diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,71113 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.999802839116719,
+  "eval_steps": 1268,
+  "global_step": 10144,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0001971608832807571,
+      "grad_norm": 3.145049268342521,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.6851,
+      "step": 1
+    },
+    {
+      "epoch": 0.0001971608832807571,
+      "eval_loss": 0.6705650091171265,
+      "eval_runtime": 342.719,
+      "eval_samples_per_second": 23.722,
+      "eval_steps_per_second": 1.485,
+      "step": 1
+    },
+    {
+      "epoch": 0.0003943217665615142,
+      "grad_norm": 2.928706161869534,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.6183,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005914826498422713,
+      "grad_norm": 2.8063413659680485,
+      "learning_rate": 3e-06,
+      "loss": 0.6447,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007886435331230284,
+      "grad_norm": 4.474770355859649,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6615,
+      "step": 4
+    },
+    {
+      "epoch": 0.0009858044164037854,
+      "grad_norm": 2.83892152250444,
+      "learning_rate": 5e-06,
+      "loss": 0.5571,
+      "step": 5
+    },
+    {
+      "epoch": 0.0011829652996845426,
+      "grad_norm": 3.174661998811544,
+      "learning_rate": 6e-06,
+      "loss": 0.6092,
+      "step": 6
+    },
+    {
+      "epoch": 0.0013801261829652998,
+      "grad_norm": 2.086667997610238,
+      "learning_rate": 7e-06,
+      "loss": 0.5469,
+      "step": 7
+    },
+    {
+      "epoch": 0.0015772870662460567,
+      "grad_norm": 2.4093348319308454,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.6155,
+      "step": 8
+    },
+    {
+      "epoch": 0.001774447949526814,
+      "grad_norm": 1.753355191522501,
+      "learning_rate": 9e-06,
+      "loss": 0.5135,
+      "step": 9
+    },
+    {
+      "epoch": 0.001971608832807571,
+      "grad_norm": 1.813162138301549,
+      "learning_rate": 1e-05,
+      "loss": 0.5444,
+      "step": 10
+    },
+    {
+      "epoch": 0.002168769716088328,
+      "grad_norm": 1.5430930747728506,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.5032,
+      "step": 11
+    },
+    {
+      "epoch": 0.002365930599369085,
+      "grad_norm": 1.4956193189962286,
+      "learning_rate": 1.2e-05,
+      "loss": 0.5294,
+      "step": 12
+    },
+    {
+      "epoch": 0.0025630914826498424,
+      "grad_norm": 1.7344594420796293,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.5647,
+      "step": 13
+    },
+    {
+      "epoch": 0.0027602523659305996,
+      "grad_norm": 1.1966218841914706,
+      "learning_rate": 1.4e-05,
+      "loss": 0.5103,
+      "step": 14
+    },
+    {
+      "epoch": 0.0029574132492113563,
+      "grad_norm": 1.432228734486573,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.5,
+      "step": 15
+    },
+    {
+      "epoch": 0.0031545741324921135,
+      "grad_norm": 1.4645984670917185,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.4978,
+      "step": 16
+    },
+    {
+      "epoch": 0.0033517350157728706,
+      "grad_norm": 1.3563922951463974,
+      "learning_rate": 1.7e-05,
+      "loss": 0.506,
+      "step": 17
+    },
+    {
+      "epoch": 0.003548895899053628,
+      "grad_norm": 1.3871081225739526,
+      "learning_rate": 1.8e-05,
+      "loss": 0.5311,
+      "step": 18
+    },
+    {
+      "epoch": 0.003746056782334385,
+      "grad_norm": 1.280010346938333,
+      "learning_rate": 1.9e-05,
+      "loss": 0.5433,
+      "step": 19
+    },
+    {
+      "epoch": 0.003943217665615142,
+      "grad_norm": 2.284117117524146,
+      "learning_rate": 2e-05,
+      "loss": 0.5448,
+      "step": 20
+    },
+    {
+      "epoch": 0.004140378548895899,
+      "grad_norm": 1.6225424151126402,
+      "learning_rate": 1.9999999879870974e-05,
+      "loss": 0.5044,
+      "step": 21
+    },
+    {
+      "epoch": 0.004337539432176656,
+      "grad_norm": 1.6087069713987625,
+      "learning_rate": 1.99999995194839e-05,
+      "loss": 0.4712,
+      "step": 22
+    },
+    {
+      "epoch": 0.004534700315457414,
+      "grad_norm": 1.4815279243249726,
+      "learning_rate": 1.9999998918838782e-05,
+      "loss": 0.5043,
+      "step": 23
+    },
+    {
+      "epoch": 0.00473186119873817,
+      "grad_norm": 1.2421866810695292,
+      "learning_rate": 1.9999998077935636e-05,
+      "loss": 0.4933,
+      "step": 24
+    },
+    {
+      "epoch": 0.004929022082018927,
+      "grad_norm": 1.392319279479948,
+      "learning_rate": 1.9999996996774485e-05,
+      "loss": 0.4657,
+      "step": 25
+    },
+    {
+      "epoch": 0.005126182965299685,
+      "grad_norm": 1.613538190971824,
+      "learning_rate": 1.9999995675355352e-05,
+      "loss": 0.5023,
+      "step": 26
+    },
+    {
+      "epoch": 0.0053233438485804415,
+      "grad_norm": 1.3274753395366978,
+      "learning_rate": 1.999999411367827e-05,
+      "loss": 0.486,
+      "step": 27
+    },
+    {
+      "epoch": 0.005520504731861199,
+      "grad_norm": 1.4785198394383092,
+      "learning_rate": 1.9999992311743276e-05,
+      "loss": 0.4713,
+      "step": 28
+    },
+    {
+      "epoch": 0.005717665615141956,
+      "grad_norm": 1.4857486979381507,
+      "learning_rate": 1.9999990269550415e-05,
+      "loss": 0.4881,
+      "step": 29
+    },
+    {
+      "epoch": 0.005914826498422713,
+      "grad_norm": 1.36552629938149,
+      "learning_rate": 1.9999987987099734e-05,
+      "loss": 0.5463,
+      "step": 30
+    },
+    {
+      "epoch": 0.00611198738170347,
+      "grad_norm": 1.2130083488975558,
+      "learning_rate": 1.999998546439129e-05,
+      "loss": 0.5119,
+      "step": 31
+    },
+    {
+      "epoch": 0.006309148264984227,
+      "grad_norm": 1.4571092650286936,
+      "learning_rate": 1.999998270142514e-05,
+      "loss": 0.5107,
+      "step": 32
+    },
+    {
+      "epoch": 0.0065063091482649845,
+      "grad_norm": 1.3046426394488964,
+      "learning_rate": 1.9999979698201355e-05,
+      "loss": 0.5163,
+      "step": 33
+    },
+    {
+      "epoch": 0.006703470031545741,
+      "grad_norm": 1.3429233230107818,
+      "learning_rate": 1.999997645472e-05,
+      "loss": 0.4949,
+      "step": 34
+    },
+    {
+      "epoch": 0.006900630914826498,
+      "grad_norm": 1.1865584873872308,
+      "learning_rate": 1.9999972970981164e-05,
+      "loss": 0.5042,
+      "step": 35
+    },
+    {
+      "epoch": 0.007097791798107256,
+      "grad_norm": 1.2089421648447298,
+      "learning_rate": 1.999996924698492e-05,
+      "loss": 0.4996,
+      "step": 36
+    },
+    {
+      "epoch": 0.007294952681388012,
+      "grad_norm": 1.2171992786206751,
+      "learning_rate": 1.9999965282731364e-05,
+      "loss": 0.5232,
+      "step": 37
+    },
+    {
+      "epoch": 0.00749211356466877,
+      "grad_norm": 1.5903110532514266,
+      "learning_rate": 1.9999961078220587e-05,
+      "loss": 0.5206,
+      "step": 38
+    },
+    {
+      "epoch": 0.007689274447949527,
+      "grad_norm": 1.2539942673764422,
+      "learning_rate": 1.9999956633452696e-05,
+      "loss": 0.4656,
+      "step": 39
+    },
+    {
+      "epoch": 0.007886435331230283,
+      "grad_norm": 1.1722354147042853,
+      "learning_rate": 1.9999951948427793e-05,
+      "loss": 0.4811,
+      "step": 40
+    },
+    {
+      "epoch": 0.008083596214511041,
+      "grad_norm": 1.520380985973137,
+      "learning_rate": 1.9999947023145992e-05,
+      "loss": 0.4745,
+      "step": 41
+    },
+    {
+      "epoch": 0.008280757097791799,
+      "grad_norm": 1.181201745034852,
+      "learning_rate": 1.999994185760741e-05,
+      "loss": 0.5026,
+      "step": 42
+    },
+    {
+      "epoch": 0.008477917981072555,
+      "grad_norm": 1.2315559894723096,
+      "learning_rate": 1.9999936451812168e-05,
+      "loss": 0.5071,
+      "step": 43
+    },
+    {
+      "epoch": 0.008675078864353312,
+      "grad_norm": 1.0027266084772204,
+      "learning_rate": 1.9999930805760403e-05,
+      "loss": 0.4293,
+      "step": 44
+    },
+    {
+      "epoch": 0.00887223974763407,
+      "grad_norm": 1.5355646992423921,
+      "learning_rate": 1.999992491945225e-05,
+      "loss": 0.5075,
+      "step": 45
+    },
+    {
+      "epoch": 0.009069400630914827,
+      "grad_norm": 1.2141404668583387,
+      "learning_rate": 1.9999918792887844e-05,
+      "loss": 0.5002,
+      "step": 46
+    },
+    {
+      "epoch": 0.009266561514195583,
+      "grad_norm": 1.1874086565333581,
+      "learning_rate": 1.9999912426067335e-05,
+      "loss": 0.5214,
+      "step": 47
+    },
+    {
+      "epoch": 0.00946372239747634,
+      "grad_norm": 1.310468730323023,
+      "learning_rate": 1.999990581899088e-05,
+      "loss": 0.4954,
+      "step": 48
+    },
+    {
+      "epoch": 0.009660883280757098,
+      "grad_norm": 1.1678068139516498,
+      "learning_rate": 1.9999898971658632e-05,
+      "loss": 0.4874,
+      "step": 49
+    },
+    {
+      "epoch": 0.009858044164037854,
+      "grad_norm": 1.2436720803185994,
+      "learning_rate": 1.9999891884070764e-05,
+      "loss": 0.5053,
+      "step": 50
+    },
+    {
+      "epoch": 0.010055205047318612,
+      "grad_norm": 1.191814384498826,
+      "learning_rate": 1.999988455622744e-05,
+      "loss": 0.4437,
+      "step": 51
+    },
+    {
+      "epoch": 0.01025236593059937,
+      "grad_norm": 1.0535554735149129,
+      "learning_rate": 1.9999876988128832e-05,
+      "loss": 0.4572,
+      "step": 52
+    },
+    {
+      "epoch": 0.010449526813880125,
+      "grad_norm": 1.5861895667849393,
+      "learning_rate": 1.9999869179775126e-05,
+      "loss": 0.5274,
+      "step": 53
+    },
+    {
+      "epoch": 0.010646687697160883,
+      "grad_norm": 1.3410047214951284,
+      "learning_rate": 1.9999861131166513e-05,
+      "loss": 0.5134,
