
+      "grad_norm": 0.43335718768494447,
+      "learning_rate": 5.842490842490844e-06,
+      "loss": 0.5282,
+      "step": 320
+    },
+    {
+      "epoch": 0.0588127519237816,
+      "grad_norm": 0.4649559782288625,
+      "learning_rate": 5.860805860805862e-06,
+      "loss": 0.5163,
+      "step": 321
+    },
+    {
+      "epoch": 0.05899596921949432,
+      "grad_norm": 0.44623794630893965,
+      "learning_rate": 5.8791208791208794e-06,
+      "loss": 0.514,
+      "step": 322
+    },
+    {
+      "epoch": 0.05917918651520704,
+      "grad_norm": 0.4521538486948526,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 0.5248,
+      "step": 323
+    },
+    {
+      "epoch": 0.05936240381091975,
+      "grad_norm": 0.453853173278213,
+      "learning_rate": 5.915750915750916e-06,
+      "loss": 0.5105,
+      "step": 324
+    },
+    {
+      "epoch": 0.05954562110663247,
+      "grad_norm": 0.45615939787874893,
+      "learning_rate": 5.934065934065935e-06,
+      "loss": 0.5321,
+      "step": 325
+    },
+    {
+      "epoch": 0.059728838402345184,
+      "grad_norm": 0.5103507598238842,
+      "learning_rate": 5.9523809523809525e-06,
+      "loss": 0.4889,
+      "step": 326
+    },
+    {
+      "epoch": 0.059912055698057895,
+      "grad_norm": 0.4755868373014201,
+      "learning_rate": 5.970695970695971e-06,
+      "loss": 0.5233,
+      "step": 327
+    },
+    {
+      "epoch": 0.06009527299377061,
+      "grad_norm": 0.48186711326734255,
+      "learning_rate": 5.989010989010989e-06,
+      "loss": 0.5216,
+      "step": 328
+    },
+    {
+      "epoch": 0.06027849028948333,
+      "grad_norm": 0.406290487409726,
+      "learning_rate": 6.007326007326008e-06,
+      "loss": 0.5048,
+      "step": 329
+    },
+    {
+      "epoch": 0.06046170758519604,
+      "grad_norm": 0.47965200235867606,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 0.513,
+      "step": 330
+    },
+    {
+      "epoch": 0.06064492488090876,
+      "grad_norm": 0.492103246179344,
+      "learning_rate": 6.043956043956044e-06,
+      "loss": 0.5553,
+      "step": 331
+    },
+    {
+      "epoch": 0.060828142176621476,
+      "grad_norm": 0.49171785130479284,
+      "learning_rate": 6.062271062271062e-06,
+      "loss": 0.5289,
+      "step": 332
+    },
+    {
+      "epoch": 0.06101135947233419,
+      "grad_norm": 0.4659535600554528,
+      "learning_rate": 6.080586080586081e-06,
+      "loss": 0.5143,
+      "step": 333
+    },
+    {
+      "epoch": 0.061194576768046904,
+      "grad_norm": 0.43136677550682173,
+      "learning_rate": 6.0989010989011e-06,
+      "loss": 0.469,
+      "step": 334
+    },
+    {
+      "epoch": 0.06137779406375962,
+      "grad_norm": 0.6094192173431031,
+      "learning_rate": 6.117216117216118e-06,
+      "loss": 0.5121,
+      "step": 335
+    },
+    {
+      "epoch": 0.06156101135947233,
+      "grad_norm": 0.45352827270619606,
+      "learning_rate": 6.135531135531136e-06,
+      "loss": 0.5097,
+      "step": 336
+    },
+    {
+      "epoch": 0.06174422865518505,
+      "grad_norm": 0.42626268219917746,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.5234,
+      "step": 337
+    },
+    {
+      "epoch": 0.06192744595089777,
+      "grad_norm": 0.4438479757326601,
+      "learning_rate": 6.172161172161173e-06,
+      "loss": 0.4852,
+      "step": 338
+    },
+    {
+      "epoch": 0.06211066324661048,
+      "grad_norm": 0.688188614062373,
+      "learning_rate": 6.1904761904761914e-06,
+      "loss": 0.502,
+      "step": 339
+    },
+    {
+      "epoch": 0.062293880542323196,
+      "grad_norm": 0.5160627484540854,
+      "learning_rate": 6.208791208791209e-06,
+      "loss": 0.521,
+      "step": 340
+    },
+    {
+      "epoch": 0.062477097838035914,
+      "grad_norm": 0.4356067335955304,
+      "learning_rate": 6.227106227106228e-06,
+      "loss": 0.4918,
+      "step": 341
+    },
+    {
+      "epoch": 0.06266031513374863,
+      "grad_norm": 0.5001791021027777,
+      "learning_rate": 6.245421245421246e-06,
+      "loss": 0.5173,
+      "step": 342
+    },
+    {
+      "epoch": 0.06284353242946134,
+      "grad_norm": 0.45621279623031163,
+      "learning_rate": 6.2637362637362645e-06,
+      "loss": 0.5532,
+      "step": 343
+    },
+    {
+      "epoch": 0.06302674972517405,
+      "grad_norm": 0.4686583276600699,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 0.5544,
+      "step": 344
+    },
+    {
+      "epoch": 0.06320996702088677,
+      "grad_norm": 0.4500457038114704,
+      "learning_rate": 6.300366300366301e-06,
+      "loss": 0.5094,
+      "step": 345
+    },
+    {
+      "epoch": 0.06339318431659949,
+      "grad_norm": 0.4707435610591325,
+      "learning_rate": 6.318681318681319e-06,
+      "loss": 0.5317,
+      "step": 346
+    },
+    {
+      "epoch": 0.0635764016123122,
+      "grad_norm": 0.44910822533973516,
+      "learning_rate": 6.336996336996337e-06,
+      "loss": 0.5063,
+      "step": 347
+    },
+    {
+      "epoch": 0.06375961890802492,
+      "grad_norm": 0.4262957774336457,
+      "learning_rate": 6.3553113553113555e-06,
+      "loss": 0.5389,
+      "step": 348
+    },
+    {
+      "epoch": 0.06394283620373763,
+      "grad_norm": 0.4958715530578741,
+      "learning_rate": 6.373626373626373e-06,
+      "loss": 0.5253,
+      "step": 349
+    },
+    {
+      "epoch": 0.06412605349945034,
+      "grad_norm": 0.4706503863353741,
+      "learning_rate": 6.391941391941392e-06,
+      "loss": 0.5078,
+      "step": 350
+    },
+    {
+      "epoch": 0.06430927079516306,
+      "grad_norm": 0.5025375698465898,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 0.5366,
+      "step": 351
+    },
+    {
+      "epoch": 0.06449248809087578,
+      "grad_norm": 0.43307402952194485,
+      "learning_rate": 6.4285714285714295e-06,
+      "loss": 0.5053,
+      "step": 352
+    },
+    {
+      "epoch": 0.0646757053865885,
+      "grad_norm": 0.4145957842766705,
+      "learning_rate": 6.446886446886448e-06,
+      "loss": 0.5134,
+      "step": 353
+    },
+    {
+      "epoch": 0.06485892268230121,
+      "grad_norm": 0.42558146560441634,
+      "learning_rate": 6.465201465201466e-06,
+      "loss": 0.498,
+      "step": 354
+    },
+    {
+      "epoch": 0.06504213997801392,
+      "grad_norm": 0.6808736114735602,
+      "learning_rate": 6.483516483516485e-06,
+      "loss": 0.4865,
+      "step": 355
+    },
+    {
+      "epoch": 0.06522535727372664,
+      "grad_norm": 0.4858578720351213,
+      "learning_rate": 6.5018315018315026e-06,
+      "loss": 0.4977,
+      "step": 356
+    },
+    {
+      "epoch": 0.06540857456943935,
+      "grad_norm": 0.5523209851617275,
+      "learning_rate": 6.5201465201465204e-06,
+      "loss": 0.5278,
+      "step": 357
+    },
+    {
+      "epoch": 0.06559179186515207,
+      "grad_norm": 0.4756243917905379,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 0.5018,
+      "step": 358
+    },
+    {
+      "epoch": 0.06577500916086479,
+      "grad_norm": 0.5270789359516691,
+      "learning_rate": 6.556776556776557e-06,
+      "loss": 0.5127,
+      "step": 359
+    },
+    {
+      "epoch": 0.0659582264565775,
+      "grad_norm": 0.5213313487503423,
+      "learning_rate": 6.575091575091576e-06,
+      "loss": 0.504,
+      "step": 360
+    },
+    {
+      "epoch": 0.06614144375229021,
+      "grad_norm": 0.47699985237826076,
+      "learning_rate": 6.5934065934065935e-06,
+      "loss": 0.5356,
+      "step": 361
+    },
+    {
+      "epoch": 0.06632466104800293,
+      "grad_norm": 0.4398115760336562,
+      "learning_rate": 6.611721611721612e-06,
+      "loss": 0.5038,
+      "step": 362
+    },
+    {
+      "epoch": 0.06650787834371565,
+      "grad_norm": 0.5848808647247892,
+      "learning_rate": 6.63003663003663e-06,
+      "loss": 0.5393,
+      "step": 363
+    },
+    {
+      "epoch": 0.06669109563942836,
+      "grad_norm": 0.40297449884724584,
+      "learning_rate": 6.648351648351649e-06,
+      "loss": 0.5231,
+      "step": 364
+    },
+    {
+      "epoch": 0.06687431293514108,
+      "grad_norm": 0.4615954062532406,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.5017,
+      "step": 365
+    },
+    {
+      "epoch": 0.0670575302308538,
+      "grad_norm": 0.4204668889117905,
+      "learning_rate": 6.684981684981685e-06,
+      "loss": 0.5651,
+      "step": 366
+    },
+    {
+      "epoch": 0.0672407475265665,
+      "grad_norm": 0.4765525618556606,
+      "learning_rate": 6.703296703296703e-06,
+      "loss": 0.5253,
+      "step": 367
+    },
+    {
+      "epoch": 0.06742396482227922,
+      "grad_norm": 0.431663431335195,
+      "learning_rate": 6.721611721611723e-06,
+      "loss": 0.5398,
+      "step": 368
+    },
+    {
+      "epoch": 0.06760718211799194,
+      "grad_norm": 0.4403652862248983,
+      "learning_rate": 6.739926739926741e-06,
+      "loss": 0.5252,
+      "step": 369
+    },
+    {
+      "epoch": 0.06779039941370466,
+      "grad_norm": 0.483725795531489,
+      "learning_rate": 6.758241758241759e-06,
+      "loss": 0.5349,
+      "step": 370
+    },
+    {
+      "epoch": 0.06797361670941737,
+      "grad_norm": 0.8892216780264888,
+      "learning_rate": 6.776556776556777e-06,
+      "loss": 0.5184,
+      "step": 371
+    },
+    {
+      "epoch": 0.06815683400513009,
+      "grad_norm": 0.48775432933817636,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 0.5275,
+      "step": 372
+    },
+    {
+      "epoch": 0.0683400513008428,
+      "grad_norm": 0.4772597017011689,
+      "learning_rate": 6.813186813186814e-06,
+      "loss": 0.5517,
+      "step": 373
+    },
+    {
+      "epoch": 0.06852326859655551,
+      "grad_norm": 0.46418441159832313,
+      "learning_rate": 6.831501831501832e-06,
+      "loss": 0.5429,
+      "step": 374
+    },
+    {
+      "epoch": 0.06870648589226823,
+      "grad_norm": 0.45474945807496187,
+      "learning_rate": 6.84981684981685e-06,
+      "loss": 0.4999,
+      "step": 375
+    },
+    {
+      "epoch": 0.06888970318798095,
+      "grad_norm": 0.5576640330118253,
+      "learning_rate": 6.868131868131869e-06,
+      "loss": 0.5198,
+      "step": 376
+    },
+    {
+      "epoch": 0.06907292048369366,
+      "grad_norm": 0.4921371781399045,
+      "learning_rate": 6.886446886446887e-06,
+      "loss": 0.5379,
+      "step": 377
+    },
+    {
+      "epoch": 0.06925613777940638,
+      "grad_norm": 0.41639097386056484,
+      "learning_rate": 6.9047619047619055e-06,
+      "loss": 0.519,
+      "step": 378
+    },
+    {
+      "epoch": 0.06943935507511909,
+      "grad_norm": 0.4778909271586564,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.5336,
+      "step": 379
+    },
+    {
+      "epoch": 0.0696225723708318,
+      "grad_norm": 0.4735414706699025,
+      "learning_rate": 6.941391941391942e-06,
+      "loss": 0.5,
+      "step": 380
+    },
+    {
+      "epoch": 0.06980578966654452,
+      "grad_norm": 0.5646322357431656,
+      "learning_rate": 6.95970695970696e-06,
+      "loss": 0.5073,
+      "step": 381
+    },
+    {
+      "epoch": 0.06998900696225724,
+      "grad_norm": 0.48231469572609614,
+      "learning_rate": 6.978021978021979e-06,
+      "loss": 0.519,
+      "step": 382
+    },
+    {
+      "epoch": 0.07017222425796996,
+      "grad_norm": 0.5051213072897166,
+      "learning_rate": 6.9963369963369965e-06,
+      "loss": 0.5315,
+      "step": 383
+    },
+    {
+      "epoch": 0.07035544155368267,
+      "grad_norm": 0.5166250853237789,
+      "learning_rate": 7.014652014652014e-06,
+      "loss": 0.4978,
+      "step": 384
+    },
+    {
+      "epoch": 0.07053865884939538,
+      "grad_norm": 0.43900982727967125,
+      "learning_rate": 7.032967032967034e-06,
+      "loss": 0.4913,
+      "step": 385
+    },
+    {
+      "epoch": 0.0707218761451081,
+      "grad_norm": 0.5262819691743885,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 0.5189,
+      "step": 386
+    },
+    {
+      "epoch": 0.07090509344082081,
+      "grad_norm": 0.4906825009780104,
+      "learning_rate": 7.0695970695970705e-06,
+      "loss": 0.5586,
+      "step": 387
+    },
+    {
+      "epoch": 0.07108831073653353,
+      "grad_norm": 0.48825547184252527,
+      "learning_rate": 7.087912087912089e-06,
+      "loss": 0.5218,
+      "step": 388
+    },
+    {
+      "epoch": 0.07127152803224625,
+      "grad_norm": 0.5005437189224704,
+      "learning_rate": 7.106227106227107e-06,
+      "loss": 0.5088,
+      "step": 389
+    },
+    {
+      "epoch": 0.07145474532795897,
+      "grad_norm": 0.4820090004987099,
+      "learning_rate": 7.124542124542126e-06,
+      "loss": 0.5132,
+      "step": 390
+    },
+    {
+      "epoch": 0.07163796262367167,
+      "grad_norm": 0.46585258023856246,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.5198,
+      "step": 391
+    },
+    {
+      "epoch": 0.07182117991938439,
+      "grad_norm": 0.5351346796257885,
+      "learning_rate": 7.161172161172162e-06,
+      "loss": 0.5174,
+      "step": 392
+    },
+    {
+      "epoch": 0.0720043972150971,
+      "grad_norm": 0.505980784032641,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.517,
+      "step": 393
+    },
+    {
+      "epoch": 0.07218761451080982,
+      "grad_norm": 0.3991493735258601,
+      "learning_rate": 7.197802197802198e-06,
+      "loss": 0.5193,
+      "step": 394
+    },
+    {
+      "epoch": 0.07237083180652254,
+      "grad_norm": 0.47050897190858826,
+      "learning_rate": 7.216117216117217e-06,
+      "loss": 0.5278,
+      "step": 395
+    },
+    {
+      "epoch": 0.07255404910223526,
+      "grad_norm": 0.42709223237684557,
+      "learning_rate": 7.2344322344322345e-06,
+      "loss": 0.5233,
+      "step": 396
+    },
+    {
+      "epoch": 0.07273726639794796,
+      "grad_norm": 0.40701893214503493,
+      "learning_rate": 7.252747252747253e-06,
+      "loss": 0.4594,
+      "step": 397
+    },
+    {
+      "epoch": 0.07292048369366068,
+      "grad_norm": 0.4793723507064191,
+      "learning_rate": 7.271062271062271e-06,
+      "loss": 0.5022,
+      "step": 398
+    },
+    {
+      "epoch": 0.0731037009893734,
+      "grad_norm": 0.4084855491847111,
+      "learning_rate": 7.28937728937729e-06,
+      "loss": 0.5177,
+      "step": 399
+    },
+    {
+      "epoch": 0.07328691828508611,
+      "grad_norm": 0.48791204362548757,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 0.5336,
+      "step": 400
+    },
+    {
+      "epoch": 0.07347013558079883,
+      "grad_norm": 0.4721917787188149,
+      "learning_rate": 7.326007326007326e-06,
+      "loss": 0.5291,
+      "step": 401
+    },
+    {
+      "epoch": 0.07365335287651155,
+      "grad_norm": 0.4482564665450733,
+      "learning_rate": 7.344322344322346e-06,
+      "loss": 0.516,
+      "step": 402
+    },
+    {
+      "epoch": 0.07383657017222425,
+      "grad_norm": 0.503350703275625,
+      "learning_rate": 7.362637362637364e-06,
+      "loss": 0.5232,
+      "step": 403
+    },
+    {
+      "epoch": 0.07401978746793697,
+      "grad_norm": 0.5152368271826678,
+      "learning_rate": 7.380952380952382e-06,
+      "loss": 0.5258,
+      "step": 404
+    },
+    {
+      "epoch": 0.07420300476364969,
+      "grad_norm": 0.4523690267979874,
+      "learning_rate": 7.3992673992674e-06,
+      "loss": 0.5007,
+      "step": 405
+    },
+    {
+      "epoch": 0.0743862220593624,
+      "grad_norm": 0.44122338596098665,
+      "learning_rate": 7.417582417582418e-06,
+      "loss": 0.5325,
+      "step": 406
+    },
+    {
+      "epoch": 0.07456943935507512,
+      "grad_norm": 0.418021168922302,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 0.5048,
+      "step": 407
+    },
+    {
+      "epoch": 0.07475265665078784,
+      "grad_norm": 0.44537098554090276,
+      "learning_rate": 7.454212454212455e-06,
+      "loss": 0.5294,
+      "step": 408
+    },
+    {
+      "epoch": 0.07493587394650054,
+      "grad_norm": 0.41384101894590325,
+      "learning_rate": 7.472527472527473e-06,
+      "loss": 0.5119,
+      "step": 409
+    },
+    {
+      "epoch": 0.07511909124221326,
+      "grad_norm": 0.4647470180215214,
+      "learning_rate": 7.490842490842491e-06,
+      "loss": 0.5202,
+      "step": 410
+    },
+    {
+      "epoch": 0.07530230853792598,
+      "grad_norm": 0.4836409216489063,
+      "learning_rate": 7.50915750915751e-06,
+      "loss": 0.5056,
+      "step": 411
+    },
+    {
+      "epoch": 0.0754855258336387,
+      "grad_norm": 0.45195196209865357,
+      "learning_rate": 7.527472527472528e-06,
+      "loss": 0.5213,
+      "step": 412
+    },
+    {
+      "epoch": 0.07566874312935142,
+      "grad_norm": 0.5036829282617328,
+      "learning_rate": 7.5457875457875465e-06,
+      "loss": 0.5347,
+      "step": 413
+    },
+    {
+      "epoch": 0.07585196042506413,
+      "grad_norm": 0.4666038283310089,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 0.5151,
+      "step": 414
+    },
+    {
+      "epoch": 0.07603517772077684,
+      "grad_norm": 0.5067769344340564,
+      "learning_rate": 7.582417582417583e-06,
+      "loss": 0.5171,
+      "step": 415
+    },
+    {
+      "epoch": 0.07621839501648955,
+      "grad_norm": 0.4093230258713508,
+      "learning_rate": 7.600732600732601e-06,
+      "loss": 0.5039,
+      "step": 416
+    },
+    {
+      "epoch": 0.07640161231220227,
+      "grad_norm": 0.6498006485071868,
+      "learning_rate": 7.61904761904762e-06,
+      "loss": 0.5232,
+      "step": 417
+    },
+    {
+      "epoch": 0.07658482960791499,
+      "grad_norm": 0.4558152414900924,
+      "learning_rate": 7.637362637362638e-06,
+      "loss": 0.5232,
+      "step": 418
+    },
+    {
+      "epoch": 0.0767680469036277,
+      "grad_norm": 0.4804930639066785,
+      "learning_rate": 7.655677655677656e-06,
+      "loss": 0.5088,
+      "step": 419
+    },
+    {
+      "epoch": 0.07695126419934042,
+      "grad_norm": 0.5081120095319347,
+      "learning_rate": 7.673992673992676e-06,
+      "loss": 0.5308,
+      "step": 420
+    },
+    {
+      "epoch": 0.07713448149505313,
+      "grad_norm": 0.42393212923875745,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.5084,
+      "step": 421
+    },
+    {
+      "epoch": 0.07731769879076585,
+      "grad_norm": 0.44672075335716155,
+      "learning_rate": 7.710622710622711e-06,
+      "loss": 0.483,
+      "step": 422
+    },
+    {
+      "epoch": 0.07750091608647856,
+      "grad_norm": 0.4942211324726324,
+      "learning_rate": 7.72893772893773e-06,
+      "loss": 0.5236,
+      "step": 423
+    },
+    {
+      "epoch": 0.07768413338219128,
+      "grad_norm": 0.39817279842841763,
+      "learning_rate": 7.747252747252749e-06,
+      "loss": 0.5332,
+      "step": 424
+    },
+    {
+      "epoch": 0.077867350677904,
+      "grad_norm": 0.4608671155196199,
+      "learning_rate": 7.765567765567767e-06,
+      "loss": 0.523,
+      "step": 425
+    },
+    {
+      "epoch": 0.07805056797361672,
+      "grad_norm": 0.5225991254379231,
+      "learning_rate": 7.783882783882785e-06,
+      "loss": 0.5171,
+      "step": 426
+    },
+    {
+      "epoch": 0.07823378526932942,
+      "grad_norm": 0.5423490196220466,
+      "learning_rate": 7.802197802197802e-06,
+      "loss": 0.4985,
+      "step": 427
+    },
+    {
+      "epoch": 0.07841700256504214,
+      "grad_norm": 0.502419506962013,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 0.5087,
+      "step": 428
+    },
+    {
+      "epoch": 0.07860021986075486,
+      "grad_norm": 0.5102984182578104,
+      "learning_rate": 7.83882783882784e-06,
+      "loss": 0.5358,
+      "step": 429
+    },
+    {
+      "epoch": 0.07878343715646757,
+      "grad_norm": 0.4742126281062651,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 0.5585,
+      "step": 430
+    },
+    {
+      "epoch": 0.07896665445218029,
+      "grad_norm": 0.508787409495283,
+      "learning_rate": 7.875457875457876e-06,
+      "loss": 0.5411,
+      "step": 431
+    },
+    {
+      "epoch": 0.079149871747893,
+      "grad_norm": 0.4559258370924022,
+      "learning_rate": 7.893772893772893e-06,
+      "loss": 0.5229,
+      "step": 432
+    },
+    {
+      "epoch": 0.07933308904360571,
+      "grad_norm": 0.6320276655690735,
+      "learning_rate": 7.912087912087913e-06,
+      "loss": 0.4688,
+      "step": 433
+    },
+    {
+      "epoch": 0.07951630633931843,
+      "grad_norm": 0.9797165625621964,
+      "learning_rate": 7.93040293040293e-06,
+      "loss": 0.5372,
+      "step": 434
+    },
+    {
+      "epoch": 0.07969952363503115,
+      "grad_norm": 0.43648526458810954,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 0.4528,
+      "step": 435
+    },
+    {
+      "epoch": 0.07988274093074386,
+      "grad_norm": 0.46101717931557235,
+      "learning_rate": 7.967032967032966e-06,
+      "loss": 0.4836,
+      "step": 436
+    },
+    {
+      "epoch": 0.08006595822645658,
+      "grad_norm": 0.46019219132505085,
+      "learning_rate": 7.985347985347986e-06,
+      "loss": 0.4874,
+      "step": 437
+    },
+    {
+      "epoch": 0.08024917552216929,
+      "grad_norm": 0.4228905699110607,
+      "learning_rate": 8.003663003663006e-06,
+      "loss": 0.483,
+      "step": 438
+    },
+    {
+      "epoch": 0.080432392817882,
+      "grad_norm": 0.4920935432281776,
+      "learning_rate": 8.021978021978023e-06,
+      "loss": 0.489,
+      "step": 439
+    },
+    {
+      "epoch": 0.08061561011359472,
+      "grad_norm": 0.45385267234753407,
+      "learning_rate": 8.040293040293041e-06,
+      "loss": 0.5262,
+      "step": 440
+    },
+    {
+      "epoch": 0.08079882740930744,
+      "grad_norm": 0.48390990285633756,
+      "learning_rate": 8.058608058608059e-06,
+      "loss": 0.4969,
+      "step": 441
+    },
+    {
+      "epoch": 0.08098204470502016,
+      "grad_norm": 0.4424118373260341,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 0.5217,
+      "step": 442
+    },
+    {
+      "epoch": 0.08116526200073287,
+      "grad_norm": 0.4647327426734162,
+      "learning_rate": 8.095238095238097e-06,
+      "loss": 0.5103,
+      "step": 443
+    },
+    {
+      "epoch": 0.08134847929644558,
+      "grad_norm": 0.42019112334691044,
+      "learning_rate": 8.113553113553114e-06,
+      "loss": 0.5378,
+      "step": 444
+    },
+    {
+      "epoch": 0.0815316965921583,
+      "grad_norm": 0.40127483204178854,
+      "learning_rate": 8.131868131868132e-06,
+      "loss": 0.4813,
+      "step": 445
+    },
+    {
+      "epoch": 0.08171491388787101,
+      "grad_norm": 0.4606579465800615,
+      "learning_rate": 8.15018315018315e-06,
+      "loss": 0.5067,
+      "step": 446
+    },
+    {
+      "epoch": 0.08189813118358373,
+      "grad_norm": 0.4715535246345655,
+      "learning_rate": 8.16849816849817e-06,
+      "loss": 0.5181,
+      "step": 447
+    },
+    {
+      "epoch": 0.08208134847929645,
+      "grad_norm": 0.5386430386856308,
+      "learning_rate": 8.186813186813188e-06,
+      "loss": 0.5188,
+      "step": 448
+    },
+    {
+      "epoch": 0.08226456577500917,
+      "grad_norm": 0.5389030998275688,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.516,
+      "step": 449
+    },
+    {
+      "epoch": 0.08244778307072187,
+      "grad_norm": 0.5181954163327931,
+      "learning_rate": 8.223443223443223e-06,
+      "loss": 0.5099,
+      "step": 450
+    },
+    {
+      "epoch": 0.08263100036643459,
+      "grad_norm": 0.5425277965290457,
+      "learning_rate": 8.241758241758243e-06,
+      "loss": 0.5508,
+      "step": 451
+    },
+    {
+      "epoch": 0.0828142176621473,
+      "grad_norm": 0.4366152129754919,
+      "learning_rate": 8.26007326007326e-06,
+      "loss": 0.4772,
+      "step": 452
+    },
+    {
+      "epoch": 0.08299743495786002,
+      "grad_norm": 0.45607475441878936,
+      "learning_rate": 8.278388278388278e-06,
+      "loss": 0.5319,
+      "step": 453
+    },
+    {
+      "epoch": 0.08318065225357274,
+      "grad_norm": 0.4712644997156548,
+      "learning_rate": 8.296703296703298e-06,
+      "loss": 0.5477,
+      "step": 454
+    },
+    {
+      "epoch": 0.08336386954928546,
+      "grad_norm": 0.6856852588986854,
+      "learning_rate": 8.315018315018316e-06,
+      "loss": 0.5235,
+      "step": 455
+    },
+    {
+      "epoch": 0.08354708684499816,
+      "grad_norm": 0.5019345602931549,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.5101,
+      "step": 456
+    },
+    {
+      "epoch": 0.08373030414071088,
+      "grad_norm": 0.4155972768165925,
+      "learning_rate": 8.351648351648353e-06,
+      "loss": 0.4986,
+      "step": 457
+    },
+    {
+      "epoch": 0.0839135214364236,
+      "grad_norm": 0.45725631484653695,
+      "learning_rate": 8.369963369963371e-06,
+      "loss": 0.5233,
+      "step": 458
+    },
+    {
+      "epoch": 0.08409673873213631,
+      "grad_norm": 0.5500464841633228,
+      "learning_rate": 8.388278388278389e-06,
+      "loss": 0.5225,
+      "step": 459
+    },
+    {
+      "epoch": 0.08427995602784903,
+      "grad_norm": 0.5210381864086956,
+      "learning_rate": 8.406593406593407e-06,
+      "loss": 0.5509,
+      "step": 460
+    },
+    {
+      "epoch": 0.08446317332356175,
+      "grad_norm": 0.4559821107303101,
+      "learning_rate": 8.424908424908426e-06,
+      "loss": 0.5118,
+      "step": 461
+    },
+    {
+      "epoch": 0.08464639061927445,
+      "grad_norm": 0.48323545775926524,
+      "learning_rate": 8.443223443223444e-06,
+      "loss": 0.5049,
+      "step": 462
+    },
+    {
+      "epoch": 0.08482960791498717,
+      "grad_norm": 0.4812859935236779,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.5204,
+      "step": 463
+    },
+    {
+      "epoch": 0.08501282521069989,
+      "grad_norm": 0.454474341410942,
+      "learning_rate": 8.47985347985348e-06,
+      "loss": 0.5387,
+      "step": 464
+    },
+    {
+      "epoch": 0.0851960425064126,
+      "grad_norm": 0.6615650181808158,
+      "learning_rate": 8.498168498168498e-06,
+      "loss": 0.5393,
+      "step": 465
+    },
+    {
+      "epoch": 0.08537925980212532,
+      "grad_norm": 0.4355325124699567,
+      "learning_rate": 8.516483516483517e-06,
+      "loss": 0.5027,
+      "step": 466
+    },
+    {
+      "epoch": 0.08556247709783804,
+      "grad_norm": 0.43977659231205646,
+      "learning_rate": 8.534798534798535e-06,
+      "loss": 0.4851,
+      "step": 467
+    },
+    {
+      "epoch": 0.08574569439355074,
+      "grad_norm": 0.5490378411527629,
+      "learning_rate": 8.553113553113553e-06,
+      "loss": 0.5087,
+      "step": 468
+    },
+    {
+      "epoch": 0.08592891168926346,
+      "grad_norm": 0.4334113007101785,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.4874,
+      "step": 469
+    },
+    {
+      "epoch": 0.08611212898497618,
+      "grad_norm": 0.48093462345535853,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 0.5103,
+      "step": 470
+    },
+    {
+      "epoch": 0.0862953462806889,
+      "grad_norm": 0.509104778628553,
+      "learning_rate": 8.60805860805861e-06,
+      "loss": 0.5125,
+      "step": 471
+    },
+    {
+      "epoch": 0.08647856357640162,
+      "grad_norm": 0.46329459466937095,
+      "learning_rate": 8.626373626373628e-06,
+      "loss": 0.4862,
+      "step": 472
+    },
+    {
+      "epoch": 0.08666178087211433,
+      "grad_norm": 0.41221910536829426,
+      "learning_rate": 8.644688644688646e-06,
+      "loss": 0.5058,
+      "step": 473
+    },
+    {
+      "epoch": 0.08684499816782704,
+      "grad_norm": 0.4965957595147397,
+      "learning_rate": 8.663003663003664e-06,
+      "loss": 0.5045,
+      "step": 474
+    },
+    {
+      "epoch": 0.08702821546353975,
+      "grad_norm": 0.6831024596734047,
+      "learning_rate": 8.681318681318681e-06,
+      "loss": 0.533,
+      "step": 475
+    },
+    {
+      "epoch": 0.08721143275925247,
+      "grad_norm": 0.44117689667679433,
+      "learning_rate": 8.699633699633701e-06,
+      "loss": 0.494,
+      "step": 476
+    },
+    {
+      "epoch": 0.08739465005496519,
+      "grad_norm": 0.49552902332130444,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 0.5295,
+      "step": 477
+    },
+    {
+      "epoch": 0.08757786735067791,
+      "grad_norm": 0.5202864223018397,
+      "learning_rate": 8.736263736263737e-06,
+      "loss": 0.5315,
+      "step": 478
+    },
+    {
+      "epoch": 0.08776108464639062,
+      "grad_norm": 0.4531456591891388,
+      "learning_rate": 8.754578754578755e-06,
+      "loss": 0.5177,
+      "step": 479
+    },
+    {
+      "epoch": 0.08794430194210333,
+      "grad_norm": 0.5060531781359298,
+      "learning_rate": 8.772893772893774e-06,
+      "loss": 0.5422,
+      "step": 480
+    },
+    {
+      "epoch": 0.08812751923781605,
+      "grad_norm": 0.44222377285427994,
+      "learning_rate": 8.791208791208792e-06,
+      "loss": 0.4679,
+      "step": 481
+    },
+    {
+      "epoch": 0.08831073653352876,
+      "grad_norm": 0.4308135949373974,
+      "learning_rate": 8.80952380952381e-06,
+      "loss": 0.5057,
+      "step": 482
+    },
+    {
+      "epoch": 0.08849395382924148,
+      "grad_norm": 0.5107147323097014,
+      "learning_rate": 8.827838827838828e-06,
+      "loss": 0.5147,
+      "step": 483
+    },
+    {
+      "epoch": 0.0886771711249542,
+      "grad_norm": 0.4722359038239213,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 0.5069,
+      "step": 484
+    },
+    {
+      "epoch": 0.08886038842066692,
+      "grad_norm": 0.4913307139634766,
+      "learning_rate": 8.864468864468865e-06,
+      "loss": 0.4943,
+      "step": 485
+    },
+    {
+      "epoch": 0.08904360571637962,
+      "grad_norm": 0.47381589638250404,
+      "learning_rate": 8.882783882783883e-06,
+      "loss": 0.5432,
+      "step": 486
+    },
+    {
+      "epoch": 0.08922682301209234,
+      "grad_norm": 0.4763727220823097,
+      "learning_rate": 8.9010989010989e-06,
+      "loss": 0.4765,
+      "step": 487
+    },
+    {
+      "epoch": 0.08941004030780506,
+      "grad_norm": 0.48581942940784734,
+      "learning_rate": 8.91941391941392e-06,
+      "loss": 0.4918,
+      "step": 488
+    },
+    {
+      "epoch": 0.08959325760351777,
+      "grad_norm": 0.4771435540271998,
+      "learning_rate": 8.937728937728938e-06,
+      "loss": 0.5146,
+      "step": 489
+    },
+    {
+      "epoch": 0.08977647489923049,
+      "grad_norm": 0.5554558590150881,
+      "learning_rate": 8.956043956043958e-06,
+      "loss": 0.5139,
+      "step": 490
+    },
+    {
+      "epoch": 0.08995969219494321,
+      "grad_norm": 0.4409852140988399,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 0.494,
+      "step": 491
+    },
+    {
+      "epoch": 0.09014290949065591,
+      "grad_norm": 0.470617109732078,
+      "learning_rate": 8.992673992673993e-06,
+      "loss": 0.5293,
+      "step": 492
+    },
+    {
+      "epoch": 0.09032612678636863,
+      "grad_norm": 0.48719044521659705,
+      "learning_rate": 9.010989010989011e-06,
+      "loss": 0.5111,
+      "step": 493
+    },
+    {
+      "epoch": 0.09050934408208135,
+      "grad_norm": 0.46876439756978366,
+      "learning_rate": 9.02930402930403e-06,
+      "loss": 0.4942,
+      "step": 494
+    },
+    {
+      "epoch": 0.09069256137779406,
+      "grad_norm": 0.46671821534033575,
+      "learning_rate": 9.047619047619049e-06,
+      "loss": 0.5617,
+      "step": 495
+    },
+    {
+      "epoch": 0.09087577867350678,
+      "grad_norm": 0.4868372988815388,
+      "learning_rate": 9.065934065934067e-06,
+      "loss": 0.5195,
+      "step": 496
+    },
+    {
+      "epoch": 0.0910589959692195,
+      "grad_norm": 0.47526685976460153,
+      "learning_rate": 9.084249084249084e-06,
+      "loss": 0.4561,
+      "step": 497
+    },
+    {
+      "epoch": 0.0912422132649322,
+      "grad_norm": 0.526795639153496,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 0.4905,
+      "step": 498
+    },
+    {
+      "epoch": 0.09142543056064492,
+      "grad_norm": 0.47935527575760656,
+      "learning_rate": 9.120879120879122e-06,
+      "loss": 0.5129,
+      "step": 499
+    },
+    {
+      "epoch": 0.09160864785635764,
+      "grad_norm": 0.47020582683735346,
+      "learning_rate": 9.13919413919414e-06,
+      "loss": 0.511,
+      "step": 500
+    },
+    {
+      "epoch": 0.09179186515207036,
+      "grad_norm": 0.46760420885699033,
+      "learning_rate": 9.157509157509158e-06,
+      "loss": 0.4789,
+      "step": 501
+    },
+    {
+      "epoch": 0.09197508244778307,
+      "grad_norm": 0.4695251283934706,
+      "learning_rate": 9.175824175824175e-06,
+      "loss": 0.5351,
+      "step": 502
+    },
+    {
+      "epoch": 0.09215829974349579,
+      "grad_norm": 0.5153360795669553,
+      "learning_rate": 9.194139194139195e-06,
+      "loss": 0.5522,
+      "step": 503
+    },
+    {
+      "epoch": 0.0923415170392085,
+      "grad_norm": 0.44806081150815513,
+      "learning_rate": 9.212454212454213e-06,
+      "loss": 0.5173,
+      "step": 504
+    },
+    {
+      "epoch": 0.09252473433492121,
+      "grad_norm": 0.451536504838503,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.5082,
+      "step": 505
+    },
+    {
+      "epoch": 0.09270795163063393,
+      "grad_norm": 0.48436384147084766,
+      "learning_rate": 9.24908424908425e-06,
+      "loss": 0.5388,
+      "step": 506
+    },
+    {
+      "epoch": 0.09289116892634665,
+      "grad_norm": 0.4609113225702784,
+      "learning_rate": 9.267399267399268e-06,
+      "loss": 0.4892,
+      "step": 507
+    },
+    {
+      "epoch": 0.09307438622205937,
+      "grad_norm": 0.4098711492197863,
+      "learning_rate": 9.285714285714288e-06,
+      "loss": 0.4856,
+      "step": 508
+    },
+    {
+      "epoch": 0.09325760351777208,
+      "grad_norm": 0.4715675907322317,
+      "learning_rate": 9.304029304029305e-06,
+      "loss": 0.53,
+      "step": 509
+    },
+    {
+      "epoch": 0.09344082081348479,
+      "grad_norm": 0.5026574189450673,
+      "learning_rate": 9.322344322344323e-06,
+      "loss": 0.5314,
+      "step": 510
+    },
+    {
+      "epoch": 0.0936240381091975,
+      "grad_norm": 0.4253199666429807,
+      "learning_rate": 9.340659340659341e-06,
+      "loss": 0.5068,
+      "step": 511
+    },
+    {
+      "epoch": 0.09380725540491022,
+      "grad_norm": 0.45894485399853974,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 0.5235,
+      "step": 512
+    },
+    {
+      "epoch": 0.09399047270062294,
+      "grad_norm": 0.7513225377427756,
+      "learning_rate": 9.377289377289379e-06,
+      "loss": 0.5071,
+      "step": 513
+    },
+    {
+      "epoch": 0.09417368999633566,
+      "grad_norm": 0.5231017027547264,
+      "learning_rate": 9.395604395604396e-06,
+      "loss": 0.5339,
+      "step": 514
+    },
+    {
+      "epoch": 0.09435690729204838,
+      "grad_norm": 0.5146997140217462,
+      "learning_rate": 9.413919413919414e-06,
+      "loss": 0.5052,
+      "step": 515
+    },
+    {
+      "epoch": 0.09454012458776108,
+      "grad_norm": 0.47760493192836756,
+      "learning_rate": 9.432234432234432e-06,
+      "loss": 0.5102,
+      "step": 516
+    },
+    {
+      "epoch": 0.0947233418834738,
+      "grad_norm": 0.44117590454697736,
+      "learning_rate": 9.450549450549452e-06,
+      "loss": 0.5237,
+      "step": 517
+    },
+    {
+      "epoch": 0.09490655917918651,
+      "grad_norm": 0.5004680185091399,
+      "learning_rate": 9.46886446886447e-06,
+      "loss": 0.5243,
+      "step": 518
+    },
+    {
+      "epoch": 0.09508977647489923,
+      "grad_norm": 0.4549800276149351,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 0.5139,
+      "step": 519
+    },
+    {
+      "epoch": 0.09527299377061195,
+      "grad_norm": 0.4263327588755031,
+      "learning_rate": 9.505494505494505e-06,
+      "loss": 0.4877,
+      "step": 520
+    },
+    {
+      "epoch": 0.09545621106632467,
+      "grad_norm": 0.4484968491218323,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 0.4986,
+      "step": 521
+    },
+    {
+      "epoch": 0.09563942836203737,
+      "grad_norm": 0.39390345192807547,
+      "learning_rate": 9.542124542124543e-06,
+      "loss": 0.5116,
+      "step": 522
+    },
+    {
+      "epoch": 0.09582264565775009,
+      "grad_norm": 0.4529147798983982,
+      "learning_rate": 9.560439560439562e-06,
+      "loss": 0.5312,
+      "step": 523
+    },
+    {
+      "epoch": 0.0960058629534628,
+      "grad_norm": 0.49701803503982106,
+      "learning_rate": 9.57875457875458e-06,
+      "loss": 0.529,
+      "step": 524
+    },
+    {
+      "epoch": 0.09618908024917552,
+      "grad_norm": 0.4097214793479841,
+      "learning_rate": 9.597069597069598e-06,
+      "loss": 0.5005,
+      "step": 525
+    },
+    {
+      "epoch": 0.09637229754488824,
+      "grad_norm": 0.5148151884655592,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 0.53,
+      "step": 526
+    },
+    {
+      "epoch": 0.09655551484060096,
+      "grad_norm": 0.4264857545219357,
+      "learning_rate": 9.633699633699635e-06,
+      "loss": 0.5136,
+      "step": 527
+    },
+    {
+      "epoch": 0.09673873213631366,
+      "grad_norm": 0.45426464252638443,
+      "learning_rate": 9.652014652014653e-06,
+      "loss": 0.5117,
+      "step": 528
+    },
+    {
+      "epoch": 0.09692194943202638,
+      "grad_norm": 0.47571176071645493,
+      "learning_rate": 9.670329670329671e-06,
+      "loss": 0.5185,
+      "step": 529
+    },
+    {
+      "epoch": 0.0971051667277391,
+      "grad_norm": 0.42418752426373113,
+      "learning_rate": 9.688644688644689e-06,
+      "loss": 0.5285,
+      "step": 530
+    },
+    {
+      "epoch": 0.09728838402345182,
+      "grad_norm": 0.48590348616099827,
+      "learning_rate": 9.706959706959708e-06,
+      "loss": 0.5339,
+      "step": 531
+    },
+    {