+      "step": 54
+    },
+    {
+      "epoch": 0.01084384858044164,
+      "grad_norm": 1.5169630762672912,
+      "learning_rate": 1.9999852842303183e-05,
+      "loss": 0.498,
+      "step": 55
+    },
+    {
+      "epoch": 0.011041009463722398,
+      "grad_norm": 1.221925524795734,
+      "learning_rate": 1.9999844313185335e-05,
+      "loss": 0.5275,
+      "step": 56
+    },
+    {
+      "epoch": 0.011238170347003154,
+      "grad_norm": 5.091704408593217,
+      "learning_rate": 1.9999835543813174e-05,
+      "loss": 0.4799,
+      "step": 57
+    },
+    {
+      "epoch": 0.011435331230283912,
+      "grad_norm": 2.4153909798169337,
+      "learning_rate": 1.9999826534186914e-05,
+      "loss": 0.4921,
+      "step": 58
+    },
+    {
+      "epoch": 0.01163249211356467,
+      "grad_norm": 1.547979859587359,
+      "learning_rate": 1.9999817284306766e-05,
+      "loss": 0.5229,
+      "step": 59
+    },
+    {
+      "epoch": 0.011829652996845425,
+      "grad_norm": 1.5690337913664598,
+      "learning_rate": 1.9999807794172955e-05,
+      "loss": 0.4735,
+      "step": 60
+    },
+    {
+      "epoch": 0.012026813880126183,
+      "grad_norm": 1.6033315849616043,
+      "learning_rate": 1.999979806378571e-05,
+      "loss": 0.5152,
+      "step": 61
+    },
+    {
+      "epoch": 0.01222397476340694,
+      "grad_norm": 1.2835879687259206,
+      "learning_rate": 1.9999788093145264e-05,
+      "loss": 0.5271,
+      "step": 62
+    },
+    {
+      "epoch": 0.012421135646687698,
+      "grad_norm": 1.2507778090991872,
+      "learning_rate": 1.9999777882251857e-05,
+      "loss": 0.4735,
+      "step": 63
+    },
+    {
+      "epoch": 0.012618296529968454,
+      "grad_norm": 1.0918550876107425,
+      "learning_rate": 1.999976743110573e-05,
+      "loss": 0.489,
+      "step": 64
+    },
+    {
+      "epoch": 0.012815457413249211,
+      "grad_norm": 1.3178360947947898,
+      "learning_rate": 1.999975673970714e-05,
+      "loss": 0.5017,
+      "step": 65
+    },
+    {
+      "epoch": 0.013012618296529969,
+      "grad_norm": 1.226739326525634,
+      "learning_rate": 1.9999745808056344e-05,
+      "loss": 0.4785,
+      "step": 66
+    },
+    {
+      "epoch": 0.013209779179810725,
+      "grad_norm": 1.1354158482899859,
+      "learning_rate": 1.99997346361536e-05,
+      "loss": 0.4727,
+      "step": 67
+    },
+    {
+      "epoch": 0.013406940063091483,
+      "grad_norm": 1.069961387810125,
+      "learning_rate": 1.9999723223999178e-05,
+      "loss": 0.4692,
+      "step": 68
+    },
+    {
+      "epoch": 0.01360410094637224,
+      "grad_norm": 0.9762520384137559,
+      "learning_rate": 1.999971157159335e-05,
+      "loss": 0.4922,
+      "step": 69
+    },
+    {
+      "epoch": 0.013801261829652996,
+      "grad_norm": 1.1048692949441026,
+      "learning_rate": 1.99996996789364e-05,
+      "loss": 0.4762,
+      "step": 70
+    },
+    {
+      "epoch": 0.013998422712933754,
+      "grad_norm": 1.4422307357408406,
+      "learning_rate": 1.9999687546028617e-05,
+      "loss": 0.517,
+      "step": 71
+    },
+    {
+      "epoch": 0.014195583596214511,
+      "grad_norm": 1.2130637644825035,
+      "learning_rate": 1.9999675172870286e-05,
+      "loss": 0.4549,
+      "step": 72
+    },
+    {
+      "epoch": 0.014392744479495269,
+      "grad_norm": 1.3283546807946351,
+      "learning_rate": 1.9999662559461704e-05,
+      "loss": 0.4912,
+      "step": 73
+    },
+    {
+      "epoch": 0.014589905362776025,
+      "grad_norm": 1.2835417017942559,
+      "learning_rate": 1.9999649705803178e-05,
+      "loss": 0.5079,
+      "step": 74
+    },
+    {
+      "epoch": 0.014787066246056782,
+      "grad_norm": 1.0779965377716199,
+      "learning_rate": 1.9999636611895018e-05,
+      "loss": 0.4633,
+      "step": 75
+    },
+    {
+      "epoch": 0.01498422712933754,
+      "grad_norm": 1.0321678770877212,
+      "learning_rate": 1.999962327773753e-05,
+      "loss": 0.4854,
+      "step": 76
+    },
+    {
+      "epoch": 0.015181388012618296,
+      "grad_norm": 1.2808800261718734,
+      "learning_rate": 1.9999609703331045e-05,
+      "loss": 0.4931,
+      "step": 77
+    },
+    {
+      "epoch": 0.015378548895899053,
+      "grad_norm": 1.3610125610716681,
+      "learning_rate": 1.999959588867588e-05,
+      "loss": 0.5009,
+      "step": 78
+    },
+    {
+      "epoch": 0.015575709779179811,
+      "grad_norm": 1.0655940435477698,
+      "learning_rate": 1.999958183377237e-05,
+      "loss": 0.4816,
+      "step": 79
+    },
+    {
+      "epoch": 0.015772870662460567,
+      "grad_norm": 1.2930806912423924,
+      "learning_rate": 1.999956753862086e-05,
+      "loss": 0.4807,
+      "step": 80
+    },
+    {
+      "epoch": 0.015970031545741326,
+      "grad_norm": 1.287162958203915,
+      "learning_rate": 1.9999553003221682e-05,
+      "loss": 0.4588,
+      "step": 81
+    },
+    {
+      "epoch": 0.016167192429022082,
+      "grad_norm": 1.1339051396074737,
+      "learning_rate": 1.9999538227575196e-05,
+      "loss": 0.5081,
+      "step": 82
+    },
+    {
+      "epoch": 0.016364353312302838,
+      "grad_norm": 1.1265294750507961,
+      "learning_rate": 1.9999523211681746e-05,
+      "loss": 0.4669,
+      "step": 83
+    },
+    {
+      "epoch": 0.016561514195583597,
+      "grad_norm": 1.1931313211356958,
+      "learning_rate": 1.99995079555417e-05,
+      "loss": 0.48,
+      "step": 84
+    },
+    {
+      "epoch": 0.016758675078864353,
+      "grad_norm": 0.999555485327999,
+      "learning_rate": 1.9999492459155424e-05,
+      "loss": 0.4786,
+      "step": 85
+    },
+    {
+      "epoch": 0.01695583596214511,
+      "grad_norm": 1.1669083683078765,
+      "learning_rate": 1.9999476722523287e-05,
+      "loss": 0.4581,
+      "step": 86
+    },
+    {
+      "epoch": 0.01715299684542587,
+      "grad_norm": 1.0996260118567778,
+      "learning_rate": 1.9999460745645673e-05,
+      "loss": 0.4871,
+      "step": 87
+    },
+    {
+      "epoch": 0.017350157728706624,
+      "grad_norm": 1.1438449589174153,
+      "learning_rate": 1.999944452852296e-05,
+      "loss": 0.4852,
+      "step": 88
+    },
+    {
+      "epoch": 0.01754731861198738,
+      "grad_norm": 1.0031606047710635,
+      "learning_rate": 1.9999428071155535e-05,
+      "loss": 0.5007,
+      "step": 89
+    },
+    {
+      "epoch": 0.01774447949526814,
+      "grad_norm": 1.3469859967905786,
+      "learning_rate": 1.9999411373543804e-05,
+      "loss": 0.5102,
+      "step": 90
+    },
+    {
+      "epoch": 0.017941640378548895,
+      "grad_norm": 1.1936672885104658,
+      "learning_rate": 1.9999394435688158e-05,
+      "loss": 0.5077,
+      "step": 91
+    },
+    {
+      "epoch": 0.018138801261829655,
+      "grad_norm": 1.3174882849640317,
+      "learning_rate": 1.9999377257589012e-05,
+      "loss": 0.4961,
+      "step": 92
+    },
+    {
+      "epoch": 0.01833596214511041,
+      "grad_norm": 1.2572224290689644,
+      "learning_rate": 1.9999359839246775e-05,
+      "loss": 0.498,
+      "step": 93
+    },
+    {
+      "epoch": 0.018533123028391166,
+      "grad_norm": 1.1310411895258263,
+      "learning_rate": 1.9999342180661863e-05,
+      "loss": 0.5175,
+      "step": 94
+    },
+    {
+      "epoch": 0.018730283911671926,
+      "grad_norm": 1.2365159429858565,
+      "learning_rate": 1.9999324281834705e-05,
+      "loss": 0.4588,
+      "step": 95
+    },
+    {
+      "epoch": 0.01892744479495268,
+      "grad_norm": 1.3269862941988586,
+      "learning_rate": 1.9999306142765726e-05,
+      "loss": 0.485,
+      "step": 96
+    },
+    {
+      "epoch": 0.019124605678233438,
+      "grad_norm": 1.1672780497391835,
+      "learning_rate": 1.9999287763455367e-05,
+      "loss": 0.4964,
+      "step": 97
+    },
+    {
+      "epoch": 0.019321766561514197,
+      "grad_norm": 1.2280978200722572,
+      "learning_rate": 1.9999269143904066e-05,
+      "loss": 0.5031,
+      "step": 98
+    },
+    {
+      "epoch": 0.019518927444794953,
+      "grad_norm": 1.1756264159886103,
+      "learning_rate": 1.999925028411227e-05,
+      "loss": 0.4558,
+      "step": 99
+    },
+    {
+      "epoch": 0.01971608832807571,
+      "grad_norm": 1.0797299658749118,
+      "learning_rate": 1.9999231184080434e-05,
+      "loss": 0.509,
+      "step": 100
+    },
+    {
+      "epoch": 0.019913249211356468,
+      "grad_norm": 1.5073253405939093,
+      "learning_rate": 1.9999211843809018e-05,
+      "loss": 0.4829,
+      "step": 101
+    },
+    {
+      "epoch": 0.020110410094637224,
+      "grad_norm": 1.22115492589763,
+      "learning_rate": 1.9999192263298485e-05,
+      "loss": 0.4954,
+      "step": 102
+    },
+    {
+      "epoch": 0.02030757097791798,
+      "grad_norm": 1.215526599323687,
+      "learning_rate": 1.9999172442549307e-05,
+      "loss": 0.4973,
+      "step": 103
+    },
+    {
+      "epoch": 0.02050473186119874,
+      "grad_norm": 1.0092097441473407,
+      "learning_rate": 1.9999152381561955e-05,
+      "loss": 0.461,
+      "step": 104
+    },
+    {
+      "epoch": 0.020701892744479495,
+      "grad_norm": 1.1251707920134784,
+      "learning_rate": 1.9999132080336915e-05,
+      "loss": 0.503,
+      "step": 105
+    },
+    {
+      "epoch": 0.02089905362776025,
+      "grad_norm": 1.0138460258654038,
+      "learning_rate": 1.9999111538874677e-05,
+      "loss": 0.4765,
+      "step": 106
+    },
+    {
+      "epoch": 0.02109621451104101,
+      "grad_norm": 1.2072707360767827,
+      "learning_rate": 1.999909075717573e-05,
+      "loss": 0.4894,
+      "step": 107
+    },
+    {
+      "epoch": 0.021293375394321766,
+      "grad_norm": 1.0013269764600548,
+      "learning_rate": 1.9999069735240578e-05,
+      "loss": 0.4957,
+      "step": 108
+    },
+    {
+      "epoch": 0.021490536277602525,
+      "grad_norm": 1.1353117571038165,
+      "learning_rate": 1.999904847306972e-05,
+      "loss": 0.5045,
+      "step": 109
+    },
+    {
+      "epoch": 0.02168769716088328,
+      "grad_norm": 1.057179146627742,
+      "learning_rate": 1.999902697066367e-05,
+      "loss": 0.4928,
+      "step": 110
+    },
+    {
+      "epoch": 0.021884858044164037,
+      "grad_norm": 0.9423530793571464,
+      "learning_rate": 1.999900522802295e-05,
+      "loss": 0.5128,
+      "step": 111
+    },
+    {
+      "epoch": 0.022082018927444796,
+      "grad_norm": 0.968781602059733,
+      "learning_rate": 1.9998983245148072e-05,
+      "loss": 0.4547,
+      "step": 112
+    },
+    {
+      "epoch": 0.022279179810725552,
+      "grad_norm": 0.9113728069244939,
+      "learning_rate": 1.999896102203957e-05,
+      "loss": 0.4814,
+      "step": 113
+    },
+    {
+      "epoch": 0.022476340694006308,
+      "grad_norm": 1.1042295670503781,
+      "learning_rate": 1.999893855869798e-05,
+      "loss": 0.4613,
+      "step": 114
+    },
+    {
+      "epoch": 0.022673501577287068,
+      "grad_norm": 1.034777939325516,
+      "learning_rate": 1.999891585512384e-05,
+      "loss": 0.5244,
+      "step": 115
+    },
+    {
+      "epoch": 0.022870662460567823,
+      "grad_norm": 0.9820709661002166,
+      "learning_rate": 1.999889291131769e-05,
+      "loss": 0.5079,
+      "step": 116
+    },
+    {
+      "epoch": 0.02306782334384858,
+      "grad_norm": 2.571070138444174,
+      "learning_rate": 1.9998869727280088e-05,
+      "loss": 0.4896,
+      "step": 117
+    },
+    {
+      "epoch": 0.02326498422712934,
+      "grad_norm": 1.5658204423287487,
+      "learning_rate": 1.9998846303011588e-05,
+      "loss": 0.4715,
+      "step": 118
+    },
+    {
+      "epoch": 0.023462145110410094,
+      "grad_norm": 0.9704506073467836,
+      "learning_rate": 1.9998822638512757e-05,
+      "loss": 0.4948,
+      "step": 119
+    },
+    {
+      "epoch": 0.02365930599369085,
+      "grad_norm": 1.8219918715734673,
+      "learning_rate": 1.9998798733784155e-05,
+      "loss": 0.4449,
+      "step": 120
+    },
+    {
+      "epoch": 0.02385646687697161,
+      "grad_norm": 1.3674996874754528,
+      "learning_rate": 1.9998774588826362e-05,
+      "loss": 0.4736,
+      "step": 121
+    },
+    {
+      "epoch": 0.024053627760252366,
+      "grad_norm": 1.0825809239978976,
+      "learning_rate": 1.999875020363996e-05,
+      "loss": 0.4462,
+      "step": 122
+    },
+    {
+      "epoch": 0.02425078864353312,
+      "grad_norm": 1.2402946359904268,
+      "learning_rate": 1.999872557822553e-05,
+      "loss": 0.464,
+      "step": 123
+    },
+    {
+      "epoch": 0.02444794952681388,
+      "grad_norm": 0.9734568557016353,
+      "learning_rate": 1.999870071258367e-05,
+      "loss": 0.5062,
+      "step": 124
+    },
+    {
+      "epoch": 0.024645110410094637,
+      "grad_norm": 1.32265424416242,
+      "learning_rate": 1.999867560671497e-05,
+      "loss": 0.458,
+      "step": 125
+    },
+    {
+      "epoch": 0.024842271293375396,
+      "grad_norm": 1.0916211714042101,
+      "learning_rate": 1.999865026062004e-05,
+      "loss": 0.4639,
+      "step": 126
+    },
+    {
+      "epoch": 0.025039432176656152,
+      "grad_norm": 1.0101120690939442,
+      "learning_rate": 1.999862467429948e-05,
+      "loss": 0.5054,
+      "step": 127
+    },
+    {
+      "epoch": 0.025236593059936908,
+      "grad_norm": 1.07583395846967,
+      "learning_rate": 1.9998598847753918e-05,
+      "loss": 0.4147,
+      "step": 128
+    },
+    {
+      "epoch": 0.025433753943217667,
+      "grad_norm": 1.0153450766886587,
+      "learning_rate": 1.999857278098396e-05,
+      "loss": 0.4608,
+      "step": 129
+    },
+    {
+      "epoch": 0.025630914826498423,
+      "grad_norm": 0.9636929458831386,
+      "learning_rate": 1.999854647399024e-05,
+      "loss": 0.4435,
+      "step": 130
+    },
+    {
+      "epoch": 0.02582807570977918,
+      "grad_norm": 0.976995842942242,
+      "learning_rate": 1.999851992677339e-05,
+      "loss": 0.468,
+      "step": 131
+    },
+    {
+      "epoch": 0.026025236593059938,
+      "grad_norm": 1.2900795634559068,
+      "learning_rate": 1.999849313933405e-05,
+      "loss": 0.4996,
+      "step": 132
+    },
+    {
+      "epoch": 0.026222397476340694,
+      "grad_norm": 1.053599579751552,
+      "learning_rate": 1.9998466111672856e-05,
+      "loss": 0.4973,
+      "step": 133
+    },
+    {
+      "epoch": 0.02641955835962145,
+      "grad_norm": 1.018304999333016,
+      "learning_rate": 1.9998438843790463e-05,
+      "loss": 0.4754,
+      "step": 134
+    },
+    {
+      "epoch": 0.02661671924290221,
+      "grad_norm": 0.9962199755386415,
+      "learning_rate": 1.9998411335687527e-05,
+      "loss": 0.5074,
+      "step": 135
+    },
+    {
+      "epoch": 0.026813880126182965,
+      "grad_norm": 1.0381641170143454,
+      "learning_rate": 1.9998383587364706e-05,
+      "loss": 0.5516,
+      "step": 136
+    },
+    {
+      "epoch": 0.02701104100946372,
+      "grad_norm": 1.0875760040456148,
+      "learning_rate": 1.999835559882267e-05,
+      "loss": 0.4803,
+      "step": 137
+    },
+    {
+      "epoch": 0.02720820189274448,
+      "grad_norm": 1.5281532146270271,
+      "learning_rate": 1.9998327370062086e-05,
+      "loss": 0.4919,
+      "step": 138
+    },
+    {
+      "epoch": 0.027405362776025236,
+      "grad_norm": 0.9062501284108182,
+      "learning_rate": 1.9998298901083637e-05,
+      "loss": 0.4639,
+      "step": 139
+    },
+    {
+      "epoch": 0.027602523659305992,
+      "grad_norm": 0.9073298345394162,
+      "learning_rate": 1.9998270191888002e-05,
+      "loss": 0.476,
+      "step": 140
+    },
+    {
+      "epoch": 0.02779968454258675,
+      "grad_norm": 0.8859849922750372,
+      "learning_rate": 1.9998241242475876e-05,
+      "loss": 0.4808,
+      "step": 141
+    },
+    {
+      "epoch": 0.027996845425867507,
+      "grad_norm": 9.232525438376628,
+      "learning_rate": 1.9998212052847955e-05,
+      "loss": 0.5524,
+      "step": 142
+    },
+    {
+      "epoch": 0.028194006309148267,
+      "grad_norm": 1.1971641286106405,
+      "learning_rate": 1.9998182623004935e-05,
+      "loss": 0.4861,
+      "step": 143
+    },
+    {
+      "epoch": 0.028391167192429023,
+      "grad_norm": 1.1014546150503517,
+      "learning_rate": 1.9998152952947526e-05,
+      "loss": 0.4994,
+      "step": 144
+    },
+    {
+      "epoch": 0.02858832807570978,
+      "grad_norm": 1.1251735110198886,
+      "learning_rate": 1.9998123042676444e-05,
+      "loss": 0.5235,
+      "step": 145
+    },
+    {
+      "epoch": 0.028785488958990538,
+      "grad_norm": 1.0757931397923415,
+      "learning_rate": 1.9998092892192403e-05,
+      "loss": 0.4805,
+      "step": 146
+    },
+    {
+      "epoch": 0.028982649842271294,
+      "grad_norm": 1.0601740904924608,
+      "learning_rate": 1.9998062501496126e-05,
+      "loss": 0.4652,
+      "step": 147
+    },
+    {
+      "epoch": 0.02917981072555205,
+      "grad_norm": 1.0857317459697904,
+      "learning_rate": 1.999803187058835e-05,
+      "loss": 0.4949,
+      "step": 148
+    },
+    {
+      "epoch": 0.02937697160883281,
+      "grad_norm": 1.2154921400920446,
+      "learning_rate": 1.99980009994698e-05,
+      "loss": 0.4988,
+      "step": 149
+    },
+    {
+      "epoch": 0.029574132492113565,
+      "grad_norm": 1.0127571620302098,
+      "learning_rate": 1.999796988814123e-05,
+      "loss": 0.4898,
+      "step": 150
+    },
+    {
+      "epoch": 0.02977129337539432,
+      "grad_norm": 0.9897255202956721,
+      "learning_rate": 1.9997938536603386e-05,
+      "loss": 0.5307,
+      "step": 151
+    },
+    {
+      "epoch": 0.02996845425867508,
+      "grad_norm": 20.686417789329063,
+      "learning_rate": 1.999790694485701e-05,
+      "loss": 0.551,
+      "step": 152
+    },
+    {
+      "epoch": 0.030165615141955836,
+      "grad_norm": 1.7162332300485896,
+      "learning_rate": 1.999787511290287e-05,
+      "loss": 0.5087,
+      "step": 153
+    },
+    {
+      "epoch": 0.03036277602523659,
+      "grad_norm": 1.3514560362581123,
+      "learning_rate": 1.999784304074173e-05,
+      "loss": 0.5216,
+      "step": 154
+    },
+    {
+      "epoch": 0.03055993690851735,
+      "grad_norm": 1.5606869153465206,
+      "learning_rate": 1.9997810728374362e-05,
+      "loss": 0.4768,
+      "step": 155
+    },
+    {
+      "epoch": 0.030757097791798107,
+      "grad_norm": 1.3648443694718702,
+      "learning_rate": 1.999777817580154e-05,
+      "loss": 0.494,
+      "step": 156
+    },
+    {
+      "epoch": 0.030954258675078863,
+      "grad_norm": 1.132691943500696,
+      "learning_rate": 1.9997745383024043e-05,
+      "loss": 0.4662,
+      "step": 157
+    },
+    {
+      "epoch": 0.031151419558359622,
+      "grad_norm": 1.2793599540460188,
+      "learning_rate": 1.9997712350042663e-05,
+      "loss": 0.4834,
+      "step": 158
+    },
+    {
+      "epoch": 0.03134858044164038,
+      "grad_norm": 1.1425965940998726,
+      "learning_rate": 1.9997679076858193e-05,
+      "loss": 0.4975,
+      "step": 159
+    },
+    {
+      "epoch": 0.031545741324921134,
+      "grad_norm": 1.1861645099171103,
+      "learning_rate": 1.9997645563471432e-05,
+      "loss": 0.4793,
+      "step": 160
+    },
+    {
+      "epoch": 0.03174290220820189,
+      "grad_norm": 1.0972295116283801,
+      "learning_rate": 1.9997611809883187e-05,
+      "loss": 0.4746,
+      "step": 161
+    },
+    {
+      "epoch": 0.03194006309148265,
+      "grad_norm": 1.066451222541386,
+      "learning_rate": 1.9997577816094266e-05,
+      "loss": 0.4907,
+      "step": 162
+    },
+    {
+      "epoch": 0.03213722397476341,
+      "grad_norm": 1.0553209971214357,
+      "learning_rate": 1.9997543582105484e-05,
+      "loss": 0.4623,
+      "step": 163