+      "epoch": 0.09747160131916453,
+      "grad_norm": 0.463774513888958,
+      "learning_rate": 9.725274725274726e-06,
+      "loss": 0.5386,
+      "step": 532
+    },
+    {
+      "epoch": 0.09765481861487725,
+      "grad_norm": 0.4891808396070614,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 0.5326,
+      "step": 533
+    },
+    {
+      "epoch": 0.09783803591058995,
+      "grad_norm": 0.40841004082693305,
+      "learning_rate": 9.761904761904762e-06,
+      "loss": 0.5085,
+      "step": 534
+    },
+    {
+      "epoch": 0.09802125320630267,
+      "grad_norm": 0.42494728872814735,
+      "learning_rate": 9.780219780219781e-06,
+      "loss": 0.4751,
+      "step": 535
+    },
+    {
+      "epoch": 0.09820447050201539,
+      "grad_norm": 0.4594283822364781,
+      "learning_rate": 9.7985347985348e-06,
+      "loss": 0.5275,
+      "step": 536
+    },
+    {
+      "epoch": 0.09838768779772811,
+      "grad_norm": 0.4064302259117676,
+      "learning_rate": 9.816849816849817e-06,
+      "loss": 0.5089,
+      "step": 537
+    },
+    {
+      "epoch": 0.09857090509344082,
+      "grad_norm": 0.4860027724973198,
+      "learning_rate": 9.835164835164835e-06,
+      "loss": 0.505,
+      "step": 538
+    },
+    {
+      "epoch": 0.09875412238915354,
+      "grad_norm": 0.47625362870716265,
+      "learning_rate": 9.853479853479855e-06,
+      "loss": 0.4954,
+      "step": 539
+    },
+    {
+      "epoch": 0.09893733968486625,
+      "grad_norm": 0.4592940212317861,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 0.4871,
+      "step": 540
+    },
+    {
+      "epoch": 0.09912055698057896,
+      "grad_norm": 0.42717127053329573,
+      "learning_rate": 9.890109890109892e-06,
+      "loss": 0.5261,
+      "step": 541
+    },
+    {
+      "epoch": 0.09930377427629168,
+      "grad_norm": 0.43970535809042904,
+      "learning_rate": 9.90842490842491e-06,
+      "loss": 0.4866,
+      "step": 542
+    },
+    {
+      "epoch": 0.0994869915720044,
+      "grad_norm": 0.42143360797263807,
+      "learning_rate": 9.926739926739928e-06,
+      "loss": 0.5113,
+      "step": 543
+    },
+    {
+      "epoch": 0.09967020886771712,
+      "grad_norm": 0.4416711016318173,
+      "learning_rate": 9.945054945054946e-06,
+      "loss": 0.5116,
+      "step": 544
+    },
+    {
+      "epoch": 0.09985342616342983,
+      "grad_norm": 0.45325267149265236,
+      "learning_rate": 9.963369963369965e-06,
+      "loss": 0.5283,
+      "step": 545
+    },
+    {
+      "epoch": 0.10003664345914254,
+      "grad_norm": 0.3948707773420248,
+      "learning_rate": 9.981684981684983e-06,
+      "loss": 0.5114,
+      "step": 546
+    },
+    {
+      "epoch": 0.10021986075485526,
+      "grad_norm": 0.42983426021717475,
+      "learning_rate": 1e-05,
+      "loss": 0.5169,
+      "step": 547
+    },
+    {
+      "epoch": 0.10040307805056797,
+      "grad_norm": 0.5119731461361727,
+      "learning_rate": 9.999998977359419e-06,
+      "loss": 0.518,
+      "step": 548
+    },
+    {
+      "epoch": 0.10058629534628069,
+      "grad_norm": 0.48479775231872485,
+      "learning_rate": 9.999995909438092e-06,
+      "loss": 0.5207,
+      "step": 549
+    },
+    {
+      "epoch": 0.10076951264199341,
+      "grad_norm": 0.45005420614602115,
+      "learning_rate": 9.999990796237274e-06,
+      "loss": 0.4878,
+      "step": 550
+    },
+    {
+      "epoch": 0.10095272993770613,
+      "grad_norm": 0.40595587220813967,
+      "learning_rate": 9.999983637759059e-06,
+      "loss": 0.4712,
+      "step": 551
+    },
+    {
+      "epoch": 0.10113594723341883,
+      "grad_norm": 0.43893870910937327,
+      "learning_rate": 9.999974434006372e-06,
+      "loss": 0.4964,
+      "step": 552
+    },
+    {
+      "epoch": 0.10131916452913155,
+      "grad_norm": 0.4638079942473788,
+      "learning_rate": 9.99996318498298e-06,
+      "loss": 0.4834,
+      "step": 553
+    },
+    {
+      "epoch": 0.10150238182484426,
+      "grad_norm": 0.45078724585599195,
+      "learning_rate": 9.999949890693484e-06,
+      "loss": 0.4828,
+      "step": 554
+    },
+    {
+      "epoch": 0.10168559912055698,
+      "grad_norm": 0.4968230440871285,
+      "learning_rate": 9.999934551143319e-06,
+      "loss": 0.5146,
+      "step": 555
+    },
+    {
+      "epoch": 0.1018688164162697,
+      "grad_norm": 0.4431824586567993,
+      "learning_rate": 9.999917166338767e-06,
+      "loss": 0.5117,
+      "step": 556
+    },
+    {
+      "epoch": 0.10205203371198242,
+      "grad_norm": 0.40809319706771474,
+      "learning_rate": 9.999897736286932e-06,
+      "loss": 0.4942,
+      "step": 557
+    },
+    {
+      "epoch": 0.10223525100769512,
+      "grad_norm": 0.46904649406632837,
+      "learning_rate": 9.999876260995767e-06,
+      "loss": 0.5223,
+      "step": 558
+    },
+    {
+      "epoch": 0.10241846830340784,
+      "grad_norm": 0.45993616230767786,
+      "learning_rate": 9.999852740474054e-06,
+      "loss": 0.498,
+      "step": 559
+    },
+    {
+      "epoch": 0.10260168559912056,
+      "grad_norm": 0.41328723379038895,
+      "learning_rate": 9.999827174731414e-06,
+      "loss": 0.5294,
+      "step": 560
+    },
+    {
+      "epoch": 0.10278490289483327,
+      "grad_norm": 0.4584774156426854,
+      "learning_rate": 9.999799563778307e-06,
+      "loss": 0.5138,
+      "step": 561
+    },
+    {
+      "epoch": 0.10296812019054599,
+      "grad_norm": 0.4626480547701558,
+      "learning_rate": 9.999769907626024e-06,
+      "loss": 0.5122,
+      "step": 562
+    },
+    {
+      "epoch": 0.10315133748625871,
+      "grad_norm": 0.4647555669122866,
+      "learning_rate": 9.9997382062867e-06,
+      "loss": 0.5276,
+      "step": 563
+    },
+    {
+      "epoch": 0.10333455478197141,
+      "grad_norm": 0.41824885855037686,
+      "learning_rate": 9.9997044597733e-06,
+      "loss": 0.4976,
+      "step": 564
+    },
+    {
+      "epoch": 0.10351777207768413,
+      "grad_norm": 0.41187305924650414,
+      "learning_rate": 9.999668668099628e-06,
+      "loss": 0.4953,
+      "step": 565
+    },
+    {
+      "epoch": 0.10370098937339685,
+      "grad_norm": 0.42660595756816655,
+      "learning_rate": 9.999630831280329e-06,
+      "loss": 0.514,
+      "step": 566
+    },
+    {
+      "epoch": 0.10388420666910957,
+      "grad_norm": 0.5099786508796531,
+      "learning_rate": 9.999590949330876e-06,
+      "loss": 0.5038,
+      "step": 567
+    },
+    {
+      "epoch": 0.10406742396482228,
+      "grad_norm": 0.4926825490754442,
+      "learning_rate": 9.999549022267582e-06,
+      "loss": 0.5535,
+      "step": 568
+    },
+    {
+      "epoch": 0.104250641260535,
+      "grad_norm": 0.4578490443823955,
+      "learning_rate": 9.999505050107601e-06,
+      "loss": 0.5075,
+      "step": 569
+    },
+    {
+      "epoch": 0.1044338585562477,
+      "grad_norm": 0.4733323605035433,
+      "learning_rate": 9.99945903286892e-06,
+      "loss": 0.4747,
+      "step": 570
+    },
+    {
+      "epoch": 0.10461707585196042,
+      "grad_norm": 0.44893796479946446,
+      "learning_rate": 9.999410970570358e-06,
+      "loss": 0.4711,
+      "step": 571
+    },
+    {
+      "epoch": 0.10480029314767314,
+      "grad_norm": 0.5254914900315981,
+      "learning_rate": 9.99936086323158e-06,
+      "loss": 0.5358,
+      "step": 572
+    },
+    {
+      "epoch": 0.10498351044338586,
+      "grad_norm": 0.537923540220477,
+      "learning_rate": 9.999308710873083e-06,
+      "loss": 0.5079,
+      "step": 573
+    },
+    {
+      "epoch": 0.10516672773909858,
+      "grad_norm": 0.45536567675754,
+      "learning_rate": 9.999254513516196e-06,
+      "loss": 0.4908,
+      "step": 574
+    },
+    {
+      "epoch": 0.10534994503481128,
+      "grad_norm": 0.38323869490085266,
+      "learning_rate": 9.999198271183094e-06,
+      "loss": 0.4983,
+      "step": 575
+    },
+    {
+      "epoch": 0.105533162330524,
+      "grad_norm": 0.43580085679521197,
+      "learning_rate": 9.999139983896779e-06,
+      "loss": 0.5031,
+      "step": 576
+    },
+    {
+      "epoch": 0.10571637962623671,
+      "grad_norm": 0.4193926406628939,
+      "learning_rate": 9.999079651681096e-06,
+      "loss": 0.5026,
+      "step": 577
+    },
+    {
+      "epoch": 0.10589959692194943,
+      "grad_norm": 0.5149327291476714,
+      "learning_rate": 9.999017274560722e-06,
+      "loss": 0.5414,
+      "step": 578
+    },
+    {
+      "epoch": 0.10608281421766215,
+      "grad_norm": 0.44098660936462153,
+      "learning_rate": 9.998952852561176e-06,
+      "loss": 0.5032,
+      "step": 579
+    },
+    {
+      "epoch": 0.10626603151337487,
+      "grad_norm": 0.4413320145497695,
+      "learning_rate": 9.998886385708807e-06,
+      "loss": 0.497,
+      "step": 580
+    },
+    {
+      "epoch": 0.10644924880908757,
+      "grad_norm": 0.45575842795199323,
+      "learning_rate": 9.998817874030808e-06,
+      "loss": 0.5207,
+      "step": 581
+    },
+    {
+      "epoch": 0.10663246610480029,
+      "grad_norm": 0.532654166955093,
+      "learning_rate": 9.9987473175552e-06,
+      "loss": 0.5212,
+      "step": 582
+    },
+    {
+      "epoch": 0.106815683400513,
+      "grad_norm": 0.4157619606008366,
+      "learning_rate": 9.998674716310846e-06,
+      "loss": 0.5085,
+      "step": 583
+    },
+    {
+      "epoch": 0.10699890069622572,
+      "grad_norm": 0.4868388397165033,
+      "learning_rate": 9.998600070327444e-06,
+      "loss": 0.5306,
+      "step": 584
+    },
+    {
+      "epoch": 0.10718211799193844,
+      "grad_norm": 0.49401937562080056,
+      "learning_rate": 9.998523379635527e-06,
+      "loss": 0.5429,
+      "step": 585
+    },
+    {
+      "epoch": 0.10736533528765116,
+      "grad_norm": 0.3986626914030523,
+      "learning_rate": 9.99844464426647e-06,
+      "loss": 0.4695,
+      "step": 586
+    },
+    {
+      "epoch": 0.10754855258336386,
+      "grad_norm": 0.49524254606268836,
+      "learning_rate": 9.998363864252474e-06,
+      "loss": 0.5226,
+      "step": 587
+    },
+    {
+      "epoch": 0.10773176987907658,
+      "grad_norm": 0.46186447850660334,
+      "learning_rate": 9.998281039626588e-06,
+      "loss": 0.5294,
+      "step": 588
+    },
+    {
+      "epoch": 0.1079149871747893,
+      "grad_norm": 0.4207082230286084,
+      "learning_rate": 9.99819617042269e-06,
+      "loss": 0.5096,
+      "step": 589
+    },
+    {
+      "epoch": 0.10809820447050202,
+      "grad_norm": 0.3955565788721964,
+      "learning_rate": 9.998109256675496e-06,
+      "loss": 0.4931,
+      "step": 590
+    },
+    {
+      "epoch": 0.10828142176621473,
+      "grad_norm": 0.36723747707336185,
+      "learning_rate": 9.998020298420559e-06,
+      "loss": 0.4964,
+      "step": 591
+    },
+    {
+      "epoch": 0.10846463906192745,
+      "grad_norm": 0.4185544526751163,
+      "learning_rate": 9.997929295694266e-06,
+      "loss": 0.4801,
+      "step": 592
+    },
+    {
+      "epoch": 0.10864785635764015,
+      "grad_norm": 0.42948593775837285,
+      "learning_rate": 9.997836248533845e-06,
+      "loss": 0.5197,
+      "step": 593
+    },
+    {
+      "epoch": 0.10883107365335287,
+      "grad_norm": 0.43966759837587943,
+      "learning_rate": 9.997741156977356e-06,
+      "loss": 0.5254,
+      "step": 594
+    },
+    {
+      "epoch": 0.10901429094906559,
+      "grad_norm": 0.4632445570564666,
+      "learning_rate": 9.997644021063698e-06,
+      "loss": 0.4802,
+      "step": 595
+    },
+    {
+      "epoch": 0.10919750824477831,
+      "grad_norm": 0.46480621716730125,
+      "learning_rate": 9.997544840832604e-06,
+      "loss": 0.5219,
+      "step": 596
+    },
+    {
+      "epoch": 0.10938072554049103,
+      "grad_norm": 0.44567173311984926,
+      "learning_rate": 9.997443616324645e-06,
+      "loss": 0.5105,
+      "step": 597
+    },
+    {
+      "epoch": 0.10956394283620374,
+      "grad_norm": 0.4534263401515789,
+      "learning_rate": 9.997340347581226e-06,
+      "loss": 0.5212,
+      "step": 598
+    },
+    {
+      "epoch": 0.10974716013191645,
+      "grad_norm": 0.40359525147680225,
+      "learning_rate": 9.99723503464459e-06,
+      "loss": 0.5261,
+      "step": 599
+    },
+    {
+      "epoch": 0.10993037742762916,
+      "grad_norm": 0.8305073137318505,
+      "learning_rate": 9.99712767755782e-06,
+      "loss": 0.5173,
+      "step": 600
+    },
+    {
+      "epoch": 0.11011359472334188,
+      "grad_norm": 0.5039765167575108,
+      "learning_rate": 9.997018276364825e-06,
+      "loss": 0.4983,
+      "step": 601
+    },
+    {
+      "epoch": 0.1102968120190546,
+      "grad_norm": 0.4000444070826362,
+      "learning_rate": 9.99690683111036e-06,
+      "loss": 0.4876,
+      "step": 602
+    },
+    {
+      "epoch": 0.11048002931476732,
+      "grad_norm": 0.45535979366815893,
+      "learning_rate": 9.99679334184001e-06,
+      "loss": 0.5116,
+      "step": 603
+    },
+    {
+      "epoch": 0.11066324661048003,
+      "grad_norm": 0.5208092576973751,
+      "learning_rate": 9.996677808600202e-06,
+      "loss": 0.5079,
+      "step": 604
+    },
+    {
+      "epoch": 0.11084646390619274,
+      "grad_norm": 0.47721138790862244,
+      "learning_rate": 9.996560231438193e-06,
+      "loss": 0.5031,
+      "step": 605
+    },
+    {
+      "epoch": 0.11102968120190546,
+      "grad_norm": 0.4106250945716929,
+      "learning_rate": 9.996440610402078e-06,
+      "loss": 0.4789,
+      "step": 606
+    },
+    {
+      "epoch": 0.11121289849761817,
+      "grad_norm": 0.4981443340009068,
+      "learning_rate": 9.996318945540792e-06,
+      "loss": 0.4749,
+      "step": 607
+    },
+    {
+      "epoch": 0.11139611579333089,
+      "grad_norm": 0.4365748987869686,
+      "learning_rate": 9.996195236904098e-06,
+      "loss": 0.503,
+      "step": 608
+    },
+    {
+      "epoch": 0.11157933308904361,
+      "grad_norm": 0.42769357659836243,
+      "learning_rate": 9.996069484542605e-06,
+      "loss": 0.4883,
+      "step": 609
+    },
+    {
+      "epoch": 0.11176255038475633,
+      "grad_norm": 0.4468425715852285,
+      "learning_rate": 9.995941688507749e-06,
+      "loss": 0.528,
+      "step": 610
+    },
+    {
+      "epoch": 0.11194576768046903,
+      "grad_norm": 0.44305501127316244,
+      "learning_rate": 9.995811848851807e-06,
+      "loss": 0.5192,
+      "step": 611
+    },
+    {
+      "epoch": 0.11212898497618175,
+      "grad_norm": 0.4430509247761743,
+      "learning_rate": 9.995679965627891e-06,
+      "loss": 0.4879,
+      "step": 612
+    },
+    {
+      "epoch": 0.11231220227189447,
+      "grad_norm": 0.4330736232909904,
+      "learning_rate": 9.995546038889948e-06,
+      "loss": 0.5315,
+      "step": 613
+    },
+    {
+      "epoch": 0.11249541956760718,
+      "grad_norm": 0.38536887827486094,
+      "learning_rate": 9.995410068692763e-06,
+      "loss": 0.4789,
+      "step": 614
+    },
+    {
+      "epoch": 0.1126786368633199,
+      "grad_norm": 0.5420414741691238,
+      "learning_rate": 9.995272055091954e-06,
+      "loss": 0.4863,
+      "step": 615
+    },
+    {
+      "epoch": 0.11286185415903262,
+      "grad_norm": 0.4166977498062619,
+      "learning_rate": 9.995131998143976e-06,
+      "loss": 0.4863,
+      "step": 616
+    },
+    {
+      "epoch": 0.11304507145474532,
+      "grad_norm": 0.40681536551456327,
+      "learning_rate": 9.99498989790612e-06,
+      "loss": 0.5158,
+      "step": 617
+    },
+    {
+      "epoch": 0.11322828875045804,
+      "grad_norm": 0.4164498167643854,
+      "learning_rate": 9.994845754436516e-06,
+      "loss": 0.5287,
+      "step": 618
+    },
+    {
+      "epoch": 0.11341150604617076,
+      "grad_norm": 0.430875770675555,
+      "learning_rate": 9.994699567794123e-06,
+      "loss": 0.4949,
+      "step": 619
+    },
+    {
+      "epoch": 0.11359472334188347,
+      "grad_norm": 0.47209089186963143,
+      "learning_rate": 9.994551338038742e-06,
+      "loss": 0.5212,
+      "step": 620
+    },
+    {
+      "epoch": 0.11377794063759619,
+      "grad_norm": 0.4693460118731517,
+      "learning_rate": 9.994401065231008e-06,
+      "loss": 0.5218,
+      "step": 621
+    },
+    {
+      "epoch": 0.11396115793330891,
+      "grad_norm": 0.43645941737097255,
+      "learning_rate": 9.994248749432388e-06,
+      "loss": 0.5355,
+      "step": 622
+    },
+    {
+      "epoch": 0.11414437522902161,
+      "grad_norm": 0.45174829183811943,
+      "learning_rate": 9.994094390705189e-06,
+      "loss": 0.4941,
+      "step": 623
+    },
+    {
+      "epoch": 0.11432759252473433,
+      "grad_norm": 0.4609885421062077,
+      "learning_rate": 9.993937989112554e-06,
+      "loss": 0.5424,
+      "step": 624
+    },
+    {
+      "epoch": 0.11451080982044705,
+      "grad_norm": 0.4711205796072519,
+      "learning_rate": 9.993779544718459e-06,
+      "loss": 0.5054,
+      "step": 625
+    },
+    {
+      "epoch": 0.11469402711615977,
+      "grad_norm": 0.4692458224302816,
+      "learning_rate": 9.993619057587714e-06,
+      "loss": 0.5013,
+      "step": 626
+    },
+    {
+      "epoch": 0.11487724441187248,
+      "grad_norm": 0.44543459536907126,
+      "learning_rate": 9.99345652778597e-06,
+      "loss": 0.4882,
+      "step": 627
+    },
+    {
+      "epoch": 0.1150604617075852,
+      "grad_norm": 0.43461884283611496,
+      "learning_rate": 9.993291955379713e-06,
+      "loss": 0.5352,
+      "step": 628
+    },
+    {
+      "epoch": 0.1152436790032979,
+      "grad_norm": 0.41500076390959145,
+      "learning_rate": 9.993125340436258e-06,
+      "loss": 0.4895,
+      "step": 629
+    },
+    {
+      "epoch": 0.11542689629901062,
+      "grad_norm": 0.4307377636082482,
+      "learning_rate": 9.992956683023762e-06,
+      "loss": 0.5149,
+      "step": 630
+    },
+    {
+      "epoch": 0.11561011359472334,
+      "grad_norm": 0.45085961488644083,
+      "learning_rate": 9.992785983211214e-06,
+      "loss": 0.5061,
+      "step": 631
+    },
+    {
+      "epoch": 0.11579333089043606,
+      "grad_norm": 0.45702457348234676,
+      "learning_rate": 9.992613241068444e-06,
+      "loss": 0.4947,
+      "step": 632
+    },
+    {
+      "epoch": 0.11597654818614878,
+      "grad_norm": 0.46046219980195896,
+      "learning_rate": 9.992438456666108e-06,
+      "loss": 0.4727,
+      "step": 633
+    },
+    {
+      "epoch": 0.1161597654818615,
+      "grad_norm": 0.5109084834664002,
+      "learning_rate": 9.992261630075704e-06,
+      "loss": 0.5015,
+      "step": 634
+    },
+    {
+      "epoch": 0.1163429827775742,
+      "grad_norm": 0.4605028589793491,
+      "learning_rate": 9.992082761369567e-06,
+      "loss": 0.518,
+      "step": 635
+    },
+    {
+      "epoch": 0.11652620007328691,
+      "grad_norm": 0.4135254020313275,
+      "learning_rate": 9.991901850620861e-06,
+      "loss": 0.4833,
+      "step": 636
+    },
+    {
+      "epoch": 0.11670941736899963,
+      "grad_norm": 0.5226822189268372,
+      "learning_rate": 9.99171889790359e-06,
+      "loss": 0.5118,
+      "step": 637
+    },
+    {
+      "epoch": 0.11689263466471235,
+      "grad_norm": 0.4061164579366039,
+      "learning_rate": 9.991533903292592e-06,
+      "loss": 0.4953,
+      "step": 638
+    },
+    {
+      "epoch": 0.11707585196042507,
+      "grad_norm": 2.3765281557608486,
+      "learning_rate": 9.99134686686354e-06,
+      "loss": 0.4708,
+      "step": 639
+    },
+    {
+      "epoch": 0.11725906925613779,
+      "grad_norm": 0.4675538148921928,
+      "learning_rate": 9.991157788692942e-06,
+      "loss": 0.4994,
+      "step": 640
+    },
+    {
+      "epoch": 0.11744228655185049,
+      "grad_norm": 0.4467597960418954,
+      "learning_rate": 9.990966668858144e-06,
+      "loss": 0.4982,
+      "step": 641
+    },
+    {
+      "epoch": 0.1176255038475632,
+      "grad_norm": 0.46222076234046117,
+      "learning_rate": 9.99077350743732e-06,
+      "loss": 0.5208,
+      "step": 642
+    },
+    {
+      "epoch": 0.11780872114327592,
+      "grad_norm": 0.45856899421216063,
+      "learning_rate": 9.990578304509488e-06,
+      "loss": 0.5109,
+      "step": 643
+    },
+    {
+      "epoch": 0.11799193843898864,
+      "grad_norm": 0.5006843127431095,
+      "learning_rate": 9.990381060154496e-06,
+      "loss": 0.4941,
+      "step": 644
+    },
+    {
+      "epoch": 0.11817515573470136,
+      "grad_norm": 0.4638055815431298,
+      "learning_rate": 9.990181774453028e-06,
+      "loss": 0.509,
+      "step": 645
+    },
+    {
+      "epoch": 0.11835837303041408,
+      "grad_norm": 0.4429725714841303,
+      "learning_rate": 9.989980447486601e-06,
+      "loss": 0.5196,
+      "step": 646
+    },
+    {
+      "epoch": 0.11854159032612678,
+      "grad_norm": 0.44051330732830757,
+      "learning_rate": 9.989777079337572e-06,
+      "loss": 0.5316,
+      "step": 647
+    },
+    {
+      "epoch": 0.1187248076218395,
+      "grad_norm": 0.41555588890083767,
+      "learning_rate": 9.989571670089129e-06,
+      "loss": 0.5126,
+      "step": 648
+    },
+    {
+      "epoch": 0.11890802491755222,
+      "grad_norm": 0.46239837428682623,
+      "learning_rate": 9.989364219825295e-06,
+      "loss": 0.506,
+      "step": 649
+    },
+    {
+      "epoch": 0.11909124221326493,
+      "grad_norm": 0.416093035128269,
+      "learning_rate": 9.98915472863093e-06,
+      "loss": 0.5194,
+      "step": 650
+    },
+    {
+      "epoch": 0.11927445950897765,
+      "grad_norm": 0.43847114193541153,
+      "learning_rate": 9.988943196591727e-06,
+      "loss": 0.4994,
+      "step": 651
+    },
+    {
+      "epoch": 0.11945767680469037,
+      "grad_norm": 0.44451525143619486,
+      "learning_rate": 9.988729623794215e-06,
+      "loss": 0.5283,
+      "step": 652
+    },
+    {
+      "epoch": 0.11964089410040307,
+      "grad_norm": 0.4041297930697904,
+      "learning_rate": 9.988514010325758e-06,
+      "loss": 0.5049,
+      "step": 653
+    },
+    {
+      "epoch": 0.11982411139611579,
+      "grad_norm": 0.4075716530480146,
+      "learning_rate": 9.988296356274551e-06,
+      "loss": 0.4921,
+      "step": 654
+    },
+    {
+      "epoch": 0.12000732869182851,
+      "grad_norm": 0.40736195443308915,
+      "learning_rate": 9.988076661729631e-06,
+      "loss": 0.4805,
+      "step": 655
+    },
+    {
+      "epoch": 0.12019054598754123,
+      "grad_norm": 0.4545077167148167,
+      "learning_rate": 9.987854926780863e-06,
+      "loss": 0.4789,
+      "step": 656
+    },
+    {
+      "epoch": 0.12037376328325394,
+      "grad_norm": 0.5314832542562551,
+      "learning_rate": 9.987631151518948e-06,
+      "loss": 0.5263,
+      "step": 657
+    },
+    {
+      "epoch": 0.12055698057896666,
+      "grad_norm": 0.4967524623940482,
+      "learning_rate": 9.987405336035425e-06,
+      "loss": 0.5106,
+      "step": 658
+    },
+    {
+      "epoch": 0.12074019787467936,
+      "grad_norm": 0.4306664623952609,
+      "learning_rate": 9.987177480422663e-06,
+      "loss": 0.5376,
+      "step": 659
+    },
+    {
+      "epoch": 0.12092341517039208,
+      "grad_norm": 0.4158801663482348,
+      "learning_rate": 9.98694758477387e-06,
+      "loss": 0.4838,
+      "step": 660
+    },
+    {
+      "epoch": 0.1211066324661048,
+      "grad_norm": 0.40605820771792417,
+      "learning_rate": 9.986715649183084e-06,
+      "loss": 0.505,
+      "step": 661
+    },
+    {
+      "epoch": 0.12128984976181752,
+      "grad_norm": 0.4085721106839356,
+      "learning_rate": 9.986481673745183e-06,
+      "loss": 0.4993,
+      "step": 662
+    },
+    {
+      "epoch": 0.12147306705753023,
+      "grad_norm": 0.4785768918832099,
+      "learning_rate": 9.986245658555873e-06,
+      "loss": 0.5387,
+      "step": 663
+    },
+    {
+      "epoch": 0.12165628435324295,
+      "grad_norm": 0.4175528791505364,
+      "learning_rate": 9.986007603711698e-06,
+      "loss": 0.4976,
+      "step": 664
+    },
+    {
+      "epoch": 0.12183950164895566,
+      "grad_norm": 0.4323331836919794,
+      "learning_rate": 9.985767509310035e-06,
+      "loss": 0.5346,
+      "step": 665
+    },
+    {
+      "epoch": 0.12202271894466837,
+      "grad_norm": 0.4868011056584651,
+      "learning_rate": 9.9855253754491e-06,
+      "loss": 0.5309,
+      "step": 666
+    },
+    {
+      "epoch": 0.12220593624038109,
+      "grad_norm": 0.46372133085485157,
+      "learning_rate": 9.985281202227936e-06,
+      "loss": 0.5053,
+      "step": 667
+    },
+    {
+      "epoch": 0.12238915353609381,
+      "grad_norm": 0.48326269099644514,
+      "learning_rate": 9.985034989746423e-06,
+      "loss": 0.4941,
+      "step": 668
+    },
+    {
+      "epoch": 0.12257237083180653,
+      "grad_norm": 0.4557552947362274,
+      "learning_rate": 9.984786738105279e-06,
+      "loss": 0.5121,
+      "step": 669
+    },
+    {
+      "epoch": 0.12275558812751924,
+      "grad_norm": 0.39704238565295197,
+      "learning_rate": 9.98453644740605e-06,
+      "loss": 0.4962,
+      "step": 670
+    },
+    {
+      "epoch": 0.12293880542323195,
+      "grad_norm": 0.40645243334504044,
+      "learning_rate": 9.98428411775112e-06,
+      "loss": 0.5046,
+      "step": 671
+    },
+    {
+      "epoch": 0.12312202271894467,
+      "grad_norm": 0.42629021258457467,
+      "learning_rate": 9.984029749243707e-06,
+      "loss": 0.5084,
+      "step": 672
+    },
+    {
+      "epoch": 0.12330524001465738,
+      "grad_norm": 0.45965819318406503,
+      "learning_rate": 9.98377334198786e-06,
+      "loss": 0.4759,
+      "step": 673
+    },
+    {
+      "epoch": 0.1234884573103701,
+      "grad_norm": 0.41967629717267835,
+      "learning_rate": 9.983514896088466e-06,
+      "loss": 0.498,
+      "step": 674
+    },
+    {
+      "epoch": 0.12367167460608282,
+      "grad_norm": 0.4295939596150219,
+      "learning_rate": 9.983254411651242e-06,
+      "loss": 0.5243,
+      "step": 675
+    },
+    {
+      "epoch": 0.12385489190179554,
+      "grad_norm": 0.4808089074794197,
+      "learning_rate": 9.982991888782742e-06,
+      "loss": 0.5,
+      "step": 676
+    },
+    {
+      "epoch": 0.12403810919750824,
+      "grad_norm": 0.48728725704379316,
+      "learning_rate": 9.982727327590352e-06,
+      "loss": 0.5459,
+      "step": 677
+    },
+    {
+      "epoch": 0.12422132649322096,
+      "grad_norm": 0.42799437772375803,
+      "learning_rate": 9.982460728182292e-06,
+      "loss": 0.5488,
+      "step": 678
+    },
+    {
+      "epoch": 0.12440454378893367,
+      "grad_norm": 0.4334349975707624,
+      "learning_rate": 9.982192090667618e-06,
+      "loss": 0.501,
+      "step": 679
+    },
+    {
+      "epoch": 0.12458776108464639,
+      "grad_norm": 0.45671378695636383,
+      "learning_rate": 9.981921415156217e-06,
+      "loss": 0.5236,
+      "step": 680
+    },
+    {
+      "epoch": 0.12477097838035911,
+      "grad_norm": 0.49531406109254905,
+      "learning_rate": 9.98164870175881e-06,
+      "loss": 0.5224,
+      "step": 681
+    },
+    {
+      "epoch": 0.12495419567607183,
+      "grad_norm": 0.3902067301024602,
+      "learning_rate": 9.981373950586952e-06,
+      "loss": 0.4835,
+      "step": 682
+    },
+    {
+      "epoch": 0.12513741297178455,
+      "grad_norm": 0.46995820446689096,
+      "learning_rate": 9.981097161753032e-06,
+      "loss": 0.527,
+      "step": 683
+    },
+    {
+      "epoch": 0.12532063026749726,
+      "grad_norm": 0.43642001976560174,
+      "learning_rate": 9.980818335370273e-06,
+      "loss": 0.514,
+      "step": 684
+    },
+    {
+      "epoch": 0.12550384756320998,
+      "grad_norm": 0.5074521053811396,
+      "learning_rate": 9.980537471552728e-06,
+      "loss": 0.4897,
+      "step": 685
+    },
+    {
+      "epoch": 0.12568706485892267,
+      "grad_norm": 0.43625566798430576,
+      "learning_rate": 9.98025457041529e-06,
+      "loss": 0.5174,
+      "step": 686
+    },
+    {
+      "epoch": 0.1258702821546354,
+      "grad_norm": 0.4129578447344159,
+      "learning_rate": 9.979969632073678e-06,
+      "loss": 0.5135,
+      "step": 687
+    },
+    {
+      "epoch": 0.1260534994503481,
+      "grad_norm": 0.4553566655837448,
+      "learning_rate": 9.97968265664445e-06,
+      "loss": 0.5228,
+      "step": 688
+    },
+    {
+      "epoch": 0.12623671674606082,
+      "grad_norm": 0.4682536997157449,
+      "learning_rate": 9.979393644244992e-06,
+      "loss": 0.5039,
+      "step": 689
+    },
+    {
+      "epoch": 0.12641993404177354,
+      "grad_norm": 0.42839595687214543,
+      "learning_rate": 9.979102594993533e-06,
+      "loss": 0.5128,
+      "step": 690
+    },
+    {
+      "epoch": 0.12660315133748626,
+      "grad_norm": 0.46092632925079513,
+      "learning_rate": 9.978809509009121e-06,
+      "loss": 0.5011,
+      "step": 691
+    },
+    {
+      "epoch": 0.12678636863319898,
+      "grad_norm": 0.4486176186332823,
+      "learning_rate": 9.97851438641165e-06,
+      "loss": 0.5169,
+      "step": 692
+    },
+    {
+      "epoch": 0.1269695859289117,
+      "grad_norm": 0.47419826995019465,
+      "learning_rate": 9.978217227321837e-06,
+      "loss": 0.5131,
+      "step": 693
+    },
+    {
+      "epoch": 0.1271528032246244,
+      "grad_norm": 0.4004091408746546,
+      "learning_rate": 9.97791803186124e-06,
+      "loss": 0.4743,
+      "step": 694
+    },
+    {
+      "epoch": 0.12733602052033713,
+      "grad_norm": 0.38849935615066844,
+      "learning_rate": 9.977616800152248e-06,
+      "loss": 0.4944,
+      "step": 695
+    },
+    {
+      "epoch": 0.12751923781604985,
+      "grad_norm": 0.44216099670793213,
+      "learning_rate": 9.977313532318078e-06,
+      "loss": 0.5072,
+      "step": 696
+    },
+    {
+      "epoch": 0.12770245511176256,
+      "grad_norm": 0.4462646919434802,
+      "learning_rate": 9.977008228482785e-06,
+      "loss": 0.4901,
+      "step": 697
+    },
+    {
+      "epoch": 0.12788567240747525,
+      "grad_norm": 0.44769942596462214,
+      "learning_rate": 9.976700888771259e-06,
+      "loss": 0.5263,
+      "step": 698
+    },
+    {
+      "epoch": 0.12806888970318797,
+      "grad_norm": 0.5113829977752261,
+      "learning_rate": 9.976391513309212e-06,
+      "loss": 0.5202,
+      "step": 699
+    },
+    {
+      "epoch": 0.1282521069989007,
+      "grad_norm": 0.4245344457429005,
+      "learning_rate": 9.976080102223202e-06,
+      "loss": 0.5182,
+      "step": 700
+    },
+    {
+      "epoch": 0.1284353242946134,
+      "grad_norm": 0.4305381223047699,
+      "learning_rate": 9.97576665564061e-06,
+      "loss": 0.507,
+      "step": 701
+    },
+    {
+      "epoch": 0.12861854159032612,
+      "grad_norm": 0.47072785404283757,
+      "learning_rate": 9.975451173689658e-06,
+      "loss": 0.5166,
+      "step": 702
+    },
+    {
+      "epoch": 0.12880175888603884,
+      "grad_norm": 0.6675845713373886,
+      "learning_rate": 9.975133656499392e-06,
+      "loss": 0.47,
+      "step": 703
+    },
+    {
+      "epoch": 0.12898497618175156,
+      "grad_norm": 0.4533261672443661,
+      "learning_rate": 9.974814104199694e-06,
+      "loss": 0.5362,
+      "step": 704
+    },
+    {
+      "epoch": 0.12916819347746428,
+      "grad_norm": 0.4503943455070683,
+      "learning_rate": 9.97449251692128e-06,
+      "loss": 0.5134,
+      "step": 705
+    },
+    {
+      "epoch": 0.129351410773177,
+      "grad_norm": 0.44545859946020006,
+      "learning_rate": 9.974168894795698e-06,
+      "loss": 0.4981,
+      "step": 706
+    },
+    {
+      "epoch": 0.1295346280688897,
+      "grad_norm": 0.430864283858029,
+      "learning_rate": 9.973843237955328e-06,
+      "loss": 0.5329,
+      "step": 707
+    },
+    {
+      "epoch": 0.12971784536460243,
+      "grad_norm": 0.7865986917626206,
+      "learning_rate": 9.973515546533379e-06,
+      "loss": 0.504,
+      "step": 708
+    },
+    {
+      "epoch": 0.12990106266031515,
+      "grad_norm": 0.44976211133712124,
+      "learning_rate": 9.973185820663897e-06,
+      "loss": 0.4737,
+      "step": 709
+    },
+    {
+      "epoch": 0.13008427995602784,
+      "grad_norm": 0.4549304394815992,
+      "learning_rate": 9.97285406048176e-06,
+      "loss": 0.5425,
+      "step": 710
+    },
+    {
+      "epoch": 0.13026749725174055,
+      "grad_norm": 0.5049990473394447,
+      "learning_rate": 9.972520266122676e-06,
+      "loss": 0.5117,
+      "step": 711
+    },
+    {
+      "epoch": 0.13045071454745327,
+      "grad_norm": 0.42657669286902156,
+      "learning_rate": 9.972184437723182e-06,
+      "loss": 0.5057,
+      "step": 712
+    },
+    {
+      "epoch": 0.130633931843166,
+      "grad_norm": 0.5299863700854671,
+      "learning_rate": 9.971846575420656e-06,
+      "loss": 0.5262,
+      "step": 713
+    },
+    {
+      "epoch": 0.1308171491388787,
+      "grad_norm": 0.45345740582998295,
+      "learning_rate": 9.9715066793533e-06,
+      "loss": 0.4997,
+      "step": 714
+    },
+    {
+      "epoch": 0.13100036643459143,
+      "grad_norm": 0.39393086616308026,
+      "learning_rate": 9.971164749660149e-06,
+      "loss": 0.4996,
+      "step": 715
+    },
+    {
+      "epoch": 0.13118358373030414,
+      "grad_norm": 0.4577611248522203,
+      "learning_rate": 9.970820786481075e-06,
+      "loss": 0.5018,
+      "step": 716
+    },
+    {
+      "epoch": 0.13136680102601686,
+      "grad_norm": 0.43279352533859367,
+      "learning_rate": 9.970474789956775e-06,
+      "loss": 0.4914,
+      "step": 717
+    },
+    {
+      "epoch": 0.13155001832172958,
+      "grad_norm": 0.4366448132250816,
+      "learning_rate": 9.970126760228785e-06,
+      "loss": 0.4864,
+      "step": 718
+    },
+    {
+      "epoch": 0.1317332356174423,
+      "grad_norm": 0.43998158412052296,
+      "learning_rate": 9.969776697439463e-06,
+      "loss": 0.4834,
+      "step": 719
+    },
+    {
+      "epoch": 0.131916452913155,
+      "grad_norm": 0.46940772392002544,
+      "learning_rate": 9.969424601732011e-06,
+      "loss": 0.501,
+      "step": 720
+    },
+    {
+      "epoch": 0.1320996702088677,
+      "grad_norm": 0.43117110167391715,
+      "learning_rate": 9.96907047325045e-06,
+      "loss": 0.5562,
+      "step": 721
+    },
+    {
+      "epoch": 0.13228288750458042,
+      "grad_norm": 0.4119352967972661,
+      "learning_rate": 9.968714312139642e-06,
+      "loss": 0.4748,
+      "step": 722
+    },
+    {
+      "epoch": 0.13246610480029314,
+      "grad_norm": 0.5028443530663289,
+      "learning_rate": 9.968356118545277e-06,
+      "loss": 0.5344,
+      "step": 723
+    },
+    {
+      "epoch": 0.13264932209600586,
+      "grad_norm": 0.43089419048593447,
+      "learning_rate": 9.967995892613875e-06,
+      "loss": 0.4704,
+      "step": 724
+    },
+    {
+      "epoch": 0.13283253939171857,
+      "grad_norm": 0.5044933769783081,
+      "learning_rate": 9.967633634492788e-06,
+      "loss": 0.4996,
+      "step": 725
+    },
+    {
+      "epoch": 0.1330157566874313,
+      "grad_norm": 0.442554097189932,
+      "learning_rate": 9.967269344330201e-06,
+      "loss": 0.5278,
+      "step": 726
+    },
+    {
+      "epoch": 0.133198973983144,
+      "grad_norm": 0.431094495901828,
+      "learning_rate": 9.966903022275131e-06,
+      "loss": 0.4943,
+      "step": 727
+    },
+    {
+      "epoch": 0.13338219127885673,
+      "grad_norm": 0.6222105275898999,
+      "learning_rate": 9.966534668477421e-06,
+      "loss": 0.5215,
+      "step": 728
+    },
+    {
+      "epoch": 0.13356540857456944,
+      "grad_norm": 0.4562449049230116,
+      "learning_rate": 9.96616428308775e-06,
+      "loss": 0.5112,
+      "step": 729
+    },
+    {
+      "epoch": 0.13374862587028216,
+      "grad_norm": 0.4249667668518143,
+      "learning_rate": 9.965791866257626e-06,
+      "loss": 0.5083,
+      "step": 730
+    },
+    {
+      "epoch": 0.13393184316599488,
+      "grad_norm": 0.42736565388331654,
+      "learning_rate": 9.96541741813939e-06,
+      "loss": 0.5078,
+      "step": 731
+    },
+    {
+      "epoch": 0.1341150604617076,
+      "grad_norm": 0.42789645341508464,
+      "learning_rate": 9.96504093888621e-06,
+      "loss": 0.5148,
+      "step": 732
+    },
+    {
+      "epoch": 0.1342982777574203,
+      "grad_norm": 0.5018533252539279,
+      "learning_rate": 9.964662428652088e-06,
+      "loss": 0.4994,
+      "step": 733
+    },
+    {
+      "epoch": 0.134481495053133,
+      "grad_norm": 0.4561559749816332,
+      "learning_rate": 9.964281887591856e-06,
+      "loss": 0.5036,
+      "step": 734
+    },
+    {
+      "epoch": 0.13466471234884572,
+      "grad_norm": 0.43874828747054045,
+      "learning_rate": 9.963899315861176e-06,
+      "loss": 0.4976,
+      "step": 735
+    },
+    {
+      "epoch": 0.13484792964455844,
+      "grad_norm": 0.44683273370213666,
+      "learning_rate": 9.963514713616544e-06,
+      "loss": 0.4837,
+      "step": 736
+    },
+    {
+      "epoch": 0.13503114694027116,
+      "grad_norm": 0.4575940082218496,
+      "learning_rate": 9.963128081015282e-06,
+      "loss": 0.531,
+      "step": 737
+    },
+    {
+      "epoch": 0.13521436423598387,
+      "grad_norm": 0.4190671986873532,
+      "learning_rate": 9.962739418215545e-06,
+      "loss": 0.525,
+      "step": 738
+    },
+    {
+      "epoch": 0.1353975815316966,
+      "grad_norm": 0.41284775197321,
+      "learning_rate": 9.962348725376318e-06,
+      "loss": 0.519,
+      "step": 739
+    },
+    {
+      "epoch": 0.1355807988274093,
+      "grad_norm": 0.41220938724190687,
+      "learning_rate": 9.961956002657414e-06,
+      "loss": 0.535,
+      "step": 740
+    },
+    {
+      "epoch": 0.13576401612312203,
+      "grad_norm": 0.4428272636340964,
+      "learning_rate": 9.961561250219482e-06,
+      "loss": 0.5287,
+      "step": 741
+    },
+    {
+      "epoch": 0.13594723341883475,
+      "grad_norm": 0.42565025293852765,