+    },
+    {
+      "epoch": 0.032334384858044164,
+      "grad_norm": 0.978588088915193,
+      "learning_rate": 1.999750910791767e-05,
+      "loss": 0.4758,
+      "step": 164
+    },
+    {
+      "epoch": 0.03253154574132492,
+      "grad_norm": 1.1746969879352085,
+      "learning_rate": 1.9997474393531648e-05,
+      "loss": 0.5184,
+      "step": 165
+    },
+    {
+      "epoch": 0.032728706624605676,
+      "grad_norm": 1.1684805326278327,
+      "learning_rate": 1.999743943894825e-05,
+      "loss": 0.5299,
+      "step": 166
+    },
+    {
+      "epoch": 0.03292586750788644,
+      "grad_norm": 1.0632876477663906,
+      "learning_rate": 1.999740424416832e-05,
+      "loss": 0.4858,
+      "step": 167
+    },
+    {
+      "epoch": 0.033123028391167195,
+      "grad_norm": 1.0392183455349258,
+      "learning_rate": 1.9997368809192704e-05,
+      "loss": 0.4637,
+      "step": 168
+    },
+    {
+      "epoch": 0.03332018927444795,
+      "grad_norm": 1.0283561237168466,
+      "learning_rate": 1.999733313402225e-05,
+      "loss": 0.5204,
+      "step": 169
+    },
+    {
+      "epoch": 0.033517350157728706,
+      "grad_norm": 1.1695473033049606,
+      "learning_rate": 1.999729721865782e-05,
+      "loss": 0.5182,
+      "step": 170
+    },
+    {
+      "epoch": 0.03371451104100946,
+      "grad_norm": 0.9872390177704469,
+      "learning_rate": 1.999726106310027e-05,
+      "loss": 0.4961,
+      "step": 171
+    },
+    {
+      "epoch": 0.03391167192429022,
+      "grad_norm": 2.3231197797384,
+      "learning_rate": 1.9997224667350474e-05,
+      "loss": 0.4806,
+      "step": 172
+    },
+    {
+      "epoch": 0.03410883280757098,
+      "grad_norm": 1.1146751680169542,
+      "learning_rate": 1.9997188031409302e-05,
+      "loss": 0.4741,
+      "step": 173
+    },
+    {
+      "epoch": 0.03430599369085174,
+      "grad_norm": 0.896181610550275,
+      "learning_rate": 1.9997151155277638e-05,
+      "loss": 0.4462,
+      "step": 174
+    },
+    {
+      "epoch": 0.03450315457413249,
+      "grad_norm": 0.9768565430868928,
+      "learning_rate": 1.9997114038956367e-05,
+      "loss": 0.4981,
+      "step": 175
+    },
+    {
+      "epoch": 0.03470031545741325,
+      "grad_norm": 0.9551378209936853,
+      "learning_rate": 1.999707668244638e-05,
+      "loss": 0.4674,
+      "step": 176
+    },
+    {
+      "epoch": 0.034897476340694004,
+      "grad_norm": 0.9851958907905455,
+      "learning_rate": 1.9997039085748576e-05,
+      "loss": 0.4604,
+      "step": 177
+    },
+    {
+      "epoch": 0.03509463722397476,
+      "grad_norm": 1.0161464160279716,
+      "learning_rate": 1.9997001248863858e-05,
+      "loss": 0.4541,
+      "step": 178
+    },
+    {
+      "epoch": 0.03529179810725552,
+      "grad_norm": 1.050476492771695,
+      "learning_rate": 1.9996963171793132e-05,
+      "loss": 0.4949,
+      "step": 179
+    },
+    {
+      "epoch": 0.03548895899053628,
+      "grad_norm": 1.1900294403214005,
+      "learning_rate": 1.999692485453732e-05,
+      "loss": 0.4815,
+      "step": 180
+    },
+    {
+      "epoch": 0.035686119873817035,
+      "grad_norm": 1.1031147141223552,
+      "learning_rate": 1.9996886297097335e-05,
+      "loss": 0.4862,
+      "step": 181
+    },
+    {
+      "epoch": 0.03588328075709779,
+      "grad_norm": 1.083256023575967,
+      "learning_rate": 1.9996847499474102e-05,
+      "loss": 0.4587,
+      "step": 182
+    },
+    {
+      "epoch": 0.03608044164037855,
+      "grad_norm": 0.9670193439952784,
+      "learning_rate": 1.9996808461668565e-05,
+      "loss": 0.4513,
+      "step": 183
+    },
+    {
+      "epoch": 0.03627760252365931,
+      "grad_norm": 0.8979692827898681,
+      "learning_rate": 1.999676918368165e-05,
+      "loss": 0.4517,
+      "step": 184
+    },
+    {
+      "epoch": 0.036474763406940065,
+      "grad_norm": 0.9969877135191793,
+      "learning_rate": 1.9996729665514306e-05,
+      "loss": 0.4933,
+      "step": 185
+    },
+    {
+      "epoch": 0.03667192429022082,
+      "grad_norm": 0.9692177718636533,
+      "learning_rate": 1.999668990716748e-05,
+      "loss": 0.5174,
+      "step": 186
+    },
+    {
+      "epoch": 0.03686908517350158,
+      "grad_norm": 0.8842091129334964,
+      "learning_rate": 1.999664990864213e-05,
+      "loss": 0.4742,
+      "step": 187
+    },
+    {
+      "epoch": 0.03706624605678233,
+      "grad_norm": 0.9359713100345171,
+      "learning_rate": 1.9996609669939214e-05,
+      "loss": 0.4754,
+      "step": 188
+    },
+    {
+      "epoch": 0.03726340694006309,
+      "grad_norm": 0.9191517387573663,
+      "learning_rate": 1.9996569191059705e-05,
+      "loss": 0.4416,
+      "step": 189
+    },
+    {
+      "epoch": 0.03746056782334385,
+      "grad_norm": 0.9570764893044853,
+      "learning_rate": 1.9996528472004567e-05,
+      "loss": 0.5044,
+      "step": 190
+    },
+    {
+      "epoch": 0.03765772870662461,
+      "grad_norm": 0.9708144686903228,
+      "learning_rate": 1.999648751277478e-05,
+      "loss": 0.4986,
+      "step": 191
+    },
+    {
+      "epoch": 0.03785488958990536,
+      "grad_norm": 0.9143041364090021,
+      "learning_rate": 1.9996446313371334e-05,
+      "loss": 0.4715,
+      "step": 192
+    },
+    {
+      "epoch": 0.03805205047318612,
+      "grad_norm": 1.0756442068587027,
+      "learning_rate": 1.9996404873795216e-05,
+      "loss": 0.466,
+      "step": 193
+    },
+    {
+      "epoch": 0.038249211356466875,
+      "grad_norm": 0.8509637591107831,
+      "learning_rate": 1.999636319404742e-05,
+      "loss": 0.4419,
+      "step": 194
+    },
+    {
+      "epoch": 0.03844637223974763,
+      "grad_norm": 0.8874791124952598,
+      "learning_rate": 1.9996321274128947e-05,
+      "loss": 0.4899,
+      "step": 195
+    },
+    {
+      "epoch": 0.038643533123028394,
+      "grad_norm": 0.9741286408649374,
+      "learning_rate": 1.9996279114040806e-05,
+      "loss": 0.4753,
+      "step": 196
+    },
+    {
+      "epoch": 0.03884069400630915,
+      "grad_norm": 0.9085224492157351,
+      "learning_rate": 1.999623671378401e-05,
+      "loss": 0.4701,
+      "step": 197
+    },
+    {
+      "epoch": 0.039037854889589906,
+      "grad_norm": 1.0258910989454444,
+      "learning_rate": 1.9996194073359576e-05,
+      "loss": 0.4888,
+      "step": 198
+    },
+    {
+      "epoch": 0.03923501577287066,
+      "grad_norm": 0.8126732806815207,
+      "learning_rate": 1.999615119276853e-05,
+      "loss": 0.4555,
+      "step": 199
+    },
+    {
+      "epoch": 0.03943217665615142,
+      "grad_norm": 0.8218884414629809,
+      "learning_rate": 1.99961080720119e-05,
+      "loss": 0.4769,
+      "step": 200
+    },
+    {
+      "epoch": 0.03962933753943218,
+      "grad_norm": 0.8991946813956051,
+      "learning_rate": 1.9996064711090727e-05,
+      "loss": 0.4588,
+      "step": 201
+    },
+    {
+      "epoch": 0.039826498422712936,
+      "grad_norm": 0.8566916147286883,
+      "learning_rate": 1.9996021110006046e-05,
+      "loss": 0.5019,
+      "step": 202
+    },
+    {
+      "epoch": 0.04002365930599369,
+      "grad_norm": 0.9756060213649459,
+      "learning_rate": 1.9995977268758912e-05,
+      "loss": 0.4746,
+      "step": 203
+    },
+    {
+      "epoch": 0.04022082018927445,
+      "grad_norm": 2.1550975546469253,
+      "learning_rate": 1.9995933187350372e-05,
+      "loss": 0.5151,
+      "step": 204
+    },
+    {
+      "epoch": 0.040417981072555204,
+      "grad_norm": 2.0312951209754386,
+      "learning_rate": 1.999588886578149e-05,
+      "loss": 0.4713,
+      "step": 205
+    },
+    {
+      "epoch": 0.04061514195583596,
+      "grad_norm": 1.0090851939886327,
+      "learning_rate": 1.9995844304053325e-05,
+      "loss": 0.4607,
+      "step": 206
+    },
+    {
+      "epoch": 0.04081230283911672,
+      "grad_norm": 1.2675954246551175,
+      "learning_rate": 1.9995799502166952e-05,
+      "loss": 0.4975,
+      "step": 207
+    },
+    {
+      "epoch": 0.04100946372239748,
+      "grad_norm": 1.218002669237388,
+      "learning_rate": 1.9995754460123445e-05,
+      "loss": 0.4877,
+      "step": 208
+    },
+    {
+      "epoch": 0.041206624605678234,
+      "grad_norm": 1.621757149431934,
+      "learning_rate": 1.999570917792389e-05,
+      "loss": 0.487,
+      "step": 209
+    },
+    {
+      "epoch": 0.04140378548895899,
+      "grad_norm": 1.0341497665720387,
+      "learning_rate": 1.999566365556937e-05,
+      "loss": 0.5016,
+      "step": 210
+    },
+    {
+      "epoch": 0.041600946372239746,
+      "grad_norm": 0.882195008783823,
+      "learning_rate": 1.9995617893060984e-05,
+      "loss": 0.4523,
+      "step": 211
+    },
+    {
+      "epoch": 0.0417981072555205,
+      "grad_norm": 0.9026396173054654,
+      "learning_rate": 1.9995571890399827e-05,
+      "loss": 0.4561,
+      "step": 212
+    },
+    {
+      "epoch": 0.041995268138801264,
+      "grad_norm": 1.2144739906269006,
+      "learning_rate": 1.9995525647587005e-05,
+      "loss": 0.5047,
+      "step": 213
+    },
+    {
+      "epoch": 0.04219242902208202,
+      "grad_norm": 0.975053328028896,
+      "learning_rate": 1.9995479164623633e-05,
+      "loss": 0.4736,
+      "step": 214
+    },
+    {
+      "epoch": 0.042389589905362776,
+      "grad_norm": 0.9231877842298208,
+      "learning_rate": 1.9995432441510824e-05,
+      "loss": 0.4898,
+      "step": 215
+    },
+    {
+      "epoch": 0.04258675078864353,
+      "grad_norm": 0.9144587574515857,
+      "learning_rate": 1.9995385478249697e-05,
+      "loss": 0.4796,
+      "step": 216
+    },
+    {
+      "epoch": 0.04278391167192429,
+      "grad_norm": 17.62666522642842,
+      "learning_rate": 1.999533827484139e-05,
+      "loss": 0.6332,
+      "step": 217
+    },
+    {
+      "epoch": 0.04298107255520505,
+      "grad_norm": 1.471356897407039,