+      "learning_rate": 9.961164468223996e-06,
+      "loss": 0.5329,
+      "step": 742
+    },
+    {
+      "epoch": 0.13613045071454746,
+      "grad_norm": 0.5963216630193675,
+      "learning_rate": 9.960765656833263e-06,
+      "loss": 0.5036,
+      "step": 743
+    },
+    {
+      "epoch": 0.13631366801026018,
+      "grad_norm": 0.42494573499796934,
+      "learning_rate": 9.96036481621042e-06,
+      "loss": 0.4936,
+      "step": 744
+    },
+    {
+      "epoch": 0.13649688530597287,
+      "grad_norm": 0.39224716002216886,
+      "learning_rate": 9.959961946519431e-06,
+      "loss": 0.4704,
+      "step": 745
+    },
+    {
+      "epoch": 0.1366801026016856,
+      "grad_norm": 0.42574265198042266,
+      "learning_rate": 9.959557047925095e-06,
+      "loss": 0.4964,
+      "step": 746
+    },
+    {
+      "epoch": 0.1368633198973983,
+      "grad_norm": 0.457270415221301,
+      "learning_rate": 9.959150120593035e-06,
+      "loss": 0.5322,
+      "step": 747
+    },
+    {
+      "epoch": 0.13704653719311102,
+      "grad_norm": 0.4245718410150897,
+      "learning_rate": 9.95874116468971e-06,
+      "loss": 0.4952,
+      "step": 748
+    },
+    {
+      "epoch": 0.13722975448882374,
+      "grad_norm": 0.47589556246220993,
+      "learning_rate": 9.958330180382405e-06,
+      "loss": 0.5124,
+      "step": 749
+    },
+    {
+      "epoch": 0.13741297178453646,
+      "grad_norm": 0.45440270321018955,
+      "learning_rate": 9.957917167839238e-06,
+      "loss": 0.5109,
+      "step": 750
+    },
+    {
+      "epoch": 0.13759618908024918,
+      "grad_norm": 0.5232685566183007,
+      "learning_rate": 9.95750212722915e-06,
+      "loss": 0.5174,
+      "step": 751
+    },
+    {
+      "epoch": 0.1377794063759619,
+      "grad_norm": 0.538012343254673,
+      "learning_rate": 9.957085058721918e-06,
+      "loss": 0.5105,
+      "step": 752
+    },
+    {
+      "epoch": 0.1379626236716746,
+      "grad_norm": 0.38959097348139504,
+      "learning_rate": 9.956665962488148e-06,
+      "loss": 0.4533,
+      "step": 753
+    },
+    {
+      "epoch": 0.13814584096738733,
+      "grad_norm": 0.4404610898092818,
+      "learning_rate": 9.956244838699271e-06,
+      "loss": 0.4978,
+      "step": 754
+    },
+    {
+      "epoch": 0.13832905826310005,
+      "grad_norm": 0.47384922621236214,
+      "learning_rate": 9.955821687527554e-06,
+      "loss": 0.5309,
+      "step": 755
+    },
+    {
+      "epoch": 0.13851227555881276,
+      "grad_norm": 0.45266354792605684,
+      "learning_rate": 9.955396509146084e-06,
+      "loss": 0.5151,
+      "step": 756
+    },
+    {
+      "epoch": 0.13869549285452545,
+      "grad_norm": 0.4098036061670406,
+      "learning_rate": 9.95496930372879e-06,
+      "loss": 0.5,
+      "step": 757
+    },
+    {
+      "epoch": 0.13887871015023817,
+      "grad_norm": 0.4520393902857722,
+      "learning_rate": 9.954540071450418e-06,
+      "loss": 0.4766,
+      "step": 758
+    },
+    {
+      "epoch": 0.1390619274459509,
+      "grad_norm": 0.41592431105518124,
+      "learning_rate": 9.95410881248655e-06,
+      "loss": 0.5279,
+      "step": 759
+    },
+    {
+      "epoch": 0.1392451447416636,
+      "grad_norm": 0.9055926932881602,
+      "learning_rate": 9.953675527013594e-06,
+      "loss": 0.4544,
+      "step": 760
+    },
+    {
+      "epoch": 0.13942836203737632,
+      "grad_norm": 0.4230083797544526,
+      "learning_rate": 9.953240215208787e-06,
+      "loss": 0.5174,
+      "step": 761
+    },
+    {
+      "epoch": 0.13961157933308904,
+      "grad_norm": 0.46167105853029766,
+      "learning_rate": 9.9528028772502e-06,
+      "loss": 0.5087,
+      "step": 762
+    },
+    {
+      "epoch": 0.13979479662880176,
+      "grad_norm": 0.4614988799168283,
+      "learning_rate": 9.952363513316727e-06,
+      "loss": 0.5035,
+      "step": 763
+    },
+    {
+      "epoch": 0.13997801392451448,
+      "grad_norm": 0.4599378559524754,
+      "learning_rate": 9.951922123588091e-06,
+      "loss": 0.4982,
+      "step": 764
+    },
+    {
+      "epoch": 0.1401612312202272,
+      "grad_norm": 0.4169187563233499,
+      "learning_rate": 9.951478708244847e-06,
+      "loss": 0.5205,
+      "step": 765
+    },
+    {
+      "epoch": 0.1403444485159399,
+      "grad_norm": 1.4346541301623688,
+      "learning_rate": 9.951033267468375e-06,
+      "loss": 0.4704,
+      "step": 766
+    },
+    {
+      "epoch": 0.14052766581165263,
+      "grad_norm": 0.4477016515011828,
+      "learning_rate": 9.950585801440889e-06,
+      "loss": 0.489,
+      "step": 767
+    },
+    {
+      "epoch": 0.14071088310736535,
+      "grad_norm": 0.4178701034776549,
+      "learning_rate": 9.950136310345425e-06,
+      "loss": 0.5147,
+      "step": 768
+    },
+    {
+      "epoch": 0.14089410040307804,
+      "grad_norm": 0.4893104037393506,
+      "learning_rate": 9.949684794365848e-06,
+      "loss": 0.4913,
+      "step": 769
+    },
+    {
+      "epoch": 0.14107731769879076,
+      "grad_norm": 0.4018838415277579,
+      "learning_rate": 9.949231253686857e-06,
+      "loss": 0.464,
+      "step": 770
+    },
+    {
+      "epoch": 0.14126053499450347,
+      "grad_norm": 0.38608379924790914,
+      "learning_rate": 9.948775688493974e-06,
+      "loss": 0.5173,
+      "step": 771
+    },
+    {
+      "epoch": 0.1414437522902162,
+      "grad_norm": 0.4162388206396457,
+      "learning_rate": 9.948318098973552e-06,
+      "loss": 0.4917,
+      "step": 772
+    },
+    {
+      "epoch": 0.1416269695859289,
+      "grad_norm": 0.3781950328417496,
+      "learning_rate": 9.947858485312772e-06,
+      "loss": 0.4446,
+      "step": 773
+    },
+    {
+      "epoch": 0.14181018688164163,
+      "grad_norm": 0.45959051057998834,
+      "learning_rate": 9.947396847699638e-06,
+      "loss": 0.4918,
+      "step": 774
+    },
+    {
+      "epoch": 0.14199340417735434,
+      "grad_norm": 0.480317102541819,
+      "learning_rate": 9.946933186322988e-06,
+      "loss": 0.5369,
+      "step": 775
+    },
+    {
+      "epoch": 0.14217662147306706,
+      "grad_norm": 0.5008985493720574,
+      "learning_rate": 9.946467501372485e-06,
+      "loss": 0.4992,
+      "step": 776
+    },
+    {
+      "epoch": 0.14235983876877978,
+      "grad_norm": 0.4117591563183844,
+      "learning_rate": 9.94599979303862e-06,
+      "loss": 0.4818,
+      "step": 777
+    },
+    {
+      "epoch": 0.1425430560644925,
+      "grad_norm": 0.531108956987235,
+      "learning_rate": 9.945530061512714e-06,
+      "loss": 0.517,
+      "step": 778
+    },
+    {
+      "epoch": 0.1427262733602052,
+      "grad_norm": 0.43817523582155865,
+      "learning_rate": 9.945058306986911e-06,
+      "loss": 0.5473,
+      "step": 779
+    },
+    {
+      "epoch": 0.14290949065591793,
+      "grad_norm": 0.4747080227792923,
+      "learning_rate": 9.944584529654187e-06,
+      "loss": 0.4819,
+      "step": 780
+    },
+    {
+      "epoch": 0.14309270795163062,
+      "grad_norm": 0.4249215515551791,
+      "learning_rate": 9.944108729708342e-06,
+      "loss": 0.4979,
+      "step": 781
+    },
+    {
+      "epoch": 0.14327592524734334,
+      "grad_norm": 0.43558257752578505,
+      "learning_rate": 9.943630907344008e-06,
+      "loss": 0.484,
+      "step": 782
+    },
+    {
+      "epoch": 0.14345914254305606,
+      "grad_norm": 0.4750219883418816,
+      "learning_rate": 9.943151062756638e-06,
+      "loss": 0.4825,
+      "step": 783
+    },
+    {
+      "epoch": 0.14364235983876877,
+      "grad_norm": 0.4332476417990451,
+      "learning_rate": 9.942669196142516e-06,
+      "loss": 0.4929,
+      "step": 784
+    },
+    {
+      "epoch": 0.1438255771344815,
+      "grad_norm": 0.48307720479460303,
+      "learning_rate": 9.942185307698754e-06,
+      "loss": 0.5099,
+      "step": 785
+    },
+    {
+      "epoch": 0.1440087944301942,
+      "grad_norm": 0.43004630110605213,
+      "learning_rate": 9.941699397623289e-06,
+      "loss": 0.4435,
+      "step": 786
+    },
+    {
+      "epoch": 0.14419201172590693,
+      "grad_norm": 0.5526090978976514,
+      "learning_rate": 9.941211466114883e-06,
+      "loss": 0.4831,
+      "step": 787
+    },
+    {
+      "epoch": 0.14437522902161964,
+      "grad_norm": 0.4299548315161719,
+      "learning_rate": 9.94072151337313e-06,
+      "loss": 0.5055,
+      "step": 788
+    },
+    {
+      "epoch": 0.14455844631733236,
+      "grad_norm": 0.46488795405005734,
+      "learning_rate": 9.940229539598449e-06,
+      "loss": 0.5247,
+      "step": 789
+    },
+    {
+      "epoch": 0.14474166361304508,
+      "grad_norm": 0.45711196453956754,
+      "learning_rate": 9.939735544992084e-06,
+      "loss": 0.5155,
+      "step": 790
+    },
+    {
+      "epoch": 0.1449248809087578,
+      "grad_norm": 0.39377908902960373,
+      "learning_rate": 9.939239529756106e-06,
+      "loss": 0.4861,
+      "step": 791
+    },
+    {
+      "epoch": 0.14510809820447051,
+      "grad_norm": 0.3780973854580167,
+      "learning_rate": 9.938741494093413e-06,
+      "loss": 0.4743,
+      "step": 792
+    },
+    {
+      "epoch": 0.1452913155001832,
+      "grad_norm": 0.4442667746426281,
+      "learning_rate": 9.93824143820773e-06,
+      "loss": 0.5124,
+      "step": 793
+    },
+    {
+      "epoch": 0.14547453279589592,
+      "grad_norm": 0.47139079982459514,
+      "learning_rate": 9.93773936230361e-06,
+      "loss": 0.5095,
+      "step": 794
+    },
+    {
+      "epoch": 0.14565775009160864,
+      "grad_norm": 0.47478113201389704,
+      "learning_rate": 9.937235266586425e-06,
+      "loss": 0.5242,
+      "step": 795
+    },
+    {
+      "epoch": 0.14584096738732136,
+      "grad_norm": 0.4253566981856297,
+      "learning_rate": 9.936729151262383e-06,
+      "loss": 0.4816,
+      "step": 796
+    },
+    {
+      "epoch": 0.14602418468303408,
+      "grad_norm": 0.3995444614750483,
+      "learning_rate": 9.936221016538514e-06,
+      "loss": 0.5114,
+      "step": 797
+    },
+    {
+      "epoch": 0.1462074019787468,
+      "grad_norm": 0.3958644449360531,
+      "learning_rate": 9.935710862622671e-06,
+      "loss": 0.5044,
+      "step": 798
+    },
+    {
+      "epoch": 0.1463906192744595,
+      "grad_norm": 0.39672728856696343,
+      "learning_rate": 9.935198689723537e-06,
+      "loss": 0.5079,
+      "step": 799
+    },
+    {
+      "epoch": 0.14657383657017223,
+      "grad_norm": 0.4401216989939036,
+      "learning_rate": 9.934684498050619e-06,
+      "loss": 0.5169,
+      "step": 800
+    },
+    {
+      "epoch": 0.14675705386588495,
+      "grad_norm": 0.42531985045283716,
+      "learning_rate": 9.93416828781425e-06,
+      "loss": 0.5039,
+      "step": 801
+    },
+    {
+      "epoch": 0.14694027116159766,
+      "grad_norm": 0.43641771981673,
+      "learning_rate": 9.93365005922559e-06,
+      "loss": 0.4705,
+      "step": 802
+    },
+    {
+      "epoch": 0.14712348845731038,
+      "grad_norm": 0.4154214727248315,
+      "learning_rate": 9.933129812496623e-06,
+      "loss": 0.4801,
+      "step": 803
+    },
+    {
+      "epoch": 0.1473067057530231,
+      "grad_norm": 0.4668765888893056,
+      "learning_rate": 9.932607547840161e-06,
+      "loss": 0.5262,
+      "step": 804
+    },
+    {
+      "epoch": 0.1474899230487358,
+      "grad_norm": 0.45163409733644266,
+      "learning_rate": 9.932083265469836e-06,
+      "loss": 0.5009,
+      "step": 805
+    },
+    {
+      "epoch": 0.1476731403444485,
+      "grad_norm": 0.44602336917824525,
+      "learning_rate": 9.93155696560011e-06,
+      "loss": 0.5071,
+      "step": 806
+    },
+    {
+      "epoch": 0.14785635764016122,
+      "grad_norm": 0.40571563214546336,
+      "learning_rate": 9.931028648446273e-06,
+      "loss": 0.4867,
+      "step": 807
+    },
+    {
+      "epoch": 0.14803957493587394,
+      "grad_norm": 0.412866633306789,
+      "learning_rate": 9.930498314224433e-06,
+      "loss": 0.5204,
+      "step": 808
+    },
+    {
+      "epoch": 0.14822279223158666,
+      "grad_norm": 0.4491417286505596,
+      "learning_rate": 9.929965963151526e-06,
+      "loss": 0.4959,
+      "step": 809
+    },
+    {
+      "epoch": 0.14840600952729938,
+      "grad_norm": 0.483698524685917,
+      "learning_rate": 9.929431595445315e-06,
+      "loss": 0.4945,
+      "step": 810
+    },
+    {
+      "epoch": 0.1485892268230121,
+      "grad_norm": 0.45809625895317874,
+      "learning_rate": 9.928895211324387e-06,
+      "loss": 0.4795,
+      "step": 811
+    },
+    {
+      "epoch": 0.1487724441187248,
+      "grad_norm": 0.4671855759216309,
+      "learning_rate": 9.928356811008153e-06,
+      "loss": 0.502,
+      "step": 812
+    },
+    {
+      "epoch": 0.14895566141443753,
+      "grad_norm": 0.49991252806247843,
+      "learning_rate": 9.927816394716847e-06,
+      "loss": 0.4989,
+      "step": 813
+    },
+    {
+      "epoch": 0.14913887871015025,
+      "grad_norm": 0.4930888660574426,
+      "learning_rate": 9.92727396267153e-06,
+      "loss": 0.5256,
+      "step": 814
+    },
+    {
+      "epoch": 0.14932209600586296,
+      "grad_norm": 0.4301245288190715,
+      "learning_rate": 9.926729515094092e-06,
+      "loss": 0.5062,
+      "step": 815
+    },
+    {
+      "epoch": 0.14950531330157568,
+      "grad_norm": 0.4254345495226345,
+      "learning_rate": 9.926183052207235e-06,
+      "loss": 0.4885,
+      "step": 816
+    },
+    {
+      "epoch": 0.14968853059728837,
+      "grad_norm": 0.40854658960332646,
+      "learning_rate": 9.925634574234499e-06,
+      "loss": 0.4909,
+      "step": 817
+    },
+    {
+      "epoch": 0.1498717478930011,
+      "grad_norm": 0.4307966660913799,
+      "learning_rate": 9.925084081400241e-06,
+      "loss": 0.4974,
+      "step": 818
+    },
+    {
+      "epoch": 0.1500549651887138,
+      "grad_norm": 0.44710983697734263,
+      "learning_rate": 9.924531573929641e-06,
+      "loss": 0.5107,
+      "step": 819
+    },
+    {
+      "epoch": 0.15023818248442652,
+      "grad_norm": 0.47247983608563787,
+      "learning_rate": 9.923977052048708e-06,
+      "loss": 0.5185,
+      "step": 820
+    },
+    {
+      "epoch": 0.15042139978013924,
+      "grad_norm": 0.43408666873611645,
+      "learning_rate": 9.923420515984272e-06,
+      "loss": 0.5256,
+      "step": 821
+    },
+    {
+      "epoch": 0.15060461707585196,
+      "grad_norm": 0.3618361050633503,
+      "learning_rate": 9.922861965963987e-06,
+      "loss": 0.4835,
+      "step": 822
+    },
+    {
+      "epoch": 0.15078783437156468,
+      "grad_norm": 0.478572088704784,
+      "learning_rate": 9.922301402216334e-06,
+      "loss": 0.4971,
+      "step": 823
+    },
+    {
+      "epoch": 0.1509710516672774,
+      "grad_norm": 0.4956723559981853,
+      "learning_rate": 9.921738824970611e-06,
+      "loss": 0.4664,
+      "step": 824
+    },
+    {
+      "epoch": 0.1511542689629901,
+      "grad_norm": 0.43104833444065105,
+      "learning_rate": 9.921174234456947e-06,
+      "loss": 0.5056,
+      "step": 825
+    },
+    {
+      "epoch": 0.15133748625870283,
+      "grad_norm": 0.38636058698792763,
+      "learning_rate": 9.920607630906289e-06,
+      "loss": 0.509,
+      "step": 826
+    },
+    {
+      "epoch": 0.15152070355441555,
+      "grad_norm": 0.4638108326048624,
+      "learning_rate": 9.920039014550413e-06,
+      "loss": 0.4928,
+      "step": 827
+    },
+    {
+      "epoch": 0.15170392085012827,
+      "grad_norm": 0.38962500919500936,
+      "learning_rate": 9.919468385621912e-06,
+      "loss": 0.5025,
+      "step": 828
+    },
+    {
+      "epoch": 0.15188713814584096,
+      "grad_norm": 0.423466480397551,
+      "learning_rate": 9.918895744354204e-06,
+      "loss": 0.4903,
+      "step": 829
+    },
+    {
+      "epoch": 0.15207035544155367,
+      "grad_norm": 0.5945183810824295,
+      "learning_rate": 9.918321090981537e-06,
+      "loss": 0.4961,
+      "step": 830
+    },
+    {
+      "epoch": 0.1522535727372664,
+      "grad_norm": 0.3857346095853992,
+      "learning_rate": 9.917744425738971e-06,
+      "loss": 0.4903,
+      "step": 831
+    },
+    {
+      "epoch": 0.1524367900329791,
+      "grad_norm": 0.4368019305898112,
+      "learning_rate": 9.917165748862398e-06,
+      "loss": 0.4985,
+      "step": 832
+    },
+    {
+      "epoch": 0.15262000732869183,
+      "grad_norm": 0.4622792781214721,
+      "learning_rate": 9.916585060588526e-06,
+      "loss": 0.5298,
+      "step": 833
+    },
+    {
+      "epoch": 0.15280322462440454,
+      "grad_norm": 0.43464073300377387,
+      "learning_rate": 9.916002361154894e-06,
+      "loss": 0.4956,
+      "step": 834
+    },
+    {
+      "epoch": 0.15298644192011726,
+      "grad_norm": 0.37914154472744427,
+      "learning_rate": 9.915417650799855e-06,
+      "loss": 0.4914,
+      "step": 835
+    },
+    {
+      "epoch": 0.15316965921582998,
+      "grad_norm": 0.41110768627195465,
+      "learning_rate": 9.914830929762588e-06,
+      "loss": 0.4933,
+      "step": 836
+    },
+    {
+      "epoch": 0.1533528765115427,
+      "grad_norm": 0.3922989216974703,
+      "learning_rate": 9.914242198283099e-06,
+      "loss": 0.5,
+      "step": 837
+    },
+    {
+      "epoch": 0.1535360938072554,
+      "grad_norm": 0.4310098182316404,
+      "learning_rate": 9.91365145660221e-06,
+      "loss": 0.4817,
+      "step": 838
+    },
+    {
+      "epoch": 0.15371931110296813,
+      "grad_norm": 0.40690680117071937,
+      "learning_rate": 9.913058704961565e-06,
+      "loss": 0.4778,
+      "step": 839
+    },
+    {
+      "epoch": 0.15390252839868085,
+      "grad_norm": 0.4584415585967629,
+      "learning_rate": 9.912463943603635e-06,
+      "loss": 0.4585,
+      "step": 840
+    },
+    {
+      "epoch": 0.15408574569439354,
+      "grad_norm": 0.48734349658323495,
+      "learning_rate": 9.911867172771711e-06,
+      "loss": 0.5345,
+      "step": 841
+    },
+    {
+      "epoch": 0.15426896299010626,
+      "grad_norm": 0.442686833828428,
+      "learning_rate": 9.911268392709908e-06,
+      "loss": 0.5134,
+      "step": 842
+    },
+    {
+      "epoch": 0.15445218028581897,
+      "grad_norm": 0.44506293028939,
+      "learning_rate": 9.910667603663156e-06,
+      "loss": 0.5158,
+      "step": 843
+    },
+    {
+      "epoch": 0.1546353975815317,
+      "grad_norm": 0.4311992637968541,
+      "learning_rate": 9.910064805877214e-06,
+      "loss": 0.4837,
+      "step": 844
+    },
+    {
+      "epoch": 0.1548186148772444,
+      "grad_norm": 0.4493099076015729,
+      "learning_rate": 9.90945999959866e-06,
+      "loss": 0.5108,
+      "step": 845
+    },
+    {
+      "epoch": 0.15500183217295713,
+      "grad_norm": 0.45478365363321205,
+      "learning_rate": 9.908853185074896e-06,
+      "loss": 0.4998,
+      "step": 846
+    },
+    {
+      "epoch": 0.15518504946866984,
+      "grad_norm": 0.46301414888311854,
+      "learning_rate": 9.90824436255414e-06,
+      "loss": 0.5225,
+      "step": 847
+    },
+    {
+      "epoch": 0.15536826676438256,
+      "grad_norm": 0.40118617155548936,
+      "learning_rate": 9.907633532285435e-06,
+      "loss": 0.4927,
+      "step": 848
+    },
+    {
+      "epoch": 0.15555148406009528,
+      "grad_norm": 0.49481295874709574,
+      "learning_rate": 9.907020694518646e-06,
+      "loss": 0.4933,
+      "step": 849
+    },
+    {
+      "epoch": 0.155734701355808,
+      "grad_norm": 0.4102626498510104,
+      "learning_rate": 9.90640584950446e-06,
+      "loss": 0.5111,
+      "step": 850
+    },
+    {
+      "epoch": 0.15591791865152071,
+      "grad_norm": 0.46514690031713113,
+      "learning_rate": 9.905788997494377e-06,
+      "loss": 0.4728,
+      "step": 851
+    },
+    {
+      "epoch": 0.15610113594723343,
+      "grad_norm": 0.4054441718781369,
+      "learning_rate": 9.905170138740732e-06,
+      "loss": 0.4828,
+      "step": 852
+    },
+    {
+      "epoch": 0.15628435324294612,
+      "grad_norm": 0.4400252577977126,
+      "learning_rate": 9.904549273496666e-06,
+      "loss": 0.501,
+      "step": 853
+    },
+    {
+      "epoch": 0.15646757053865884,
+      "grad_norm": 0.42167618939817325,
+      "learning_rate": 9.903926402016153e-06,
+      "loss": 0.4922,
+      "step": 854
+    },
+    {
+      "epoch": 0.15665078783437156,
+      "grad_norm": 0.4093707700968775,
+      "learning_rate": 9.90330152455398e-06,
+      "loss": 0.5014,
+      "step": 855
+    },
+    {
+      "epoch": 0.15683400513008428,
+      "grad_norm": 0.4523522809487888,
+      "learning_rate": 9.902674641365756e-06,
+      "loss": 0.4829,
+      "step": 856
+    },
+    {
+      "epoch": 0.157017222425797,
+      "grad_norm": 0.49143897815029397,
+      "learning_rate": 9.902045752707916e-06,
+      "loss": 0.5074,
+      "step": 857
+    },
+    {
+      "epoch": 0.1572004397215097,
+      "grad_norm": 0.43621740114283203,
+      "learning_rate": 9.901414858837706e-06,
+      "loss": 0.5182,
+      "step": 858
+    },
+    {
+      "epoch": 0.15738365701722243,
+      "grad_norm": 0.40732498942379025,
+      "learning_rate": 9.9007819600132e-06,
+      "loss": 0.4674,
+      "step": 859
+    },
+    {
+      "epoch": 0.15756687431293515,
+      "grad_norm": 0.43675476681733244,
+      "learning_rate": 9.900147056493285e-06,
+      "loss": 0.5374,
+      "step": 860
+    },
+    {
+      "epoch": 0.15775009160864786,
+      "grad_norm": 0.43060074322101793,
+      "learning_rate": 9.89951014853768e-06,
+      "loss": 0.4635,
+      "step": 861
+    },
+    {
+      "epoch": 0.15793330890436058,
+      "grad_norm": 0.41179202097945616,
+      "learning_rate": 9.898871236406907e-06,
+      "loss": 0.4951,
+      "step": 862
+    },
+    {
+      "epoch": 0.1581165262000733,
+      "grad_norm": 0.371370018002044,
+      "learning_rate": 9.898230320362323e-06,
+      "loss": 0.479,
+      "step": 863
+    },
+    {
+      "epoch": 0.158299743495786,
+      "grad_norm": 0.4300134296591519,
+      "learning_rate": 9.897587400666097e-06,
+      "loss": 0.4868,
+      "step": 864
+    },
+    {
+      "epoch": 0.1584829607914987,
+      "grad_norm": 0.43589209027697184,
+      "learning_rate": 9.896942477581221e-06,
+      "loss": 0.5124,
+      "step": 865
+    },
+    {
+      "epoch": 0.15866617808721142,
+      "grad_norm": 0.575247632832188,
+      "learning_rate": 9.896295551371503e-06,
+      "loss": 0.5053,
+      "step": 866
+    },
+    {
+      "epoch": 0.15884939538292414,
+      "grad_norm": 0.4273872012355264,
+      "learning_rate": 9.89564662230157e-06,
+      "loss": 0.5138,
+      "step": 867
+    },
+    {
+      "epoch": 0.15903261267863686,
+      "grad_norm": 0.4543163505662005,
+      "learning_rate": 9.894995690636874e-06,
+      "loss": 0.5229,
+      "step": 868
+    },
+    {
+      "epoch": 0.15921582997434958,
+      "grad_norm": 0.4491410597219586,
+      "learning_rate": 9.894342756643684e-06,
+      "loss": 0.5045,
+      "step": 869
+    },
+    {
+      "epoch": 0.1593990472700623,
+      "grad_norm": 0.47087708750396473,
+      "learning_rate": 9.893687820589081e-06,
+      "loss": 0.5087,
+      "step": 870
+    },
+    {
+      "epoch": 0.159582264565775,
+      "grad_norm": 0.4240236693967998,
+      "learning_rate": 9.893030882740977e-06,
+      "loss": 0.5052,
+      "step": 871
+    },
+    {
+      "epoch": 0.15976548186148773,
+      "grad_norm": 0.4287525963131784,
+      "learning_rate": 9.892371943368092e-06,
+      "loss": 0.527,
+      "step": 872
+    },
+    {
+      "epoch": 0.15994869915720045,
+      "grad_norm": 0.4131563755872441,
+      "learning_rate": 9.891711002739971e-06,
+      "loss": 0.5074,
+      "step": 873
+    },
+    {
+      "epoch": 0.16013191645291316,
+      "grad_norm": 0.39231143806918656,
+      "learning_rate": 9.891048061126975e-06,
+      "loss": 0.5018,
+      "step": 874
+    },
+    {
+      "epoch": 0.16031513374862588,
+      "grad_norm": 0.4391493046066672,
+      "learning_rate": 9.890383118800287e-06,
+      "loss": 0.518,
+      "step": 875
+    },
+    {
+      "epoch": 0.16049835104433857,
+      "grad_norm": 0.468016398258625,
+      "learning_rate": 9.889716176031903e-06,
+      "loss": 0.5191,
+      "step": 876
+    },
+    {
+      "epoch": 0.1606815683400513,
+      "grad_norm": 0.4619425954928644,
+      "learning_rate": 9.88904723309464e-06,
+      "loss": 0.5114,
+      "step": 877
+    },
+    {
+      "epoch": 0.160864785635764,
+      "grad_norm": 0.4581492770476195,
+      "learning_rate": 9.888376290262134e-06,
+      "loss": 0.5349,
+      "step": 878
+    },
+    {
+      "epoch": 0.16104800293147672,
+      "grad_norm": 0.390022001202663,
+      "learning_rate": 9.887703347808838e-06,
+      "loss": 0.507,
+      "step": 879
+    },
+    {
+      "epoch": 0.16123122022718944,
+      "grad_norm": 0.38271876665722593,
+      "learning_rate": 9.887028406010026e-06,
+      "loss": 0.4987,
+      "step": 880
+    },
+    {
+      "epoch": 0.16141443752290216,
+      "grad_norm": 0.39991523006604696,
+      "learning_rate": 9.886351465141785e-06,
+      "loss": 0.4807,
+      "step": 881
+    },
+    {
+      "epoch": 0.16159765481861488,
+      "grad_norm": 0.3867195675970219,
+      "learning_rate": 9.88567252548102e-06,
+      "loss": 0.515,
+      "step": 882
+    },
+    {
+      "epoch": 0.1617808721143276,
+      "grad_norm": 0.47566001886735476,
+      "learning_rate": 9.884991587305459e-06,
+      "loss": 0.4461,
+      "step": 883
+    },
+    {
+      "epoch": 0.1619640894100403,
+      "grad_norm": 0.4453653597243864,
+      "learning_rate": 9.884308650893642e-06,
+      "loss": 0.5493,
+      "step": 884
+    },
+    {
+      "epoch": 0.16214730670575303,
+      "grad_norm": 0.38787040868110323,
+      "learning_rate": 9.883623716524929e-06,
+      "loss": 0.4998,
+      "step": 885
+    },
+    {
+      "epoch": 0.16233052400146575,
+      "grad_norm": 0.4570258592705414,
+      "learning_rate": 9.882936784479498e-06,
+      "loss": 0.5226,
+      "step": 886
+    },
+    {
+      "epoch": 0.16251374129717847,
+      "grad_norm": 0.4438200453897428,
+      "learning_rate": 9.882247855038339e-06,
+      "loss": 0.5073,
+      "step": 887
+    },
+    {
+      "epoch": 0.16269695859289116,
+      "grad_norm": 0.3887912773501902,
+      "learning_rate": 9.881556928483266e-06,
+      "loss": 0.5213,
+      "step": 888
+    },
+    {
+      "epoch": 0.16288017588860387,
+      "grad_norm": 0.44048355093202696,
+      "learning_rate": 9.880864005096906e-06,
+      "loss": 0.511,
+      "step": 889
+    },
+    {
+      "epoch": 0.1630633931843166,
+      "grad_norm": 0.44235079469227967,
+      "learning_rate": 9.880169085162703e-06,
+      "loss": 0.5223,
+      "step": 890
+    },
+    {
+      "epoch": 0.1632466104800293,
+      "grad_norm": 0.3877832289424,
+      "learning_rate": 9.87947216896492e-06,
+      "loss": 0.4537,
+      "step": 891
+    },
+    {
+      "epoch": 0.16342982777574203,
+      "grad_norm": 0.44869938145307137,
+      "learning_rate": 9.878773256788635e-06,
+      "loss": 0.5013,
+      "step": 892
+    },
+    {
+      "epoch": 0.16361304507145474,
+      "grad_norm": 0.46375171983760244,
+      "learning_rate": 9.878072348919738e-06,
+      "loss": 0.489,
+      "step": 893
+    },
+    {
+      "epoch": 0.16379626236716746,
+      "grad_norm": 0.49608852880970994,
+      "learning_rate": 9.877369445644945e-06,
+      "loss": 0.5177,
+      "step": 894
+    },
+    {
+      "epoch": 0.16397947966288018,
+      "grad_norm": 0.4082183549684993,
+      "learning_rate": 9.876664547251781e-06,
+      "loss": 0.5091,
+      "step": 895
+    },
+    {
+      "epoch": 0.1641626969585929,
+      "grad_norm": 0.4347968628628821,
+      "learning_rate": 9.875957654028588e-06,
+      "loss": 0.5242,
+      "step": 896
+    },
+    {
+      "epoch": 0.16434591425430561,
+      "grad_norm": 0.45545255399303036,
+      "learning_rate": 9.875248766264527e-06,
+      "loss": 0.5406,
+      "step": 897
+    },
+    {
+      "epoch": 0.16452913155001833,
+      "grad_norm": 0.41358644063630506,
+      "learning_rate": 9.874537884249574e-06,
+      "loss": 0.511,
+      "step": 898
+    },
+    {
+      "epoch": 0.16471234884573105,
+      "grad_norm": 0.4576667619206321,
+      "learning_rate": 9.873825008274514e-06,
+      "loss": 0.5073,
+      "step": 899
+    },
+    {
+      "epoch": 0.16489556614144374,
+      "grad_norm": 0.47721781865779744,
+      "learning_rate": 9.87311013863096e-06,
+      "loss": 0.5044,
+      "step": 900
+    },
+    {
+      "epoch": 0.16507878343715646,
+      "grad_norm": 0.4622288619794578,
+      "learning_rate": 9.872393275611329e-06,
+      "loss": 0.5096,
+      "step": 901
+    },
+    {
+      "epoch": 0.16526200073286917,
+      "grad_norm": 0.41177688817899083,
+      "learning_rate": 9.871674419508864e-06,
+      "loss": 0.523,
+      "step": 902
+    },
+    {
+      "epoch": 0.1654452180285819,
+      "grad_norm": 0.371054810469729,
+      "learning_rate": 9.87095357061761e-06,
+      "loss": 0.4897,
+      "step": 903
+    },
+    {
+      "epoch": 0.1656284353242946,
+      "grad_norm": 0.41574608839188226,
+      "learning_rate": 9.87023072923244e-06,
+      "loss": 0.4823,
+      "step": 904
+    },
+    {
+      "epoch": 0.16581165262000733,
+      "grad_norm": 0.4218915251343208,
+      "learning_rate": 9.869505895649036e-06,
+      "loss": 0.5079,
+      "step": 905
+    },
+    {
+      "epoch": 0.16599486991572004,
+      "grad_norm": 0.5549831381708766,
+      "learning_rate": 9.868779070163895e-06,
+      "loss": 0.484,
+      "step": 906
+    },
+    {
+      "epoch": 0.16617808721143276,
+      "grad_norm": 0.3891473564827362,
+      "learning_rate": 9.868050253074328e-06,
+      "loss": 0.4903,
+      "step": 907
+    },
+    {
+      "epoch": 0.16636130450714548,
+      "grad_norm": 0.4268945705632106,
+      "learning_rate": 9.867319444678465e-06,
+      "loss": 0.5024,
+      "step": 908
+    },
+    {
+      "epoch": 0.1665445218028582,
+      "grad_norm": 0.40332206359603545,
+      "learning_rate": 9.866586645275247e-06,
+      "loss": 0.4477,
+      "step": 909
+    },
+    {
+      "epoch": 0.16672773909857092,
+      "grad_norm": 0.3720747589583317,
+      "learning_rate": 9.86585185516443e-06,
+      "loss": 0.469,
+      "step": 910
+    },
+    {
+      "epoch": 0.16691095639428363,
+      "grad_norm": 0.451113552639029,
+      "learning_rate": 9.865115074646583e-06,
+      "loss": 0.5221,
+      "step": 911
+    },
+    {
+      "epoch": 0.16709417368999632,
+      "grad_norm": 0.4454998332020719,
+      "learning_rate": 9.864376304023092e-06,
+      "loss": 0.495,
+      "step": 912
+    },
+    {
+      "epoch": 0.16727739098570904,
+      "grad_norm": 0.5363910151131063,
+      "learning_rate": 9.863635543596156e-06,
+      "loss": 0.5233,
+      "step": 913
+    },
+    {
+      "epoch": 0.16746060828142176,
+      "grad_norm": 0.4392403856880424,
+      "learning_rate": 9.862892793668787e-06,
+      "loss": 0.4998,
+      "step": 914
+    },
+    {
+      "epoch": 0.16764382557713448,
+      "grad_norm": 0.42093143672832734,
+      "learning_rate": 9.862148054544812e-06,
+      "loss": 0.4977,
+      "step": 915
+    },
+    {
+      "epoch": 0.1678270428728472,
+      "grad_norm": 0.39275787265485074,
+      "learning_rate": 9.86140132652887e-06,
+      "loss": 0.4694,
+      "step": 916
+    },
+    {
+      "epoch": 0.1680102601685599,
+      "grad_norm": 0.4159908861548446,
+      "learning_rate": 9.860652609926417e-06,
+      "loss": 0.4886,
+      "step": 917
+    },
+    {
+      "epoch": 0.16819347746427263,
+      "grad_norm": 0.38579180273261776,
+      "learning_rate": 9.859901905043718e-06,
+      "loss": 0.48,
+      "step": 918
+    },
+    {
+      "epoch": 0.16837669475998535,
+      "grad_norm": 0.4273621763370794,
+      "learning_rate": 9.859149212187855e-06,
+      "loss": 0.512,
+      "step": 919
+    },
+    {
+      "epoch": 0.16855991205569806,
+      "grad_norm": 0.45913559298162226,
+      "learning_rate": 9.85839453166672e-06,
+      "loss": 0.4877,
+      "step": 920
+    },
+    {
+      "epoch": 0.16874312935141078,
+      "grad_norm": 0.3847363797846655,
+      "learning_rate": 9.85763786378902e-06,
+      "loss": 0.4586,
+      "step": 921
+    },
+    {
+      "epoch": 0.1689263466471235,
+      "grad_norm": 0.4872256895625686,
+      "learning_rate": 9.856879208864277e-06,
+      "loss": 0.5035,
+      "step": 922
+    },
+    {
+      "epoch": 0.16910956394283622,
+      "grad_norm": 0.47756329412725707,
+      "learning_rate": 9.85611856720282e-06,
+      "loss": 0.5031,
+      "step": 923
+    },
+    {
+      "epoch": 0.1692927812385489,
+      "grad_norm": 0.474431940864528,
+      "learning_rate": 9.8553559391158e-06,
+      "loss": 0.512,
+      "step": 924
+    },
+    {
+      "epoch": 0.16947599853426162,
+      "grad_norm": 0.41972238109665655,
+      "learning_rate": 9.85459132491517e-06,
+      "loss": 0.5386,
+      "step": 925
+    },
+    {
+      "epoch": 0.16965921582997434,
+      "grad_norm": 0.4369771091332424,
+      "learning_rate": 9.8538247249137e-06,
+      "loss": 0.4909,
+      "step": 926
+    },
+    {
+      "epoch": 0.16984243312568706,
+      "grad_norm": 0.4523772832392917,
+      "learning_rate": 9.853056139424974e-06,
+      "loss": 0.4951,
+      "step": 927
+    },
+    {
+      "epoch": 0.17002565042139978,
+      "grad_norm": 0.40235086982371876,
+      "learning_rate": 9.852285568763387e-06,
+      "loss": 0.4828,
+      "step": 928
+    },
+    {
+      "epoch": 0.1702088677171125,
+      "grad_norm": 0.4150844402567988,
+      "learning_rate": 9.851513013244144e-06,
+      "loss": 0.4933,
+      "step": 929
+    },
+    {
+      "epoch": 0.1703920850128252,
+      "grad_norm": 0.3941602147072389,
+      "learning_rate": 9.850738473183266e-06,
+      "loss": 0.4855,
+      "step": 930
+    },
+    {
+      "epoch": 0.17057530230853793,
+      "grad_norm": 0.47118394280764797,
+      "learning_rate": 9.849961948897582e-06,
+      "loss": 0.5024,
+      "step": 931
+    },
+    {
+      "epoch": 0.17075851960425065,
+      "grad_norm": 0.3912878663413955,
+      "learning_rate": 9.849183440704735e-06,
+      "loss": 0.5041,
+      "step": 932
+    },
+    {
+      "epoch": 0.17094173689996336,
+      "grad_norm": 0.4001961484144372,
+      "learning_rate": 9.848402948923177e-06,
+      "loss": 0.4753,
+      "step": 933
+    },
+    {
+      "epoch": 0.17112495419567608,
+      "grad_norm": 0.4441137892348275,
+      "learning_rate": 9.847620473872172e-06,
+      "loss": 0.5102,
+      "step": 934
+    },
+    {
+      "epoch": 0.1713081714913888,
+      "grad_norm": 0.40101758764349804,
+      "learning_rate": 9.846836015871802e-06,
+      "loss": 0.4824,
+      "step": 935
+    },
+    {
+      "epoch": 0.1714913887871015,
+      "grad_norm": 0.41075944565503664,
+      "learning_rate": 9.846049575242949e-06,
+      "loss": 0.5235,
+      "step": 936
+    },
+    {
+      "epoch": 0.1716746060828142,
+      "grad_norm": 0.3780837940291423,
+      "learning_rate": 9.845261152307312e-06,
+      "loss": 0.5211,
+      "step": 937
+    },
+    {
+      "epoch": 0.17185782337852692,
+      "grad_norm": 0.45070991069476113,
+      "learning_rate": 9.844470747387403e-06,
+      "loss": 0.4926,
+      "step": 938
+    },
+    {
+      "epoch": 0.17204104067423964,
+      "grad_norm": 0.40836878453605646,
+      "learning_rate": 9.843678360806542e-06,
+      "loss": 0.5173,
+      "step": 939
+    },
+    {
+      "epoch": 0.17222425796995236,
+      "grad_norm": 0.36676941914948374,
+      "learning_rate": 9.842883992888855e-06,
+      "loss": 0.5101,
+      "step": 940
+    },
+    {
+      "epoch": 0.17240747526566508,
+      "grad_norm": 0.42894488109200807,
+      "learning_rate": 9.842087643959288e-06,
+      "loss": 0.5398,
+      "step": 941
+    },
+    {
+      "epoch": 0.1725906925613778,
+      "grad_norm": 0.4004768449690441,
+      "learning_rate": 9.841289314343591e-06,
+      "loss": 0.4967,
+      "step": 942
+    },
+    {
+      "epoch": 0.1727739098570905,
+      "grad_norm": 0.3781076276341992,
+      "learning_rate": 9.840489004368325e-06,
+      "loss": 0.4731,
+      "step": 943
+    },
+    {
+      "epoch": 0.17295712715280323,
+      "grad_norm": 0.45307945315267256,
+      "learning_rate": 9.839686714360864e-06,
+      "loss": 0.4737,
+      "step": 944
+    },
+    {
+      "epoch": 0.17314034444851595,
+      "grad_norm": 0.4079247794136399,
+      "learning_rate": 9.838882444649387e-06,
+      "loss": 0.4927,
+      "step": 945
+    },
+    {
+      "epoch": 0.17332356174422867,
+      "grad_norm": 0.47171018664243136,
+      "learning_rate": 9.838076195562886e-06,
+      "loss": 0.5009,
+      "step": 946
+    },
+    {
+      "epoch": 0.17350677903994138,
+      "grad_norm": 0.44572108382734166,
+      "learning_rate": 9.837267967431164e-06,
+      "loss": 0.4858,
+      "step": 947
+    },
+    {
+      "epoch": 0.17368999633565407,
+      "grad_norm": 0.40649730883742624,
+      "learning_rate": 9.83645776058483e-06,
+      "loss": 0.4943,
+      "step": 948
+    },
+    {
+      "epoch": 0.1738732136313668,
+      "grad_norm": 0.43002759908477306,
+      "learning_rate": 9.835645575355304e-06,
+      "loss": 0.4872,
+      "step": 949
+    },
+    {
+      "epoch": 0.1740564309270795,
+      "grad_norm": 0.39064612647120167,
+      "learning_rate": 9.834831412074816e-06,
+      "loss": 0.4409,
+      "step": 950
+    },
+    {
+      "epoch": 0.17423964822279223,
+      "grad_norm": 0.3761573603114394,
+      "learning_rate": 9.834015271076405e-06,
+      "loss": 0.459,
+      "step": 951
+    },
+    {
+      "epoch": 0.17442286551850494,
+      "grad_norm": 0.42796519720566795,
+      "learning_rate": 9.83319715269392e-06,
+      "loss": 0.4906,
+      "step": 952
+    },
+    {
+      "epoch": 0.17460608281421766,
+      "grad_norm": 0.5091807084893948,
+      "learning_rate": 9.832377057262015e-06,
+      "loss": 0.5516,
+      "step": 953
+    },
+    {
+      "epoch": 0.17478930010993038,
+      "grad_norm": 0.48065067696295527,
+      "learning_rate": 9.831554985116155e-06,
+      "loss": 0.5097,
+      "step": 954
+    },
+    {
+      "epoch": 0.1749725174056431,
+      "grad_norm": 0.4310364985277951,
+      "learning_rate": 9.830730936592615e-06,
+      "loss": 0.5025,
+      "step": 955
+    },
+    {
+      "epoch": 0.17515573470135581,
+      "grad_norm": 0.39964086093662415,
+      "learning_rate": 9.829904912028477e-06,
+      "loss": 0.494,
+      "step": 956
+    },
+    {
+      "epoch": 0.17533895199706853,
+      "grad_norm": 0.40935380850458175,
+      "learning_rate": 9.829076911761631e-06,
+      "loss": 0.5182,
+      "step": 957
+    },
+    {
+      "epoch": 0.17552216929278125,
+      "grad_norm": 0.454766087870644,
+      "learning_rate": 9.828246936130777e-06,
+      "loss": 0.5113,
+      "step": 958
+    },
+    {
+      "epoch": 0.17570538658849397,
+      "grad_norm": 0.5087972352515657,
+      "learning_rate": 9.827414985475419e-06,
+      "loss": 0.4797,
+      "step": 959
+    },
+    {
+      "epoch": 0.17588860388420666,
+      "grad_norm": 0.4255562501722622,
+      "learning_rate": 9.826581060135873e-06,
+      "loss": 0.5045,
+      "step": 960
+    },
+    {
+      "epoch": 0.17607182117991937,
+      "grad_norm": 0.45583787528367603,
+      "learning_rate": 9.825745160453264e-06,
+      "loss": 0.4755,
+      "step": 961
+    },
+    {
+      "epoch": 0.1762550384756321,
+      "grad_norm": 0.45648883029913284,
+      "learning_rate": 9.824907286769519e-06,
+      "loss": 0.4736,
+      "step": 962
+    },
+    {
+      "epoch": 0.1764382557713448,
+      "grad_norm": 0.436071698669034,
+      "learning_rate": 9.824067439427374e-06,
+      "loss": 0.5036,
+      "step": 963
+    },
+    {
+      "epoch": 0.17662147306705753,
+      "grad_norm": 0.41502000197263017,
+      "learning_rate": 9.823225618770378e-06,
+      "loss": 0.5121,
+      "step": 964
+    },
+    {
+      "epoch": 0.17680469036277024,
+      "grad_norm": 0.450252173902017,
+      "learning_rate": 9.822381825142879e-06,
+      "loss": 0.506,
+      "step": 965
+    },
+    {
+      "epoch": 0.17698790765848296,
+      "grad_norm": 0.43561778776659554,
+      "learning_rate": 9.82153605889004e-06,
+      "loss": 0.5304,
+      "step": 966
+    },
+    {
+      "epoch": 0.17717112495419568,
+      "grad_norm": 0.4306520452468297,
+      "learning_rate": 9.820688320357823e-06,
+      "loss": 0.5112,
+      "step": 967
+    },
+    {
+      "epoch": 0.1773543422499084,
+      "grad_norm": 0.39680977809710444,
+      "learning_rate": 9.819838609893005e-06,
+      "loss": 0.4904,
+      "step": 968
+    },
+    {
+      "epoch": 0.17753755954562112,
+      "grad_norm": 0.4225898008101114,
+      "learning_rate": 9.81898692784316e-06,
+      "loss": 0.5137,
+      "step": 969
+    },
+    {
+      "epoch": 0.17772077684133383,
+      "grad_norm": 0.4267062187327317,
+      "learning_rate": 9.818133274556679e-06,
+      "loss": 0.517,
+      "step": 970
+    },
+    {
+      "epoch": 0.17790399413704655,
+      "grad_norm": 0.4450452397955529,
+      "learning_rate": 9.81727765038275e-06,
+      "loss": 0.4949,
+      "step": 971
+    },
+    {
+      "epoch": 0.17808721143275924,
+      "grad_norm": 0.42730228271929965,
+      "learning_rate": 9.816420055671374e-06,
+      "loss": 0.5013,
+      "step": 972
+    },
+    {
+      "epoch": 0.17827042872847196,
+      "grad_norm": 0.4421383626296405,
+      "learning_rate": 9.815560490773356e-06,
+      "loss": 0.4679,
+      "step": 973
+    },
+    {
+      "epoch": 0.17845364602418468,
+      "grad_norm": 0.43307265096730885,
+      "learning_rate": 9.814698956040305e-06,
+      "loss": 0.4976,
+      "step": 974
+    },
+    {
+      "epoch": 0.1786368633198974,
+      "grad_norm": 0.40376464412932705,
+      "learning_rate": 9.813835451824636e-06,
+      "loss": 0.5021,
+      "step": 975
+    },
+    {
+      "epoch": 0.1788200806156101,
+      "grad_norm": 0.44775776339538953,
+      "learning_rate": 9.812969978479573e-06,
+      "loss": 0.4824,
+      "step": 976
+    },
+    {
+      "epoch": 0.17900329791132283,
+      "grad_norm": 0.468036097947629,
+      "learning_rate": 9.812102536359142e-06,
+      "loss": 0.4966,
+      "step": 977
+    },
+    {
+      "epoch": 0.17918651520703555,
+      "grad_norm": 0.49832149697786493,
+      "learning_rate": 9.811233125818176e-06,
+      "loss": 0.4853,
+      "step": 978
+    },
+    {
+      "epoch": 0.17936973250274826,
+      "grad_norm": 0.44754723833693727,
+      "learning_rate": 9.810361747212313e-06,
+      "loss": 0.4999,
+      "step": 979
+    },
+    {
+      "epoch": 0.17955294979846098,
+      "grad_norm": 0.4747011696315986,
+      "learning_rate": 9.809488400897996e-06,
+      "loss": 0.5108,
+      "step": 980
+    },
+    {
+      "epoch": 0.1797361670941737,
+      "grad_norm": 0.4055842519238836,
+      "learning_rate": 9.808613087232473e-06,
+      "loss": 0.4786,
+      "step": 981
+    },
+    {
+      "epoch": 0.17991938438988642,
+      "grad_norm": 0.3971701443501286,
+      "learning_rate": 9.807735806573795e-06,
+      "loss": 0.4925,
+      "step": 982
+    },
+    {
+      "epoch": 0.18010260168559913,
+      "grad_norm": 0.3784480421793096,
+      "learning_rate": 9.806856559280819e-06,
+      "loss": 0.4762,
+      "step": 983
+    },
+    {
+      "epoch": 0.18028581898131182,
+      "grad_norm": 0.41465183173286063,
+      "learning_rate": 9.80597534571321e-06,
+      "loss": 0.5037,
+      "step": 984
+    },
+    {
+      "epoch": 0.18046903627702454,
+      "grad_norm": 0.4713550462839489,
+      "learning_rate": 9.80509216623143e-06,
+      "loss": 0.4816,
+      "step": 985
+    },
+    {
+      "epoch": 0.18065225357273726,
+      "grad_norm": 0.34923786309805893,
+      "learning_rate": 9.804207021196751e-06,
+      "loss": 0.4903,
+      "step": 986
+    },
+    {
+      "epoch": 0.18083547086844998,
+      "grad_norm": 0.4359203385963395,
+      "learning_rate": 9.803319910971248e-06,
+      "loss": 0.5252,
+      "step": 987
+    },
+    {
+      "epoch": 0.1810186881641627,
+      "grad_norm": 0.40286824933567683,
+      "learning_rate": 9.802430835917796e-06,
+      "loss": 0.5248,
+      "step": 988
+    },
+    {
+      "epoch": 0.1812019054598754,
+      "grad_norm": 0.44314680454994215,
+      "learning_rate": 9.801539796400078e-06,
+      "loss": 0.5071,
+      "step": 989
+    },
+    {
+      "epoch": 0.18138512275558813,
+      "grad_norm": 0.44074612784921385,
+      "learning_rate": 9.80064679278258e-06,
+      "loss": 0.5057,
+      "step": 990
+    },
+    {
+      "epoch": 0.18156834005130085,
+      "grad_norm": 0.40968360376086815,
+      "learning_rate": 9.799751825430592e-06,
+      "loss": 0.4919,
+      "step": 991
+    },
+    {
+      "epoch": 0.18175155734701356,
+      "grad_norm": 0.4204735967781832,
+      "learning_rate": 9.798854894710202e-06,
+      "loss": 0.546,
+      "step": 992
+    },
+    {
+      "epoch": 0.18193477464272628,
+      "grad_norm": 0.36210624498761007,
+      "learning_rate": 9.79795600098831e-06,
+      "loss": 0.4904,
+      "step": 993
+    },
+    {
+      "epoch": 0.182117991938439,
+      "grad_norm": 0.45193717949202633,
+      "learning_rate": 9.797055144632609e-06,
+      "loss": 0.5292,
+      "step": 994
+    },
+    {
+      "epoch": 0.18230120923415172,
+      "grad_norm": 0.415630496229026,
+      "learning_rate": 9.796152326011604e-06,
+      "loss": 0.533,
+      "step": 995
+    },
+    {
+      "epoch": 0.1824844265298644,
+      "grad_norm": 0.39372729416385926,
+      "learning_rate": 9.795247545494594e-06,
+      "loss": 0.5147,
+      "step": 996
+    },
+    {
+      "epoch": 0.18266764382557713,
+      "grad_norm": 0.43540226077312505,
+      "learning_rate": 9.794340803451692e-06,
+      "loss": 0.49,
+      "step": 997
+    },
+    {
+      "epoch": 0.18285086112128984,
+      "grad_norm": 0.39372791981341043,
+      "learning_rate": 9.7934321002538e-06,
+      "loss": 0.4776,
+      "step": 998
+    },
+    {
+      "epoch": 0.18303407841700256,
+      "grad_norm": 0.40381508343023165,
+      "learning_rate": 9.792521436272633e-06,
+      "loss": 0.4566,
+      "step": 999
+    },
+    {
+      "epoch": 0.18321729571271528,
+      "grad_norm": 0.47729065322343966,
+      "learning_rate": 9.791608811880702e-06,
+      "loss": 0.5013,
+      "step": 1000
+    },
+    {
+      "epoch": 0.183400513008428,
+      "grad_norm": 0.4569898838129068,
+      "learning_rate": 9.79069422745132e-06,
+      "loss": 0.4834,
+      "step": 1001
+    },
+    {
+      "epoch": 0.1835837303041407,
+      "grad_norm": 0.41137738030360954,
+      "learning_rate": 9.789777683358607e-06,
+      "loss": 0.4982,
+      "step": 1002
+    },
+    {
+      "epoch": 0.18376694759985343,
+      "grad_norm": 0.443462511432857,
+      "learning_rate": 9.788859179977478e-06,
+      "loss": 0.4822,
+      "step": 1003
+    },
+    {
+      "epoch": 0.18395016489556615,
+      "grad_norm": 0.3803483898936361,
+      "learning_rate": 9.787938717683654e-06,
+      "loss": 0.4781,
+      "step": 1004
+    },
+    {
+      "epoch": 0.18413338219127887,
+      "grad_norm": 0.4711405888417879,
+      "learning_rate": 9.787016296853657e-06,
+      "loss": 0.5104,
+      "step": 1005
+    },
+    {
+      "epoch": 0.18431659948699158,
+      "grad_norm": 0.42018309024133216,
+      "learning_rate": 9.786091917864807e-06,
+      "loss": 0.4922,
+      "step": 1006
+    },
+    {
+      "epoch": 0.18449981678270427,
+      "grad_norm": 0.46053661876076,
+      "learning_rate": 9.785165581095225e-06,
+      "loss": 0.4777,
+      "step": 1007
+    },
+    {
+      "epoch": 0.184683034078417,
+      "grad_norm": 0.38543837395940844,
+      "learning_rate": 9.78423728692384e-06,
+      "loss": 0.5053,
+      "step": 1008
+    },
+    {
+      "epoch": 0.1848662513741297,
+      "grad_norm": 0.3681107156627339,
+      "learning_rate": 9.783307035730375e-06,
+      "loss": 0.463,
+      "step": 1009
+    },
+    {
+      "epoch": 0.18504946866984243,
+      "grad_norm": 0.42988658038887384,
+      "learning_rate": 9.782374827895353e-06,
+      "loss": 0.5179,
+      "step": 1010
+    },
+    {
+      "epoch": 0.18523268596555514,
+      "grad_norm": 0.3985975808728344,
+      "learning_rate": 9.781440663800099e-06,
+      "loss": 0.4979,
+      "step": 1011
+    },
+    {
+      "epoch": 0.18541590326126786,
+      "grad_norm": 0.419869548881785,
+      "learning_rate": 9.780504543826741e-06,
+      "loss": 0.4952,
+      "step": 1012
+    },
+    {
+      "epoch": 0.18559912055698058,
+      "grad_norm": 0.4348607167141751,
+      "learning_rate": 9.779566468358205e-06,
+      "loss": 0.4999,
+      "step": 1013
+    },
+    {
+      "epoch": 0.1857823378526933,
+      "grad_norm": 0.43518629406794973,
+      "learning_rate": 9.778626437778214e-06,
+      "loss": 0.5028,
+      "step": 1014
+    },
+    {
+      "epoch": 0.18596555514840601,
+      "grad_norm": 0.4475968637904264,
+      "learning_rate": 9.777684452471296e-06,
+      "loss": 0.4997,
+      "step": 1015
+    },
+    {
+      "epoch": 0.18614877244411873,
+      "grad_norm": 0.4834533700606862,
+      "learning_rate": 9.776740512822773e-06,
+      "loss": 0.5139,
+      "step": 1016
+    },
+    {
+      "epoch": 0.18633198973983145,
+      "grad_norm": 0.49443428074486423,
+      "learning_rate": 9.775794619218773e-06,
+      "loss": 0.4818,
+      "step": 1017
+    },
+    {
+      "epoch": 0.18651520703554417,
+      "grad_norm": 0.42771715112146097,
+      "learning_rate": 9.774846772046216e-06,
+      "loss": 0.507,
+      "step": 1018
+    },
+    {
+      "epoch": 0.18669842433125686,
+      "grad_norm": 0.43776726147874706,
+      "learning_rate": 9.77389697169283e-06,
+      "loss": 0.5002,
+      "step": 1019
+    },
+    {
+      "epoch": 0.18688164162696957,
+      "grad_norm": 0.44345646316302056,
+      "learning_rate": 9.77294521854713e-06,
+      "loss": 0.4953,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1870648589226823,
+      "grad_norm": 0.4199805356687226,
+      "learning_rate": 9.77199151299844e-06,
+      "loss": 0.491,
+      "step": 1021
+    },
+    {
+      "epoch": 0.187248076218395,
+      "grad_norm": 0.4157991523056247,
+      "learning_rate": 9.77103585543688e-06,
+      "loss": 0.4809,
+      "step": 1022
+    },
+    {
+      "epoch": 0.18743129351410773,
+      "grad_norm": 0.3883786524287937,
+      "learning_rate": 9.770078246253367e-06,
+      "loss": 0.4811,
+      "step": 1023
+    },
+    {
+      "epoch": 0.18761451080982045,
+      "grad_norm": 0.4460460362572771,
+      "learning_rate": 9.769118685839616e-06,
+      "loss": 0.4869,
+      "step": 1024
+    },
+    {
+      "epoch": 0.18779772810553316,
+      "grad_norm": 0.4153757045736682,
+      "learning_rate": 9.768157174588144e-06,
+      "loss": 0.4464,
+      "step": 1025
+    },
+    {
+      "epoch": 0.18798094540124588,
+      "grad_norm": 0.4146331433621528,
+      "learning_rate": 9.767193712892259e-06,
+      "loss": 0.4588,
+      "step": 1026
+    },
+    {
+      "epoch": 0.1881641626969586,
+      "grad_norm": 0.42852199415357106,
+      "learning_rate": 9.766228301146074e-06,
+      "loss": 0.5132,
+      "step": 1027
+    },
+    {
+      "epoch": 0.18834737999267132,
+      "grad_norm": 0.40617799593038223,
+      "learning_rate": 9.765260939744496e-06,
+      "loss": 0.4715,
+      "step": 1028
+    },
+    {
+      "epoch": 0.18853059728838403,
+      "grad_norm": 0.43057345885222986,
+      "learning_rate": 9.76429162908323e-06,
+      "loss": 0.4613,
+      "step": 1029
+    },
+    {
+      "epoch": 0.18871381458409675,
+      "grad_norm": 0.4287852170015339,
+      "learning_rate": 9.763320369558777e-06,
+      "loss": 0.5029,
+      "step": 1030
+    },
+    {
+      "epoch": 0.18889703187980944,
+      "grad_norm": 0.4338612990572842,
+      "learning_rate": 9.762347161568441e-06,
+      "loss": 0.51,
+      "step": 1031
+    },
+    {
+      "epoch": 0.18908024917552216,
+      "grad_norm": 0.6099286344029112,
+      "learning_rate": 9.761372005510315e-06,
+      "loss": 0.5176,
+      "step": 1032
+    },
+    {
+      "epoch": 0.18926346647123488,
+      "grad_norm": 0.4139699553573524,
+      "learning_rate": 9.760394901783294e-06,
+      "loss": 0.4803,
+      "step": 1033
+    },
+    {
+      "epoch": 0.1894466837669476,
+      "grad_norm": 0.4613287614032801,
+      "learning_rate": 9.759415850787068e-06,
+      "loss": 0.5114,
+      "step": 1034
+    },
+    {
+      "epoch": 0.1896299010626603,
+      "grad_norm": 0.43042214487008984,
+      "learning_rate": 9.758434852922124e-06,
+      "loss": 0.5049,
+      "step": 1035
+    },
+    {
+      "epoch": 0.18981311835837303,
+      "grad_norm": 0.4708254422812291,
+      "learning_rate": 9.757451908589746e-06,
+      "loss": 0.5466,
+      "step": 1036
+    },
+    {
+      "epoch": 0.18999633565408575,
+      "grad_norm": 0.4823877518080949,
+      "learning_rate": 9.756467018192013e-06,
+      "loss": 0.4851,
+      "step": 1037
+    },
+    {
+      "epoch": 0.19017955294979846,
+      "grad_norm": 0.49290249361142535,
+      "learning_rate": 9.7554801821318e-06,
+      "loss": 0.4915,
+      "step": 1038
+    },
+    {
+      "epoch": 0.19036277024551118,
+      "grad_norm": 0.5097570694075713,
+      "learning_rate": 9.754491400812779e-06,
+      "loss": 0.5195,
+      "step": 1039
+    },
+    {
+      "epoch": 0.1905459875412239,
+      "grad_norm": 0.41886567376911943,
+      "learning_rate": 9.753500674639417e-06,
+      "loss": 0.5133,
+      "step": 1040
+    },
+    {
+      "epoch": 0.19072920483693662,
+      "grad_norm": 0.48924775085841404,
+      "learning_rate": 9.752508004016976e-06,
+      "loss": 0.5174,
+      "step": 1041
+    },
+    {
+      "epoch": 0.19091242213264933,
+      "grad_norm": 0.44418461742112236,
+      "learning_rate": 9.751513389351517e-06,
+      "loss": 0.4711,
+      "step": 1042
+    },
+    {
+      "epoch": 0.19109563942836202,
+      "grad_norm": 0.4103673664900802,
+      "learning_rate": 9.75051683104989e-06,
+      "loss": 0.478,
+      "step": 1043
+    },
+    {
+      "epoch": 0.19127885672407474,
+      "grad_norm": 0.45394849049437136,
+      "learning_rate": 9.749518329519745e-06,
+      "loss": 0.5069,
+      "step": 1044
+    },
+    {
+      "epoch": 0.19146207401978746,
+      "grad_norm": 0.4841087058964896,
+      "learning_rate": 9.748517885169525e-06,
+      "loss": 0.5226,
+      "step": 1045
+    },
+    {
+      "epoch": 0.19164529131550018,
+      "grad_norm": 0.4138362206735341,
+      "learning_rate": 9.747515498408466e-06,
+      "loss": 0.5085,
+      "step": 1046
+    },
+    {
+      "epoch": 0.1918285086112129,
+      "grad_norm": 0.4267417097417006,
+      "learning_rate": 9.746511169646604e-06,
+      "loss": 0.5074,
+      "step": 1047
+    },
+    {
+      "epoch": 0.1920117259069256,
+      "grad_norm": 0.4397686656970245,
+      "learning_rate": 9.745504899294764e-06,
+      "loss": 0.5242,
+      "step": 1048
+    },
+    {
+      "epoch": 0.19219494320263833,
+      "grad_norm": 0.6027364595972364,
+      "learning_rate": 9.744496687764568e-06,
+      "loss": 0.5046,
+      "step": 1049
+    },
+    {
+      "epoch": 0.19237816049835105,
+      "grad_norm": 0.45130752478163744,
+      "learning_rate": 9.743486535468431e-06,
+      "loss": 0.4699,
+      "step": 1050
+    },
+    {
+      "epoch": 0.19256137779406376,
+      "grad_norm": 0.4971542747077622,
+      "learning_rate": 9.742474442819561e-06,
+      "loss": 0.4446,
+      "step": 1051
+    },
+    {
+      "epoch": 0.19274459508977648,
+      "grad_norm": 0.566649600088142,
+      "learning_rate": 9.741460410231962e-06,
+      "loss": 0.5531,
+      "step": 1052
+    },
+    {
+      "epoch": 0.1929278123854892,
+      "grad_norm": 0.44162327591540923,
+      "learning_rate": 9.74044443812043e-06,
+      "loss": 0.5194,
+      "step": 1053
+    },
+    {
+      "epoch": 0.19311102968120192,
+      "grad_norm": 0.40323605263226747,
+      "learning_rate": 9.739426526900555e-06,
+      "loss": 0.5013,
+      "step": 1054
+    },
+    {
+      "epoch": 0.1932942469769146,
+      "grad_norm": 0.45567902759630347,
+      "learning_rate": 9.738406676988721e-06,
+      "loss": 0.4397,
+      "step": 1055
+    },
+    {
+      "epoch": 0.19347746427262733,
+      "grad_norm": 0.4604601552443739,
+      "learning_rate": 9.737384888802099e-06,
+      "loss": 0.4888,
+      "step": 1056
+    },
+    {
+      "epoch": 0.19366068156834004,
+      "grad_norm": 0.3927639599518843,
+      "learning_rate": 9.736361162758665e-06,
+      "loss": 0.4632,
+      "step": 1057
+    },
+    {
+      "epoch": 0.19384389886405276,
+      "grad_norm": 0.40880550818933825,
+      "learning_rate": 9.735335499277176e-06,
+      "loss": 0.484,
+      "step": 1058
+    },
+    {
+      "epoch": 0.19402711615976548,
+      "grad_norm": 0.4813436727815402,
+      "learning_rate": 9.734307898777187e-06,
+      "loss": 0.5351,
+      "step": 1059
+    },
+    {
+      "epoch": 0.1942103334554782,
+      "grad_norm": 0.44626763914485296,
+      "learning_rate": 9.733278361679042e-06,
+      "loss": 0.4847,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1943935507511909,
+      "grad_norm": 0.3999429153352251,
+      "learning_rate": 9.732246888403885e-06,
+      "loss": 0.4888,
+      "step": 1061
+    },
+    {
+      "epoch": 0.19457676804690363,
+      "grad_norm": 0.39498667028752193,
+      "learning_rate": 9.731213479373643e-06,
+      "loss": 0.4947,
+      "step": 1062
+    },
+    {
+      "epoch": 0.19475998534261635,
+      "grad_norm": 0.4096597841615662,
+      "learning_rate": 9.730178135011036e-06,
+      "loss": 0.4747,
+      "step": 1063
+    },
+    {
+      "epoch": 0.19494320263832907,
+      "grad_norm": 0.4279641584926225,
+      "learning_rate": 9.729140855739585e-06,
+      "loss": 0.5013,
+      "step": 1064
+    },
+    {
+      "epoch": 0.19512641993404178,
+      "grad_norm": 0.5191493361942445,
+      "learning_rate": 9.72810164198359e-06,
+      "loss": 0.4947,
+      "step": 1065
+    },
+    {
+      "epoch": 0.1953096372297545,
+      "grad_norm": 0.4196384047011208,
+      "learning_rate": 9.727060494168148e-06,
+      "loss": 0.4825,
+      "step": 1066
+    },
+    {
+      "epoch": 0.1954928545254672,
+      "grad_norm": 0.4345875246850262,
+      "learning_rate": 9.726017412719151e-06,
+      "loss": 0.4734,
+      "step": 1067
+    },
+    {
+      "epoch": 0.1956760718211799,
+      "grad_norm": 0.45563595275942326,
+      "learning_rate": 9.724972398063273e-06,
+      "loss": 0.5153,
+      "step": 1068
+    },
+    {
+      "epoch": 0.19585928911689263,
+      "grad_norm": 0.44886022596575603,
+      "learning_rate": 9.723925450627988e-06,
+      "loss": 0.478,
+      "step": 1069
+    },
+    {
+      "epoch": 0.19604250641260534,
+      "grad_norm": 0.413358196754176,
+      "learning_rate": 9.722876570841554e-06,
+      "loss": 0.5132,
+      "step": 1070
+    },
+    {
+      "epoch": 0.19622572370831806,
+      "grad_norm": 0.41859249060654946,
+      "learning_rate": 9.721825759133022e-06,
+      "loss": 0.4911,
+      "step": 1071
+    },
+    {
+      "epoch": 0.19640894100403078,
+      "grad_norm": 0.4261104024613703,
+      "learning_rate": 9.720773015932234e-06,
+      "loss": 0.512,
+      "step": 1072
+    },
+    {
+      "epoch": 0.1965921582997435,
+      "grad_norm": 0.41207466582794494,
+      "learning_rate": 9.71971834166982e-06,
+      "loss": 0.4739,
+      "step": 1073
+    },
+    {
+      "epoch": 0.19677537559545621,
+      "grad_norm": 0.4191753315427551,
+      "learning_rate": 9.718661736777202e-06,
+      "loss": 0.4906,
+      "step": 1074
+    },
+    {
+      "epoch": 0.19695859289116893,
+      "grad_norm": 0.42085342925013736,
+      "learning_rate": 9.71760320168659e-06,
+      "loss": 0.5164,
+      "step": 1075
+    },
+    {
+      "epoch": 0.19714181018688165,
+      "grad_norm": 0.4268987095250944,
+      "learning_rate": 9.716542736830985e-06,
+      "loss": 0.4861,
+      "step": 1076
+    },
+    {
+      "epoch": 0.19732502748259437,
+      "grad_norm": 0.42610930294181376,
+      "learning_rate": 9.715480342644177e-06,
+      "loss": 0.475,
+      "step": 1077
+    },
+    {
+      "epoch": 0.19750824477830708,
+      "grad_norm": 0.46458365169769283,
+      "learning_rate": 9.714416019560744e-06,
+      "loss": 0.4949,
+      "step": 1078
+    },
+    {
+      "epoch": 0.19769146207401977,
+      "grad_norm": 0.40388410318318285,
+      "learning_rate": 9.713349768016055e-06,
+      "loss": 0.4886,
+      "step": 1079
+    },
+    {
+      "epoch": 0.1978746793697325,
+      "grad_norm": 0.47359324461508207,
+      "learning_rate": 9.712281588446267e-06,
+      "loss": 0.5166,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1980578966654452,
+      "grad_norm": 0.3996969974243518,
+      "learning_rate": 9.711211481288323e-06,
+      "loss": 0.5208,
+      "step": 1081
+    },
+    {
+      "epoch": 0.19824111396115793,
+      "grad_norm": 0.37701213979445153,
+      "learning_rate": 9.710139446979961e-06,
+      "loss": 0.5,
+      "step": 1082
+    },
+    {
+      "epoch": 0.19842433125687065,
+      "grad_norm": 0.43491339118601396,
+      "learning_rate": 9.7090654859597e-06,
+      "loss": 0.4899,
+      "step": 1083
+    },
+    {
+      "epoch": 0.19860754855258336,
+      "grad_norm": 0.4651922376126236,
+      "learning_rate": 9.707989598666852e-06,
+      "loss": 0.4684,
+      "step": 1084
+    },
+    {
+      "epoch": 0.19879076584829608,
+      "grad_norm": 0.41200327863289965,
+      "learning_rate": 9.706911785541515e-06,
+      "loss": 0.4408,
+      "step": 1085
+    },
+    {
+      "epoch": 0.1989739831440088,
+      "grad_norm": 0.48300341584043166,
+      "learning_rate": 9.705832047024574e-06,
+      "loss": 0.53,
+      "step": 1086
+    },
+    {
+      "epoch": 0.19915720043972152,
+      "grad_norm": 0.4403619842708917,
+      "learning_rate": 9.704750383557707e-06,
+      "loss": 0.5361,
+      "step": 1087
+    },
+    {
+      "epoch": 0.19934041773543423,
+      "grad_norm": 0.3719557224033517,
+      "learning_rate": 9.70366679558337e-06,
+      "loss": 0.4948,
+      "step": 1088
+    },
+    {
+      "epoch": 0.19952363503114695,
+      "grad_norm": 0.4176787500801207,
+      "learning_rate": 9.702581283544813e-06,
+      "loss": 0.5145,
+      "step": 1089
+    },
+    {
+      "epoch": 0.19970685232685967,
+      "grad_norm": 0.4221402861265115,
+      "learning_rate": 9.701493847886075e-06,
+      "loss": 0.5128,
+      "step": 1090
+    },
+    {
+      "epoch": 0.19989006962257236,
+      "grad_norm": 0.40690517337963633,
+      "learning_rate": 9.700404489051974e-06,
+      "loss": 0.4911,
+      "step": 1091
+    },
+    {
+      "epoch": 0.20007328691828508,
+      "grad_norm": 0.46227138813892915,
+      "learning_rate": 9.69931320748812e-06,
+      "loss": 0.5037,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2002565042139978,
+      "grad_norm": 0.3860882899117114,
+      "learning_rate": 9.69822000364091e-06,
+      "loss": 0.5063,
+      "step": 1093
+    },
+    {
+      "epoch": 0.2004397215097105,
+      "grad_norm": 0.512213232549586,
+      "learning_rate": 9.697124877957524e-06,
+      "loss": 0.5019,
+      "step": 1094
+    },
+    {
+      "epoch": 0.20062293880542323,
+      "grad_norm": 0.455116276042089,
+      "learning_rate": 9.69602783088593e-06,
+      "loss": 0.5145,
+      "step": 1095
+    },
+    {
+      "epoch": 0.20080615610113595,
+      "grad_norm": 0.4251117291337367,
+      "learning_rate": 9.694928862874883e-06,
+      "loss": 0.5096,
+      "step": 1096
+    },
+    {
+      "epoch": 0.20098937339684866,
+      "grad_norm": 0.46240748219044453,
+      "learning_rate": 9.693827974373925e-06,
+      "loss": 0.4506,
+      "step": 1097
+    },
+    {
+      "epoch": 0.20117259069256138,
+      "grad_norm": 0.35673376148210156,
+      "learning_rate": 9.692725165833377e-06,
+      "loss": 0.4688,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2013558079882741,
+      "grad_norm": 0.469083523866663,
+      "learning_rate": 9.69162043770435e-06,
+      "loss": 0.5049,
+      "step": 1099
+    },
+    {
+      "epoch": 0.20153902528398682,
+      "grad_norm": 0.42797732688089807,
+      "learning_rate": 9.690513790438743e-06,
+      "loss": 0.4696,
+      "step": 1100
+    },
+    {
+      "epoch": 0.20172224257969953,
+      "grad_norm": 0.4281881317554375,
+      "learning_rate": 9.689405224489235e-06,
+      "loss": 0.4973,
+      "step": 1101
+    },
+    {
+      "epoch": 0.20190545987541225,
+      "grad_norm": 0.4513286311031276,
+      "learning_rate": 9.688294740309292e-06,
+      "loss": 0.4694,
+      "step": 1102
+    },
+    {
+      "epoch": 0.20208867717112494,
+      "grad_norm": 0.3940963878288531,
+      "learning_rate": 9.687182338353166e-06,
+      "loss": 0.4765,
+      "step": 1103
+    },
+    {
+      "epoch": 0.20227189446683766,
+      "grad_norm": 0.3902749441842877,
+      "learning_rate": 9.68606801907589e-06,
+      "loss": 0.4759,
+      "step": 1104
+    },
+    {
+      "epoch": 0.20245511176255038,
+      "grad_norm": 0.4789649740688253,
+      "learning_rate": 9.684951782933282e-06,
+      "loss": 0.4948,
+      "step": 1105
+    },
+    {
+      "epoch": 0.2026383290582631,
+      "grad_norm": 0.47833293392492787,
+      "learning_rate": 9.683833630381949e-06,
+      "loss": 0.5239,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2028215463539758,
+      "grad_norm": 0.38231892821814056,
+      "learning_rate": 9.682713561879275e-06,
+      "loss": 0.4958,
+      "step": 1107
+    },
+    {
+      "epoch": 0.20300476364968853,
+      "grad_norm": 0.37675479403578893,
+      "learning_rate": 9.681591577883433e-06,
+      "loss": 0.4778,
+      "step": 1108
+    },
+    {
+      "epoch": 0.20318798094540125,
+      "grad_norm": 0.3570523937587689,
+      "learning_rate": 9.680467678853378e-06,
+      "loss": 0.484,
+      "step": 1109
+    },
+    {
+      "epoch": 0.20337119824111397,
+      "grad_norm": 0.4245386369867795,
+      "learning_rate": 9.679341865248848e-06,
+      "loss": 0.4866,
+      "step": 1110
+    },
+    {
+      "epoch": 0.20355441553682668,
+      "grad_norm": 0.4405417241764804,
+      "learning_rate": 9.678214137530361e-06,
+      "loss": 0.4532,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2037376328325394,
+      "grad_norm": 0.3965079787279517,
+      "learning_rate": 9.677084496159224e-06,
+      "loss": 0.4731,
+      "step": 1112
+    },
+    {
+      "epoch": 0.20392085012825212,
+      "grad_norm": 0.38208118269449826,
+      "learning_rate": 9.675952941597522e-06,
+      "loss": 0.4397,
+      "step": 1113
+    },
+    {
+      "epoch": 0.20410406742396484,
+      "grad_norm": 0.39488634039550174,
+      "learning_rate": 9.674819474308126e-06,
+      "loss": 0.4928,
+      "step": 1114
+    },
+    {
+      "epoch": 0.20428728471967753,
+      "grad_norm": 0.3739080933909842,
+      "learning_rate": 9.673684094754686e-06,
+      "loss": 0.486,
+      "step": 1115
+    },
+    {
+      "epoch": 0.20447050201539024,
+      "grad_norm": 0.3958666571427383,
+      "learning_rate": 9.672546803401638e-06,
+      "loss": 0.5102,
+      "step": 1116
+    },
+    {
+      "epoch": 0.20465371931110296,
+      "grad_norm": 1.2610206438333627,
+      "learning_rate": 9.671407600714197e-06,
+      "loss": 0.4555,
+      "step": 1117
+    },
+    {
+      "epoch": 0.20483693660681568,
+      "grad_norm": 0.4461796609741918,
+      "learning_rate": 9.670266487158363e-06,
+      "loss": 0.4718,
+      "step": 1118
+    },
+    {
+      "epoch": 0.2050201539025284,
+      "grad_norm": 0.43024748470939017,
+      "learning_rate": 9.669123463200914e-06,
+      "loss": 0.4953,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2052033711982411,
+      "grad_norm": 0.44154226794099993,
+      "learning_rate": 9.66797852930941e-06,
+      "loss": 0.5181,
+      "step": 1120
+    },
+    {
+      "epoch": 0.20538658849395383,
+      "grad_norm": 0.41502499575208784,
+      "learning_rate": 9.666831685952194e-06,
+      "loss": 0.4772,
+      "step": 1121
+    },
+    {
+      "epoch": 0.20556980578966655,
+      "grad_norm": 0.4084436936730005,
+      "learning_rate": 9.66568293359839e-06,
+      "loss": 0.5047,
+      "step": 1122
+    },
+    {
+      "epoch": 0.20575302308537927,
+      "grad_norm": 0.43479037966133816,
+      "learning_rate": 9.664532272717902e-06,
+      "loss": 0.4698,
+      "step": 1123
+    },
+    {
+      "epoch": 0.20593624038109198,
+      "grad_norm": 0.7606951229520282,
+      "learning_rate": 9.663379703781414e-06,
+      "loss": 0.4841,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2061194576768047,
+      "grad_norm": 0.4006994667526495,
+      "learning_rate": 9.662225227260395e-06,
+      "loss": 0.4963,
+      "step": 1125
+    },
+    {
+      "epoch": 0.20630267497251742,
+      "grad_norm": 0.39866531966563096,
+      "learning_rate": 9.661068843627088e-06,
+      "loss": 0.4777,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2064858922682301,
+      "grad_norm": 0.4035785610200985,
+      "learning_rate": 9.65991055335452e-06,
+      "loss": 0.4666,
+      "step": 1127
+    },
+    {
+      "epoch": 0.20666910956394283,
+      "grad_norm": 0.438045344209569,
+      "learning_rate": 9.658750356916494e-06,
+      "loss": 0.4956,
+      "step": 1128
+    },
+    {
+      "epoch": 0.20685232685965554,
+      "grad_norm": 0.3835030416878391,
+      "learning_rate": 9.657588254787598e-06,
+      "loss": 0.4886,
+      "step": 1129
+    },
+    {
+      "epoch": 0.20703554415536826,
+      "grad_norm": 0.4044887736362754,
+      "learning_rate": 9.6564242474432e-06,
+      "loss": 0.4831,
+      "step": 1130
+    },
+    {
+      "epoch": 0.20721876145108098,
+      "grad_norm": 0.40816269155213675,
+      "learning_rate": 9.655258335359438e-06,
+      "loss": 0.497,
+      "step": 1131
+    },
+    {
+      "epoch": 0.2074019787467937,
+      "grad_norm": 0.39413831022845913,
+      "learning_rate": 9.65409051901324e-06,
+      "loss": 0.4588,
+      "step": 1132
+    },
+    {
+      "epoch": 0.20758519604250641,
+      "grad_norm": 0.5053059172212088,
+      "learning_rate": 9.652920798882307e-06,
+      "loss": 0.4985,
+      "step": 1133
+    },
+    {
+      "epoch": 0.20776841333821913,
+      "grad_norm": 0.3669036545586826,
+      "learning_rate": 9.651749175445123e-06,
+      "loss": 0.4917,
+      "step": 1134
+    },
+    {
+      "epoch": 0.20795163063393185,
+      "grad_norm": 0.4268516645401171,
+      "learning_rate": 9.650575649180944e-06,
+      "loss": 0.4849,
+      "step": 1135
+    },
+    {
+      "epoch": 0.20813484792964457,
+      "grad_norm": 0.4262534436550349,
+      "learning_rate": 9.649400220569812e-06,
+      "loss": 0.5128,
+      "step": 1136
+    },
+    {
+      "epoch": 0.20831806522535729,
+      "grad_norm": 0.3944076446197719,
+      "learning_rate": 9.648222890092538e-06,
+      "loss": 0.5014,
+      "step": 1137
+    },
+    {
+      "epoch": 0.20850128252107,
+      "grad_norm": 0.40237247723443686,
+      "learning_rate": 9.647043658230723e-06,
+      "loss": 0.5162,
+      "step": 1138
+    },
+    {
+      "epoch": 0.2086844998167827,
+      "grad_norm": 0.40871120340037176,
+      "learning_rate": 9.645862525466734e-06,
+      "loss": 0.5021,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2088677171124954,
+      "grad_norm": 0.40979376980666626,
+      "learning_rate": 9.644679492283723e-06,
+      "loss": 0.5104,
+      "step": 1140
+    },
+    {
+      "epoch": 0.20905093440820813,
+      "grad_norm": 0.41263650327798096,
+      "learning_rate": 9.643494559165616e-06,
+      "loss": 0.4764,
+      "step": 1141
+    },
+    {
+      "epoch": 0.20923415170392085,
+      "grad_norm": 0.37127003400941044,
+      "learning_rate": 9.64230772659712e-06,
+      "loss": 0.4858,
+      "step": 1142
+    },
+    {
+      "epoch": 0.20941736899963356,
+      "grad_norm": 0.4556209336445142,
+      "learning_rate": 9.641118995063712e-06,
+      "loss": 0.4668,
+      "step": 1143
+    },
+    {
+      "epoch": 0.20960058629534628,
+      "grad_norm": 0.37966915346806135,
+      "learning_rate": 9.639928365051655e-06,
+      "loss": 0.4733,
+      "step": 1144
+    },
+    {
+      "epoch": 0.209783803591059,
+      "grad_norm": 0.5527182868106905,
+      "learning_rate": 9.638735837047976e-06,
+      "loss": 0.5182,
+      "step": 1145
+    },
+    {
+      "epoch": 0.20996702088677172,
+      "grad_norm": 0.37939914018444165,
+      "learning_rate": 9.637541411540496e-06,
+      "loss": 0.4866,
+      "step": 1146
+    },
+    {
+      "epoch": 0.21015023818248443,
+      "grad_norm": 0.44598863023348195,
+      "learning_rate": 9.636345089017795e-06,
+      "loss": 0.5115,
+      "step": 1147
+    },
+    {
+      "epoch": 0.21033345547819715,
+      "grad_norm": 0.4509291233105335,
+      "learning_rate": 9.635146869969239e-06,
+      "loss": 0.5099,
+      "step": 1148
+    },
+    {
+      "epoch": 0.21051667277390987,
+      "grad_norm": 0.43540452743790453,
+      "learning_rate": 9.633946754884963e-06,
+      "loss": 0.4971,
+      "step": 1149
+    },
+    {
+      "epoch": 0.21069989006962256,
+      "grad_norm": 0.43503175511293474,
+      "learning_rate": 9.63274474425589e-06,
+      "loss": 0.535,
+      "step": 1150
+    },
+    {
+      "epoch": 0.21088310736533528,
+      "grad_norm": 0.4284909564211116,
+      "learning_rate": 9.6315408385737e-06,
+      "loss": 0.5195,
+      "step": 1151
+    },
+    {
+      "epoch": 0.211066324661048,
+      "grad_norm": 0.4144953761451503,
+      "learning_rate": 9.630335038330867e-06,
+      "loss": 0.4917,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2112495419567607,
+      "grad_norm": 0.4897343272100308,
+      "learning_rate": 9.629127344020625e-06,
+      "loss": 0.5315,
+      "step": 1153
+    },
+    {
+      "epoch": 0.21143275925247343,
+      "grad_norm": 0.422768220232635,
+      "learning_rate": 9.627917756136991e-06,
+      "loss": 0.4751,
+      "step": 1154
+    },
+    {
+      "epoch": 0.21161597654818615,
+      "grad_norm": 0.4834794267838247,
+      "learning_rate": 9.626706275174754e-06,
+      "loss": 0.49,
+      "step": 1155
+    },
+    {
+      "epoch": 0.21179919384389886,
+      "grad_norm": 0.43277895942455413,
+      "learning_rate": 9.625492901629478e-06,
+      "loss": 0.4942,
+      "step": 1156
+    },
+    {
+      "epoch": 0.21198241113961158,
+      "grad_norm": 0.4500403514457418,
+      "learning_rate": 9.624277635997503e-06,
+      "loss": 0.4794,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2121656284353243,
+      "grad_norm": 0.4099295617842837,
+      "learning_rate": 9.62306047877594e-06,
+      "loss": 0.5079,
+      "step": 1158
+    },
+    {
+      "epoch": 0.21234884573103702,
+      "grad_norm": 0.39697880470360636,
+      "learning_rate": 9.62184143046267e-06,
+      "loss": 0.4718,
+      "step": 1159
+    },
+    {
+      "epoch": 0.21253206302674973,
+      "grad_norm": 0.45285900320929984,
+      "learning_rate": 9.620620491556359e-06,
+      "loss": 0.4815,
+      "step": 1160
+    },
+    {
+      "epoch": 0.21271528032246245,
+      "grad_norm": 0.4616778523748189,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 0.5072,
+      "step": 1161
+    },
+    {
+      "epoch": 0.21289849761817514,
+      "grad_norm": 0.42472446401669417,