+      "learning_rate": 1.9995290831287032e-05,
+      "loss": 0.4722,
+      "step": 218
+    },
+    {
+      "epoch": 0.04317823343848581,
+      "grad_norm": 11.254738838618625,
+      "learning_rate": 1.9995243147587758e-05,
+      "loss": 0.5691,
+      "step": 219
+    },
+    {
+      "epoch": 0.04337539432176656,
+      "grad_norm": 2.055362498950427,
+      "learning_rate": 1.999519522374472e-05,
+      "loss": 0.4621,
+      "step": 220
+    },
+    {
+      "epoch": 0.04357255520504732,
+      "grad_norm": 8.194523341432308,
+      "learning_rate": 1.999514705975907e-05,
+      "loss": 0.4652,
+      "step": 221
+    },
+    {
+      "epoch": 0.043769716088328074,
+      "grad_norm": 1.9182748476507985,
+      "learning_rate": 1.9995098655631957e-05,
+      "loss": 0.5171,
+      "step": 222
+    },
+    {
+      "epoch": 0.04396687697160883,
+      "grad_norm": 1.0480239186671922,
+      "learning_rate": 1.9995050011364557e-05,
+      "loss": 0.4751,
+      "step": 223
+    },
+    {
+      "epoch": 0.04416403785488959,
+      "grad_norm": 1.665247357622267,
+      "learning_rate": 1.9995001126958025e-05,
+      "loss": 0.4874,
+      "step": 224
+    },
+    {
+      "epoch": 0.04436119873817035,
+      "grad_norm": 1.2047598455545891,
+      "learning_rate": 1.999495200241355e-05,
+      "loss": 0.4712,
+      "step": 225
+    },
+    {
+      "epoch": 0.044558359621451105,
+      "grad_norm": 1.4064682020754653,
+      "learning_rate": 1.9994902637732295e-05,
+      "loss": 0.4641,
+      "step": 226
+    },
+    {
+      "epoch": 0.04475552050473186,
+      "grad_norm": 1.354286339165095,
+      "learning_rate": 1.999485303291546e-05,
+      "loss": 0.4793,
+      "step": 227
+    },
+    {
+      "epoch": 0.044952681388012616,
+      "grad_norm": 1.5121544057772933,
+      "learning_rate": 1.9994803187964233e-05,
+      "loss": 0.5025,
+      "step": 228
+    },
+    {
+      "epoch": 0.04514984227129337,
+      "grad_norm": 1.0667226038238984,
+      "learning_rate": 1.9994753102879807e-05,
+      "loss": 0.4352,
+      "step": 229
+    },
+    {
+      "epoch": 0.045347003154574135,
+      "grad_norm": 2.1283935213841714,
+      "learning_rate": 1.999470277766339e-05,
+      "loss": 0.4862,
+      "step": 230
+    },
+    {
+      "epoch": 0.04554416403785489,
+      "grad_norm": 0.9323160156698969,
+      "learning_rate": 1.9994652212316193e-05,
+      "loss": 0.4586,
+      "step": 231
+    },
+    {
+      "epoch": 0.04574132492113565,
+      "grad_norm": 1.2774644395318477,
+      "learning_rate": 1.9994601406839428e-05,
+      "loss": 0.5293,
+      "step": 232
+    },
+    {
+      "epoch": 0.0459384858044164,
+      "grad_norm": 0.8563504697504617,
+      "learning_rate": 1.9994550361234314e-05,
+      "loss": 0.4671,
+      "step": 233
+    },
+    {
+      "epoch": 0.04613564668769716,
+      "grad_norm": 1.2981705612870134,
+      "learning_rate": 1.9994499075502078e-05,
+      "loss": 0.4774,
+      "step": 234
+    },
+    {
+      "epoch": 0.04633280757097792,
+      "grad_norm": 0.9650905057900812,
+      "learning_rate": 1.999444754964395e-05,
+      "loss": 0.4785,
+      "step": 235
+    },
+    {
+      "epoch": 0.04652996845425868,
+      "grad_norm": 0.9892482279995838,
+      "learning_rate": 1.9994395783661177e-05,
+      "loss": 0.478,
+      "step": 236
+    },
+    {
+      "epoch": 0.04672712933753943,
+      "grad_norm": 1.0372935377294314,
+      "learning_rate": 1.9994343777554995e-05,
+      "loss": 0.4965,
+      "step": 237
+    },
+    {
+      "epoch": 0.04692429022082019,
+      "grad_norm": 1.2079564548732702,
+      "learning_rate": 1.9994291531326656e-05,
+      "loss": 0.4444,
+      "step": 238
+    },
+    {
+      "epoch": 0.047121451104100945,
+      "grad_norm": 0.9768363136576569,
+      "learning_rate": 1.999423904497741e-05,
+      "loss": 0.4985,
+      "step": 239
+    },
+    {
+      "epoch": 0.0473186119873817,
+      "grad_norm": 1.7614504439224654,
+      "learning_rate": 1.999418631850853e-05,
+      "loss": 0.4653,
+      "step": 240
+    },
+    {
+      "epoch": 0.047515772870662464,
+      "grad_norm": 1.1032925804972276,
+      "learning_rate": 1.9994133351921274e-05,
+      "loss": 0.4659,
+      "step": 241
+    },
+    {
+      "epoch": 0.04771293375394322,
+      "grad_norm": 1.2804412541262307,
+      "learning_rate": 1.9994080145216908e-05,
+      "loss": 0.4483,
+      "step": 242
+    },
+    {
+      "epoch": 0.047910094637223975,
+      "grad_norm": 2.294484377346188,
+      "learning_rate": 1.9994026698396727e-05,
+      "loss": 0.4887,
+      "step": 243
+    },
+    {
+      "epoch": 0.04810725552050473,
+      "grad_norm": 1.118613544316812,
+      "learning_rate": 1.9993973011462004e-05,
+      "loss": 0.4775,
+      "step": 244
+    },
+    {
+      "epoch": 0.04830441640378549,
+      "grad_norm": 0.8403959866676638,
+      "learning_rate": 1.999391908441403e-05,
+      "loss": 0.4442,
+      "step": 245
+    },
+    {
+      "epoch": 0.04850157728706624,
+      "grad_norm": 1.0300870576969843,
+      "learning_rate": 1.9993864917254103e-05,
+      "loss": 0.4462,
+      "step": 246
+    },
+    {
+      "epoch": 0.048698738170347006,
+      "grad_norm": 1.1077983845276296,
+      "learning_rate": 1.9993810509983524e-05,
+      "loss": 0.4789,
+      "step": 247
+    },
+    {
+      "epoch": 0.04889589905362776,
+      "grad_norm": 0.9138468109346035,
+      "learning_rate": 1.9993755862603597e-05,
+      "loss": 0.4638,
+      "step": 248
+    },
+    {
+      "epoch": 0.04909305993690852,
+      "grad_norm": 0.9955167957819017,
+      "learning_rate": 1.9993700975115636e-05,
+      "loss": 0.5144,
+      "step": 249
+    },
+    {
+      "epoch": 0.04929022082018927,
+      "grad_norm": 0.9520927314114271,
+      "learning_rate": 1.9993645847520965e-05,
+      "loss": 0.4976,
+      "step": 250
+    },
+    {
+      "epoch": 0.04948738170347003,
+      "grad_norm": 1.2526305722815652,
+      "learning_rate": 1.9993590479820906e-05,
+      "loss": 0.5092,
+      "step": 251
+    },
+    {
+      "epoch": 0.04968454258675079,
+      "grad_norm": 0.8493530635842205,
+      "learning_rate": 1.9993534872016784e-05,
+      "loss": 0.4517,
+      "step": 252
+    },
+    {
+      "epoch": 0.04988170347003155,
+      "grad_norm": 0.9146559563429385,
+      "learning_rate": 1.999347902410994e-05,
+      "loss": 0.5143,
+      "step": 253
+    },
+    {
+      "epoch": 0.050078864353312304,
+      "grad_norm": 1.2331946908319966,
+      "learning_rate": 1.9993422936101715e-05,
+      "loss": 0.5273,
+      "step": 254
+    },
+    {
+      "epoch": 0.05027602523659306,
+      "grad_norm": 0.9053104001937419,
+      "learning_rate": 1.9993366607993457e-05,
+      "loss": 0.4623,
+      "step": 255
+    },
+    {
+      "epoch": 0.050473186119873815,
+      "grad_norm": 1.0114462146919174,
+      "learning_rate": 1.999331003978652e-05,
+      "loss": 0.484,
+      "step": 256
+    },
+    {
+      "epoch": 0.05067034700315457,
+      "grad_norm": 0.8468019047992964,
+      "learning_rate": 1.9993253231482258e-05,
+      "loss": 0.477,
+      "step": 257
+    },
+    {
+      "epoch": 0.050867507886435334,
+      "grad_norm": 0.9113246734228752,
+      "learning_rate": 1.999319618308204e-05,
+      "loss": 0.4913,
+      "step": 258
+    },
+    {
+      "epoch": 0.05106466876971609,
+      "grad_norm": 0.8740898510256843,
+      "learning_rate": 1.999313889458724e-05,
+      "loss": 0.504,
+      "step": 259
+    },
+    {
+      "epoch": 0.051261829652996846,
+      "grad_norm": 0.8543091738206727,
+      "learning_rate": 1.9993081365999228e-05,
+      "loss": 0.4752,
+      "step": 260
+    },
+    {
+      "epoch": 0.0514589905362776,
+      "grad_norm": 0.8447390016252053,
+      "learning_rate": 1.9993023597319387e-05,
+      "loss": 0.4582,
+      "step": 261
+    },
+    {
+      "epoch": 0.05165615141955836,
+      "grad_norm": 0.8559956935990739,
+      "learning_rate": 1.999296558854911e-05,
+      "loss": 0.4567,
+      "step": 262
+    },
+    {
+      "epoch": 0.051853312302839114,
+      "grad_norm": 0.8497707119769089,
+      "learning_rate": 1.9992907339689786e-05,
+      "loss": 0.4681,
+      "step": 263
+    },
+    {
+      "epoch": 0.052050473186119876,
+      "grad_norm": 1.2959384495585367,
+      "learning_rate": 1.9992848850742817e-05,
+      "loss": 0.4612,
+      "step": 264
+    },
+    {
+      "epoch": 0.05224763406940063,
+      "grad_norm": 0.9252677800758691,
+      "learning_rate": 1.9992790121709604e-05,
+      "loss": 0.4969,
+      "step": 265
+    },
+    {
+      "epoch": 0.05244479495268139,
+      "grad_norm": 0.9986026073347362,
+      "learning_rate": 1.9992731152591563e-05,
+      "loss": 0.483,
+      "step": 266
+    },
+    {
+      "epoch": 0.052641955835962144,
+      "grad_norm": 0.8766211673811323,
+      "learning_rate": 1.999267194339011e-05,
+      "loss": 0.4952,
+      "step": 267
+    },
+    {
+      "epoch": 0.0528391167192429,
+      "grad_norm": 0.9062026471143431,
+      "learning_rate": 1.9992612494106666e-05,
+      "loss": 0.4922,
+      "step": 268
+    },
+    {
+      "epoch": 0.05303627760252366,
+      "grad_norm": 1.2051311308802126,
+      "learning_rate": 1.999255280474266e-05,
+      "loss": 0.4528,
+      "step": 269
+    },
+    {
+      "epoch": 0.05323343848580442,
+      "grad_norm": 0.9481095486211566,
+      "learning_rate": 1.9992492875299528e-05,
+      "loss": 0.4676,
+      "step": 270
+    },
+    {
+      "epoch": 0.053430599369085174,
+      "grad_norm": 1.0581042227575974,
+      "learning_rate": 1.9992432705778703e-05,
+      "loss": 0.5275,
+      "step": 271
+    },
+    {
+      "epoch": 0.05362776025236593,
+      "grad_norm": 1.0862738326883588,
+      "learning_rate": 1.9992372296181637e-05,
+      "loss": 0.4802,
+      "step": 272