+      "learning_rate": 9.618172943963107e-06,
+      "loss": 0.5205,
+      "step": 1162
+    },
+    {
+      "epoch": 0.21308171491388786,
+      "grad_norm": 1.1202188747857207,
+      "learning_rate": 9.61694633627735e-06,
+      "loss": 0.4941,
+      "step": 1163
+    },
+    {
+      "epoch": 0.21326493220960058,
+      "grad_norm": 0.4238555504122449,
+      "learning_rate": 9.615717840000922e-06,
+      "loss": 0.5259,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2134481495053133,
+      "grad_norm": 0.4331940724685078,
+      "learning_rate": 9.614487455636341e-06,
+      "loss": 0.4792,
+      "step": 1165
+    },
+    {
+      "epoch": 0.213631366801026,
+      "grad_norm": 0.42864744785979175,
+      "learning_rate": 9.613255183686906e-06,
+      "loss": 0.4737,
+      "step": 1166
+    },
+    {
+      "epoch": 0.21381458409673873,
+      "grad_norm": 0.43365153166481973,
+      "learning_rate": 9.612021024656685e-06,
+      "loss": 0.4709,
+      "step": 1167
+    },
+    {
+      "epoch": 0.21399780139245145,
+      "grad_norm": 0.43747080354215073,
+      "learning_rate": 9.610784979050519e-06,
+      "loss": 0.5012,
+      "step": 1168
+    },
+    {
+      "epoch": 0.21418101868816417,
+      "grad_norm": 0.4204470389686437,
+      "learning_rate": 9.609547047374018e-06,
+      "loss": 0.4905,
+      "step": 1169
+    },
+    {
+      "epoch": 0.21436423598387688,
+      "grad_norm": 0.42512506549883267,
+      "learning_rate": 9.608307230133566e-06,
+      "loss": 0.4672,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2145474532795896,
+      "grad_norm": 0.4326161922702061,
+      "learning_rate": 9.607065527836324e-06,
+      "loss": 0.5035,
+      "step": 1171
+    },
+    {
+      "epoch": 0.21473067057530232,
+      "grad_norm": 0.4012535242609492,
+      "learning_rate": 9.60582194099021e-06,
+      "loss": 0.5042,
+      "step": 1172
+    },
+    {
+      "epoch": 0.21491388787101504,
+      "grad_norm": 0.407044902400079,
+      "learning_rate": 9.604576470103923e-06,
+      "loss": 0.4751,
+      "step": 1173
+    },
+    {
+      "epoch": 0.21509710516672773,
+      "grad_norm": 0.420735767335274,
+      "learning_rate": 9.603329115686934e-06,
+      "loss": 0.4872,
+      "step": 1174
+    },
+    {
+      "epoch": 0.21528032246244044,
+      "grad_norm": 0.41386495522739447,
+      "learning_rate": 9.60207987824948e-06,
+      "loss": 0.4554,
+      "step": 1175
+    },
+    {
+      "epoch": 0.21546353975815316,
+      "grad_norm": 0.3993847396510676,
+      "learning_rate": 9.600828758302568e-06,
+      "loss": 0.4769,
+      "step": 1176
+    },
+    {
+      "epoch": 0.21564675705386588,
+      "grad_norm": 0.43412675596114314,
+      "learning_rate": 9.599575756357974e-06,
+      "loss": 0.4604,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2158299743495786,
+      "grad_norm": 0.42400411471398847,
+      "learning_rate": 9.598320872928251e-06,
+      "loss": 0.4644,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2160131916452913,
+      "grad_norm": 0.43220524123609294,
+      "learning_rate": 9.597064108526715e-06,
+      "loss": 0.4931,
+      "step": 1179
+    },
+    {
+      "epoch": 0.21619640894100403,
+      "grad_norm": 0.4139513466321143,
+      "learning_rate": 9.595805463667452e-06,
+      "loss": 0.502,
+      "step": 1180
+    },
+    {
+      "epoch": 0.21637962623671675,
+      "grad_norm": 0.5080108758300018,
+      "learning_rate": 9.59454493886532e-06,
+      "loss": 0.4825,
+      "step": 1181
+    },
+    {
+      "epoch": 0.21656284353242947,
+      "grad_norm": 0.47000114005402754,
+      "learning_rate": 9.593282534635945e-06,
+      "loss": 0.4678,
+      "step": 1182
+    },
+    {
+      "epoch": 0.21674606082814218,
+      "grad_norm": 0.42032814714737154,
+      "learning_rate": 9.59201825149572e-06,
+      "loss": 0.5086,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2169292781238549,
+      "grad_norm": 0.4011425053622946,
+      "learning_rate": 9.590752089961811e-06,
+      "loss": 0.5067,
+      "step": 1184
+    },
+    {
+      "epoch": 0.21711249541956762,
+      "grad_norm": 0.38415361503458767,
+      "learning_rate": 9.589484050552142e-06,
+      "loss": 0.5192,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2172957127152803,
+      "grad_norm": 0.4099189657711175,
+      "learning_rate": 9.588214133785421e-06,
+      "loss": 0.5087,
+      "step": 1186
+    },
+    {
+      "epoch": 0.21747893001099303,
+      "grad_norm": 0.4259010610337274,
+      "learning_rate": 9.58694234018111e-06,
+      "loss": 0.4931,
+      "step": 1187
+    },
+    {
+      "epoch": 0.21766214730670574,
+      "grad_norm": 0.4709131450449139,
+      "learning_rate": 9.585668670259446e-06,
+      "loss": 0.4949,
+      "step": 1188
+    },
+    {
+      "epoch": 0.21784536460241846,
+      "grad_norm": 0.44547955151410057,
+      "learning_rate": 9.584393124541431e-06,
+      "loss": 0.4771,
+      "step": 1189
+    },
+    {
+      "epoch": 0.21802858189813118,
+      "grad_norm": 0.43579531600042903,
+      "learning_rate": 9.583115703548835e-06,
+      "loss": 0.5121,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2182117991938439,
+      "grad_norm": 0.3959951122458681,
+      "learning_rate": 9.581836407804196e-06,
+      "loss": 0.4739,
+      "step": 1191
+    },
+    {
+      "epoch": 0.21839501648955661,
+      "grad_norm": 0.39913718386402464,
+      "learning_rate": 9.580555237830817e-06,
+      "loss": 0.476,
+      "step": 1192
+    },
+    {
+      "epoch": 0.21857823378526933,
+      "grad_norm": 0.3981984945355583,
+      "learning_rate": 9.579272194152767e-06,
+      "loss": 0.5042,
+      "step": 1193
+    },
+    {
+      "epoch": 0.21876145108098205,
+      "grad_norm": 0.4369820172983135,
+      "learning_rate": 9.577987277294887e-06,
+      "loss": 0.5072,
+      "step": 1194
+    },
+    {
+      "epoch": 0.21894466837669477,
+      "grad_norm": 0.43452616563791835,
+      "learning_rate": 9.576700487782775e-06,
+      "loss": 0.5111,
+      "step": 1195
+    },
+    {
+      "epoch": 0.21912788567240749,
+      "grad_norm": 0.45908350014787985,
+      "learning_rate": 9.575411826142806e-06,
+      "loss": 0.5284,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2193111029681202,
+      "grad_norm": 0.40541576693311915,
+      "learning_rate": 9.57412129290211e-06,
+      "loss": 0.465,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2194943202638329,
+      "grad_norm": 0.3894120792636639,
+      "learning_rate": 9.572828888588593e-06,
+      "loss": 0.502,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2196775375595456,
+      "grad_norm": 0.41214065106611036,
+      "learning_rate": 9.571534613730915e-06,
+      "loss": 0.4999,
+      "step": 1199
+    },
+    {
+      "epoch": 0.21986075485525833,
+      "grad_norm": 0.38064251771626856,
+      "learning_rate": 9.57023846885851e-06,
+      "loss": 0.485,
+      "step": 1200
+    },
+    {
+      "epoch": 0.22004397215097105,
+      "grad_norm": 0.3546653255179806,
+      "learning_rate": 9.568940454501578e-06,
+      "loss": 0.4575,
+      "step": 1201
+    },
+    {
+      "epoch": 0.22022718944668376,
+      "grad_norm": 0.40806335023375157,
+      "learning_rate": 9.567640571191073e-06,
+      "loss": 0.4516,
+      "step": 1202
+    },
+    {
+      "epoch": 0.22041040674239648,
+      "grad_norm": 0.4688612954629896,
+      "learning_rate": 9.566338819458726e-06,
+      "loss": 0.5134,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2205936240381092,
+      "grad_norm": 0.39969548266649896,
+      "learning_rate": 9.56503519983702e-06,
+      "loss": 0.4847,
+      "step": 1204
+    },
+    {
+      "epoch": 0.22077684133382192,
+      "grad_norm": 0.4383715060312841,
+      "learning_rate": 9.563729712859216e-06,
+      "loss": 0.5254,
+      "step": 1205
+    },
+    {
+      "epoch": 0.22096005862953463,
+      "grad_norm": 0.4352285842560572,
+      "learning_rate": 9.562422359059328e-06,
+      "loss": 0.5016,
+      "step": 1206
+    },
+    {
+      "epoch": 0.22114327592524735,
+      "grad_norm": 0.39691391609356236,
+      "learning_rate": 9.561113138972138e-06,
+      "loss": 0.5089,
+      "step": 1207
+    },
+    {
+      "epoch": 0.22132649322096007,
+      "grad_norm": 0.3960087971203996,
+      "learning_rate": 9.55980205313319e-06,
+      "loss": 0.4977,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2215097105166728,
+      "grad_norm": 0.4287362367803256,
+      "learning_rate": 9.558489102078792e-06,
+      "loss": 0.4641,
+      "step": 1209
+    },
+    {
+      "epoch": 0.22169292781238548,
+      "grad_norm": 0.3811089768117096,
+      "learning_rate": 9.557174286346014e-06,
+      "loss": 0.5213,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2218761451080982,
+      "grad_norm": 0.4593900206484617,
+      "learning_rate": 9.555857606472692e-06,
+      "loss": 0.514,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2220593624038109,
+      "grad_norm": 0.3711599555884036,
+      "learning_rate": 9.554539062997421e-06,
+      "loss": 0.5003,
+      "step": 1212
+    },
+    {
+      "epoch": 0.22224257969952363,
+      "grad_norm": 0.42622144015650515,
+      "learning_rate": 9.553218656459558e-06,
+      "loss": 0.5486,
+      "step": 1213
+    },
+    {
+      "epoch": 0.22242579699523635,
+      "grad_norm": 0.4167867779019379,
+      "learning_rate": 9.551896387399226e-06,
+      "loss": 0.5184,
+      "step": 1214
+    },
+    {
+      "epoch": 0.22260901429094906,
+      "grad_norm": 0.3809013711065984,
+      "learning_rate": 9.550572256357305e-06,
+      "loss": 0.5022,
+      "step": 1215
+    },
+    {
+      "epoch": 0.22279223158666178,
+      "grad_norm": 0.44157858433876945,
+      "learning_rate": 9.54924626387544e-06,
+      "loss": 0.4605,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2229754488823745,
+      "grad_norm": 0.38145210565349935,
+      "learning_rate": 9.547918410496037e-06,
+      "loss": 0.497,
+      "step": 1217
+    },
+    {
+      "epoch": 0.22315866617808722,
+      "grad_norm": 0.4245361578656831,
+      "learning_rate": 9.546588696762262e-06,
+      "loss": 0.5249,
+      "step": 1218
+    },
+    {
+      "epoch": 0.22334188347379993,
+      "grad_norm": 0.3806441240590457,
+      "learning_rate": 9.545257123218043e-06,
+      "loss": 0.489,
+      "step": 1219
+    },
+    {
+      "epoch": 0.22352510076951265,
+      "grad_norm": 0.3771756284454525,
+      "learning_rate": 9.54392369040807e-06,
+      "loss": 0.4835,
+      "step": 1220
+    },
+    {
+      "epoch": 0.22370831806522537,
+      "grad_norm": 0.42014861145100263,
+      "learning_rate": 9.542588398877787e-06,
+      "loss": 0.4908,
+      "step": 1221
+    },
+    {
+      "epoch": 0.22389153536093806,
+      "grad_norm": 0.40603250481690667,
+      "learning_rate": 9.54125124917341e-06,
+      "loss": 0.4936,
+      "step": 1222
+    },
+    {
+      "epoch": 0.22407475265665078,
+      "grad_norm": 0.4061177784088164,
+      "learning_rate": 9.539912241841904e-06,
+      "loss": 0.4827,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2242579699523635,
+      "grad_norm": 0.3861564878611609,
+      "learning_rate": 9.538571377431e-06,
+      "loss": 0.4676,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2244411872480762,
+      "grad_norm": 0.42382201486067383,
+      "learning_rate": 9.537228656489187e-06,
+      "loss": 0.4602,
+      "step": 1225
+    },
+    {
+      "epoch": 0.22462440454378893,
+      "grad_norm": 0.3822471479641513,
+      "learning_rate": 9.53588407956571e-06,
+      "loss": 0.4508,
+      "step": 1226
+    },
+    {
+      "epoch": 0.22480762183950165,
+      "grad_norm": 0.3882035970294104,
+      "learning_rate": 9.534537647210582e-06,
+      "loss": 0.4723,
+      "step": 1227
+    },
+    {
+      "epoch": 0.22499083913521437,
+      "grad_norm": 0.46953161111998076,
+      "learning_rate": 9.533189359974564e-06,
+      "loss": 0.5037,
+      "step": 1228
+    },
+    {
+      "epoch": 0.22517405643092708,
+      "grad_norm": 0.3742186217573921,
+      "learning_rate": 9.531839218409186e-06,
+      "loss": 0.4821,
+      "step": 1229
+    },
+    {
+      "epoch": 0.2253572737266398,
+      "grad_norm": 0.42732489656890116,
+      "learning_rate": 9.53048722306673e-06,
+      "loss": 0.4959,
+      "step": 1230
+    },
+    {
+      "epoch": 0.22554049102235252,
+      "grad_norm": 0.36370123103326807,
+      "learning_rate": 9.52913337450024e-06,
+      "loss": 0.509,
+      "step": 1231
+    },
+    {
+      "epoch": 0.22572370831806524,
+      "grad_norm": 0.3842157959300981,
+      "learning_rate": 9.527777673263512e-06,
+      "loss": 0.4776,
+      "step": 1232
+    },
+    {
+      "epoch": 0.22590692561377795,
+      "grad_norm": 0.39958584358419963,
+      "learning_rate": 9.526420119911109e-06,
+      "loss": 0.4937,
+      "step": 1233
+    },
+    {
+      "epoch": 0.22609014290949064,
+      "grad_norm": 0.3795874511893427,
+      "learning_rate": 9.525060714998341e-06,
+      "loss": 0.4673,
+      "step": 1234
+    },
+    {
+      "epoch": 0.22627336020520336,
+      "grad_norm": 0.443615180266169,
+      "learning_rate": 9.523699459081285e-06,
+      "loss": 0.497,
+      "step": 1235
+    },
+    {
+      "epoch": 0.22645657750091608,
+      "grad_norm": 0.4126187255162815,
+      "learning_rate": 9.52233635271677e-06,
+      "loss": 0.5026,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2266397947966288,
+      "grad_norm": 0.3898731986202734,
+      "learning_rate": 9.520971396462383e-06,
+      "loss": 0.5142,
+      "step": 1237
+    },
+    {
+      "epoch": 0.2268230120923415,
+      "grad_norm": 0.4303787774334329,
+      "learning_rate": 9.519604590876471e-06,
+      "loss": 0.5027,
+      "step": 1238
+    },
+    {
+      "epoch": 0.22700622938805423,
+      "grad_norm": 0.4881712350764513,
+      "learning_rate": 9.51823593651813e-06,
+      "loss": 0.5194,
+      "step": 1239
+    },
+    {
+      "epoch": 0.22718944668376695,
+      "grad_norm": 0.3958516733915845,
+      "learning_rate": 9.516865433947218e-06,
+      "loss": 0.4873,
+      "step": 1240
+    },
+    {
+      "epoch": 0.22737266397947967,
+      "grad_norm": 0.37588183496475464,
+      "learning_rate": 9.515493083724348e-06,
+      "loss": 0.483,
+      "step": 1241
+    },
+    {
+      "epoch": 0.22755588127519238,
+      "grad_norm": 0.42594083794952703,
+      "learning_rate": 9.514118886410889e-06,
+      "loss": 0.515,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2277390985709051,
+      "grad_norm": 0.42418243334298944,
+      "learning_rate": 9.512742842568964e-06,
+      "loss": 0.5057,
+      "step": 1243
+    },
+    {
+      "epoch": 0.22792231586661782,
+      "grad_norm": 0.4402484021876601,
+      "learning_rate": 9.511364952761453e-06,
+      "loss": 0.4961,
+      "step": 1244
+    },
+    {
+      "epoch": 0.22810553316233054,
+      "grad_norm": 0.39304351029537393,
+      "learning_rate": 9.509985217551989e-06,
+      "loss": 0.4803,
+      "step": 1245
+    },
+    {
+      "epoch": 0.22828875045804323,
+      "grad_norm": 0.40248532246553786,
+      "learning_rate": 9.508603637504962e-06,
+      "loss": 0.4801,
+      "step": 1246
+    },
+    {
+      "epoch": 0.22847196775375594,
+      "grad_norm": 0.4302259083201261,
+      "learning_rate": 9.507220213185517e-06,
+      "loss": 0.5027,
+      "step": 1247
+    },
+    {
+      "epoch": 0.22865518504946866,
+      "grad_norm": 0.40571061563755584,
+      "learning_rate": 9.505834945159552e-06,
+      "loss": 0.5162,
+      "step": 1248
+    },
+    {
+      "epoch": 0.22883840234518138,
+      "grad_norm": 0.4032017209651049,
+      "learning_rate": 9.504447833993717e-06,
+      "loss": 0.4948,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2290216196408941,
+      "grad_norm": 0.4195307039259119,
+      "learning_rate": 9.503058880255423e-06,
+      "loss": 0.5007,
+      "step": 1250
+    },
+    {
+      "epoch": 0.22920483693660682,
+      "grad_norm": 0.4677433036389937,
+      "learning_rate": 9.501668084512827e-06,
+      "loss": 0.4676,
+      "step": 1251
+    },
+    {
+      "epoch": 0.22938805423231953,
+      "grad_norm": 0.44203412318706997,
+      "learning_rate": 9.500275447334843e-06,
+      "loss": 0.5178,
+      "step": 1252
+    },
+    {
+      "epoch": 0.22957127152803225,
+      "grad_norm": 0.3745369415002761,
+      "learning_rate": 9.498880969291138e-06,
+      "loss": 0.4705,
+      "step": 1253
+    },
+    {
+      "epoch": 0.22975448882374497,
+      "grad_norm": 0.3760740011440762,
+      "learning_rate": 9.497484650952133e-06,
+      "loss": 0.4982,
+      "step": 1254
+    },
+    {
+      "epoch": 0.22993770611945769,
+      "grad_norm": 0.38594463821493263,
+      "learning_rate": 9.496086492889001e-06,
+      "loss": 0.4887,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2301209234151704,
+      "grad_norm": 0.4212966089256434,
+      "learning_rate": 9.494686495673665e-06,
+      "loss": 0.4837,
+      "step": 1256
+    },
+    {
+      "epoch": 0.23030414071088312,
+      "grad_norm": 0.4210112942058116,
+      "learning_rate": 9.493284659878802e-06,
+      "loss": 0.517,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2304873580065958,
+      "grad_norm": 0.41531125869001356,
+      "learning_rate": 9.491880986077846e-06,
+      "loss": 0.5263,
+      "step": 1258
+    },
+    {
+      "epoch": 0.23067057530230853,
+      "grad_norm": 0.46209059357171517,
+      "learning_rate": 9.490475474844976e-06,
+      "loss": 0.5232,
+      "step": 1259
+    },
+    {
+      "epoch": 0.23085379259802125,
+      "grad_norm": 0.4297726055183601,
+      "learning_rate": 9.489068126755124e-06,
+      "loss": 0.4613,
+      "step": 1260
+    },
+    {
+      "epoch": 0.23103700989373396,
+      "grad_norm": 0.46903664754890034,
+      "learning_rate": 9.487658942383975e-06,
+      "loss": 0.5076,
+      "step": 1261
+    },
+    {
+      "epoch": 0.23122022718944668,
+      "grad_norm": 0.48896965014560756,
+      "learning_rate": 9.486247922307967e-06,
+      "loss": 0.487,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2314034444851594,
+      "grad_norm": 0.4218926363173936,
+      "learning_rate": 9.484835067104285e-06,
+      "loss": 0.4914,
+      "step": 1263
+    },
+    {
+      "epoch": 0.23158666178087212,
+      "grad_norm": 0.40940946307427944,
+      "learning_rate": 9.483420377350865e-06,
+      "loss": 0.4984,
+      "step": 1264
+    },
+    {
+      "epoch": 0.23176987907658483,
+      "grad_norm": 0.45293257509435014,
+      "learning_rate": 9.482003853626396e-06,
+      "loss": 0.5082,
+      "step": 1265
+    },
+    {
+      "epoch": 0.23195309637229755,
+      "grad_norm": 0.3935023699694669,
+      "learning_rate": 9.480585496510315e-06,
+      "loss": 0.5099,
+      "step": 1266
+    },
+    {
+      "epoch": 0.23213631366801027,
+      "grad_norm": 0.39198925832901294,
+      "learning_rate": 9.479165306582811e-06,
+      "loss": 0.4682,
+      "step": 1267
+    },
+    {
+      "epoch": 0.232319530963723,
+      "grad_norm": 0.34440630702814135,
+      "learning_rate": 9.47774328442482e-06,
+      "loss": 0.4835,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2325027482594357,
+      "grad_norm": 0.4696525716200109,
+      "learning_rate": 9.476319430618033e-06,
+      "loss": 0.5098,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2326859655551484,
+      "grad_norm": 0.4172646920959579,
+      "learning_rate": 9.47489374574488e-06,
+      "loss": 0.5023,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2328691828508611,
+      "grad_norm": 0.3756156161424675,
+      "learning_rate": 9.473466230388552e-06,
+      "loss": 0.4754,
+      "step": 1271
+    },
+    {
+      "epoch": 0.23305240014657383,
+      "grad_norm": 1.9398173346161394,
+      "learning_rate": 9.472036885132979e-06,
+      "loss": 0.5082,
+      "step": 1272
+    },
+    {
+      "epoch": 0.23323561744228655,
+      "grad_norm": 0.47073412743230747,
+      "learning_rate": 9.470605710562845e-06,
+      "loss": 0.5285,
+      "step": 1273
+    },
+    {
+      "epoch": 0.23341883473799926,
+      "grad_norm": 0.38789836660988597,
+      "learning_rate": 9.469172707263582e-06,
+      "loss": 0.4925,
+      "step": 1274
+    },
+    {
+      "epoch": 0.23360205203371198,
+      "grad_norm": 0.3825601936653152,
+      "learning_rate": 9.467737875821368e-06,
+      "loss": 0.4659,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2337852693294247,
+      "grad_norm": 0.3973414450558941,
+      "learning_rate": 9.46630121682313e-06,
+      "loss": 0.4772,
+      "step": 1276
+    },
+    {
+      "epoch": 0.23396848662513742,
+      "grad_norm": 0.40322202930522744,
+      "learning_rate": 9.464862730856542e-06,
+      "loss": 0.4832,
+      "step": 1277
+    },
+    {
+      "epoch": 0.23415170392085013,
+      "grad_norm": 0.42699100304138604,
+      "learning_rate": 9.463422418510024e-06,
+      "loss": 0.5212,
+      "step": 1278
+    },
+    {
+      "epoch": 0.23433492121656285,
+      "grad_norm": 0.42176298384733996,
+      "learning_rate": 9.461980280372748e-06,
+      "loss": 0.4669,
+      "step": 1279
+    },
+    {
+      "epoch": 0.23451813851227557,
+      "grad_norm": 0.4081443681151287,
+      "learning_rate": 9.460536317034627e-06,
+      "loss": 0.5113,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2347013558079883,
+      "grad_norm": 0.38795147083032466,
+      "learning_rate": 9.459090529086325e-06,
+      "loss": 0.4392,
+      "step": 1281
+    },
+    {
+      "epoch": 0.23488457310370098,
+      "grad_norm": 0.5217298258213892,
+      "learning_rate": 9.457642917119249e-06,
+      "loss": 0.5181,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2350677903994137,
+      "grad_norm": 0.42457976380897766,
+      "learning_rate": 9.456193481725555e-06,
+      "loss": 0.5034,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2352510076951264,
+      "grad_norm": 0.37057304524699214,
+      "learning_rate": 9.454742223498145e-06,
+      "loss": 0.4502,
+      "step": 1284
+    },
+    {
+      "epoch": 0.23543422499083913,
+      "grad_norm": 0.3946439030149264,
+      "learning_rate": 9.453289143030662e-06,
+      "loss": 0.4777,
+      "step": 1285
+    },
+    {
+      "epoch": 0.23561744228655185,
+      "grad_norm": 0.44885227772255887,
+      "learning_rate": 9.451834240917498e-06,
+      "loss": 0.4847,
+      "step": 1286
+    },
+    {
+      "epoch": 0.23580065958226457,
+      "grad_norm": 0.43519145234243767,
+      "learning_rate": 9.45037751775379e-06,
+      "loss": 0.5088,
+      "step": 1287
+    },
+    {
+      "epoch": 0.23598387687797728,
+      "grad_norm": 0.4309093065470007,
+      "learning_rate": 9.448918974135424e-06,
+      "loss": 0.4677,
+      "step": 1288
+    },
+    {
+      "epoch": 0.23616709417369,
+      "grad_norm": 0.47978496271668625,
+      "learning_rate": 9.447458610659019e-06,
+      "loss": 0.5116,
+      "step": 1289
+    },
+    {
+      "epoch": 0.23635031146940272,
+      "grad_norm": 0.41021471445709895,
+      "learning_rate": 9.445996427921951e-06,
+      "loss": 0.523,
+      "step": 1290
+    },
+    {
+      "epoch": 0.23653352876511544,
+      "grad_norm": 0.4296171869446803,
+      "learning_rate": 9.444532426522334e-06,
+      "loss": 0.4691,
+      "step": 1291
+    },
+    {
+      "epoch": 0.23671674606082815,
+      "grad_norm": 0.4077954317933853,
+      "learning_rate": 9.443066607059026e-06,
+      "loss": 0.4772,
+      "step": 1292
+    },
+    {
+      "epoch": 0.23689996335654084,
+      "grad_norm": 0.35793769775976464,
+      "learning_rate": 9.44159897013163e-06,
+      "loss": 0.4924,
+      "step": 1293
+    },
+    {
+      "epoch": 0.23708318065225356,
+      "grad_norm": 0.39909841974854016,
+      "learning_rate": 9.440129516340492e-06,
+      "loss": 0.4823,
+      "step": 1294
+    },
+    {
+      "epoch": 0.23726639794796628,
+      "grad_norm": 0.3675254686648848,
+      "learning_rate": 9.4386582462867e-06,
+      "loss": 0.506,
+      "step": 1295
+    },
+    {
+      "epoch": 0.237449615243679,
+      "grad_norm": 0.37951723029275103,
+      "learning_rate": 9.43718516057209e-06,
+      "loss": 0.4996,
+      "step": 1296
+    },
+    {
+      "epoch": 0.23763283253939171,
+      "grad_norm": 0.4395625646258237,
+      "learning_rate": 9.435710259799234e-06,
+      "loss": 0.4841,
+      "step": 1297
+    },
+    {
+      "epoch": 0.23781604983510443,
+      "grad_norm": 0.4409131115143356,
+      "learning_rate": 9.434233544571446e-06,
+      "loss": 0.524,
+      "step": 1298
+    },
+    {
+      "epoch": 0.23799926713081715,
+      "grad_norm": 0.38035953245650445,
+      "learning_rate": 9.432755015492794e-06,
+      "loss": 0.4833,
+      "step": 1299
+    },
+    {
+      "epoch": 0.23818248442652987,
+      "grad_norm": 0.37603352556682307,
+      "learning_rate": 9.431274673168072e-06,
+      "loss": 0.4771,
+      "step": 1300
+    },
+    {
+      "epoch": 0.23836570172224258,
+      "grad_norm": 0.3973590263816894,
+      "learning_rate": 9.429792518202826e-06,
+      "loss": 0.4519,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2385489190179553,
+      "grad_norm": 0.46279982818002663,
+      "learning_rate": 9.428308551203342e-06,
+      "loss": 0.5067,
+      "step": 1302
+    },
+    {
+      "epoch": 0.23873213631366802,
+      "grad_norm": 0.44423083147602355,
+      "learning_rate": 9.426822772776645e-06,
+      "loss": 0.4955,
+      "step": 1303
+    },
+    {
+      "epoch": 0.23891535360938074,
+      "grad_norm": 0.3765473227250332,
+      "learning_rate": 9.425335183530501e-06,
+      "loss": 0.4688,
+      "step": 1304
+    },
+    {
+      "epoch": 0.23909857090509343,
+      "grad_norm": 0.38352159262743835,
+      "learning_rate": 9.42384578407342e-06,
+      "loss": 0.5013,
+      "step": 1305
+    },
+    {
+      "epoch": 0.23928178820080614,
+      "grad_norm": 0.38474924269669325,
+      "learning_rate": 9.422354575014644e-06,
+      "loss": 0.5259,
+      "step": 1306
+    },
+    {
+      "epoch": 0.23946500549651886,
+      "grad_norm": 0.3764815533565891,
+      "learning_rate": 9.42086155696417e-06,
+      "loss": 0.4907,
+      "step": 1307
+    },
+    {
+      "epoch": 0.23964822279223158,
+      "grad_norm": 0.40120663442717996,
+      "learning_rate": 9.41936673053272e-06,
+      "loss": 0.5082,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2398314400879443,
+      "grad_norm": 0.38410734753609116,
+      "learning_rate": 9.417870096331764e-06,
+      "loss": 0.5235,
+      "step": 1309
+    },
+    {
+      "epoch": 0.24001465738365702,
+      "grad_norm": 0.4824543008653062,
+      "learning_rate": 9.416371654973513e-06,
+      "loss": 0.5306,
+      "step": 1310
+    },
+    {
+      "epoch": 0.24019787467936973,
+      "grad_norm": 0.4448880086308854,
+      "learning_rate": 9.414871407070906e-06,
+      "loss": 0.5045,
+      "step": 1311
+    },
+    {
+      "epoch": 0.24038109197508245,
+      "grad_norm": 0.3818507137941972,
+      "learning_rate": 9.413369353237637e-06,
+      "loss": 0.4903,
+      "step": 1312
+    },
+    {
+      "epoch": 0.24056430927079517,
+      "grad_norm": 0.42284475981294684,
+      "learning_rate": 9.411865494088124e-06,
+      "loss": 0.4964,
+      "step": 1313
+    },
+    {
+      "epoch": 0.24074752656650789,
+      "grad_norm": 0.39875491330187324,
+      "learning_rate": 9.410359830237534e-06,
+      "loss": 0.5081,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2409307438622206,
+      "grad_norm": 0.3905340354403746,
+      "learning_rate": 9.408852362301768e-06,
+      "loss": 0.5006,
+      "step": 1315
+    },
+    {
+      "epoch": 0.24111396115793332,
+      "grad_norm": 0.5340945724213095,
+      "learning_rate": 9.407343090897464e-06,
+      "loss": 0.5028,
+      "step": 1316
+    },
+    {
+      "epoch": 0.241297178453646,
+      "grad_norm": 0.43175601981581013,
+      "learning_rate": 9.405832016641997e-06,
+      "loss": 0.5053,
+      "step": 1317
+    },
+    {
+      "epoch": 0.24148039574935873,
+      "grad_norm": 0.4128194078123402,
+      "learning_rate": 9.404319140153484e-06,
+      "loss": 0.446,
+      "step": 1318
+    },
+    {
+      "epoch": 0.24166361304507145,
+      "grad_norm": 0.3673682428697322,
+      "learning_rate": 9.402804462050776e-06,
+      "loss": 0.4664,
+      "step": 1319
+    },
+    {
+      "epoch": 0.24184683034078416,
+      "grad_norm": 0.35927179937888154,
+      "learning_rate": 9.401287982953462e-06,
+      "loss": 0.4772,
+      "step": 1320
+    },
+    {
+      "epoch": 0.24203004763649688,
+      "grad_norm": 0.436824106029447,
+      "learning_rate": 9.399769703481865e-06,
+      "loss": 0.4973,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2422132649322096,
+      "grad_norm": 0.44788915497504966,
+      "learning_rate": 9.39824962425705e-06,
+      "loss": 0.5297,
+      "step": 1322
+    },
+    {
+      "epoch": 0.24239648222792232,
+      "grad_norm": 0.43045640486065784,
+      "learning_rate": 9.396727745900811e-06,
+      "loss": 0.4906,
+      "step": 1323
+    },
+    {
+      "epoch": 0.24257969952363503,
+      "grad_norm": 0.5421905192174398,
+      "learning_rate": 9.395204069035686e-06,
+      "loss": 0.4659,
+      "step": 1324
+    },
+    {
+      "epoch": 0.24276291681934775,
+      "grad_norm": 0.4219323385690796,
+      "learning_rate": 9.39367859428494e-06,
+      "loss": 0.5003,
+      "step": 1325
+    },
+    {
+      "epoch": 0.24294613411506047,
+      "grad_norm": 0.4119758712615104,
+      "learning_rate": 9.392151322272583e-06,
+      "loss": 0.499,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2431293514107732,
+      "grad_norm": 0.4287819235082752,
+      "learning_rate": 9.390622253623353e-06,
+      "loss": 0.4939,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2433125687064859,
+      "grad_norm": 0.38055263240566445,
+      "learning_rate": 9.389091388962723e-06,
+      "loss": 0.4712,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2434957860021986,
+      "grad_norm": 0.36009711673291733,
+      "learning_rate": 9.387558728916905e-06,
+      "loss": 0.4728,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2436790032979113,
+      "grad_norm": 0.43571235014716525,
+      "learning_rate": 9.386024274112842e-06,
+      "loss": 0.4821,
+      "step": 1330
+    },
+    {
+      "epoch": 0.24386222059362403,
+      "grad_norm": 0.3659773293799275,
+      "learning_rate": 9.384488025178214e-06,
+      "loss": 0.4572,
+      "step": 1331
+    },
+    {
+      "epoch": 0.24404543788933675,
+      "grad_norm": 0.37704886297840073,
+      "learning_rate": 9.382949982741429e-06,
+      "loss": 0.4336,
+      "step": 1332
+    },
+    {
+      "epoch": 0.24422865518504946,
+      "grad_norm": 0.4586883760808847,
+      "learning_rate": 9.38141014743164e-06,
+      "loss": 0.4783,
+      "step": 1333
+    },
+    {
+      "epoch": 0.24441187248076218,
+      "grad_norm": 0.5054332600007021,
+      "learning_rate": 9.379868519878718e-06,
+      "loss": 0.4883,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2445950897764749,
+      "grad_norm": 0.3980887593736299,
+      "learning_rate": 9.378325100713283e-06,
+      "loss": 0.4865,
+      "step": 1335
+    },
+    {
+      "epoch": 0.24477830707218762,
+      "grad_norm": 0.4215392454714871,
+      "learning_rate": 9.376779890566675e-06,
+      "loss": 0.5154,
+      "step": 1336
+    },
+    {
+      "epoch": 0.24496152436790034,
+      "grad_norm": 0.42292016972151214,
+      "learning_rate": 9.375232890070973e-06,
+      "loss": 0.5184,
+      "step": 1337
+    },
+    {
+      "epoch": 0.24514474166361305,
+      "grad_norm": 0.38508367690050965,
+      "learning_rate": 9.373684099858989e-06,
+      "loss": 0.5014,
+      "step": 1338
+    },
+    {
+      "epoch": 0.24532795895932577,
+      "grad_norm": 0.4481685963998327,
+      "learning_rate": 9.372133520564264e-06,
+      "loss": 0.4831,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2455111762550385,
+      "grad_norm": 0.5247732605935221,
+      "learning_rate": 9.37058115282107e-06,
+      "loss": 0.5238,
+      "step": 1340
+    },
+    {
+      "epoch": 0.24569439355075118,
+      "grad_norm": 0.3817180851569393,
+      "learning_rate": 9.369026997264417e-06,
+      "loss": 0.4513,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2458776108464639,
+      "grad_norm": 0.41437756480910404,
+      "learning_rate": 9.36747105453004e-06,
+      "loss": 0.4692,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2460608281421766,
+      "grad_norm": 0.45961904721533503,
+      "learning_rate": 9.365913325254406e-06,
+      "loss": 0.4824,
+      "step": 1343
+    },
+    {
+      "epoch": 0.24624404543788933,
+      "grad_norm": 0.411323007740071,
+      "learning_rate": 9.364353810074716e-06,
+      "loss": 0.4767,
+      "step": 1344
+    },
+    {
+      "epoch": 0.24642726273360205,
+      "grad_norm": 0.40884023238234113,
+      "learning_rate": 9.362792509628897e-06,
+      "loss": 0.5129,
+      "step": 1345
+    },
+    {
+      "epoch": 0.24661048002931477,
+      "grad_norm": 0.4143220748874932,
+      "learning_rate": 9.361229424555609e-06,
+      "loss": 0.4969,
+      "step": 1346
+    },
+    {
+      "epoch": 0.24679369732502748,
+      "grad_norm": 0.42312617400069397,
+      "learning_rate": 9.359664555494244e-06,
+      "loss": 0.4586,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2469769146207402,
+      "grad_norm": 0.476793297997145,
+      "learning_rate": 9.35809790308492e-06,
+      "loss": 0.4933,
+      "step": 1348
+    },
+    {
+      "epoch": 0.24716013191645292,
+      "grad_norm": 0.42441332538018967,
+      "learning_rate": 9.356529467968485e-06,
+      "loss": 0.4588,
+      "step": 1349
+    },
+    {
+      "epoch": 0.24734334921216564,
+      "grad_norm": 0.39138693081487125,
+      "learning_rate": 9.354959250786519e-06,
+      "loss": 0.4602,
+      "step": 1350
+    },
+    {
+      "epoch": 0.24752656650787835,
+      "grad_norm": 0.44700626066197013,
+      "learning_rate": 9.353387252181328e-06,
+      "loss": 0.4954,
+      "step": 1351
+    },
+    {
+      "epoch": 0.24770978380359107,
+      "grad_norm": 0.5055157069210112,
+      "learning_rate": 9.351813472795947e-06,
+      "loss": 0.4796,
+      "step": 1352
+    },
+    {
+      "epoch": 0.24789300109930376,
+      "grad_norm": 0.45338574174735713,
+      "learning_rate": 9.350237913274143e-06,
+      "loss": 0.5119,
+      "step": 1353
+    },
+    {
+      "epoch": 0.24807621839501648,
+      "grad_norm": 0.392746885183511,
+      "learning_rate": 9.348660574260406e-06,
+      "loss": 0.514,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2482594356907292,
+      "grad_norm": 0.4164788422614105,
+      "learning_rate": 9.347081456399958e-06,
+      "loss": 0.5089,
+      "step": 1355
+    },
+    {
+      "epoch": 0.24844265298644191,
+      "grad_norm": 0.7070530239591771,
+      "learning_rate": 9.345500560338745e-06,
+      "loss": 0.4702,
+      "step": 1356
+    },
+    {
+      "epoch": 0.24862587028215463,
+      "grad_norm": 0.4394194345754478,
+      "learning_rate": 9.343917886723444e-06,
+      "loss": 0.477,
+      "step": 1357
+    },
+    {
+      "epoch": 0.24880908757786735,
+      "grad_norm": 0.42911702918417705,
+      "learning_rate": 9.342333436201457e-06,
+      "loss": 0.4923,
+      "step": 1358
+    },
+    {
+      "epoch": 0.24899230487358007,
+      "grad_norm": 0.4470370063218885,
+      "learning_rate": 9.340747209420913e-06,
+      "loss": 0.5228,
+      "step": 1359
+    },
+    {
+      "epoch": 0.24917552216929278,
+      "grad_norm": 0.4599577217037793,
+      "learning_rate": 9.339159207030668e-06,
+      "loss": 0.5182,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2493587394650055,
+      "grad_norm": 0.4405760361683581,
+      "learning_rate": 9.337569429680306e-06,
+      "loss": 0.5319,
+      "step": 1361
+    },
+    {
+      "epoch": 0.24954195676071822,
+      "grad_norm": 0.4164050414490823,
+      "learning_rate": 9.335977878020132e-06,
+      "loss": 0.5231,
+      "step": 1362
+    },
+    {
+      "epoch": 0.24972517405643094,
+      "grad_norm": 0.4968041900926753,
+      "learning_rate": 9.334384552701183e-06,
+      "loss": 0.5057,
+      "step": 1363
+    },
+    {
+      "epoch": 0.24990839135214366,
+      "grad_norm": 0.40916476625199594,
+      "learning_rate": 9.332789454375219e-06,
+      "loss": 0.4936,
+      "step": 1364
+    },
+    {
+      "epoch": 0.25009160864785634,
+      "grad_norm": 0.40707879703628913,
+      "learning_rate": 9.331192583694722e-06,
+      "loss": 0.4911,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2502748259435691,
+      "grad_norm": 0.36742835933618884,
+      "learning_rate": 9.329593941312904e-06,
+      "loss": 0.4697,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2504580432392818,
+      "grad_norm": 0.417024876859677,
+      "learning_rate": 9.327993527883698e-06,
+      "loss": 0.4726,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2506412605349945,
+      "grad_norm": 0.4039071282595343,
+      "learning_rate": 9.326391344061767e-06,
+      "loss": 0.4545,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2508244778307072,
+      "grad_norm": 0.5042927537180502,
+      "learning_rate": 9.32478739050249e-06,
+      "loss": 0.4994,
+      "step": 1369
+    },
+    {
+      "epoch": 0.25100769512641996,
+      "grad_norm": 0.43157899355551005,
+      "learning_rate": 9.323181667861976e-06,
+      "loss": 0.5033,
+      "step": 1370
+    },
+    {
+      "epoch": 0.25119091242213265,
+      "grad_norm": 0.4065168380531392,
+      "learning_rate": 9.321574176797055e-06,
+      "loss": 0.4767,
+      "step": 1371
+    },
+    {
+      "epoch": 0.25137412971784534,
+      "grad_norm": 0.3698342286632864,
+      "learning_rate": 9.31996491796528e-06,
+      "loss": 0.4863,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2515573470135581,
+      "grad_norm": 0.43408179894940496,
+      "learning_rate": 9.318353892024934e-06,
+      "loss": 0.505,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2517405643092708,
+      "grad_norm": 0.39467414262175576,
+      "learning_rate": 9.316741099635012e-06,
+      "loss": 0.4922,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2519237816049835,
+      "grad_norm": 0.37099615952376014,
+      "learning_rate": 9.315126541455237e-06,
+      "loss": 0.4723,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2521069989006962,
+      "grad_norm": 0.40654704513015383,
+      "learning_rate": 9.313510218146055e-06,
+      "loss": 0.4858,
+      "step": 1376
+    },
+    {
+      "epoch": 0.25229021619640896,
+      "grad_norm": 0.4320196580766476,
+      "learning_rate": 9.311892130368636e-06,
+      "loss": 0.4773,
+      "step": 1377
+    },
+    {
+      "epoch": 0.25247343349212165,
+      "grad_norm": 0.4649358511506784,
+      "learning_rate": 9.310272278784865e-06,
+      "loss": 0.506,
+      "step": 1378
+    },
+    {
+      "epoch": 0.2526566507878344,
+      "grad_norm": 0.37417265682881873,
+      "learning_rate": 9.308650664057352e-06,
+      "loss": 0.474,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2528398680835471,
+      "grad_norm": 0.3877983907225597,
+      "learning_rate": 9.307027286849434e-06,
+      "loss": 0.5005,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2530230853792598,
+      "grad_norm": 0.3604568123656264,
+      "learning_rate": 9.305402147825155e-06,
+      "loss": 0.4931,
+      "step": 1381
+    },
+    {
+      "epoch": 0.2532063026749725,
+      "grad_norm": 0.4076431370889414,
+      "learning_rate": 9.303775247649297e-06,
+      "loss": 0.4781,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2533895199706852,
+      "grad_norm": 0.4125540102911716,
+      "learning_rate": 9.302146586987347e-06,
+      "loss": 0.489,
+      "step": 1383
+    },
+    {
+      "epoch": 0.25357273726639795,
+      "grad_norm": 0.38688511296298134,
+      "learning_rate": 9.300516166505523e-06,
+      "loss": 0.4997,
+      "step": 1384
+    },
+    {
+      "epoch": 0.25375595456211064,
+      "grad_norm": 0.3901423062453873,
+      "learning_rate": 9.298883986870757e-06,
+      "loss": 0.4599,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2539391718578234,
+      "grad_norm": 0.3963777690852405,
+      "learning_rate": 9.297250048750702e-06,
+      "loss": 0.5013,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2541223891535361,
+      "grad_norm": 0.42814597983549907,
+      "learning_rate": 9.295614352813732e-06,
+      "loss": 0.4841,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2543056064492488,
+      "grad_norm": 0.4342615131111368,
+      "learning_rate": 9.293976899728936e-06,
+      "loss": 0.4893,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2544888237449615,
+      "grad_norm": 0.3801898878490571,
+      "learning_rate": 9.292337690166127e-06,
+      "loss": 0.4921,
+      "step": 1389
+    },
+    {
+      "epoch": 0.25467204104067426,
+      "grad_norm": 0.41610528591828644,
+      "learning_rate": 9.290696724795833e-06,
+      "loss": 0.5179,
+      "step": 1390
+    },
+    {
+      "epoch": 0.25485525833638695,
+      "grad_norm": 0.4071057844053913,
+      "learning_rate": 9.289054004289302e-06,
+      "loss": 0.463,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2550384756320997,
+      "grad_norm": 0.44828293293160687,
+      "learning_rate": 9.287409529318497e-06,
+      "loss": 0.4899,
+      "step": 1392
+    },
+    {
+      "epoch": 0.2552216929278124,
+      "grad_norm": 0.4044272763220796,
+      "learning_rate": 9.285763300556101e-06,
+      "loss": 0.5114,
+      "step": 1393
+    },
+    {
+      "epoch": 0.25540491022352513,
+      "grad_norm": 0.4027954029425126,
+      "learning_rate": 9.284115318675515e-06,
+      "loss": 0.4706,
+      "step": 1394
+    },
+    {
+      "epoch": 0.2555881275192378,
+      "grad_norm": 0.4318050133048906,
+      "learning_rate": 9.282465584350856e-06,
+      "loss": 0.4931,
+      "step": 1395
+    },
+    {
+      "epoch": 0.2557713448149505,
+      "grad_norm": 0.39621664861383626,
+      "learning_rate": 9.280814098256961e-06,
+      "loss": 0.4561,
+      "step": 1396
+    },
+    {
+      "epoch": 0.25595456211066325,
+      "grad_norm": 0.4720751749880995,
+      "learning_rate": 9.279160861069376e-06,
+      "loss": 0.4889,
+      "step": 1397
+    },
+    {
+      "epoch": 0.25613777940637594,
+      "grad_norm": 0.47784125077057304,
+      "learning_rate": 9.277505873464369e-06,
+      "loss": 0.5292,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2563209967020887,
+      "grad_norm": 0.48714587413048666,
+      "learning_rate": 9.275849136118926e-06,
+      "loss": 0.4785,
+      "step": 1399
+    },
+    {
+      "epoch": 0.2565042139978014,
+      "grad_norm": 0.403984192319981,
+      "learning_rate": 9.274190649710743e-06,
+      "loss": 0.5076,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2566874312935141,
+      "grad_norm": 0.39832914408918374,
+      "learning_rate": 9.272530414918236e-06,
+      "loss": 0.508,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2568706485892268,
+      "grad_norm": 0.4345591492942482,
+      "learning_rate": 9.270868432420532e-06,
+      "loss": 0.434,
+      "step": 1402
+    },
+    {
+      "epoch": 0.25705386588493956,
+      "grad_norm": 0.38575416668818974,
+      "learning_rate": 9.269204702897476e-06,
+      "loss": 0.5123,
+      "step": 1403
+    },
+    {
+      "epoch": 0.25723708318065225,
+      "grad_norm": 0.46248145640766813,
+      "learning_rate": 9.267539227029628e-06,
+      "loss": 0.462,
+      "step": 1404
+    },
+    {
+      "epoch": 0.257420300476365,
+      "grad_norm": 0.41624325170317916,
+      "learning_rate": 9.265872005498263e-06,
+      "loss": 0.5224,
+      "step": 1405
+    },
+    {
+      "epoch": 0.2576035177720777,
+      "grad_norm": 0.45537999223861836,
+      "learning_rate": 9.264203038985362e-06,
+      "loss": 0.4863,
+      "step": 1406
+    },
+    {
+      "epoch": 0.2577867350677904,
+      "grad_norm": 0.41647722499814044,
+      "learning_rate": 9.262532328173633e-06,
+      "loss": 0.498,
+      "step": 1407
+    },
+    {
+      "epoch": 0.2579699523635031,
+      "grad_norm": 0.39714530799054115,
+      "learning_rate": 9.260859873746487e-06,
+      "loss": 0.4874,
+      "step": 1408
+    },
+    {
+      "epoch": 0.2581531696592158,
+      "grad_norm": 0.3703455263000783,
+      "learning_rate": 9.259185676388053e-06,
+      "loss": 0.4676,
+      "step": 1409
+    },
+    {
+      "epoch": 0.25833638695492855,
+      "grad_norm": 0.40896486520452896,
+      "learning_rate": 9.257509736783173e-06,
+      "loss": 0.4972,
+      "step": 1410
+    },
+    {
+      "epoch": 0.25851960425064124,
+      "grad_norm": 0.38349770802111716,
+      "learning_rate": 9.2558320556174e-06,
+      "loss": 0.4986,
+      "step": 1411
+    },
+    {
+      "epoch": 0.258702821546354,
+      "grad_norm": 0.38845233640399973,
+      "learning_rate": 9.254152633576996e-06,
+      "loss": 0.4781,
+      "step": 1412
+    },
+    {
+      "epoch": 0.2588860388420667,
+      "grad_norm": 0.4414806402475634,
+      "learning_rate": 9.252471471348945e-06,
+      "loss": 0.4998,
+      "step": 1413
+    },
+    {
+      "epoch": 0.2590692561377794,
+      "grad_norm": 0.6508729529767923,
+      "learning_rate": 9.250788569620933e-06,
+      "loss": 0.4731,
+      "step": 1414
+    },
+    {
+      "epoch": 0.2592524734334921,
+      "grad_norm": 0.3943363905618656,
+      "learning_rate": 9.249103929081362e-06,
+      "loss": 0.5024,
+      "step": 1415
+    },
+    {
+      "epoch": 0.25943569072920486,
+      "grad_norm": 0.40856928097689554,
+      "learning_rate": 9.247417550419347e-06,
+      "loss": 0.5212,
+      "step": 1416
+    },
+    {
+      "epoch": 0.25961890802491755,
+      "grad_norm": 0.441062293662255,
+      "learning_rate": 9.245729434324708e-06,
+      "loss": 0.4818,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2598021253206303,
+      "grad_norm": 0.3948984673415181,
+      "learning_rate": 9.244039581487983e-06,
+      "loss": 0.4666,
+      "step": 1418
+    },
+    {
+      "epoch": 0.259985342616343,
+      "grad_norm": 0.3925260255665206,
+      "learning_rate": 9.242347992600416e-06,
+      "loss": 0.5116,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2601685599120557,
+      "grad_norm": 0.4304859621407188,
+      "learning_rate": 9.24065466835396e-06,
+      "loss": 0.4982,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2603517772077684,
+      "grad_norm": 0.40221018211406145,
+      "learning_rate": 9.23895960944128e-06,
+      "loss": 0.5005,
+      "step": 1421
+    },
+    {
+      "epoch": 0.2605349945034811,
+      "grad_norm": 0.41700277900738475,
+      "learning_rate": 9.237262816555755e-06,
+      "loss": 0.5002,
+      "step": 1422
+    },
+    {
+      "epoch": 0.26071821179919386,
+      "grad_norm": 0.4685774232060306,
+      "learning_rate": 9.235564290391461e-06,
+      "loss": 0.511,
+      "step": 1423
+    },
+    {
+      "epoch": 0.26090142909490655,
+      "grad_norm": 0.43210491917117466,
+      "learning_rate": 9.233864031643199e-06,
+      "loss": 0.4969,
+      "step": 1424
+    },
+    {
+      "epoch": 0.2610846463906193,
+      "grad_norm": 0.39899357606382013,
+      "learning_rate": 9.232162041006463e-06,
+      "loss": 0.5067,
+      "step": 1425
+    },
+    {
+      "epoch": 0.261267863686332,
+      "grad_norm": 0.3914730405010264,
+      "learning_rate": 9.230458319177467e-06,
+      "loss": 0.4567,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2614510809820447,
+      "grad_norm": 0.4224757115631004,
+      "learning_rate": 9.22875286685313e-06,
+      "loss": 0.4878,
+      "step": 1427
+    },
+    {
+      "epoch": 0.2616342982777574,
+      "grad_norm": 0.44070937438452384,
+      "learning_rate": 9.227045684731075e-06,
+      "loss": 0.4751,
+      "step": 1428
+    },
+    {
+      "epoch": 0.26181751557347016,
+      "grad_norm": 0.41874124802141355,
+      "learning_rate": 9.225336773509637e-06,
+      "loss": 0.495,
+      "step": 1429
+    },
+    {
+      "epoch": 0.26200073286918285,
+      "grad_norm": 0.39663369299149254,
+      "learning_rate": 9.223626133887856e-06,
+      "loss": 0.502,
+      "step": 1430
+    },
+    {
+      "epoch": 0.26218395016489554,
+      "grad_norm": 0.3978403037913836,
+      "learning_rate": 9.22191376656548e-06,
+      "loss": 0.5057,
+      "step": 1431
+    },
+    {
+      "epoch": 0.2623671674606083,
+      "grad_norm": 0.39621285816390917,
+      "learning_rate": 9.220199672242965e-06,
+      "loss": 0.4787,
+      "step": 1432
+    },
+    {
+      "epoch": 0.262550384756321,
+      "grad_norm": 0.36823591327830396,
+      "learning_rate": 9.218483851621469e-06,
+      "loss": 0.4759,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2627336020520337,
+      "grad_norm": 0.40956001204846254,
+      "learning_rate": 9.216766305402862e-06,
+      "loss": 0.4992,
+      "step": 1434
+    },
+    {
+      "epoch": 0.2629168193477464,
+      "grad_norm": 0.4909546794694869,
+      "learning_rate": 9.215047034289716e-06,
+      "loss": 0.4893,
+      "step": 1435
+    },
+    {
+      "epoch": 0.26310003664345916,
+      "grad_norm": 0.38305868280097505,
+      "learning_rate": 9.213326038985308e-06,
+      "loss": 0.4998,
+      "step": 1436
+    },
+    {
+      "epoch": 0.26328325393917185,
+      "grad_norm": 0.4506364554374671,
+      "learning_rate": 9.211603320193624e-06,
+      "loss": 0.5376,
+      "step": 1437
+    },
+    {
+      "epoch": 0.2634664712348846,
+      "grad_norm": 0.4532783129632112,
+      "learning_rate": 9.209878878619354e-06,
+      "loss": 0.4792,
+      "step": 1438
+    },
+    {
+      "epoch": 0.2636496885305973,
+      "grad_norm": 0.3927838537868758,
+      "learning_rate": 9.208152714967888e-06,
+      "loss": 0.4671,
+      "step": 1439
+    },
+    {
+      "epoch": 0.26383290582631,
+      "grad_norm": 0.43211843791598387,
+      "learning_rate": 9.206424829945326e-06,
+      "loss": 0.5083,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2640161231220227,
+      "grad_norm": 0.4598413576241595,
+      "learning_rate": 9.20469522425847e-06,
+      "loss": 0.4946,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2641993404177354,
+      "grad_norm": 0.4087512250232026,
+      "learning_rate": 9.202963898614825e-06,
+      "loss": 0.5002,
+      "step": 1442
+    },
+    {
+      "epoch": 0.26438255771344815,
+      "grad_norm": 0.4584659510926636,
+      "learning_rate": 9.201230853722603e-06,
+      "loss": 0.4979,
+      "step": 1443
+    },
+    {
+      "epoch": 0.26456577500916084,
+      "grad_norm": 0.4088655212775404,
+      "learning_rate": 9.199496090290713e-06,
+      "loss": 0.5083,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2647489923048736,
+      "grad_norm": 0.39750766713674895,
+      "learning_rate": 9.197759609028774e-06,
+      "loss": 0.4447,
+      "step": 1445
+    },
+    {
+      "epoch": 0.2649322096005863,
+      "grad_norm": 2.5841998747640154,
+      "learning_rate": 9.196021410647104e-06,
+      "loss": 0.4938,
+      "step": 1446
+    },
+    {
+      "epoch": 0.265115426896299,
+      "grad_norm": 0.4352826266759048,
+      "learning_rate": 9.194281495856724e-06,
+      "loss": 0.4791,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2652986441920117,
+      "grad_norm": 0.4347966702352818,
+      "learning_rate": 9.192539865369354e-06,
+      "loss": 0.4713,
+      "step": 1448
+    },
+    {
+      "epoch": 0.26548186148772446,
+      "grad_norm": 0.37418945105267554,
+      "learning_rate": 9.190796519897423e-06,
+      "loss": 0.4953,
+      "step": 1449
+    },
+    {
+      "epoch": 0.26566507878343715,
+      "grad_norm": 0.45051341568089304,
+      "learning_rate": 9.189051460154054e-06,
+      "loss": 0.4844,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2658482960791499,
+      "grad_norm": 0.4213633340607822,
+      "learning_rate": 9.187304686853078e-06,
+      "loss": 0.4927,
+      "step": 1451
+    },
+    {
+      "epoch": 0.2660315133748626,
+      "grad_norm": 0.4083049015051242,
+      "learning_rate": 9.185556200709021e-06,
+      "loss": 0.4819,
+      "step": 1452
+    },
+    {
+      "epoch": 0.26621473067057533,
+      "grad_norm": 0.7665941590007657,
+      "learning_rate": 9.183806002437112e-06,
+      "loss": 0.5237,
+      "step": 1453
+    },
+    {
+      "epoch": 0.266397947966288,
+      "grad_norm": 0.4231366399299115,
+      "learning_rate": 9.182054092753281e-06,
+      "loss": 0.5165,
+      "step": 1454
+    },
+    {
+      "epoch": 0.2665811652620007,
+      "grad_norm": 0.3819742000154287,
+      "learning_rate": 9.180300472374158e-06,
+      "loss": 0.5068,
+      "step": 1455
+    },
+    {
+      "epoch": 0.26676438255771345,
+      "grad_norm": 0.430881793871794,
+      "learning_rate": 9.178545142017073e-06,
+      "loss": 0.4855,
+      "step": 1456
+    },
+    {
+      "epoch": 0.26694759985342614,
+      "grad_norm": 0.4120790750347018,
+      "learning_rate": 9.176788102400053e-06,
+      "loss": 0.4885,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2671308171491389,
+      "grad_norm": 0.4538815285108574,
+      "learning_rate": 9.175029354241827e-06,
+      "loss": 0.4863,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2673140344448516,
+      "grad_norm": 0.3830901020270506,
+      "learning_rate": 9.173268898261822e-06,
+      "loss": 0.4863,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2674972517405643,
+      "grad_norm": 0.4245356876119995,
+      "learning_rate": 9.171506735180164e-06,
+      "loss": 0.4711,
+      "step": 1460
+    },
+    {
+      "epoch": 0.267680469036277,
+      "grad_norm": 0.47590664332011035,
+      "learning_rate": 9.169742865717675e-06,
+      "loss": 0.4902,
+      "step": 1461
+    },
+    {
+      "epoch": 0.26786368633198976,
+      "grad_norm": 0.4472826376520853,
+      "learning_rate": 9.167977290595879e-06,
+      "loss": 0.4752,
+      "step": 1462
+    },
+    {
+      "epoch": 0.26804690362770245,
+      "grad_norm": 0.45383853982527605,
+      "learning_rate": 9.166210010536996e-06,
+      "loss": 0.5041,
+      "step": 1463
+    },
+    {
+      "epoch": 0.2682301209234152,
+      "grad_norm": 0.4349556389365925,
+      "learning_rate": 9.164441026263939e-06,
+      "loss": 0.4841,
+      "step": 1464
+    },
+    {
+      "epoch": 0.2684133382191279,
+      "grad_norm": 0.3896144484943624,
+      "learning_rate": 9.162670338500323e-06,
+      "loss": 0.4691,
+      "step": 1465
+    },
+    {
+      "epoch": 0.2685965555148406,
+      "grad_norm": 0.40828579323895614,
+      "learning_rate": 9.160897947970463e-06,
+      "loss": 0.5087,
+      "step": 1466
+    },
+    {
+      "epoch": 0.2687797728105533,
+      "grad_norm": 0.44921268339324344,
+      "learning_rate": 9.159123855399364e-06,
+      "loss": 0.5205,
+      "step": 1467
+    },
+    {
+      "epoch": 0.268962990106266,
+      "grad_norm": 0.3749973142689185,
+      "learning_rate": 9.157348061512728e-06,
+      "loss": 0.5054,
+      "step": 1468
+    },
+    {
+      "epoch": 0.26914620740197875,
+      "grad_norm": 0.8460064155416173,
+      "learning_rate": 9.155570567036956e-06,
+      "loss": 0.504,
+      "step": 1469
+    },
+    {
+      "epoch": 0.26932942469769144,
+      "grad_norm": 0.40099978049747603,
+      "learning_rate": 9.153791372699141e-06,
+      "loss": 0.5169,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2695126419934042,
+      "grad_norm": 0.399695867626504,
+      "learning_rate": 9.152010479227078e-06,
+      "loss": 0.4435,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2696958592891169,
+      "grad_norm": 0.4522941668627639,
+      "learning_rate": 9.15022788734925e-06,
+      "loss": 0.4892,
+      "step": 1472
+    },
+    {
+      "epoch": 0.2698790765848296,
+      "grad_norm": 0.3907998733925797,
+      "learning_rate": 9.148443597794839e-06,
+      "loss": 0.4763,
+      "step": 1473
+    },
+    {
+      "epoch": 0.2700622938805423,
+      "grad_norm": 0.3989688222157153,
+      "learning_rate": 9.146657611293714e-06,
+      "loss": 0.4907,
+      "step": 1474
+    },
+    {
+      "epoch": 0.27024551117625506,
+      "grad_norm": 0.4158272717294187,
+      "learning_rate": 9.144869928576451e-06,
+      "loss": 0.4585,
+      "step": 1475
+    },
+    {
+      "epoch": 0.27042872847196775,
+      "grad_norm": 0.4122595114642358,
+      "learning_rate": 9.14308055037431e-06,
+      "loss": 0.5156,
+      "step": 1476
+    },
+    {
+      "epoch": 0.2706119457676805,
+      "grad_norm": 0.5428248637214301,
+      "learning_rate": 9.141289477419246e-06,
+      "loss": 0.5146,
+      "step": 1477
+    },
+    {
+      "epoch": 0.2707951630633932,
+      "grad_norm": 0.4081668756027033,
+      "learning_rate": 9.139496710443911e-06,
+      "loss": 0.4962,
+      "step": 1478
+    },
+    {
+      "epoch": 0.2709783803591059,
+      "grad_norm": 0.43919686618422127,
+      "learning_rate": 9.137702250181646e-06,
+      "loss": 0.5238,
+      "step": 1479
+    },
+    {
+      "epoch": 0.2711615976548186,
+      "grad_norm": 0.4199352917649959,
+      "learning_rate": 9.135906097366486e-06,
+      "loss": 0.5171,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2713448149505313,
+      "grad_norm": 0.3812628825329553,
+      "learning_rate": 9.134108252733159e-06,
+      "loss": 0.4919,
+      "step": 1481
+    },
+    {
+      "epoch": 0.27152803224624406,
+      "grad_norm": 0.4062350904486568,
+      "learning_rate": 9.132308717017084e-06,
+      "loss": 0.4827,
+      "step": 1482
+    },
+    {
+      "epoch": 0.27171124954195675,
+      "grad_norm": 0.37649670407672037,
+      "learning_rate": 9.130507490954375e-06,
+      "loss": 0.4862,
+      "step": 1483
+    },
+    {
+      "epoch": 0.2718944668376695,
+      "grad_norm": 0.39510364308925733,
+      "learning_rate": 9.12870457528183e-06,
+      "loss": 0.4695,
+      "step": 1484
+    },
+    {
+      "epoch": 0.2720776841333822,
+      "grad_norm": 0.4412716554236446,
+      "learning_rate": 9.126899970736947e-06,
+      "loss": 0.4929,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2722609014290949,
+      "grad_norm": 0.37777764528277363,
+      "learning_rate": 9.125093678057909e-06,
+      "loss": 0.4965,
+      "step": 1486
+    },
+    {
+      "epoch": 0.2724441187248076,
+      "grad_norm": 0.3701200088318955,
+      "learning_rate": 9.12328569798359e-06,
+      "loss": 0.4857,
+      "step": 1487
+    },
+    {
+      "epoch": 0.27262733602052036,
+      "grad_norm": 0.4323306961731567,
+      "learning_rate": 9.121476031253557e-06,
+      "loss": 0.47,
+      "step": 1488
+    },
+    {
+      "epoch": 0.27281055331623305,
+      "grad_norm": 0.40253559983281856,
+      "learning_rate": 9.119664678608067e-06,
+      "loss": 0.4474,
+      "step": 1489
+    },
+    {
+      "epoch": 0.27299377061194574,
+      "grad_norm": 0.36316272808709116,
+      "learning_rate": 9.117851640788064e-06,
+      "loss": 0.4743,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2731769879076585,
+      "grad_norm": 0.39987807717863316,
+      "learning_rate": 9.11603691853518e-06,
+      "loss": 0.5074,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2733602052033712,
+      "grad_norm": 0.36775081110408747,
+      "learning_rate": 9.11422051259174e-06,
+      "loss": 0.4805,
+      "step": 1492
+    },
+    {
+      "epoch": 0.2735434224990839,
+      "grad_norm": 0.41940289454678054,
+      "learning_rate": 9.112402423700759e-06,
+      "loss": 0.5071,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2737266397947966,
+      "grad_norm": 0.3883498469145558,
+      "learning_rate": 9.110582652605934e-06,
+      "loss": 0.4839,
+      "step": 1494
+    },
+    {
+      "epoch": 0.27390985709050936,
+      "grad_norm": 0.40736107872427696,
+      "learning_rate": 9.108761200051655e-06,
+      "loss": 0.4769,
+      "step": 1495
+    },
+    {
+      "epoch": 0.27409307438622205,
+      "grad_norm": 0.42145067923082186,
+      "learning_rate": 9.106938066782998e-06,
+      "loss": 0.5176,
+      "step": 1496
+    },
+    {
+      "epoch": 0.2742762916819348,
+      "grad_norm": 0.4436513324175086,
+      "learning_rate": 9.105113253545727e-06,
+      "loss": 0.5185,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2744595089776475,
+      "grad_norm": 0.3692944363116775,
+      "learning_rate": 9.103286761086294e-06,
+      "loss": 0.4369,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2746427262733602,
+      "grad_norm": 0.4760665480063585,
+      "learning_rate": 9.101458590151837e-06,
+      "loss": 0.5006,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2748259435690729,
+      "grad_norm": 0.4079202639977091,
+      "learning_rate": 9.099628741490179e-06,
+      "loss": 0.486,
+      "step": 1500
+    },
+    {
+      "epoch": 0.27500916086478566,
+      "grad_norm": 0.43182526673935,
+      "learning_rate": 9.097797215849834e-06,
+      "loss": 0.5081,
+      "step": 1501
+    },
+    {
+      "epoch": 0.27519237816049835,
+      "grad_norm": 0.39431567471837187,
+      "learning_rate": 9.095964013979998e-06,
+      "loss": 0.4692,
+      "step": 1502
+    },
+    {
+      "epoch": 0.27537559545621104,
+      "grad_norm": 0.4019693616622923,
+      "learning_rate": 9.094129136630552e-06,
+      "loss": 0.476,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2755588127519238,
+      "grad_norm": 0.4313498077896387,
+      "learning_rate": 9.092292584552064e-06,
+      "loss": 0.4775,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2757420300476365,
+      "grad_norm": 0.3604034259823722,
+      "learning_rate": 9.09045435849579e-06,
+      "loss": 0.497,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2759252473433492,
+      "grad_norm": 0.4981716823895217,
+      "learning_rate": 9.088614459213665e-06,
+      "loss": 0.5151,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2761084646390619,
+      "grad_norm": 0.42985286366030234,
+      "learning_rate": 9.086772887458314e-06,
+      "loss": 0.4724,
+      "step": 1507
+    },
+    {
+      "epoch": 0.27629168193477466,
+      "grad_norm": 0.40646371473453785,
+      "learning_rate": 9.08492964398304e-06,
+      "loss": 0.4875,
+      "step": 1508
+    },
+    {
+      "epoch": 0.27647489923048735,
+      "grad_norm": 0.3651071526478735,
+      "learning_rate": 9.083084729541836e-06,
+      "loss": 0.5,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2766581165262001,
+      "grad_norm": 0.49895693881998293,
+      "learning_rate": 9.081238144889373e-06,
+      "loss": 0.5184,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2768413338219128,
+      "grad_norm": 0.4128024780491911,
+      "learning_rate": 9.079389890781011e-06,
+      "loss": 0.4879,
+      "step": 1511
+    },
+    {
+      "epoch": 0.27702455111762553,
+      "grad_norm": 0.41553878397064187,
+      "learning_rate": 9.077539967972788e-06,
+      "loss": 0.5132,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2772077684133382,
+      "grad_norm": 0.5023456269550239,
+      "learning_rate": 9.075688377221428e-06,
+      "loss": 0.5316,
+      "step": 1513
+    },
+    {
+      "epoch": 0.2773909857090509,
+      "grad_norm": 0.42658543986166453,
+      "learning_rate": 9.073835119284333e-06,
+      "loss": 0.4954,
+      "step": 1514
+    },
+    {
+      "epoch": 0.27757420300476365,
+      "grad_norm": 0.4257754248428276,
+      "learning_rate": 9.071980194919592e-06,
+      "loss": 0.4739,
+      "step": 1515
+    },
+    {
+      "epoch": 0.27775742030047634,
+      "grad_norm": 0.36513061047252116,
+      "learning_rate": 9.070123604885973e-06,
+      "loss": 0.4666,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2779406375961891,
+      "grad_norm": 0.39730883326499855,
+      "learning_rate": 9.068265349942926e-06,
+      "loss": 0.4857,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2781238548919018,
+      "grad_norm": 0.4089832030333659,
+      "learning_rate": 9.06640543085058e-06,
+      "loss": 0.5033,
+      "step": 1518
+    },
+    {
+      "epoch": 0.2783070721876145,
+      "grad_norm": 0.45930382348605114,
+      "learning_rate": 9.064543848369749e-06,
+      "loss": 0.5071,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2784902894833272,
+      "grad_norm": 0.420241795630585,
+      "learning_rate": 9.062680603261923e-06,
+      "loss": 0.4966,
+      "step": 1520
+    },
+    {
+      "epoch": 0.27867350677903996,
+      "grad_norm": 0.45585947515287706,
+      "learning_rate": 9.060815696289273e-06,
+      "loss": 0.475,
+      "step": 1521
+    },
+    {
+      "epoch": 0.27885672407475265,
+      "grad_norm": 0.37205804655379004,
+      "learning_rate": 9.058949128214655e-06,
+      "loss": 0.5017,
+      "step": 1522
+    },
+    {
+      "epoch": 0.2790399413704654,
+      "grad_norm": 0.4529387207435349,
+      "learning_rate": 9.057080899801598e-06,
+      "loss": 0.482,
+      "step": 1523
+    },
+    {
+      "epoch": 0.2792231586661781,
+      "grad_norm": 0.3712655517985857,
+      "learning_rate": 9.055211011814312e-06,
+      "loss": 0.458,
+      "step": 1524
+    },
+    {
+      "epoch": 0.27940637596189083,
+      "grad_norm": 0.39642625126203734,
+      "learning_rate": 9.053339465017685e-06,
+      "loss": 0.4687,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2795895932576035,
+      "grad_norm": 0.4136774885883469,
+      "learning_rate": 9.051466260177286e-06,
+      "loss": 0.5088,
+      "step": 1526
+    },
+    {
+      "epoch": 0.2797728105533162,
+      "grad_norm": 0.41134059761022984,
+      "learning_rate": 9.049591398059364e-06,
+      "loss": 0.4934,
+      "step": 1527
+    },
+    {
+      "epoch": 0.27995602784902895,
+      "grad_norm": 0.42078398613472917,
+      "learning_rate": 9.047714879430841e-06,
+      "loss": 0.488,
+      "step": 1528
+    },
+    {
+      "epoch": 0.28013924514474164,
+      "grad_norm": 0.3893031394642896,
+      "learning_rate": 9.045836705059316e-06,
+      "loss": 0.4787,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2803224624404544,
+      "grad_norm": 0.4093693977481649,
+      "learning_rate": 9.043956875713071e-06,
+      "loss": 0.5038,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2805056797361671,
+      "grad_norm": 0.39262539132038,
+      "learning_rate": 9.042075392161062e-06,
+      "loss": 0.4789,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2806888970318798,
+      "grad_norm": 0.45335811960346867,
+      "learning_rate": 9.040192255172919e-06,
+      "loss": 0.4875,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2808721143275925,
+      "grad_norm": 0.36503912559147556,
+      "learning_rate": 9.038307465518954e-06,
+      "loss": 0.477,
+      "step": 1533
+    },
+    {
+      "epoch": 0.28105533162330526,
+      "grad_norm": 0.4257211312986204,
+      "learning_rate": 9.03642102397015e-06,
+      "loss": 0.469,
+      "step": 1534
+    },
+    {
+      "epoch": 0.28123854891901795,
+      "grad_norm": 0.4252988723490824,
+      "learning_rate": 9.034532931298169e-06,
+      "loss": 0.4881,
+      "step": 1535
+    },
+    {
+      "epoch": 0.2814217662147307,
+      "grad_norm": 0.40947075405147687,
+      "learning_rate": 9.032643188275346e-06,
+      "loss": 0.4943,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2816049835104434,
+      "grad_norm": 0.4104856496215327,
+      "learning_rate": 9.030751795674693e-06,
+      "loss": 0.5001,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2817882008061561,
+      "grad_norm": 0.4108489148032505,
+      "learning_rate": 9.028858754269893e-06,
+      "loss": 0.486,
+      "step": 1538
+    },
+    {
+      "epoch": 0.2819714181018688,
+      "grad_norm": 0.400218515841904,
+      "learning_rate": 9.026964064835312e-06,
+      "loss": 0.4875,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2821546353975815,
+      "grad_norm": 0.392307962036209,
+      "learning_rate": 9.02506772814598e-06,
+      "loss": 0.4696,
+      "step": 1540
+    },
+    {
+      "epoch": 0.28233785269329426,
+      "grad_norm": 0.39418249225011087,
+      "learning_rate": 9.023169744977607e-06,
+      "loss": 0.4944,
+      "step": 1541
+    },
+    {
+      "epoch": 0.28252106998900695,
+      "grad_norm": 0.36881071138331667,
+      "learning_rate": 9.021270116106574e-06,
+      "loss": 0.458,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2827042872847197,
+      "grad_norm": 0.3604688853323364,
+      "learning_rate": 9.019368842309937e-06,
+      "loss": 0.4612,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2828875045804324,
+      "grad_norm": 0.43502584499562924,
+      "learning_rate": 9.017465924365423e-06,
+      "loss": 0.535,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2830707218761451,
+      "grad_norm": 0.3495743983793779,
+      "learning_rate": 9.015561363051434e-06,
+      "loss": 0.4884,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2832539391718578,
+      "grad_norm": 0.38746814328257545,
+      "learning_rate": 9.01365515914704e-06,
+      "loss": 0.4684,
+      "step": 1546
+    },
+    {
+      "epoch": 0.28343715646757056,
+      "grad_norm": 0.42208621683162756,
+      "learning_rate": 9.011747313431988e-06,
+      "loss": 0.4894,
+      "step": 1547
+    },
+    {
+      "epoch": 0.28362037376328325,
+      "grad_norm": 0.40174728758795464,
+      "learning_rate": 9.009837826686693e-06,
+      "loss": 0.4892,
+      "step": 1548
+    },
+    {
+      "epoch": 0.283803591058996,
+      "grad_norm": 0.3804107018996848,
+      "learning_rate": 9.007926699692244e-06,
+      "loss": 0.4781,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2839868083547087,
+      "grad_norm": 0.4347596518542267,
+      "learning_rate": 9.006013933230398e-06,
+      "loss": 0.4735,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2841700256504214,
+      "grad_norm": 0.4121135548481947,
+      "learning_rate": 9.004099528083583e-06,
+      "loss": 0.4719,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2843532429461341,
+      "grad_norm": 0.44795396529132614,
+      "learning_rate": 9.0021834850349e-06,
+      "loss": 0.5099,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2845364602418468,
+      "grad_norm": 0.44372233437035424,
+      "learning_rate": 9.000265804868118e-06,
+      "loss": 0.4532,
+      "step": 1553
+    },
+    {
+      "epoch": 0.28471967753755956,
+      "grad_norm": 0.37122246988181606,
+      "learning_rate": 8.998346488367675e-06,
+      "loss": 0.5031,
+      "step": 1554
+    },
+    {
+      "epoch": 0.28490289483327225,
+      "grad_norm": 0.4132553844079889,
+      "learning_rate": 8.996425536318683e-06,
+      "loss": 0.4973,
+      "step": 1555
+    },
+    {
+      "epoch": 0.285086112128985,
+      "grad_norm": 0.44678572881841533,
+      "learning_rate": 8.994502949506914e-06,
+      "loss": 0.4937,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2852693294246977,
+      "grad_norm": 0.3961645213095161,
+      "learning_rate": 8.99257872871882e-06,
+      "loss": 0.4917,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2854525467204104,
+      "grad_norm": 0.4068575794251254,
+      "learning_rate": 8.99065287474151e-06,
+      "loss": 0.5059,
+      "step": 1558
+    },
+    {
+      "epoch": 0.2856357640161231,
+      "grad_norm": 0.3439421297681402,
+      "learning_rate": 8.98872538836277e-06,
+      "loss": 0.4943,
+      "step": 1559
+    },
+    {
+      "epoch": 0.28581898131183586,
+      "grad_norm": 0.4072178672567339,
+      "learning_rate": 8.986796270371047e-06,
+      "loss": 0.4884,
+      "step": 1560
+    },
+    {
+      "epoch": 0.28600219860754855,
+      "grad_norm": 0.42449273454289244,
+      "learning_rate": 8.984865521555464e-06,
+      "loss": 0.4875,
+      "step": 1561
+    },
+    {
+      "epoch": 0.28618541590326124,
+      "grad_norm": 0.37613787684490696,
+      "learning_rate": 8.9829331427058e-06,
+      "loss": 0.4857,
+      "step": 1562
+    },
+    {
+      "epoch": 0.286368633198974,
+      "grad_norm": 0.36486974332112687,
+      "learning_rate": 8.980999134612512e-06,
+      "loss": 0.5002,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2865518504946867,
+      "grad_norm": 0.38386234504756434,
+      "learning_rate": 8.979063498066714e-06,
+      "loss": 0.4885,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2867350677903994,
+      "grad_norm": 0.39062088012736546,
+      "learning_rate": 8.977126233860193e-06,
+      "loss": 0.4936,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2869182850861121,
+      "grad_norm": 0.4126575331178653,
+      "learning_rate": 8.975187342785397e-06,
+      "loss": 0.4903,
+      "step": 1566
+    },
+    {
+      "epoch": 0.28710150238182486,
+      "grad_norm": 0.46684149874863134,
+      "learning_rate": 8.973246825635441e-06,
+      "loss": 0.527,
+      "step": 1567
+    },
+    {
+      "epoch": 0.28728471967753755,
+      "grad_norm": 0.39335552791237355,
+      "learning_rate": 8.97130468320411e-06,
+      "loss": 0.4779,
+      "step": 1568
+    },
+    {
+      "epoch": 0.2874679369732503,
+      "grad_norm": 0.4246548051800614,
+      "learning_rate": 8.969360916285844e-06,
+      "loss": 0.4815,
+      "step": 1569
+    },
+    {
+      "epoch": 0.287651154268963,
+      "grad_norm": 0.44025612619503274,
+      "learning_rate": 8.967415525675756e-06,
+      "loss": 0.4371,
+      "step": 1570
+    },
+    {
+      "epoch": 0.28783437156467573,
+      "grad_norm": 0.9703871617250511,
+      "learning_rate": 8.96546851216962e-06,
+      "loss": 0.4856,
+      "step": 1571
+    },
+    {
+      "epoch": 0.2880175888603884,
+      "grad_norm": 0.49217374632039296,
+      "learning_rate": 8.96351987656387e-06,
+      "loss": 0.5087,
+      "step": 1572
+    },
+    {
+      "epoch": 0.28820080615610116,
+      "grad_norm": 0.41378220482583816,
+      "learning_rate": 8.961569619655615e-06,
+      "loss": 0.4736,
+      "step": 1573
+    },
+    {
+      "epoch": 0.28838402345181385,
+      "grad_norm": 0.4186462441272065,
+      "learning_rate": 8.959617742242615e-06,
+      "loss": 0.493,
+      "step": 1574
+    },
+    {
+      "epoch": 0.28856724074752654,
+      "grad_norm": 0.3880931424436376,
+      "learning_rate": 8.957664245123297e-06,
+      "loss": 0.4844,
+      "step": 1575
+    },
+    {
+      "epoch": 0.2887504580432393,
+      "grad_norm": 0.3741132383694949,
+      "learning_rate": 8.955709129096752e-06,
+      "loss": 0.4055,
+      "step": 1576
+    },
+    {
+      "epoch": 0.288933675338952,
+      "grad_norm": 0.44863598418647255,
+      "learning_rate": 8.953752394962736e-06,
+      "loss": 0.5278,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2891168926346647,
+      "grad_norm": 0.3876273635035489,
+      "learning_rate": 8.951794043521657e-06,
+      "loss": 0.4859,
+      "step": 1578
+    },
+    {
+      "epoch": 0.2893001099303774,
+      "grad_norm": 0.34806890820847197,
+      "learning_rate": 8.949834075574595e-06,
+      "loss": 0.4768,
+      "step": 1579
+    },
+    {
+      "epoch": 0.28948332722609016,
+      "grad_norm": 0.4072445168776299,
+      "learning_rate": 8.947872491923288e-06,
+      "loss": 0.487,
+      "step": 1580
+    },
+    {
+      "epoch": 0.28966654452180285,
+      "grad_norm": 0.4293336896066131,
+      "learning_rate": 8.94590929337013e-06,