+    },
+    {
+      "epoch": 0.053824921135646686,
+      "grad_norm": 1.1530116773068577,
+      "learning_rate": 1.999231164650978e-05,
+      "loss": 0.4963,
+      "step": 273
+    },
+    {
+      "epoch": 0.05402208201892744,
+      "grad_norm": 1.1406385581126395,
+      "learning_rate": 1.999225075676459e-05,
+      "loss": 0.5017,
+      "step": 274
+    },
+    {
+      "epoch": 0.054219242902208205,
+      "grad_norm": 0.9840133715884154,
+      "learning_rate": 1.9992189626947528e-05,
+      "loss": 0.4184,
+      "step": 275
+    },
+    {
+      "epoch": 0.05441640378548896,
+      "grad_norm": 0.8369535854434705,
+      "learning_rate": 1.9992128257060064e-05,
+      "loss": 0.4766,
+      "step": 276
+    },
+    {
+      "epoch": 0.05461356466876972,
+      "grad_norm": 0.9733589505757138,
+      "learning_rate": 1.999206664710367e-05,
+      "loss": 0.4783,
+      "step": 277
+    },
+    {
+      "epoch": 0.05481072555205047,
+      "grad_norm": 1.1016083321043935,
+      "learning_rate": 1.999200479707983e-05,
+      "loss": 0.5023,
+      "step": 278
+    },
+    {
+      "epoch": 0.05500788643533123,
+      "grad_norm": 0.9453559986774039,
+      "learning_rate": 1.9991942706990028e-05,
+      "loss": 0.4997,
+      "step": 279
+    },
+    {
+      "epoch": 0.055205047318611984,
+      "grad_norm": 1.1205398036178456,
+      "learning_rate": 1.999188037683576e-05,
+      "loss": 0.4918,
+      "step": 280
+    },
+    {
+      "epoch": 0.05540220820189275,
+      "grad_norm": 0.8547868381519064,
+      "learning_rate": 1.9991817806618512e-05,
+      "loss": 0.4691,
+      "step": 281
+    },
+    {
+      "epoch": 0.0555993690851735,
+      "grad_norm": 1.101644837851159,
+      "learning_rate": 1.99917549963398e-05,
+      "loss": 0.478,
+      "step": 282
+    },
+    {
+      "epoch": 0.05579652996845426,
+      "grad_norm": 0.9648045124396167,
+      "learning_rate": 1.9991691946001123e-05,
+      "loss": 0.451,
+      "step": 283
+    },
+    {
+      "epoch": 0.055993690851735015,
+      "grad_norm": 0.9890972584569301,
+      "learning_rate": 1.9991628655604006e-05,
+      "loss": 0.4688,
+      "step": 284
+    },
+    {
+      "epoch": 0.05619085173501577,
+      "grad_norm": 1.087262812823818,
+      "learning_rate": 1.999156512514996e-05,
+      "loss": 0.4974,
+      "step": 285
+    },
+    {
+      "epoch": 0.05638801261829653,
+      "grad_norm": 1.1458887553291872,
+      "learning_rate": 1.9991501354640517e-05,
+      "loss": 0.4982,
+      "step": 286
+    },
+    {
+      "epoch": 0.05658517350157729,
+      "grad_norm": 0.8822768555579248,
+      "learning_rate": 1.9991437344077212e-05,
+      "loss": 0.4486,
+      "step": 287
+    },
+    {
+      "epoch": 0.056782334384858045,
+      "grad_norm": 1.2313826346986292,
+      "learning_rate": 1.9991373093461574e-05,
+      "loss": 0.4873,
+      "step": 288
+    },
+    {
+      "epoch": 0.0569794952681388,
+      "grad_norm": 0.8882605229939141,
+      "learning_rate": 1.9991308602795156e-05,
+      "loss": 0.4553,
+      "step": 289
+    },
+    {
+      "epoch": 0.05717665615141956,
+      "grad_norm": 1.0115742101244216,
+      "learning_rate": 1.9991243872079495e-05,
+      "loss": 0.4561,
+      "step": 290
+    },
+    {
+      "epoch": 0.05737381703470031,
+      "grad_norm": 0.9908774101006372,
+      "learning_rate": 1.999117890131616e-05,
+      "loss": 0.4427,
+      "step": 291
+    },
+    {
+      "epoch": 0.057570977917981075,
+      "grad_norm": 0.8561287541214222,
+      "learning_rate": 1.9991113690506705e-05,
+      "loss": 0.4268,
+      "step": 292
+    },
+    {
+      "epoch": 0.05776813880126183,
+      "grad_norm": 1.0671041225310995,
+      "learning_rate": 1.99910482396527e-05,
+      "loss": 0.48,
+      "step": 293
+    },
+    {
+      "epoch": 0.05796529968454259,
+      "grad_norm": 0.9685606768815737,
+      "learning_rate": 1.9990982548755712e-05,
+      "loss": 0.5016,
+      "step": 294
+    },
+    {
+      "epoch": 0.05816246056782334,
+      "grad_norm": 0.9888136254935049,
+      "learning_rate": 1.9990916617817323e-05,
+      "loss": 0.5133,
+      "step": 295
+    },
+    {
+      "epoch": 0.0583596214511041,
+      "grad_norm": 0.9279140857322566,
+      "learning_rate": 1.9990850446839114e-05,
+      "loss": 0.4645,
+      "step": 296
+    },
+    {
+      "epoch": 0.058556782334384855,
+      "grad_norm": 0.9503914334506955,
+      "learning_rate": 1.999078403582268e-05,
+      "loss": 0.4442,
+      "step": 297
+    },
+    {
+      "epoch": 0.05875394321766562,
+      "grad_norm": 0.8179461934678155,
+      "learning_rate": 1.9990717384769617e-05,
+      "loss": 0.4266,
+      "step": 298
+    },
+    {
+      "epoch": 0.058951104100946373,
+      "grad_norm": 0.9592664796746861,
+      "learning_rate": 1.9990650493681517e-05,
+      "loss": 0.4843,
+      "step": 299
+    },
+    {
+      "epoch": 0.05914826498422713,
+      "grad_norm": 0.9196227211585776,
+      "learning_rate": 1.999058336256e-05,
+      "loss": 0.4678,
+      "step": 300
+    },
+    {
+      "epoch": 0.059345425867507885,
+      "grad_norm": 1.2796468769464633,
+      "learning_rate": 1.9990515991406666e-05,
+      "loss": 0.4968,
+      "step": 301
+    },
+    {
+      "epoch": 0.05954258675078864,
+      "grad_norm": 0.9210540437615927,
+      "learning_rate": 1.9990448380223145e-05,
+      "loss": 0.4315,
+      "step": 302
+    },
+    {
+      "epoch": 0.059739747634069404,
+      "grad_norm": 0.9150812147912519,
+      "learning_rate": 1.9990380529011056e-05,
+      "loss": 0.4992,
+      "step": 303
+    },
+    {
+      "epoch": 0.05993690851735016,
+      "grad_norm": 1.0184465463567445,
+      "learning_rate": 1.9990312437772025e-05,
+      "loss": 0.5149,
+      "step": 304
+    },
+    {
+      "epoch": 0.060134069400630916,
+      "grad_norm": 0.909544636631595,
+      "learning_rate": 1.99902441065077e-05,
+      "loss": 0.4802,
+      "step": 305
+    },
+    {
+      "epoch": 0.06033123028391167,
+      "grad_norm": 1.0603662944162957,
+      "learning_rate": 1.9990175535219708e-05,
+      "loss": 0.4475,
+      "step": 306
+    },
+    {
+      "epoch": 0.06052839116719243,
+      "grad_norm": 1.241117709798254,
+      "learning_rate": 1.999010672390971e-05,
+      "loss": 0.477,
+      "step": 307
+    },
+    {
+      "epoch": 0.06072555205047318,
+      "grad_norm": 1.035832208551567,
+      "learning_rate": 1.9990037672579347e-05,
+      "loss": 0.4822,
+      "step": 308
+    },
+    {
+      "epoch": 0.060922712933753946,
+      "grad_norm": 0.9408863869538917,
+      "learning_rate": 1.9989968381230288e-05,
+      "loss": 0.4862,
+      "step": 309
+    },
+    {
+      "epoch": 0.0611198738170347,
+      "grad_norm": 0.9296848076140912,
+      "learning_rate": 1.998989884986419e-05,
+      "loss": 0.5125,
+      "step": 310
+    },
+    {
+      "epoch": 0.06131703470031546,
+      "grad_norm": 0.8918476259653481,
+      "learning_rate": 1.998982907848273e-05,
+      "loss": 0.4937,
+      "step": 311
+    },
+    {
+      "epoch": 0.061514195583596214,
+      "grad_norm": 1.4848687664773468,
+      "learning_rate": 1.9989759067087582e-05,
+      "loss": 0.4865,
+      "step": 312
+    },
+    {
+      "epoch": 0.06171135646687697,
+      "grad_norm": 0.8732760816687123,
+      "learning_rate": 1.998968881568043e-05,
+      "loss": 0.4778,
+      "step": 313
+    },
+    {
+      "epoch": 0.061908517350157725,
+      "grad_norm": 1.1341625184193884,
+      "learning_rate": 1.998961832426295e-05,
+      "loss": 0.4798,
+      "step": 314
+    },
+    {
+      "epoch": 0.06210567823343849,
+      "grad_norm": 0.8993091141038897,
+      "learning_rate": 1.9989547592836853e-05,
+      "loss": 0.4292,
+      "step": 315
+    },
+    {
+      "epoch": 0.062302839116719244,
+      "grad_norm": 0.8850853933714552,
+      "learning_rate": 1.998947662140383e-05,
+      "loss": 0.4594,
+      "step": 316
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 1.1127472937800988,
+      "learning_rate": 1.9989405409965585e-05,
+      "loss": 0.4943,
+      "step": 317
+    },
+    {
+      "epoch": 0.06269716088328076,
+      "grad_norm": 0.866897163477894,
+      "learning_rate": 1.998933395852383e-05,
+      "loss": 0.4633,
+      "step": 318
+    },
+    {
+      "epoch": 0.06289432176656151,
+      "grad_norm": 0.9547891953542145,
+      "learning_rate": 1.9989262267080284e-05,
+      "loss": 0.5155,
+      "step": 319
+    },
+    {
+      "epoch": 0.06309148264984227,
+      "grad_norm": 5.818229069199477,
+      "learning_rate": 1.998919033563667e-05,
+      "loss": 0.5422,
+      "step": 320
+    },
+    {
+      "epoch": 0.06328864353312302,
+      "grad_norm": 1.5240274072221687,
+      "learning_rate": 1.998911816419471e-05,
+      "loss": 0.4835,
+      "step": 321
+    },
+    {
+      "epoch": 0.06348580441640378,
+      "grad_norm": 0.9717985560653184,
+      "learning_rate": 1.9989045752756145e-05,
+      "loss": 0.4848,
+      "step": 322
+    },
+    {
+      "epoch": 0.06368296529968455,
+      "grad_norm": 1.1194941379469816,
+      "learning_rate": 1.998897310132271e-05,
+      "loss": 0.465,
+      "step": 323
+    },
+    {
+      "epoch": 0.0638801261829653,
+      "grad_norm": 1.3015806776449268,
+      "learning_rate": 1.9988900209896148e-05,
+      "loss": 0.5076,
+      "step": 324
+    },
+    {
+      "epoch": 0.06407728706624606,
+      "grad_norm": 1.3495583310297246,
+      "learning_rate": 1.998882707847822e-05,
+      "loss": 0.4891,
+      "step": 325
+    },
+    {
+      "epoch": 0.06427444794952682,
+      "grad_norm": 1.094314000355448,
+      "learning_rate": 1.9988753707070675e-05,
+      "loss": 0.4603,
+      "step": 326
+    },
+    {
+      "epoch": 0.06447160883280757,
+      "grad_norm": 0.8976180746045599,
+      "learning_rate": 1.998868009567528e-05,