+      "loss": 0.5008,
+      "step": 1581
+    },
+    {
+      "epoch": 0.2898497618175156,
+      "grad_norm": 0.41128872476228806,
+      "learning_rate": 8.943944480718184e-06,
+      "loss": 0.51,
+      "step": 1582
+    },
+    {
+      "epoch": 0.2900329791132283,
+      "grad_norm": 0.36929320177047475,
+      "learning_rate": 8.941978054771165e-06,
+      "loss": 0.5116,
+      "step": 1583
+    },
+    {
+      "epoch": 0.29021619640894103,
+      "grad_norm": 0.3533069262895081,
+      "learning_rate": 8.940010016333453e-06,
+      "loss": 0.462,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2903994137046537,
+      "grad_norm": 0.4290477614884879,
+      "learning_rate": 8.938040366210088e-06,
+      "loss": 0.4886,
+      "step": 1585
+    },
+    {
+      "epoch": 0.2905826310003664,
+      "grad_norm": 0.4049126999073717,
+      "learning_rate": 8.936069105206767e-06,
+      "loss": 0.488,
+      "step": 1586
+    },
+    {
+      "epoch": 0.29076584829607915,
+      "grad_norm": 0.37625165658845383,
+      "learning_rate": 8.934096234129843e-06,
+      "loss": 0.4704,
+      "step": 1587
+    },
+    {
+      "epoch": 0.29094906559179184,
+      "grad_norm": 0.4368199590248626,
+      "learning_rate": 8.932121753786339e-06,
+      "loss": 0.5089,
+      "step": 1588
+    },
+    {
+      "epoch": 0.2911322828875046,
+      "grad_norm": 0.4293928339113946,
+      "learning_rate": 8.930145664983921e-06,
+      "loss": 0.4989,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2913155001832173,
+      "grad_norm": 0.362276762434061,
+      "learning_rate": 8.928167968530921e-06,
+      "loss": 0.4848,
+      "step": 1590
+    },
+    {
+      "epoch": 0.29149871747893,
+      "grad_norm": 0.4545018284719774,
+      "learning_rate": 8.926188665236334e-06,
+      "loss": 0.5297,
+      "step": 1591
+    },
+    {
+      "epoch": 0.2916819347746427,
+      "grad_norm": 0.4271204305866213,
+      "learning_rate": 8.924207755909801e-06,
+      "loss": 0.4635,
+      "step": 1592
+    },
+    {
+      "epoch": 0.29186515207035546,
+      "grad_norm": 0.3895989149171432,
+      "learning_rate": 8.922225241361629e-06,
+      "loss": 0.4802,
+      "step": 1593
+    },
+    {
+      "epoch": 0.29204836936606815,
+      "grad_norm": 0.36495793219351547,
+      "learning_rate": 8.920241122402773e-06,
+      "loss": 0.4963,
+      "step": 1594
+    },
+    {
+      "epoch": 0.2922315866617809,
+      "grad_norm": 0.38869432485787914,
+      "learning_rate": 8.918255399844855e-06,
+      "loss": 0.4996,
+      "step": 1595
+    },
+    {
+      "epoch": 0.2924148039574936,
+      "grad_norm": 0.4601922887324616,
+      "learning_rate": 8.916268074500141e-06,
+      "loss": 0.5025,
+      "step": 1596
+    },
+    {
+      "epoch": 0.2925980212532063,
+      "grad_norm": 0.39205663001486896,
+      "learning_rate": 8.914279147181564e-06,
+      "loss": 0.4603,
+      "step": 1597
+    },
+    {
+      "epoch": 0.292781238548919,
+      "grad_norm": 0.4013458043993565,
+      "learning_rate": 8.912288618702706e-06,
+      "loss": 0.487,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2929644558446317,
+      "grad_norm": 0.38939321801864035,
+      "learning_rate": 8.910296489877803e-06,
+      "loss": 0.5029,
+      "step": 1599
+    },
+    {
+      "epoch": 0.29314767314034446,
+      "grad_norm": 0.35395541109470974,
+      "learning_rate": 8.90830276152175e-06,
+      "loss": 0.4551,
+      "step": 1600
+    },
+    {
+      "epoch": 0.29333089043605715,
+      "grad_norm": 0.41856036352490694,
+      "learning_rate": 8.906307434450092e-06,
+      "loss": 0.4949,
+      "step": 1601
+    },
+    {
+      "epoch": 0.2935141077317699,
+      "grad_norm": 0.5275243050289077,
+      "learning_rate": 8.904310509479031e-06,
+      "loss": 0.4961,
+      "step": 1602
+    },
+    {
+      "epoch": 0.2936973250274826,
+      "grad_norm": 0.4742284206723536,
+      "learning_rate": 8.902311987425422e-06,
+      "loss": 0.5095,
+      "step": 1603
+    },
+    {
+      "epoch": 0.2938805423231953,
+      "grad_norm": 0.3390894324421048,
+      "learning_rate": 8.900311869106772e-06,
+      "loss": 0.4675,
+      "step": 1604
+    },
+    {
+      "epoch": 0.294063759618908,
+      "grad_norm": 0.39896932989796424,
+      "learning_rate": 8.898310155341245e-06,
+      "loss": 0.4819,
+      "step": 1605
+    },
+    {
+      "epoch": 0.29424697691462076,
+      "grad_norm": 0.5410338815825568,
+      "learning_rate": 8.896306846947649e-06,
+      "loss": 0.5032,
+      "step": 1606
+    },
+    {
+      "epoch": 0.29443019421033345,
+      "grad_norm": 0.4070700524177446,
+      "learning_rate": 8.894301944745453e-06,
+      "loss": 0.48,
+      "step": 1607
+    },
+    {
+      "epoch": 0.2946134115060462,
+      "grad_norm": 0.42862296543081163,
+      "learning_rate": 8.892295449554777e-06,
+      "loss": 0.4605,
+      "step": 1608
+    },
+    {
+      "epoch": 0.2947966288017589,
+      "grad_norm": 0.4287346890153464,
+      "learning_rate": 8.890287362196385e-06,
+      "loss": 0.485,
+      "step": 1609
+    },
+    {
+      "epoch": 0.2949798460974716,
+      "grad_norm": 0.43250442423486146,
+      "learning_rate": 8.888277683491701e-06,
+      "loss": 0.4854,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2951630633931843,
+      "grad_norm": 0.49215086511267214,
+      "learning_rate": 8.886266414262797e-06,
+      "loss": 0.489,
+      "step": 1611
+    },
+    {
+      "epoch": 0.295346280688897,
+      "grad_norm": 0.4186341465088225,
+      "learning_rate": 8.884253555332393e-06,
+      "loss": 0.5078,
+      "step": 1612
+    },
+    {
+      "epoch": 0.29552949798460976,
+      "grad_norm": 0.4406315027714821,
+      "learning_rate": 8.882239107523863e-06,
+      "loss": 0.4937,
+      "step": 1613
+    },
+    {
+      "epoch": 0.29571271528032245,
+      "grad_norm": 0.3726428857037365,
+      "learning_rate": 8.88022307166123e-06,
+      "loss": 0.453,
+      "step": 1614
+    },
+    {
+      "epoch": 0.2958959325760352,
+      "grad_norm": 0.42821871054675215,
+      "learning_rate": 8.878205448569163e-06,
+      "loss": 0.4934,
+      "step": 1615
+    },
+    {
+      "epoch": 0.2960791498717479,
+      "grad_norm": 0.3670484899531654,
+      "learning_rate": 8.876186239072988e-06,
+      "loss": 0.486,
+      "step": 1616
+    },
+    {
+      "epoch": 0.2962623671674606,
+      "grad_norm": 0.34461787552049733,
+      "learning_rate": 8.874165443998672e-06,
+      "loss": 0.4761,
+      "step": 1617
+    },
+    {
+      "epoch": 0.2964455844631733,
+      "grad_norm": 0.37867132917069235,
+      "learning_rate": 8.872143064172834e-06,
+      "loss": 0.4821,
+      "step": 1618
+    },
+    {
+      "epoch": 0.29662880175888606,
+      "grad_norm": 0.3712153633078572,
+      "learning_rate": 8.870119100422743e-06,
+      "loss": 0.4931,
+      "step": 1619
+    },
+    {
+      "epoch": 0.29681201905459875,
+      "grad_norm": 0.36684611413431756,
+      "learning_rate": 8.86809355357631e-06,
+      "loss": 0.4616,
+      "step": 1620
+    },
+    {
+      "epoch": 0.29699523635031144,
+      "grad_norm": 0.38108459496354996,
+      "learning_rate": 8.866066424462103e-06,
+      "loss": 0.5118,
+      "step": 1621
+    },
+    {
+      "epoch": 0.2971784536460242,
+      "grad_norm": 0.382915281692164,
+      "learning_rate": 8.864037713909329e-06,
+      "loss": 0.4564,
+      "step": 1622
+    },
+    {
+      "epoch": 0.2973616709417369,
+      "grad_norm": 0.38365339546998406,
+      "learning_rate": 8.862007422747842e-06,
+      "loss": 0.516,
+      "step": 1623
+    },
+    {
+      "epoch": 0.2975448882374496,
+      "grad_norm": 0.5453723638931661,
+      "learning_rate": 8.85997555180815e-06,
+      "loss": 0.4722,
+      "step": 1624
+    },
+    {
+      "epoch": 0.2977281055331623,
+      "grad_norm": 0.38449352037816636,
+      "learning_rate": 8.8579421019214e-06,
+      "loss": 0.4583,
+      "step": 1625
+    },
+    {
+      "epoch": 0.29791132282887506,
+      "grad_norm": 0.4091259949883011,
+      "learning_rate": 8.855907073919388e-06,
+      "loss": 0.5102,
+      "step": 1626
+    },
+    {
+      "epoch": 0.29809454012458775,
+      "grad_norm": 0.46579655657632557,
+      "learning_rate": 8.853870468634554e-06,
+      "loss": 0.504,
+      "step": 1627
+    },
+    {
+      "epoch": 0.2982777574203005,
+      "grad_norm": 0.34914615282378225,
+      "learning_rate": 8.851832286899984e-06,
+      "loss": 0.4809,
+      "step": 1628
+    },
+    {
+      "epoch": 0.2984609747160132,
+      "grad_norm": 0.3465447713300964,
+      "learning_rate": 8.84979252954941e-06,
+      "loss": 0.4714,
+      "step": 1629
+    },
+    {
+      "epoch": 0.29864419201172593,
+      "grad_norm": 0.36719763001696804,
+      "learning_rate": 8.847751197417208e-06,
+      "loss": 0.4856,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2988274093074386,
+      "grad_norm": 0.3954216747573634,
+      "learning_rate": 8.845708291338396e-06,
+      "loss": 0.4802,
+      "step": 1631
+    },
+    {
+      "epoch": 0.29901062660315136,
+      "grad_norm": 0.40271380964660825,
+      "learning_rate": 8.84366381214864e-06,
+      "loss": 0.4788,
+      "step": 1632
+    },
+    {
+      "epoch": 0.29919384389886405,
+      "grad_norm": 0.3751708327909397,
+      "learning_rate": 8.841617760684242e-06,
+      "loss": 0.4834,
+      "step": 1633
+    },
+    {
+      "epoch": 0.29937706119457674,
+      "grad_norm": 0.36638025333362195,
+      "learning_rate": 8.839570137782157e-06,
+      "loss": 0.4451,
+      "step": 1634
+    },
+    {
+      "epoch": 0.2995602784902895,
+      "grad_norm": 0.34260231840043137,
+      "learning_rate": 8.837520944279976e-06,
+      "loss": 0.4816,
+      "step": 1635
+    },
+    {
+      "epoch": 0.2997434957860022,
+      "grad_norm": 0.41835359853675946,
+      "learning_rate": 8.835470181015935e-06,
+      "loss": 0.4657,
+      "step": 1636
+    },
+    {
+      "epoch": 0.2999267130817149,
+      "grad_norm": 0.3853522103777551,
+      "learning_rate": 8.833417848828912e-06,
+      "loss": 0.4798,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3001099303774276,
+      "grad_norm": 0.4311087142185147,
+      "learning_rate": 8.831363948558422e-06,
+      "loss": 0.5055,
+      "step": 1638
+    },
+    {
+      "epoch": 0.30029314767314036,
+      "grad_norm": 0.4121988474004896,
+      "learning_rate": 8.829308481044631e-06,
+      "loss": 0.4946,
+      "step": 1639
+    },
+    {
+      "epoch": 0.30047636496885305,
+      "grad_norm": 0.39926874503164067,
+      "learning_rate": 8.82725144712834e-06,
+      "loss": 0.5105,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3006595822645658,
+      "grad_norm": 0.42930617786556713,
+      "learning_rate": 8.82519284765099e-06,
+      "loss": 0.4742,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3008427995602785,
+      "grad_norm": 0.4756153230731036,
+      "learning_rate": 8.823132683454662e-06,
+      "loss": 0.4858,
+      "step": 1642
+    },
+    {
+      "epoch": 0.30102601685599123,
+      "grad_norm": 0.4105739310138263,
+      "learning_rate": 8.821070955382082e-06,
+      "loss": 0.4748,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3012092341517039,
+      "grad_norm": 0.42461593671764813,
+      "learning_rate": 8.819007664276614e-06,
+      "loss": 0.4747,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3013924514474166,
+      "grad_norm": 0.42619757195470465,
+      "learning_rate": 8.816942810982258e-06,
+      "loss": 0.4783,
+      "step": 1645
+    },
+    {
+      "epoch": 0.30157566874312935,
+      "grad_norm": 0.441815860292698,
+      "learning_rate": 8.814876396343655e-06,
+      "loss": 0.4822,
+      "step": 1646
+    },
+    {
+      "epoch": 0.30175888603884204,
+      "grad_norm": 0.39546125127248405,
+      "learning_rate": 8.812808421206083e-06,
+      "loss": 0.4751,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3019421033345548,
+      "grad_norm": 0.4039788211552708,
+      "learning_rate": 8.810738886415464e-06,
+      "loss": 0.4774,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3021253206302675,
+      "grad_norm": 0.4195550927374702,
+      "learning_rate": 8.80866779281835e-06,
+      "loss": 0.5184,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3023085379259802,
+      "grad_norm": 0.4410446432137119,
+      "learning_rate": 8.80659514126194e-06,
+      "loss": 0.52,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3024917552216929,
+      "grad_norm": 0.3910752829108485,
+      "learning_rate": 8.804520932594061e-06,
+      "loss": 0.4734,
+      "step": 1651
+    },
+    {
+      "epoch": 0.30267497251740566,
+      "grad_norm": 0.4084027351750214,
+      "learning_rate": 8.802445167663181e-06,
+      "loss": 0.4906,
+      "step": 1652
+    },
+    {
+      "epoch": 0.30285818981311835,
+      "grad_norm": 0.3779109317009322,
+      "learning_rate": 8.800367847318407e-06,
+      "loss": 0.4661,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3030414071088311,
+      "grad_norm": 0.4434368130901037,
+      "learning_rate": 8.798288972409477e-06,
+      "loss": 0.4956,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3032246244045438,
+      "grad_norm": 0.40653821037943955,
+      "learning_rate": 8.79620854378677e-06,
+      "loss": 0.5006,
+      "step": 1655
+    },
+    {
+      "epoch": 0.30340784170025653,
+      "grad_norm": 0.40448780548308255,
+      "learning_rate": 8.794126562301298e-06,
+      "loss": 0.5131,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3035910589959692,
+      "grad_norm": 0.4185848391026118,
+      "learning_rate": 8.792043028804706e-06,
+      "loss": 0.4937,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3037742762916819,
+      "grad_norm": 0.43416404958296123,
+      "learning_rate": 8.78995794414928e-06,
+      "loss": 0.4927,
+      "step": 1658
+    },
+    {
+      "epoch": 0.30395749358739466,
+      "grad_norm": 0.3941144660162759,
+      "learning_rate": 8.787871309187936e-06,
+      "loss": 0.4719,
+      "step": 1659
+    },
+    {
+      "epoch": 0.30414071088310735,
+      "grad_norm": 0.37126342352995245,
+      "learning_rate": 8.785783124774223e-06,
+      "loss": 0.4722,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3043239281788201,
+      "grad_norm": 0.4103240007205328,
+      "learning_rate": 8.783693391762328e-06,
+      "loss": 0.4809,
+      "step": 1661
+    },
+    {
+      "epoch": 0.3045071454745328,
+      "grad_norm": 0.4249044003852677,
+      "learning_rate": 8.781602111007066e-06,
+      "loss": 0.4626,
+      "step": 1662
+    },
+    {
+      "epoch": 0.3046903627702455,
+      "grad_norm": 0.41638775694944496,
+      "learning_rate": 8.779509283363894e-06,
+      "loss": 0.4929,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3048735800659582,
+      "grad_norm": 0.3861009808325714,
+      "learning_rate": 8.77741490968889e-06,
+      "loss": 0.4706,
+      "step": 1664
+    },
+    {
+      "epoch": 0.30505679736167096,
+      "grad_norm": 0.4284459852901781,
+      "learning_rate": 8.775318990838775e-06,
+      "loss": 0.542,
+      "step": 1665
+    },
+    {
+      "epoch": 0.30524001465738365,
+      "grad_norm": 0.39067748379098055,
+      "learning_rate": 8.773221527670896e-06,
+      "loss": 0.4921,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3054232319530964,
+      "grad_norm": 0.4636902711928648,
+      "learning_rate": 8.771122521043236e-06,
+      "loss": 0.4847,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3056064492488091,
+      "grad_norm": 0.4041218634315,
+      "learning_rate": 8.769021971814401e-06,
+      "loss": 0.4865,
+      "step": 1668
+    },
+    {
+      "epoch": 0.3057896665445218,
+      "grad_norm": 0.39459279955438104,
+      "learning_rate": 8.766919880843639e-06,
+      "loss": 0.519,
+      "step": 1669
+    },
+    {
+      "epoch": 0.3059728838402345,
+      "grad_norm": 0.40128357110983964,
+      "learning_rate": 8.764816248990822e-06,
+      "loss": 0.4841,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3061561011359472,
+      "grad_norm": 0.36997768478257936,
+      "learning_rate": 8.762711077116453e-06,
+      "loss": 0.4771,
+      "step": 1671
+    },
+    {
+      "epoch": 0.30633931843165996,
+      "grad_norm": 0.3480828562287117,
+      "learning_rate": 8.760604366081665e-06,
+      "loss": 0.4876,
+      "step": 1672
+    },
+    {
+      "epoch": 0.30652253572737265,
+      "grad_norm": 0.417116733445613,
+      "learning_rate": 8.758496116748221e-06,
+      "loss": 0.4763,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3067057530230854,
+      "grad_norm": 0.38990995104925313,
+      "learning_rate": 8.756386329978518e-06,
+      "loss": 0.4952,
+      "step": 1674
+    },
+    {
+      "epoch": 0.3068889703187981,
+      "grad_norm": 0.4102790428532168,
+      "learning_rate": 8.754275006635573e-06,
+      "loss": 0.4938,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3070721876145108,
+      "grad_norm": 0.44424837968994635,
+      "learning_rate": 8.752162147583036e-06,
+      "loss": 0.4862,
+      "step": 1676
+    },
+    {
+      "epoch": 0.3072554049102235,
+      "grad_norm": 0.36829717881414353,
+      "learning_rate": 8.750047753685188e-06,
+      "loss": 0.4714,
+      "step": 1677
+    },
+    {
+      "epoch": 0.30743862220593626,
+      "grad_norm": 0.39758875548969896,
+      "learning_rate": 8.747931825806933e-06,
+      "loss": 0.4832,
+      "step": 1678
+    },
+    {
+      "epoch": 0.30762183950164895,
+      "grad_norm": 0.41272232708939577,
+      "learning_rate": 8.745814364813806e-06,
+      "loss": 0.4872,
+      "step": 1679
+    },
+    {
+      "epoch": 0.3078050567973617,
+      "grad_norm": 0.3854030542482913,
+      "learning_rate": 8.743695371571963e-06,
+      "loss": 0.484,
+      "step": 1680
+    },
+    {
+      "epoch": 0.3079882740930744,
+      "grad_norm": 0.36746498837241803,
+      "learning_rate": 8.741574846948198e-06,
+      "loss": 0.4711,
+      "step": 1681
+    },
+    {
+      "epoch": 0.3081714913887871,
+      "grad_norm": 0.411804382886457,
+      "learning_rate": 8.73945279180992e-06,
+      "loss": 0.4759,
+      "step": 1682
+    },
+    {
+      "epoch": 0.3083547086844998,
+      "grad_norm": 0.4026474390517101,
+      "learning_rate": 8.737329207025172e-06,
+      "loss": 0.502,
+      "step": 1683
+    },
+    {
+      "epoch": 0.3085379259802125,
+      "grad_norm": 0.3639512942893561,
+      "learning_rate": 8.735204093462617e-06,
+      "loss": 0.4647,
+      "step": 1684
+    },
+    {
+      "epoch": 0.30872114327592526,
+      "grad_norm": 0.4601114282657689,
+      "learning_rate": 8.733077451991546e-06,
+      "loss": 0.4794,
+      "step": 1685
+    },
+    {
+      "epoch": 0.30890436057163795,
+      "grad_norm": 0.4029114065064777,
+      "learning_rate": 8.730949283481877e-06,
+      "loss": 0.4581,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3090875778673507,
+      "grad_norm": 0.47530481298922933,
+      "learning_rate": 8.72881958880415e-06,
+      "loss": 0.5358,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3092707951630634,
+      "grad_norm": 0.46546993084658045,
+      "learning_rate": 8.72668836882953e-06,
+      "loss": 0.508,
+      "step": 1688
+    },
+    {
+      "epoch": 0.30945401245877613,
+      "grad_norm": 0.44880821238197843,
+      "learning_rate": 8.724555624429805e-06,
+      "loss": 0.494,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3096372297544888,
+      "grad_norm": 0.39081723015666026,
+      "learning_rate": 8.722421356477389e-06,
+      "loss": 0.4845,
+      "step": 1690
+    },
+    {
+      "epoch": 0.30982044705020156,
+      "grad_norm": 0.3529090891828588,
+      "learning_rate": 8.720285565845313e-06,
+      "loss": 0.4672,
+      "step": 1691
+    },
+    {
+      "epoch": 0.31000366434591425,
+      "grad_norm": 0.39898106739463174,
+      "learning_rate": 8.718148253407242e-06,
+      "loss": 0.458,
+      "step": 1692
+    },
+    {
+      "epoch": 0.31018688164162694,
+      "grad_norm": 0.3651751506557362,
+      "learning_rate": 8.716009420037452e-06,
+      "loss": 0.4978,
+      "step": 1693
+    },
+    {
+      "epoch": 0.3103700989373397,
+      "grad_norm": 0.360403523409843,
+      "learning_rate": 8.713869066610847e-06,
+      "loss": 0.489,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3105533162330524,
+      "grad_norm": 0.37968929495733417,
+      "learning_rate": 8.711727194002955e-06,
+      "loss": 0.493,
+      "step": 1695
+    },
+    {
+      "epoch": 0.3107365335287651,
+      "grad_norm": 0.38561595360319445,
+      "learning_rate": 8.709583803089919e-06,
+      "loss": 0.4864,
+      "step": 1696
+    },
+    {
+      "epoch": 0.3109197508244778,
+      "grad_norm": 0.3925407466954778,
+      "learning_rate": 8.707438894748508e-06,
+      "loss": 0.4592,
+      "step": 1697
+    },
+    {
+      "epoch": 0.31110296812019056,
+      "grad_norm": 0.38697234703046507,
+      "learning_rate": 8.70529246985611e-06,
+      "loss": 0.506,
+      "step": 1698
+    },
+    {
+      "epoch": 0.31128618541590325,
+      "grad_norm": 0.40545775127101236,
+      "learning_rate": 8.703144529290733e-06,
+      "loss": 0.498,
+      "step": 1699
+    },
+    {
+      "epoch": 0.311469402711616,
+      "grad_norm": 0.35555366552450657,
+      "learning_rate": 8.700995073931004e-06,
+      "loss": 0.4838,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3116526200073287,
+      "grad_norm": 0.4022620468333227,
+      "learning_rate": 8.698844104656175e-06,
+      "loss": 0.4859,
+      "step": 1701
+    },
+    {
+      "epoch": 0.31183583730304143,
+      "grad_norm": 0.4370267955102546,
+      "learning_rate": 8.696691622346109e-06,
+      "loss": 0.4821,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3120190545987541,
+      "grad_norm": 0.39148309791555397,
+      "learning_rate": 8.694537627881296e-06,
+      "loss": 0.4969,
+      "step": 1703
+    },
+    {
+      "epoch": 0.31220227189446687,
+      "grad_norm": 0.4149560429420508,
+      "learning_rate": 8.69238212214284e-06,
+      "loss": 0.4833,
+      "step": 1704
+    },
+    {
+      "epoch": 0.31238548919017955,
+      "grad_norm": 0.4012554167583925,
+      "learning_rate": 8.690225106012462e-06,
+      "loss": 0.4845,
+      "step": 1705
+    },
+    {
+      "epoch": 0.31256870648589224,
+      "grad_norm": 0.3880778144850858,
+      "learning_rate": 8.688066580372506e-06,
+      "loss": 0.4402,
+      "step": 1706
+    },
+    {
+      "epoch": 0.312751923781605,
+      "grad_norm": 0.3709906292066814,
+      "learning_rate": 8.685906546105925e-06,
+      "loss": 0.461,
+      "step": 1707
+    },
+    {
+      "epoch": 0.3129351410773177,
+      "grad_norm": 0.4038618996987991,
+      "learning_rate": 8.6837450040963e-06,
+      "loss": 0.4952,
+      "step": 1708
+    },
+    {
+      "epoch": 0.3131183583730304,
+      "grad_norm": 0.4498213583992615,
+      "learning_rate": 8.681581955227823e-06,
+      "loss": 0.463,
+      "step": 1709
+    },
+    {
+      "epoch": 0.3133015756687431,
+      "grad_norm": 0.6063342768243897,
+      "learning_rate": 8.6794174003853e-06,
+      "loss": 0.4692,
+      "step": 1710
+    },
+    {
+      "epoch": 0.31348479296445586,
+      "grad_norm": 0.393078000699868,
+      "learning_rate": 8.677251340454155e-06,
+      "loss": 0.477,
+      "step": 1711
+    },
+    {
+      "epoch": 0.31366801026016855,
+      "grad_norm": 0.5208615924956094,
+      "learning_rate": 8.67508377632043e-06,
+      "loss": 0.4686,
+      "step": 1712
+    },
+    {
+      "epoch": 0.3138512275558813,
+      "grad_norm": 0.4156650220868597,
+      "learning_rate": 8.672914708870782e-06,
+      "loss": 0.489,
+      "step": 1713
+    },
+    {
+      "epoch": 0.314034444851594,
+      "grad_norm": 0.39229916132917864,
+      "learning_rate": 8.67074413899248e-06,
+      "loss": 0.4854,
+      "step": 1714
+    },
+    {
+      "epoch": 0.31421766214730673,
+      "grad_norm": 0.3432761699669145,
+      "learning_rate": 8.668572067573409e-06,
+      "loss": 0.5026,
+      "step": 1715
+    },
+    {
+      "epoch": 0.3144008794430194,
+      "grad_norm": 0.4024782274517547,
+      "learning_rate": 8.666398495502068e-06,
+      "loss": 0.5145,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3145840967387321,
+      "grad_norm": 0.4193753630351778,
+      "learning_rate": 8.664223423667571e-06,
+      "loss": 0.5072,
+      "step": 1717
+    },
+    {
+      "epoch": 0.31476731403444486,
+      "grad_norm": 0.44033356668027723,
+      "learning_rate": 8.662046852959644e-06,
+      "loss": 0.4962,
+      "step": 1718
+    },
+    {
+      "epoch": 0.31495053133015755,
+      "grad_norm": 0.4301457201335492,
+      "learning_rate": 8.65986878426863e-06,
+      "loss": 0.5034,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3151337486258703,
+      "grad_norm": 0.42488867042789086,
+      "learning_rate": 8.657689218485476e-06,
+      "loss": 0.5188,
+      "step": 1720
+    },
+    {
+      "epoch": 0.315316965921583,
+      "grad_norm": 0.41341522328402974,
+      "learning_rate": 8.65550815650175e-06,
+      "loss": 0.5216,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3155001832172957,
+      "grad_norm": 0.3771662909597496,
+      "learning_rate": 8.65332559920963e-06,
+      "loss": 0.5106,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3156834005130084,
+      "grad_norm": 0.39212468702748254,
+      "learning_rate": 8.651141547501904e-06,
+      "loss": 0.4766,
+      "step": 1723
+    },
+    {
+      "epoch": 0.31586661780872116,
+      "grad_norm": 0.36654865806495635,
+      "learning_rate": 8.64895600227197e-06,
+      "loss": 0.4802,
+      "step": 1724
+    },
+    {
+      "epoch": 0.31604983510443385,
+      "grad_norm": 0.4065034981518445,
+      "learning_rate": 8.64676896441384e-06,
+      "loss": 0.5288,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3162330524001466,
+      "grad_norm": 0.3854499856808352,
+      "learning_rate": 8.644580434822136e-06,
+      "loss": 0.4699,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3164162696958593,
+      "grad_norm": 0.4057249078954745,
+      "learning_rate": 8.64239041439209e-06,
+      "loss": 0.4759,
+      "step": 1727
+    },
+    {
+      "epoch": 0.316599486991572,
+      "grad_norm": 0.44500981118266,
+      "learning_rate": 8.640198904019544e-06,
+      "loss": 0.4822,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3167827042872847,
+      "grad_norm": 0.3896776477436031,
+      "learning_rate": 8.638005904600948e-06,
+      "loss": 0.49,
+      "step": 1729
+    },
+    {
+      "epoch": 0.3169659215829974,
+      "grad_norm": 0.4432593842454478,
+      "learning_rate": 8.635811417033361e-06,
+      "loss": 0.4952,
+      "step": 1730
+    },
+    {
+      "epoch": 0.31714913887871016,
+      "grad_norm": 0.38053644866499986,
+      "learning_rate": 8.633615442214452e-06,
+      "loss": 0.4579,
+      "step": 1731
+    },
+    {
+      "epoch": 0.31733235617442285,
+      "grad_norm": 0.36140726326704303,
+      "learning_rate": 8.6314179810425e-06,
+      "loss": 0.469,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3175155734701356,
+      "grad_norm": 0.387898862687194,
+      "learning_rate": 8.62921903441639e-06,
+      "loss": 0.4903,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3176987907658483,
+      "grad_norm": 0.4221126995607179,
+      "learning_rate": 8.627018603235613e-06,
+      "loss": 0.4959,
+      "step": 1734
+    },
+    {
+      "epoch": 0.317882008061561,
+      "grad_norm": 0.46009818640085887,
+      "learning_rate": 8.624816688400271e-06,
+      "loss": 0.4563,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3180652253572737,
+      "grad_norm": 0.4685918929067162,
+      "learning_rate": 8.62261329081107e-06,
+      "loss": 0.4714,
+      "step": 1736
+    },
+    {
+      "epoch": 0.31824844265298646,
+      "grad_norm": 0.4412482439531488,
+      "learning_rate": 8.620408411369323e-06,
+      "loss": 0.5007,
+      "step": 1737
+    },
+    {
+      "epoch": 0.31843165994869915,
+      "grad_norm": 0.4547503190879881,
+      "learning_rate": 8.61820205097695e-06,
+      "loss": 0.4581,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3186148772444119,
+      "grad_norm": 0.43395651034382204,
+      "learning_rate": 8.615994210536479e-06,
+      "loss": 0.49,
+      "step": 1739
+    },
+    {
+      "epoch": 0.3187980945401246,
+      "grad_norm": 0.4428071176266843,
+      "learning_rate": 8.613784890951036e-06,
+      "loss": 0.4922,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3189813118358373,
+      "grad_norm": 0.40849265010723296,
+      "learning_rate": 8.61157409312436e-06,
+      "loss": 0.4743,
+      "step": 1741
+    },
+    {
+      "epoch": 0.31916452913155,
+      "grad_norm": 0.42591237999179565,
+      "learning_rate": 8.609361817960794e-06,
+      "loss": 0.505,
+      "step": 1742
+    },
+    {
+      "epoch": 0.3193477464272627,
+      "grad_norm": 0.46495694756695405,
+      "learning_rate": 8.607148066365278e-06,
+      "loss": 0.5111,
+      "step": 1743
+    },
+    {
+      "epoch": 0.31953096372297546,
+      "grad_norm": 0.3979181415316983,
+      "learning_rate": 8.604932839243363e-06,
+      "loss": 0.4911,
+      "step": 1744
+    },
+    {
+      "epoch": 0.31971418101868815,
+      "grad_norm": 0.47254311835047524,
+      "learning_rate": 8.602716137501201e-06,
+      "loss": 0.5059,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3198973983144009,
+      "grad_norm": 0.4110415786461455,
+      "learning_rate": 8.600497962045551e-06,
+      "loss": 0.4931,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3200806156101136,
+      "grad_norm": 0.448385298002212,
+      "learning_rate": 8.598278313783765e-06,
+      "loss": 0.4816,
+      "step": 1747
+    },
+    {
+      "epoch": 0.32026383290582633,
+      "grad_norm": 0.564876624397995,
+      "learning_rate": 8.59605719362381e-06,
+      "loss": 0.4949,
+      "step": 1748
+    },
+    {
+      "epoch": 0.320447050201539,
+      "grad_norm": 0.35471940765133153,
+      "learning_rate": 8.593834602474248e-06,
+      "loss": 0.4794,
+      "step": 1749
+    },
+    {
+      "epoch": 0.32063026749725176,
+      "grad_norm": 0.4192300945251544,
+      "learning_rate": 8.59161054124424e-06,
+      "loss": 0.4519,
+      "step": 1750
+    },
+    {
+      "epoch": 0.32081348479296445,
+      "grad_norm": 0.5038523389603609,
+      "learning_rate": 8.589385010843557e-06,
+      "loss": 0.4729,
+      "step": 1751
+    },
+    {
+      "epoch": 0.32099670208867714,
+      "grad_norm": 0.7356309775332599,
+      "learning_rate": 8.587158012182561e-06,
+      "loss": 0.5082,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3211799193843899,
+      "grad_norm": 0.3818652835002793,
+      "learning_rate": 8.584929546172224e-06,
+      "loss": 0.4756,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3213631366801026,
+      "grad_norm": 0.4435070016683062,
+      "learning_rate": 8.582699613724111e-06,
+      "loss": 0.5102,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3215463539758153,
+      "grad_norm": 0.5377483293815659,
+      "learning_rate": 8.580468215750392e-06,
+      "loss": 0.4844,
+      "step": 1755
+    },
+    {
+      "epoch": 0.321729571271528,
+      "grad_norm": 0.4451665769582748,
+      "learning_rate": 8.578235353163832e-06,
+      "loss": 0.4617,
+      "step": 1756
+    },
+    {
+      "epoch": 0.32191278856724076,
+      "grad_norm": 0.4135616911329195,
+      "learning_rate": 8.576001026877802e-06,
+      "loss": 0.4919,
+      "step": 1757
+    },
+    {
+      "epoch": 0.32209600586295345,
+      "grad_norm": 0.4319033591486567,
+      "learning_rate": 8.573765237806262e-06,
+      "loss": 0.4699,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3222792231586662,
+      "grad_norm": 0.37755085037719216,
+      "learning_rate": 8.571527986863775e-06,
+      "loss": 0.4816,
+      "step": 1759
+    },
+    {
+      "epoch": 0.3224624404543789,
+      "grad_norm": 0.3515458496554082,
+      "learning_rate": 8.569289274965507e-06,
+      "loss": 0.4742,
+      "step": 1760
+    },
+    {
+      "epoch": 0.32264565775009163,
+      "grad_norm": 0.4448770787291807,
+      "learning_rate": 8.567049103027215e-06,
+      "loss": 0.498,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3228288750458043,
+      "grad_norm": 0.3941180324286139,
+      "learning_rate": 8.564807471965253e-06,
+      "loss": 0.5405,
+      "step": 1762
+    },
+    {
+      "epoch": 0.32301209234151707,
+      "grad_norm": 0.3435676251293926,
+      "learning_rate": 8.562564382696578e-06,
+      "loss": 0.4743,
+      "step": 1763
+    },
+    {
+      "epoch": 0.32319530963722976,
+      "grad_norm": 0.37988171856202396,
+      "learning_rate": 8.560319836138737e-06,
+      "loss": 0.4674,
+      "step": 1764
+    },
+    {
+      "epoch": 0.32337852693294244,
+      "grad_norm": 0.3836135165007263,
+      "learning_rate": 8.558073833209878e-06,
+      "loss": 0.4519,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3235617442286552,
+      "grad_norm": 0.39708836154368815,
+      "learning_rate": 8.555826374828737e-06,
+      "loss": 0.4669,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3237449615243679,
+      "grad_norm": 0.3771156156458366,
+      "learning_rate": 8.553577461914658e-06,
+      "loss": 0.4518,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3239281788200806,
+      "grad_norm": 0.3598271555441401,
+      "learning_rate": 8.55132709538757e-06,
+      "loss": 0.4729,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3241113961157933,
+      "grad_norm": 0.3951522532510331,
+      "learning_rate": 8.549075276167999e-06,
+      "loss": 0.4848,
+      "step": 1769
+    },
+    {
+      "epoch": 0.32429461341150606,
+      "grad_norm": 0.4147560955415563,
+      "learning_rate": 8.546822005177065e-06,
+      "loss": 0.5051,
+      "step": 1770
+    },
+    {
+      "epoch": 0.32447783070721875,
+      "grad_norm": 0.37548763357253123,
+      "learning_rate": 8.544567283336484e-06,
+      "loss": 0.4887,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3246610480029315,
+      "grad_norm": 0.3562817695493491,
+      "learning_rate": 8.542311111568564e-06,
+      "loss": 0.4627,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3248442652986442,
+      "grad_norm": 0.3403427009030267,
+      "learning_rate": 8.540053490796204e-06,
+      "loss": 0.498,
+      "step": 1773
+    },
+    {
+      "epoch": 0.32502748259435693,
+      "grad_norm": 0.3618093246911228,
+      "learning_rate": 8.5377944219429e-06,
+      "loss": 0.4711,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3252106998900696,
+      "grad_norm": 0.37156147331448447,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.4914,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3253939171857823,
+      "grad_norm": 0.4104697698807329,
+      "learning_rate": 8.533271943690397e-06,
+      "loss": 0.4926,
+      "step": 1776
+    },
+    {
+      "epoch": 0.32557713448149506,
+      "grad_norm": 0.4221079282420862,
+      "learning_rate": 8.531008536141143e-06,
+      "loss": 0.494,
+      "step": 1777
+    },
+    {
+      "epoch": 0.32576035177720775,
+      "grad_norm": 0.44238661768749327,
+      "learning_rate": 8.528743684210842e-06,
+      "loss": 0.4679,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3259435690729205,
+      "grad_norm": 0.37351954296984535,
+      "learning_rate": 8.52647738882594e-06,
+      "loss": 0.4743,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3261267863686332,
+      "grad_norm": 0.39302981254445146,
+      "learning_rate": 8.524209650913487e-06,
+      "loss": 0.4817,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3263100036643459,
+      "grad_norm": 0.3421628958137739,
+      "learning_rate": 8.521940471401106e-06,
+      "loss": 0.4663,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3264932209600586,
+      "grad_norm": 0.5533622907062358,
+      "learning_rate": 8.519669851217028e-06,
+      "loss": 0.4918,
+      "step": 1782
+    },
+    {
+      "epoch": 0.32667643825577136,
+      "grad_norm": 0.3970061640019179,
+      "learning_rate": 8.517397791290059e-06,
+      "loss": 0.4914,
+      "step": 1783
+    },
+    {
+      "epoch": 0.32685965555148405,
+      "grad_norm": 0.435520424701832,
+      "learning_rate": 8.5151242925496e-06,
+      "loss": 0.5093,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3270428728471968,
+      "grad_norm": 0.3465454680201401,
+      "learning_rate": 8.512849355925641e-06,
+      "loss": 0.4718,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3272260901429095,
+      "grad_norm": 0.4166371179539154,
+      "learning_rate": 8.510572982348759e-06,
+      "loss": 0.508,
+      "step": 1786
+    },
+    {
+      "epoch": 0.32740930743862223,
+      "grad_norm": 0.4192663642287507,
+      "learning_rate": 8.508295172750116e-06,
+      "loss": 0.4802,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3275925247343349,
+      "grad_norm": 0.3935083715874497,
+      "learning_rate": 8.506015928061468e-06,
+      "loss": 0.4748,
+      "step": 1788
+    },
+    {
+      "epoch": 0.3277757420300476,
+      "grad_norm": 0.4149102123371367,
+      "learning_rate": 8.50373524921515e-06,
+      "loss": 0.4927,
+      "step": 1789
+    },
+    {
+      "epoch": 0.32795895932576036,
+      "grad_norm": 0.3867478956169408,
+      "learning_rate": 8.501453137144093e-06,
+      "loss": 0.5062,
+      "step": 1790
+    },
+    {
+      "epoch": 0.32814217662147305,
+      "grad_norm": 0.4017651304087876,
+      "learning_rate": 8.499169592781807e-06,
+      "loss": 0.5199,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3283253939171858,
+      "grad_norm": 0.365617732104251,
+      "learning_rate": 8.496884617062389e-06,
+      "loss": 0.4777,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3285086112128985,
+      "grad_norm": 0.5935783428527378,
+      "learning_rate": 8.494598210920522e-06,
+      "loss": 0.5327,
+      "step": 1793
+    },
+    {
+      "epoch": 0.32869182850861123,
+      "grad_norm": 0.4461319964673768,
+      "learning_rate": 8.492310375291478e-06,
+      "loss": 0.4783,
+      "step": 1794
+    },
+    {
+      "epoch": 0.3288750458043239,
+      "grad_norm": 0.4548047555505239,
+      "learning_rate": 8.490021111111108e-06,
+      "loss": 0.4793,
+      "step": 1795
+    },
+    {
+      "epoch": 0.32905826310003666,
+      "grad_norm": 0.3837539231745361,
+      "learning_rate": 8.48773041931585e-06,
+      "loss": 0.4656,
+      "step": 1796
+    },
+    {
+      "epoch": 0.32924148039574935,
+      "grad_norm": 0.38081008240354797,
+      "learning_rate": 8.485438300842725e-06,
+      "loss": 0.4538,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3294246976914621,
+      "grad_norm": 0.381006774669353,
+      "learning_rate": 8.483144756629342e-06,
+      "loss": 0.4678,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3296079149871748,
+      "grad_norm": 4.393546461902215,
+      "learning_rate": 8.480849787613883e-06,
+      "loss": 0.48,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3297911322828875,
+      "grad_norm": 0.4216378380286295,
+      "learning_rate": 8.478553394735126e-06,
+      "loss": 0.4664,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3299743495786002,
+      "grad_norm": 0.40689364633375075,
+      "learning_rate": 8.47625557893242e-06,
+      "loss": 0.4687,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3301575668743129,
+      "grad_norm": 0.4232729156542768,
+      "learning_rate": 8.473956341145706e-06,
+      "loss": 0.4684,
+      "step": 1802
+    },
+    {
+      "epoch": 0.33034078417002566,
+      "grad_norm": 0.4289741259876096,
+      "learning_rate": 8.471655682315496e-06,
+      "loss": 0.4716,
+      "step": 1803
+    },
+    {
+      "epoch": 0.33052400146573835,
+      "grad_norm": 0.5275080199063381,
+      "learning_rate": 8.469353603382892e-06,
+      "loss": 0.4757,
+      "step": 1804
+    },
+    {
+      "epoch": 0.3307072187614511,
+      "grad_norm": 0.43974415262075855,
+      "learning_rate": 8.467050105289572e-06,
+      "loss": 0.4865,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3308904360571638,
+      "grad_norm": 0.4814298309864084,
+      "learning_rate": 8.4647451889778e-06,
+      "loss": 0.4797,
+      "step": 1806
+    },
+    {
+      "epoch": 0.33107365335287653,
+      "grad_norm": 0.42391918940288414,
+      "learning_rate": 8.462438855390409e-06,
+      "loss": 0.4823,
+      "step": 1807
+    },
+    {
+      "epoch": 0.3312568706485892,
+      "grad_norm": 0.3665799179835301,
+      "learning_rate": 8.460131105470829e-06,
+      "loss": 0.4649,
+      "step": 1808
+    },
+    {
+      "epoch": 0.33144008794430196,
+      "grad_norm": 0.4035555626403075,
+      "learning_rate": 8.45782194016305e-06,
+      "loss": 0.4758,
+      "step": 1809
+    },
+    {
+      "epoch": 0.33162330524001465,
+      "grad_norm": 0.4425161709111113,
+      "learning_rate": 8.455511360411657e-06,
+      "loss": 0.4961,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3318065225357274,
+      "grad_norm": 0.4297733230595355,
+      "learning_rate": 8.453199367161804e-06,
+      "loss": 0.4801,
+      "step": 1811
+    },
+    {
+      "epoch": 0.3319897398314401,
+      "grad_norm": 0.407563982418168,
+      "learning_rate": 8.450885961359227e-06,
+      "loss": 0.5109,
+      "step": 1812
+    },
+    {
+      "epoch": 0.3321729571271528,
+      "grad_norm": 0.40167352725847427,
+      "learning_rate": 8.448571143950239e-06,
+      "loss": 0.5056,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3323561744228655,
+      "grad_norm": 0.41581106857716255,
+      "learning_rate": 8.446254915881733e-06,
+      "loss": 0.4432,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3325393917185782,
+      "grad_norm": 0.42034315457885957,
+      "learning_rate": 8.443937278101171e-06,
+      "loss": 0.5129,
+      "step": 1815
+    },
+    {
+      "epoch": 0.33272260901429096,
+      "grad_norm": 0.49697266318929034,
+      "learning_rate": 8.441618231556604e-06,
+      "loss": 0.4914,
+      "step": 1816
+    },
+    {
+      "epoch": 0.33290582631000365,
+      "grad_norm": 0.45473540523372535,
+      "learning_rate": 8.439297777196645e-06,
+      "loss": 0.5046,
+      "step": 1817
+    },
+    {
+      "epoch": 0.3330890436057164,
+      "grad_norm": 0.38498739235497825,
+      "learning_rate": 8.436975915970496e-06,
+      "loss": 0.5125,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3332722609014291,
+      "grad_norm": 0.3856310516540707,
+      "learning_rate": 8.434652648827925e-06,
+      "loss": 0.4877,
+      "step": 1819
+    },
+    {
+      "epoch": 0.33345547819714183,
+      "grad_norm": 0.3545408916699545,
+      "learning_rate": 8.432327976719281e-06,
+      "loss": 0.4819,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3336386954928545,
+      "grad_norm": 0.3816671046542913,
+      "learning_rate": 8.430001900595487e-06,
+      "loss": 0.5074,
+      "step": 1821
+    },
+    {
+      "epoch": 0.33382191278856727,
+      "grad_norm": 0.4236478252674396,
+      "learning_rate": 8.427674421408037e-06,
+      "loss": 0.4746,
+      "step": 1822
+    },
+    {
+      "epoch": 0.33400513008427996,
+      "grad_norm": 0.4093466909489272,
+      "learning_rate": 8.425345540109e-06,
+      "loss": 0.4727,
+      "step": 1823
+    },
+    {
+      "epoch": 0.33418834737999265,
+      "grad_norm": 0.37909373859606427,
+      "learning_rate": 8.423015257651022e-06,
+      "loss": 0.482,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3343715646757054,
+      "grad_norm": 0.3801638934135252,
+      "learning_rate": 8.420683574987319e-06,
+      "loss": 0.4763,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3345547819714181,
+      "grad_norm": 1.0336015888999988,
+      "learning_rate": 8.418350493071677e-06,
+      "loss": 0.4797,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3347379992671308,
+      "grad_norm": 0.35984342130185926,
+      "learning_rate": 8.41601601285846e-06,
+      "loss": 0.4868,
+      "step": 1827
+    },
+    {
+      "epoch": 0.3349212165628435,
+      "grad_norm": 0.3550613764764406,
+      "learning_rate": 8.413680135302604e-06,
+      "loss": 0.4706,
+      "step": 1828
+    },
+    {
+      "epoch": 0.33510443385855626,
+      "grad_norm": 0.3623869235380865,
+      "learning_rate": 8.411342861359612e-06,
+      "loss": 0.4602,
+      "step": 1829
+    },
+    {
+      "epoch": 0.33528765115426895,
+      "grad_norm": 0.4364870755091544,
+      "learning_rate": 8.40900419198556e-06,
+      "loss": 0.4832,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3354708684499817,
+      "grad_norm": 0.4636618121853265,
+      "learning_rate": 8.406664128137095e-06,
+      "loss": 0.453,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3356540857456944,
+      "grad_norm": 0.36221754354794844,
+      "learning_rate": 8.404322670771436e-06,
+      "loss": 0.4849,
+      "step": 1832
+    },
+    {
+      "epoch": 0.33583730304140713,
+      "grad_norm": 0.39602504845526065,
+      "learning_rate": 8.40197982084637e-06,
+      "loss": 0.4755,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3360205203371198,
+      "grad_norm": 0.4091319003361619,
+      "learning_rate": 8.399635579320259e-06,
+      "loss": 0.5015,
+      "step": 1834
+    },
+    {
+      "epoch": 0.33620373763283257,
+      "grad_norm": 0.3415400300998229,
+      "learning_rate": 8.397289947152021e-06,
+      "loss": 0.4645,
+      "step": 1835
+    },
+    {
+      "epoch": 0.33638695492854526,
+      "grad_norm": 0.4295090419145507,
+      "learning_rate": 8.394942925301155e-06,
+      "loss": 0.4385,
+      "step": 1836
+    },
+    {
+      "epoch": 0.33657017222425795,
+      "grad_norm": 0.3919892057064269,
+      "learning_rate": 8.392594514727728e-06,
+      "loss": 0.5009,
+      "step": 1837
+    },
+    {
+      "epoch": 0.3367533895199707,
+      "grad_norm": 0.4276728302114804,
+      "learning_rate": 8.390244716392369e-06,
+      "loss": 0.5113,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3369366068156834,
+      "grad_norm": 0.4027722155599858,
+      "learning_rate": 8.387893531256278e-06,
+      "loss": 0.4825,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3371198241113961,
+      "grad_norm": 0.34856854492446304,
+      "learning_rate": 8.385540960281223e-06,
+      "loss": 0.4614,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3373030414071088,
+      "grad_norm": 0.33714083093594926,
+      "learning_rate": 8.383187004429536e-06,
+      "loss": 0.4723,
+      "step": 1841
+    },
+    {
+      "epoch": 0.33748625870282156,
+      "grad_norm": 0.5015806620678309,
+      "learning_rate": 8.38083166466412e-06,
+      "loss": 0.4804,
+      "step": 1842
+    },
+    {
+      "epoch": 0.33766947599853425,
+      "grad_norm": 0.40706034865044627,
+      "learning_rate": 8.378474941948437e-06,
+      "loss": 0.4814,
+      "step": 1843
+    },
+    {
+      "epoch": 0.337852693294247,
+      "grad_norm": 0.42150894844205045,
+      "learning_rate": 8.376116837246525e-06,
+      "loss": 0.4959,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3380359105899597,
+      "grad_norm": 0.37939627732061587,
+      "learning_rate": 8.373757351522976e-06,
+      "loss": 0.4737,
+      "step": 1845
+    },
+    {
+      "epoch": 0.33821912788567243,
+      "grad_norm": 0.37334004265146,
+      "learning_rate": 8.371396485742956e-06,
+      "loss": 0.4583,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3384023451813851,
+      "grad_norm": 0.5502578818261286,
+      "learning_rate": 8.36903424087219e-06,
+      "loss": 0.5031,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3385855624770978,
+      "grad_norm": 0.4305389526404514,
+      "learning_rate": 8.366670617876969e-06,
+      "loss": 0.5081,
+      "step": 1848
+    },
+    {
+      "epoch": 0.33876877977281056,
+      "grad_norm": 0.373003871725479,
+      "learning_rate": 8.36430561772415e-06,
+      "loss": 0.4674,
+      "step": 1849
+    },
+    {
+      "epoch": 0.33895199706852325,
+      "grad_norm": 0.36910312348131236,
+      "learning_rate": 8.361939241381148e-06,
+      "loss": 0.4824,
+      "step": 1850
+    },
+    {
+      "epoch": 0.339135214364236,
+      "grad_norm": 0.38233912239924234,
+      "learning_rate": 8.359571489815946e-06,
+      "loss": 0.5155,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3393184316599487,
+      "grad_norm": 0.42080088842006386,
+      "learning_rate": 8.357202363997085e-06,
+      "loss": 0.4932,
+      "step": 1852
+    },
+    {
+      "epoch": 0.33950164895566143,
+      "grad_norm": 0.40024242741960253,
+      "learning_rate": 8.354831864893675e-06,
+      "loss": 0.4889,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3396848662513741,
+      "grad_norm": 0.3818263254638861,
+      "learning_rate": 8.352459993475379e-06,
+      "loss": 0.4489,
+      "step": 1854
+    },
+    {
+      "epoch": 0.33986808354708686,
+      "grad_norm": 0.45691270540100465,
+      "learning_rate": 8.35008675071243e-06,
+      "loss": 0.445,
+      "step": 1855
+    },
+    {
+      "epoch": 0.34005130084279955,
+      "grad_norm": 0.40956822564503537,
+      "learning_rate": 8.347712137575614e-06,
+      "loss": 0.4875,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3402345181385123,
+      "grad_norm": 0.4223915335148929,
+      "learning_rate": 8.34533615503628e-06,
+      "loss": 0.5017,
+      "step": 1857
+    },
+    {
+      "epoch": 0.340417735434225,
+      "grad_norm": 0.4339594623834613,
+      "learning_rate": 8.342958804066345e-06,
+      "loss": 0.5349,
+      "step": 1858
+    },
+    {
+      "epoch": 0.34060095272993773,
+      "grad_norm": 0.38617828743973426,
+      "learning_rate": 8.340580085638275e-06,
+      "loss": 0.4858,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3407841700256504,
+      "grad_norm": 0.39545092131685405,
+      "learning_rate": 8.3382000007251e-06,
+      "loss": 0.4844,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3409673873213631,
+      "grad_norm": 0.4257990961795386,
+      "learning_rate": 8.33581855030041e-06,
+      "loss": 0.5055,
+      "step": 1861
+    },
+    {
+      "epoch": 0.34115060461707586,
+      "grad_norm": 0.3528695250715933,
+      "learning_rate": 8.333435735338346e-06,
+      "loss": 0.4875,
+      "step": 1862
+    },
+    {
+      "epoch": 0.34133382191278855,
+      "grad_norm": 0.37091916482456505,
+      "learning_rate": 8.331051556813623e-06,
+      "loss": 0.4984,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3415170392085013,
+      "grad_norm": 0.4264847608510263,
+      "learning_rate": 8.328666015701496e-06,
+      "loss": 0.4867,
+      "step": 1864
+    },
+    {
+      "epoch": 0.341700256504214,
+      "grad_norm": 0.4099601844664771,
+      "learning_rate": 8.326279112977791e-06,
+      "loss": 0.4994,
+      "step": 1865
+    },
+    {
+      "epoch": 0.34188347379992673,
+      "grad_norm": 0.3677590723427411,
+      "learning_rate": 8.323890849618882e-06,
+      "loss": 0.5027,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3420666910956394,
+      "grad_norm": 0.4409098617591673,
+      "learning_rate": 8.321501226601702e-06,
+      "loss": 0.496,
+      "step": 1867
+    },
+    {
+      "epoch": 0.34224990839135216,
+      "grad_norm": 0.3926664857139486,
+      "learning_rate": 8.319110244903748e-06,
+      "loss": 0.5018,
+      "step": 1868
+    },
+    {
+      "epoch": 0.34243312568706485,
+      "grad_norm": 0.3981457421214149,
+      "learning_rate": 8.316717905503058e-06,
+      "loss": 0.4803,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3426163429827776,
+      "grad_norm": 0.4111803238988,
+      "learning_rate": 8.314324209378237e-06,
+      "loss": 0.4944,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3427995602784903,
+      "grad_norm": 0.3837210327181565,
+      "learning_rate": 8.31192915750844e-06,
+      "loss": 0.4637,
+      "step": 1871
+    },
+    {
+      "epoch": 0.342982777574203,
+      "grad_norm": 0.36141808942109716,
+      "learning_rate": 8.309532750873381e-06,
+      "loss": 0.5022,
+      "step": 1872
+    },
+    {
+      "epoch": 0.3431659948699157,
+      "grad_norm": 0.4135486922923667,
+      "learning_rate": 8.30713499045332e-06,
+      "loss": 0.5429,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3433492121656284,
+      "grad_norm": 0.4100838578777171,
+      "learning_rate": 8.304735877229082e-06,
+      "loss": 0.4814,
+      "step": 1874
+    },
+    {
+      "epoch": 0.34353242946134116,
+      "grad_norm": 0.4419756386117236,
+      "learning_rate": 8.302335412182034e-06,
+      "loss": 0.506,
+      "step": 1875
+    },
+    {
+      "epoch": 0.34371564675705385,
+      "grad_norm": 0.4023978864041945,
+      "learning_rate": 8.299933596294104e-06,
+      "loss": 0.4835,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3438988640527666,
+      "grad_norm": 0.37938282086115893,
+      "learning_rate": 8.297530430547767e-06,
+      "loss": 0.4802,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3440820813484793,
+      "grad_norm": 0.44808143362476516,
+      "learning_rate": 8.295125915926057e-06,
+      "loss": 0.4948,
+      "step": 1878
+    },
+    {
+      "epoch": 0.34426529864419203,
+      "grad_norm": 0.38559125969803254,
+      "learning_rate": 8.292720053412553e-06,
+      "loss": 0.4685,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3444485159399047,
+      "grad_norm": 0.33369724182793525,
+      "learning_rate": 8.290312843991388e-06,
+      "loss": 0.4395,
+      "step": 1880
+    },
+    {
+      "epoch": 0.34463173323561747,
+      "grad_norm": 0.36560725000329003,
+      "learning_rate": 8.287904288647246e-06,
+      "loss": 0.4769,
+      "step": 1881
+    },
+    {
+      "epoch": 0.34481495053133016,
+      "grad_norm": 0.40109807099240397,
+      "learning_rate": 8.285494388365364e-06,
+      "loss": 0.4896,
+      "step": 1882
+    },
+    {
+      "epoch": 0.34499816782704285,
+      "grad_norm": 0.3758614423812384,
+      "learning_rate": 8.283083144131523e-06,
+      "loss": 0.4591,
+      "step": 1883
+    },
+    {
+      "epoch": 0.3451813851227556,
+      "grad_norm": 0.4045825714686613,
+      "learning_rate": 8.28067055693206e-06,
+      "loss": 0.4983,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3453646024184683,
+      "grad_norm": 0.5653539162553293,
+      "learning_rate": 8.278256627753857e-06,
+      "loss": 0.4912,
+      "step": 1885
+    },
+    {
+      "epoch": 0.345547819714181,
+      "grad_norm": 0.4120278659351058,
+      "learning_rate": 8.27584135758435e-06,
+      "loss": 0.5053,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3457310370098937,
+      "grad_norm": 0.4087683124654251,
+      "learning_rate": 8.273424747411519e-06,
+      "loss": 0.4768,
+      "step": 1887
+    },
+    {
+      "epoch": 0.34591425430560646,
+      "grad_norm": 0.3868920894610502,
+      "learning_rate": 8.27100679822389e-06,
+      "loss": 0.4872,
+      "step": 1888
+    },
+    {
+      "epoch": 0.34609747160131915,
+      "grad_norm": 0.3672508442805955,
+      "learning_rate": 8.268587511010546e-06,
+      "loss": 0.4492,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3462806888970319,
+      "grad_norm": 0.3226405339400022,
+      "learning_rate": 8.266166886761106e-06,
+      "loss": 0.447,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3464639061927446,
+      "grad_norm": 0.3896492518387063,
+      "learning_rate": 8.263744926465744e-06,
+      "loss": 0.4952,
+      "step": 1891
+    },
+    {
+      "epoch": 0.34664712348845733,
+      "grad_norm": 0.40806210827621486,
+      "learning_rate": 8.26132163111518e-06,
+      "loss": 0.4942,
+      "step": 1892
+    },
+    {
+      "epoch": 0.34683034078417,
+      "grad_norm": 0.5064133110397161,
+      "learning_rate": 8.258897001700673e-06,
+      "loss": 0.4914,
+      "step": 1893
+    },
+    {
+      "epoch": 0.34701355807988277,
+      "grad_norm": 0.37350719921589426,
+      "learning_rate": 8.256471039214036e-06,
+      "loss": 0.4667,
+      "step": 1894
+    },
+    {
+      "epoch": 0.34719677537559546,
+      "grad_norm": 0.4472694649954096,
+      "learning_rate": 8.254043744647625e-06,
+      "loss": 0.4768,
+      "step": 1895
+    },
+    {
+      "epoch": 0.34737999267130815,
+      "grad_norm": 0.5451471608365284,
+      "learning_rate": 8.251615118994338e-06,
+      "loss": 0.4833,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3475632099670209,
+      "grad_norm": 0.43194887420538414,
+      "learning_rate": 8.249185163247621e-06,
+      "loss": 0.5041,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3477464272627336,
+      "grad_norm": 0.4440434179645274,
+      "learning_rate": 8.24675387840146e-06,
+      "loss": 0.4928,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3479296445584463,
+      "grad_norm": 0.4470574728369111,
+      "learning_rate": 8.24432126545039e-06,
+      "loss": 0.4721,
+      "step": 1899
+    },
+    {
+      "epoch": 0.348112861854159,
+      "grad_norm": 0.37811511446638624,
+      "learning_rate": 8.241887325389486e-06,
+      "loss": 0.4916,
+      "step": 1900
+    },
+    {
+      "epoch": 0.34829607914987176,
+      "grad_norm": 0.401475995772846,
+      "learning_rate": 8.239452059214367e-06,
+      "loss": 0.4922,
+      "step": 1901
+    },
+    {
+      "epoch": 0.34847929644558445,
+      "grad_norm": 0.36588179778579716,
+      "learning_rate": 8.23701546792119e-06,
+      "loss": 0.4769,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3486625137412972,
+      "grad_norm": 0.37419447204674516,
+      "learning_rate": 8.234577552506662e-06,
+      "loss": 0.4683,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3488457310370099,
+      "grad_norm": 0.44910824354183104,
+      "learning_rate": 8.232138313968025e-06,
+      "loss": 0.5002,
+      "step": 1904
+    },
+    {
+      "epoch": 0.34902894833272263,
+      "grad_norm": 0.37853030952448913,
+      "learning_rate": 8.229697753303067e-06,
+      "loss": 0.499,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3492121656284353,
+      "grad_norm": 0.3787302843131875,
+      "learning_rate": 8.227255871510111e-06,
+      "loss": 0.4652,
+      "step": 1906
+    },
+    {
+      "epoch": 0.349395382924148,
+      "grad_norm": 0.42294805152134374,
+      "learning_rate": 8.224812669588028e-06,
+      "loss": 0.4761,
+      "step": 1907
+    },
+    {
+      "epoch": 0.34957860021986076,
+      "grad_norm": 0.4251808716753822,
+      "learning_rate": 8.222368148536223e-06,
+      "loss": 0.4781,
+      "step": 1908
+    },
+    {
+      "epoch": 0.34976181751557345,
+      "grad_norm": 0.41741877899455326,
+      "learning_rate": 8.219922309354643e-06,
+      "loss": 0.4979,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3499450348112862,
+      "grad_norm": 0.4528858990840802,
+      "learning_rate": 8.217475153043772e-06,
+      "loss": 0.4841,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3501282521069989,
+      "grad_norm": 0.37317689057150794,
+      "learning_rate": 8.215026680604637e-06,
+      "loss": 0.4713,
+      "step": 1911
+    },
+    {
+      "epoch": 0.35031146940271163,
+      "grad_norm": 0.3667470219941718,
+      "learning_rate": 8.212576893038799e-06,
+      "loss": 0.4689,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3504946866984243,
+      "grad_norm": 0.3979696962375165,
+      "learning_rate": 8.21012579134836e-06,
+      "loss": 0.4538,
+      "step": 1913
+    },
+    {
+      "epoch": 0.35067790399413706,
+      "grad_norm": 0.43331562093452664,
+      "learning_rate": 8.20767337653596e-06,
+      "loss": 0.4878,
+      "step": 1914
+    },
+    {
+      "epoch": 0.35086112128984975,
+      "grad_norm": 0.36256762359906813,
+      "learning_rate": 8.20521964960477e-06,
+      "loss": 0.4462,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3510443385855625,
+      "grad_norm": 0.43213117928406414,
+      "learning_rate": 8.202764611558507e-06,
+      "loss": 0.5066,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3512275558812752,
+      "grad_norm": 0.4293633236485711,
+      "learning_rate": 8.200308263401417e-06,
+      "loss": 0.5351,
+      "step": 1917
+    },
+    {
+      "epoch": 0.35141077317698793,
+      "grad_norm": 0.5287407111719371,
+      "learning_rate": 8.197850606138286e-06,
+      "loss": 0.4705,
+      "step": 1918
+    },
+    {
+      "epoch": 0.3515939904727006,
+      "grad_norm": 0.48855534444525384,
+      "learning_rate": 8.195391640774433e-06,
+      "loss": 0.4852,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3517772077684133,
+      "grad_norm": 0.4148812235933253,
+      "learning_rate": 8.192931368315715e-06,
+      "loss": 0.4785,
+      "step": 1920
+    },
+    {
+      "epoch": 0.35196042506412606,
+      "grad_norm": 0.44919257408205365,
+      "learning_rate": 8.190469789768517e-06,
+      "loss": 0.4851,
+      "step": 1921
+    },
+    {
+      "epoch": 0.35214364235983875,
+      "grad_norm": 0.4359193627784823,
+      "learning_rate": 8.188006906139767e-06,
+      "loss": 0.5077,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3523268596555515,
+      "grad_norm": 0.43124736358934584,
+      "learning_rate": 8.185542718436923e-06,
+      "loss": 0.4705,
+      "step": 1923
+    },
+    {
+      "epoch": 0.3525100769512642,
+      "grad_norm": 0.4270332132061218,
+      "learning_rate": 8.183077227667975e-06,
+      "loss": 0.5061,
+      "step": 1924
+    },
+    {
+      "epoch": 0.35269329424697693,
+      "grad_norm": 0.3709113632371592,
+      "learning_rate": 8.180610434841448e-06,
+      "loss": 0.4926,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3528765115426896,
+      "grad_norm": 0.3686143335119551,
+      "learning_rate": 8.1781423409664e-06,
+      "loss": 0.4736,
+      "step": 1926
+    },
+    {
+      "epoch": 0.35305972883840236,
+      "grad_norm": 0.4768757941328196,
+      "learning_rate": 8.175672947052416e-06,
+      "loss": 0.4899,
+      "step": 1927
+    },
+    {
+      "epoch": 0.35324294613411505,
+      "grad_norm": 0.37435018310100715,
+      "learning_rate": 8.173202254109622e-06,
+      "loss": 0.5019,
+      "step": 1928
+    },
+    {
+      "epoch": 0.3534261634298278,
+      "grad_norm": 0.5886108613911327,
+      "learning_rate": 8.170730263148668e-06,
+      "loss": 0.5208,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3536093807255405,
+      "grad_norm": 0.42394082862501326,
+      "learning_rate": 8.168256975180737e-06,
+      "loss": 0.5124,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3537925980212532,
+      "grad_norm": 0.39473180614386977,
+      "learning_rate": 8.165782391217543e-06,
+      "loss": 0.4863,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3539758153169659,
+      "grad_norm": 0.4022544573376633,
+      "learning_rate": 8.163306512271334e-06,
+      "loss": 0.4989,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3541590326126786,
+      "grad_norm": 0.4154530765976519,
+      "learning_rate": 8.160829339354876e-06,
+      "loss": 0.4661,
+      "step": 1933
+    },
+    {
+      "epoch": 0.35434224990839136,
+      "grad_norm": 0.3921468409429114,
+      "learning_rate": 8.158350873481478e-06,
+      "loss": 0.4738,
+      "step": 1934
+    },
+    {
+      "epoch": 0.35452546720410405,
+      "grad_norm": 0.3954474064023121,
+      "learning_rate": 8.155871115664968e-06,
+      "loss": 0.51,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3547086844998168,
+      "grad_norm": 0.3882950367284415,
+      "learning_rate": 8.15339006691971e-06,
+      "loss": 0.4642,
+      "step": 1936
+    },
+    {
+      "epoch": 0.3548919017955295,
+      "grad_norm": 0.37897867702653376,
+      "learning_rate": 8.150907728260592e-06,
+      "loss": 0.5044,
+      "step": 1937
+    },
+    {
+      "epoch": 0.35507511909124223,
+      "grad_norm": 0.3859422166696926,
+      "learning_rate": 8.14842410070303e-06,
+      "loss": 0.4551,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3552583363869549,
+      "grad_norm": 0.38875341973643246,
+      "learning_rate": 8.145939185262963e-06,
+      "loss": 0.4944,
+      "step": 1939
+    },
+    {
+      "epoch": 0.35544155368266767,
+      "grad_norm": 0.3983931627212232,
+      "learning_rate": 8.143452982956866e-06,
+      "loss": 0.5008,
+      "step": 1940
+    },
+    {
+      "epoch": 0.35562477097838036,
+      "grad_norm": 0.39775757772326403,
+      "learning_rate": 8.140965494801733e-06,
+      "loss": 0.5263,
+      "step": 1941
+    },
+    {
+      "epoch": 0.3558079882740931,
+      "grad_norm": 0.3944293856829923,
+      "learning_rate": 8.13847672181509e-06,
+      "loss": 0.4771,
+      "step": 1942
+    },
+    {
+      "epoch": 0.3559912055698058,
+      "grad_norm": 0.3817860113481927,
+      "learning_rate": 8.13598666501498e-06,
+      "loss": 0.4666,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3561744228655185,
+      "grad_norm": 0.42632726063844545,
+      "learning_rate": 8.133495325419983e-06,
+      "loss": 0.4663,
+      "step": 1944
+    },
+    {
+      "epoch": 0.3563576401612312,
+      "grad_norm": 0.3572805692583185,
+      "learning_rate": 8.131002704049189e-06,
+      "loss": 0.4708,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3565408574569439,
+      "grad_norm": 0.404852300953083,
+      "learning_rate": 8.128508801922226e-06,
+      "loss": 0.5325,
+      "step": 1946
+    },
+    {
+      "epoch": 0.35672407475265666,
+      "grad_norm": 0.35876468020600716,
+      "learning_rate": 8.126013620059236e-06,
+      "loss": 0.4949,
+      "step": 1947
+    },
+    {
+      "epoch": 0.35690729204836935,
+      "grad_norm": 0.38545171520037624,
+      "learning_rate": 8.123517159480894e-06,
+      "loss": 0.4764,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3570905093440821,
+      "grad_norm": 0.40541828483545733,
+      "learning_rate": 8.12101942120839e-06,
+      "loss": 0.4807,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3572737266397948,
+      "grad_norm": 0.4243415026003622,
+      "learning_rate": 8.118520406263437e-06,
+      "loss": 0.5003,
+      "step": 1950
+    },
+    {
+      "epoch": 0.35745694393550753,
+      "grad_norm": 0.4384025995443247,
+      "learning_rate": 8.116020115668278e-06,
+      "loss": 0.4916,
+      "step": 1951
+    },
+    {
+      "epoch": 0.3576401612312202,
+      "grad_norm": 0.356464489561591,
+      "learning_rate": 8.113518550445667e-06,
+      "loss": 0.5068,
+      "step": 1952
+    },
+    {
+      "epoch": 0.35782337852693297,
+      "grad_norm": 0.36237464089967664,
+      "learning_rate": 8.111015711618888e-06,
+      "loss": 0.4691,
+      "step": 1953
+    },
+    {
+      "epoch": 0.35800659582264566,
+      "grad_norm": 0.41004625226837194,
+      "learning_rate": 8.108511600211741e-06,
+      "loss": 0.495,
+      "step": 1954
+    },
+    {
+      "epoch": 0.35818981311835835,
+      "grad_norm": 0.37519930557205516,
+      "learning_rate": 8.106006217248552e-06,
+      "loss": 0.527,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3583730304140711,
+      "grad_norm": 0.3930139446184139,
+      "learning_rate": 8.103499563754159e-06,
+      "loss": 0.49,
+      "step": 1956
+    },
+    {
+      "epoch": 0.3585562477097838,
+      "grad_norm": 0.4049543770945572,
+      "learning_rate": 8.100991640753926e-06,
+      "loss": 0.4918,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3587394650054965,
+      "grad_norm": 0.41598194124283755,
+      "learning_rate": 8.098482449273737e-06,
+      "loss": 0.4583,
+      "step": 1958
+    },
+    {
+      "epoch": 0.3589226823012092,
+      "grad_norm": 0.3497094599476565,
+      "learning_rate": 8.095971990339987e-06,
+      "loss": 0.441,
+      "step": 1959
+    },
+    {
+      "epoch": 0.35910589959692196,
+      "grad_norm": 0.857453529704147,
+      "learning_rate": 8.0934602649796e-06,
+      "loss": 0.4862,
+      "step": 1960
+    },
+    {
+      "epoch": 0.35928911689263465,
+      "grad_norm": 0.3948051526194389,
+      "learning_rate": 8.090947274220011e-06,
+      "loss": 0.4696,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3594723341883474,
+      "grad_norm": 0.4025724340675242,
+      "learning_rate": 8.088433019089174e-06,
+      "loss": 0.4784,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3596555514840601,
+      "grad_norm": 0.4250158054819026,
+      "learning_rate": 8.08591750061556e-06,
+      "loss": 0.4887,
+      "step": 1963
+    },
+    {
+      "epoch": 0.35983876877977283,
+      "grad_norm": 0.3987052277731631,
+      "learning_rate": 8.083400719828161e-06,
+      "loss": 0.486,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3600219860754855,
+      "grad_norm": 0.3818650890686896,
+      "learning_rate": 8.080882677756479e-06,
+      "loss": 0.4961,
+      "step": 1965
+    },
+    {
+      "epoch": 0.36020520337119827,
+      "grad_norm": 0.37338678994677815,
+      "learning_rate": 8.078363375430534e-06,
+      "loss": 0.4849,
+      "step": 1966
+    },
+    {
+      "epoch": 0.36038842066691096,
+      "grad_norm": 0.4033555520259992,
+      "learning_rate": 8.075842813880865e-06,
+      "loss": 0.5092,
+      "step": 1967
+    },
+    {
+      "epoch": 0.36057163796262365,
+      "grad_norm": 0.3904145734640018,
+      "learning_rate": 8.073320994138522e-06,
+      "loss": 0.498,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3607548552583364,
+      "grad_norm": 0.41968588248870525,
+      "learning_rate": 8.070797917235071e-06,
+      "loss": 0.4592,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3609380725540491,
+      "grad_norm": 0.40225333028901955,
+      "learning_rate": 8.068273584202593e-06,
+      "loss": 0.498,
+      "step": 1970
+    },
+    {
+      "epoch": 0.36112128984976183,
+      "grad_norm": 0.4779749898506442,
+      "learning_rate": 8.065747996073681e-06,
+      "loss": 0.4852,
+      "step": 1971
+    },
+    {
+      "epoch": 0.3613045071454745,
+      "grad_norm": 0.3801653362164248,
+      "learning_rate": 8.063221153881443e-06,
+      "loss": 0.4821,
+      "step": 1972
+    },
+    {
+      "epoch": 0.36148772444118726,
+      "grad_norm": 0.4360824877290448,
+      "learning_rate": 8.0606930586595e-06,
+      "loss": 0.4867,
+      "step": 1973
+    },
+    {
+      "epoch": 0.36167094173689995,
+      "grad_norm": 0.4253700547037495,
+      "learning_rate": 8.058163711441986e-06,
+      "loss": 0.4829,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3618541590326127,
+      "grad_norm": 0.4341340818854911,
+      "learning_rate": 8.055633113263543e-06,
+      "loss": 0.5169,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3620373763283254,
+      "grad_norm": 0.4424513726038396,
+      "learning_rate": 8.053101265159331e-06,
+      "loss": 0.5005,
+      "step": 1976
+    },
+    {
+      "epoch": 0.36222059362403813,
+      "grad_norm": 0.40080376851886934,
+      "learning_rate": 8.050568168165018e-06,
+      "loss": 0.4831,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3624038109197508,
+      "grad_norm": 0.4092773451489212,
+      "learning_rate": 8.048033823316784e-06,
+      "loss": 0.4886,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3625870282154635,
+      "grad_norm": 0.5750732156961504,
+      "learning_rate": 8.045498231651314e-06,
+      "loss": 0.4924,
+      "step": 1979
+    },
+    {
+      "epoch": 0.36277024551117626,
+      "grad_norm": 0.4000642911770352,
+      "learning_rate": 8.042961394205812e-06,
+      "loss": 0.4856,
+      "step": 1980
+    },
+    {
+      "epoch": 0.36295346280688895,
+      "grad_norm": 0.38938145758582615,
+      "learning_rate": 8.040423312017986e-06,
+      "loss": 0.4778,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3631366801026017,
+      "grad_norm": 0.3643427586018773,
+      "learning_rate": 8.037883986126054e-06,
+      "loss": 0.4909,
+      "step": 1982
+    },
+    {
+      "epoch": 0.3633198973983144,
+      "grad_norm": 0.3939262969339654,
+      "learning_rate": 8.035343417568742e-06,
+      "loss": 0.4714,
+      "step": 1983
+    },
+    {
+      "epoch": 0.36350311469402713,
+      "grad_norm": 0.4016181460823361,
+      "learning_rate": 8.032801607385288e-06,
+      "loss": 0.4931,
+      "step": 1984
+    },
+    {
+      "epoch": 0.3636863319897398,
+      "grad_norm": 0.44751708009914176,
+      "learning_rate": 8.030258556615433e-06,
+      "loss": 0.4894,
+      "step": 1985
+    },
+    {
+      "epoch": 0.36386954928545256,
+      "grad_norm": 0.34958387993124196,
+      "learning_rate": 8.02771426629943e-06,
+      "loss": 0.4598,
+      "step": 1986
+    },
+    {
+      "epoch": 0.36405276658116525,
+      "grad_norm": 0.348786463445316,
+      "learning_rate": 8.025168737478034e-06,
+      "loss": 0.447,
+      "step": 1987
+    },
+    {
+      "epoch": 0.364235983876878,
+      "grad_norm": 0.3854079222539858,
+      "learning_rate": 8.022621971192513e-06,
+      "loss": 0.4812,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3644192011725907,
+      "grad_norm": 0.38734200128527324,
+      "learning_rate": 8.020073968484632e-06,
+      "loss": 0.4862,
+      "step": 1989
+    },
+    {
+      "epoch": 0.36460241846830344,
+      "grad_norm": 0.411711055657578,
+      "learning_rate": 8.017524730396673e-06,
+      "loss": 0.4907,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3647856357640161,
+      "grad_norm": 0.440618505246582,
+      "learning_rate": 8.014974257971415e-06,
+      "loss": 0.4588,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3649688530597288,
+      "grad_norm": 0.4841735582737877,
+      "learning_rate": 8.012422552252148e-06,
+      "loss": 0.4988,
+      "step": 1992
+    },
+    {
+      "epoch": 0.36515207035544156,
+      "grad_norm": 0.35566001438725936,
+      "learning_rate": 8.009869614282657e-06,
+      "loss": 0.5083,
+      "step": 1993
+    },
+    {
+      "epoch": 0.36533528765115425,
+      "grad_norm": 0.3607042141519043,
+      "learning_rate": 8.007315445107242e-06,
+      "loss": 0.4634,
+      "step": 1994
+    },
+    {
+      "epoch": 0.365518504946867,
+      "grad_norm": 0.3841642715938066,
+      "learning_rate": 8.004760045770702e-06,
+      "loss": 0.4972,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3657017222425797,
+      "grad_norm": 0.39468083856708197,
+      "learning_rate": 8.002203417318335e-06,
+      "loss": 0.4779,
+      "step": 1996
+    },
+    {
+      "epoch": 0.36588493953829243,
+      "grad_norm": 0.419151699921075,
+      "learning_rate": 7.999645560795947e-06,
+      "loss": 0.5039,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3660681568340051,
+      "grad_norm": 0.4148076872613475,
+      "learning_rate": 7.99708647724985e-06,
+      "loss": 0.4969,
+      "step": 1998
+    },
+    {
+      "epoch": 0.36625137412971787,
+      "grad_norm": 0.37720877965942756,
+      "learning_rate": 7.994526167726847e-06,
+      "loss": 0.4657,
+      "step": 1999
+    },
+    {
+      "epoch": 0.36643459142543056,
+      "grad_norm": 0.3456057298421402,
+      "learning_rate": 7.991964633274255e-06,
+      "loss": 0.5015,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5458,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 357375823380480.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}