+      "loss": 0.4852,
+      "step": 327
+    },
+    {
+      "epoch": 0.06466876971608833,
+      "grad_norm": 1.176456935066941,
+      "learning_rate": 1.99886062442938e-05,
+      "loss": 0.4505,
+      "step": 328
+    },
+    {
+      "epoch": 0.06486593059936908,
+      "grad_norm": 0.8621831797378902,
+      "learning_rate": 1.9988532152928012e-05,
+      "loss": 0.4521,
+      "step": 329
+    },
+    {
+      "epoch": 0.06506309148264984,
+      "grad_norm": 1.1751922853480894,
+      "learning_rate": 1.9988457821579698e-05,
+      "loss": 0.4684,
+      "step": 330
+    },
+    {
+      "epoch": 0.0652602523659306,
+      "grad_norm": 1.214299413735334,
+      "learning_rate": 1.9988383250250636e-05,
+      "loss": 0.5136,
+      "step": 331
+    },
+    {
+      "epoch": 0.06545741324921135,
+      "grad_norm": 1.3935797324505006,
+      "learning_rate": 1.9988308438942626e-05,
+      "loss": 0.543,
+      "step": 332
+    },
+    {
+      "epoch": 0.06565457413249211,
+      "grad_norm": 1.3113325461742111,
+      "learning_rate": 1.9988233387657462e-05,
+      "loss": 0.469,
+      "step": 333
+    },
+    {
+      "epoch": 0.06585173501577288,
+      "grad_norm": 0.9832962911329318,
+      "learning_rate": 1.9988158096396945e-05,
+      "loss": 0.4977,
+      "step": 334
+    },
+    {
+      "epoch": 0.06604889589905363,
+      "grad_norm": 1.4696155006597584,
+      "learning_rate": 1.998808256516289e-05,
+      "loss": 0.5422,
+      "step": 335
+    },
+    {
+      "epoch": 0.06624605678233439,
+      "grad_norm": 1.1348175666726867,
+      "learning_rate": 1.9988006793957106e-05,
+      "loss": 0.4752,
+      "step": 336
+    },
+    {
+      "epoch": 0.06644321766561515,
+      "grad_norm": 1.5792191486101228,
+      "learning_rate": 1.9987930782781416e-05,
+      "loss": 0.5261,
+      "step": 337
+    },
+    {
+      "epoch": 0.0666403785488959,
+      "grad_norm": 1.0368934464451127,
+      "learning_rate": 1.9987854531637644e-05,
+      "loss": 0.4642,
+      "step": 338
+    },
+    {
+      "epoch": 0.06683753943217666,
+      "grad_norm": 1.0337884912885194,
+      "learning_rate": 1.9987778040527625e-05,
+      "loss": 0.4755,
+      "step": 339
+    },
+    {
+      "epoch": 0.06703470031545741,
+      "grad_norm": 1.133861117613178,
+      "learning_rate": 1.9987701309453195e-05,
+      "loss": 0.4828,
+      "step": 340
+    },
+    {
+      "epoch": 0.06723186119873817,
+      "grad_norm": 1.111075907788898,
+      "learning_rate": 1.99876243384162e-05,
+      "loss": 0.4941,
+      "step": 341
+    },
+    {
+      "epoch": 0.06742902208201892,
+      "grad_norm": 1.2284579964954567,
+      "learning_rate": 1.9987547127418485e-05,
+      "loss": 0.4999,
+      "step": 342
+    },
+    {
+      "epoch": 0.06762618296529968,
+      "grad_norm": 1.0071834128462238,
+      "learning_rate": 1.9987469676461904e-05,
+      "loss": 0.4132,
+      "step": 343
+    },
+    {
+      "epoch": 0.06782334384858044,
+      "grad_norm": 1.1346939479337024,
+      "learning_rate": 1.9987391985548326e-05,
+      "loss": 0.5326,
+      "step": 344
+    },
+    {
+      "epoch": 0.06802050473186119,
+      "grad_norm": 0.8817839707262805,
+      "learning_rate": 1.9987314054679615e-05,
+      "loss": 0.4856,
+      "step": 345
+    },
+    {
+      "epoch": 0.06821766561514196,
+      "grad_norm": 0.967996798695509,
+      "learning_rate": 1.998723588385764e-05,
+      "loss": 0.5015,
+      "step": 346
+    },
+    {
+      "epoch": 0.06841482649842272,
+      "grad_norm": 1.2535242900070593,
+      "learning_rate": 1.9987157473084276e-05,
+      "loss": 0.4721,
+      "step": 347
+    },
+    {
+      "epoch": 0.06861198738170347,
+      "grad_norm": 0.9971572384429465,
+      "learning_rate": 1.9987078822361412e-05,
+      "loss": 0.4866,
+      "step": 348
+    },
+    {
+      "epoch": 0.06880914826498423,
+      "grad_norm": 0.9880019131723238,
+      "learning_rate": 1.9986999931690937e-05,
+      "loss": 0.4217,
+      "step": 349
+    },
+    {
+      "epoch": 0.06900630914826499,
+      "grad_norm": 0.8510336475538133,
+      "learning_rate": 1.9986920801074747e-05,
+      "loss": 0.4997,
+      "step": 350
+    },
+    {
+      "epoch": 0.06920347003154574,
+      "grad_norm": 1.3697939147272762,
+      "learning_rate": 1.9986841430514743e-05,
+      "loss": 0.4791,
+      "step": 351
+    },
+    {
+      "epoch": 0.0694006309148265,
+      "grad_norm": 0.9037413372344649,
+      "learning_rate": 1.9986761820012833e-05,
+      "loss": 0.4642,
+      "step": 352
+    },
+    {
+      "epoch": 0.06959779179810725,
+      "grad_norm": 1.1537513355491777,
+      "learning_rate": 1.9986681969570924e-05,
+      "loss": 0.4693,
+      "step": 353
+    },
+    {
+      "epoch": 0.06979495268138801,
+      "grad_norm": 0.7969873717195902,
+      "learning_rate": 1.9986601879190938e-05,
+      "loss": 0.4309,
+      "step": 354
+    },
+    {
+      "epoch": 0.06999211356466876,
+      "grad_norm": 1.0506442175587962,
+      "learning_rate": 1.9986521548874802e-05,
+      "loss": 0.4575,
+      "step": 355
+    },
+    {
+      "epoch": 0.07018927444794952,
+      "grad_norm": 0.958839633619317,
+      "learning_rate": 1.9986440978624444e-05,
+      "loss": 0.4848,
+      "step": 356
+    },
+    {
+      "epoch": 0.07038643533123029,
+      "grad_norm": 0.8487815542536234,
+      "learning_rate": 1.9986360168441798e-05,
+      "loss": 0.4346,
+      "step": 357
+    },
+    {
+      "epoch": 0.07058359621451105,
+      "grad_norm": 0.9101190136837835,
+      "learning_rate": 1.998627911832881e-05,
+      "loss": 0.4798,
+      "step": 358
+    },
+    {
+      "epoch": 0.0707807570977918,
+      "grad_norm": 0.8359817460302178,
+      "learning_rate": 1.998619782828742e-05,
+      "loss": 0.4457,
+      "step": 359
+    },
+    {
+      "epoch": 0.07097791798107256,
+      "grad_norm": 0.9727976821516041,
+      "learning_rate": 1.9986116298319585e-05,
+      "loss": 0.5337,
+      "step": 360
+    },
+    {
+      "epoch": 0.07117507886435331,
+      "grad_norm": 0.7658661529985589,
+      "learning_rate": 1.998603452842727e-05,
+      "loss": 0.4475,
+      "step": 361
+    },
+    {
+      "epoch": 0.07137223974763407,
+      "grad_norm": 0.9267521801472242,
+      "learning_rate": 1.998595251861243e-05,
+      "loss": 0.5341,
+      "step": 362
+    },
+    {
+      "epoch": 0.07156940063091483,
+      "grad_norm": 1.5288633084722636,
+      "learning_rate": 1.998587026887704e-05,
+      "loss": 0.4811,
+      "step": 363
+    },
+    {
+      "epoch": 0.07176656151419558,
+      "grad_norm": 1.1549526328305526,
+      "learning_rate": 1.9985787779223073e-05,
+      "loss": 0.4785,
+      "step": 364
+    },
+    {
+      "epoch": 0.07196372239747634,
+      "grad_norm": 1.0840635546798056,
+      "learning_rate": 1.9985705049652513e-05,
+      "loss": 0.499,
+      "step": 365
+    },
+    {
+      "epoch": 0.0721608832807571,
+      "grad_norm": 0.779747962979529,
+      "learning_rate": 1.998562208016735e-05,
+      "loss": 0.4663,
+      "step": 366
+    },
+    {
+      "epoch": 0.07235804416403785,
+      "grad_norm": 1.1129287322836847,
+      "learning_rate": 1.9985538870769573e-05,
+      "loss": 0.4727,
+      "step": 367
+    },
+    {
+      "epoch": 0.07255520504731862,
+      "grad_norm": 0.853670529824468,
+      "learning_rate": 1.9985455421461183e-05,
+      "loss": 0.4947,
+      "step": 368
+    },
+    {
+      "epoch": 0.07275236593059937,
+      "grad_norm": 0.924674407680012,
+      "learning_rate": 1.9985371732244188e-05,
+      "loss": 0.4584,
+      "step": 369
+    },
+    {
+      "epoch": 0.07294952681388013,
+      "grad_norm": 0.8629571748054644,
+      "learning_rate": 1.9985287803120595e-05,
+      "loss": 0.4705,
+      "step": 370
+    },
+    {
+      "epoch": 0.07314668769716089,
+      "grad_norm": 0.8242683110246124,
+      "learning_rate": 1.998520363409242e-05,
+      "loss": 0.4579,
+      "step": 371
+    },
+    {
+      "epoch": 0.07334384858044164,
+      "grad_norm": 1.499462700491384,
+      "learning_rate": 1.9985119225161688e-05,
+      "loss": 0.4607,
+      "step": 372
+    },
+    {
+      "epoch": 0.0735410094637224,
+      "grad_norm": 0.9669915502441957,
+      "learning_rate": 1.9985034576330425e-05,
+      "loss": 0.4956,
+      "step": 373
+    },
+    {
+      "epoch": 0.07373817034700315,
+      "grad_norm": 1.0411892472276396,
+      "learning_rate": 1.998494968760067e-05,
+      "loss": 0.4557,
+      "step": 374
+    },
+    {
+      "epoch": 0.07393533123028391,
+      "grad_norm": 0.8567052653582087,
+      "learning_rate": 1.998486455897445e-05,
+      "loss": 0.5177,
+      "step": 375
+    },
+    {
+      "epoch": 0.07413249211356467,
+      "grad_norm": 1.1116739095705466,
+      "learning_rate": 1.998477919045382e-05,
+      "loss": 0.4716,
+      "step": 376
+    },
+    {
+      "epoch": 0.07432965299684542,
+      "grad_norm": 0.8626595222872029,
+      "learning_rate": 1.9984693582040834e-05,
+      "loss": 0.5426,
+      "step": 377
+    },
+    {
+      "epoch": 0.07452681388012618,
+      "grad_norm": 0.9587022633462172,
+      "learning_rate": 1.998460773373754e-05,
+      "loss": 0.462,
+      "step": 378
+    },
+    {
+      "epoch": 0.07472397476340693,
+      "grad_norm": 0.8941067614117472,
+      "learning_rate": 1.9984521645546007e-05,
+      "loss": 0.4989,
+      "step": 379
+    },
+    {
+      "epoch": 0.0749211356466877,
+      "grad_norm": 0.9295392099689723,
+      "learning_rate": 1.9984435317468298e-05,
+      "loss": 0.4745,
+      "step": 380
+    },
+    {
+      "epoch": 0.07511829652996846,
+      "grad_norm": 0.8793262590604104,
+      "learning_rate": 1.998434874950649e-05,
+      "loss": 0.4847,
+      "step": 381
+    },
+    {
+      "epoch": 0.07531545741324921,
+      "grad_norm": 0.9197548867944333,
+      "learning_rate": 1.9984261941662666e-05,
+      "loss": 0.4822,
+      "step": 382
+    },
+    {
+      "epoch": 0.07551261829652997,
+      "grad_norm": 0.918094619940462,
+      "learning_rate": 1.9984174893938908e-05,
+      "loss": 0.473,
+      "step": 383
+    },
+    {
+      "epoch": 0.07570977917981073,
+      "grad_norm": 0.834295051365838,
+      "learning_rate": 1.998408760633731e-05,
+      "loss": 0.4423,
+      "step": 384
+    },
+    {
+      "epoch": 0.07590694006309148,
+      "grad_norm": 0.8187104700252339,
+      "learning_rate": 1.998400007885996e-05,
+      "loss": 0.4582,
+      "step": 385
+    },
+    {
+      "epoch": 0.07610410094637224,
+      "grad_norm": 1.0466182710170804,
+      "learning_rate": 1.9983912311508977e-05,
+      "loss": 0.4694,
+      "step": 386
+    },
+    {
+      "epoch": 0.076301261829653,
+      "grad_norm": 0.8718063081983285,
+      "learning_rate": 1.998382430428645e-05,
+      "loss": 0.476,
+      "step": 387
+    },
+    {
+      "epoch": 0.07649842271293375,
+      "grad_norm": 0.7915900428994362,
+      "learning_rate": 1.9983736057194512e-05,
+      "loss": 0.416,
+      "step": 388
+    },
+    {
+      "epoch": 0.0766955835962145,
+      "grad_norm": 0.8158965365869196,
+      "learning_rate": 1.9983647570235274e-05,
+      "loss": 0.4667,
+      "step": 389
+    },
+    {
+      "epoch": 0.07689274447949526,
+      "grad_norm": 0.7997150055795212,
+      "learning_rate": 1.9983558843410863e-05,
+      "loss": 0.4466,
+      "step": 390
+    },
+    {
+      "epoch": 0.07708990536277603,
+      "grad_norm": 0.8940610724543107,
+      "learning_rate": 1.998346987672341e-05,
+      "loss": 0.4658,
+      "step": 391
+    },
+    {
+      "epoch": 0.07728706624605679,
+      "grad_norm": 1.294715926753004,
+      "learning_rate": 1.9983380670175054e-05,
+      "loss": 0.4642,
+      "step": 392
+    },
+    {
+      "epoch": 0.07748422712933754,
+      "grad_norm": 11.425972122218342,
+      "learning_rate": 1.998329122376794e-05,
+      "loss": 0.5511,
+      "step": 393
+    },
+    {
+      "epoch": 0.0776813880126183,
+      "grad_norm": 1.1867295493168513,
+      "learning_rate": 1.998320153750421e-05,
+      "loss": 0.4462,
+      "step": 394
+    },
+    {
+      "epoch": 0.07787854889589906,
+      "grad_norm": 0.9729316139359963,
+      "learning_rate": 1.998311161138603e-05,
+      "loss": 0.4976,
+      "step": 395
+    },
+    {
+      "epoch": 0.07807570977917981,
+      "grad_norm": 0.9157344759346018,
+      "learning_rate": 1.998302144541555e-05,
+      "loss": 0.4709,
+      "step": 396
+    },
+    {
+      "epoch": 0.07827287066246057,
+      "grad_norm": 0.934161687151528,
+      "learning_rate": 1.998293103959494e-05,
+      "loss": 0.4673,
+      "step": 397
+    },
+    {
+      "epoch": 0.07847003154574132,
+      "grad_norm": 1.4223262840464979,
+      "learning_rate": 1.9982840393926374e-05,
+      "loss": 0.5139,
+      "step": 398
+    },
+    {
+      "epoch": 0.07866719242902208,
+      "grad_norm": 1.131720070151689,
+      "learning_rate": 1.9982749508412026e-05,
+      "loss": 0.5327,
+      "step": 399
+    },
+    {
+      "epoch": 0.07886435331230283,
+      "grad_norm": 1.2074643005200005,
+      "learning_rate": 1.998265838305409e-05,
+      "loss": 0.4974,
+      "step": 400
+    },
+    {
+      "epoch": 0.07906151419558359,
+      "grad_norm": 0.9531948695115436,
+      "learning_rate": 1.998256701785474e-05,
+      "loss": 0.4841,
+      "step": 401
+    },
+    {
+      "epoch": 0.07925867507886436,
+      "grad_norm": 1.043133116018871,
+      "learning_rate": 1.998247541281618e-05,
+      "loss": 0.4506,
+      "step": 402
+    },
+    {
+      "epoch": 0.07945583596214512,
+      "grad_norm": 1.0246039260585653,
+      "learning_rate": 1.9982383567940606e-05,
+      "loss": 0.4445,
+      "step": 403
+    },
+    {
+      "epoch": 0.07965299684542587,
+      "grad_norm": 1.0452483998819864,
+      "learning_rate": 1.9982291483230232e-05,
+      "loss": 0.4851,
+      "step": 404
+    },
+    {
+      "epoch": 0.07985015772870663,
+      "grad_norm": 1.1683875566798174,
+      "learning_rate": 1.9982199158687266e-05,
+      "loss": 0.4548,
+      "step": 405
+    },
+    {
+      "epoch": 0.08004731861198738,
+      "grad_norm": 0.906125687513742,
+      "learning_rate": 1.9982106594313924e-05,
+      "loss": 0.4704,
+      "step": 406
+    },
+    {
+      "epoch": 0.08024447949526814,
+      "grad_norm": 5.213779615448294,
+      "learning_rate": 1.9982013790112437e-05,
+      "loss": 0.526,
+      "step": 407
+    },
+    {
+      "epoch": 0.0804416403785489,
+      "grad_norm": 1.9791118557774463,
+      "learning_rate": 1.9981920746085025e-05,
+      "loss": 0.4971,
+      "step": 408
+    },
+    {
+      "epoch": 0.08063880126182965,
+      "grad_norm": 8.292277341865951,
+      "learning_rate": 1.9981827462233932e-05,
+      "loss": 0.4924,
+      "step": 409
+    },
+    {
+      "epoch": 0.08083596214511041,
+      "grad_norm": 1.4694955541234114,
+      "learning_rate": 1.99817339385614e-05,
+      "loss": 0.5109,
+      "step": 410
+    },
+    {
+      "epoch": 0.08103312302839116,
+      "grad_norm": 1.0200131993418244,
+      "learning_rate": 1.9981640175069663e-05,
+      "loss": 0.4837,
+      "step": 411
+    },
+    {
+      "epoch": 0.08123028391167192,
+      "grad_norm": 1.241162294871943,
+      "learning_rate": 1.998154617176099e-05,
+      "loss": 0.4752,
+      "step": 412
+    },
+    {
+      "epoch": 0.08142744479495267,
+      "grad_norm": 1.6151902756221161,
+      "learning_rate": 1.9981451928637627e-05,
+      "loss": 0.4739,
+      "step": 413
+    },
+    {
+      "epoch": 0.08162460567823344,
+      "grad_norm": 1.2407308825596557,
+      "learning_rate": 1.9981357445701846e-05,
+      "loss": 0.4345,
+      "step": 414
+    },
+    {
+      "epoch": 0.0818217665615142,
+      "grad_norm": 0.9936264286526395,
+      "learning_rate": 1.9981262722955913e-05,
+      "loss": 0.5211,
+      "step": 415
+    },
+    {
+      "epoch": 0.08201892744479496,
+      "grad_norm": 1.3047191271197602,
+      "learning_rate": 1.9981167760402104e-05,
+      "loss": 0.4827,
+      "step": 416
+    },
+    {
+      "epoch": 0.08221608832807571,
+      "grad_norm": 0.8876022805465946,
+      "learning_rate": 1.9981072558042705e-05,
+      "loss": 0.4838,
+      "step": 417
+    },
+    {
+      "epoch": 0.08241324921135647,
+      "grad_norm": 1.3442900789849086,
+      "learning_rate": 1.9980977115879997e-05,
+      "loss": 0.4614,
+      "step": 418
+    },
+    {
+      "epoch": 0.08261041009463722,
+      "grad_norm": 1.5881293239590277,
+      "learning_rate": 1.998088143391628e-05,
+      "loss": 0.4813,
+      "step": 419
+    },
+    {
+      "epoch": 0.08280757097791798,
+      "grad_norm": 1.1153506722397977,
+      "learning_rate": 1.9980785512153846e-05,
+      "loss": 0.5095,
+      "step": 420
+    },
+    {
+      "epoch": 0.08300473186119874,
+      "grad_norm": 0.9249721852533636,
+      "learning_rate": 1.9980689350595004e-05,
+      "loss": 0.4764,
+      "step": 421
+    },
+    {
+      "epoch": 0.08320189274447949,
+      "grad_norm": 1.073860170389333,
+      "learning_rate": 1.9980592949242063e-05,
+      "loss": 0.4597,
+      "step": 422
+    },
+    {
+      "epoch": 0.08339905362776025,
+      "grad_norm": 0.8354298196477534,
+      "learning_rate": 1.998049630809734e-05,
+      "loss": 0.4805,
+      "step": 423
+    },
+    {
+      "epoch": 0.083596214511041,
+      "grad_norm": 0.9975582382283317,
+      "learning_rate": 1.9980399427163154e-05,
+      "loss": 0.4734,
+      "step": 424
+    },
+    {
+      "epoch": 0.08379337539432177,
+      "grad_norm": 0.7461081355848223,
+      "learning_rate": 1.9980302306441834e-05,
+      "loss": 0.4456,
+      "step": 425
+    },
+    {
+      "epoch": 0.08399053627760253,
+      "grad_norm": 0.8992637044263566,
+      "learning_rate": 1.9980204945935716e-05,
+      "loss": 0.4514,
+      "step": 426
+    },
+    {
+      "epoch": 0.08418769716088328,
+      "grad_norm": 0.9484738986938942,
+      "learning_rate": 1.9980107345647133e-05,
+      "loss": 0.5281,
+      "step": 427
+    },
+    {
+      "epoch": 0.08438485804416404,
+      "grad_norm": 0.8453814347866732,
+      "learning_rate": 1.9980009505578438e-05,
+      "loss": 0.4493,
+      "step": 428
+    },
+    {
+      "epoch": 0.0845820189274448,
+      "grad_norm": 0.8262276270862967,
+      "learning_rate": 1.9979911425731978e-05,
+      "loss": 0.4483,
+      "step": 429
+    },
+    {
+      "epoch": 0.08477917981072555,
+      "grad_norm": 0.856137009508442,
+      "learning_rate": 1.9979813106110108e-05,
+      "loss": 0.4635,
+      "step": 430
+    },
+    {
+      "epoch": 0.08497634069400631,
+      "grad_norm": 0.8889039446695481,
+      "learning_rate": 1.997971454671519e-05,
+      "loss": 0.4308,
+      "step": 431
+    },
+    {
+      "epoch": 0.08517350157728706,
+      "grad_norm": 8.27925993626678,
+      "learning_rate": 1.9979615747549594e-05,
+      "loss": 0.4852,
+      "step": 432
+    },
+    {
+      "epoch": 0.08537066246056782,
+      "grad_norm": 2.064710279361049,
+      "learning_rate": 1.9979516708615696e-05,
+      "loss": 0.4843,
+      "step": 433
+    },
+    {
+      "epoch": 0.08556782334384858,
+      "grad_norm": 0.9477116643465063,
+      "learning_rate": 1.997941742991587e-05,
+      "loss": 0.4534,
+      "step": 434
+    },
+    {
+      "epoch": 0.08576498422712933,
+      "grad_norm": 0.9584045524018094,
+      "learning_rate": 1.9979317911452503e-05,
+      "loss": 0.4787,
+      "step": 435
+    },
+    {
+      "epoch": 0.0859621451104101,
+      "grad_norm": 0.8809961453962474,
+      "learning_rate": 1.997921815322799e-05,
+      "loss": 0.4742,
+      "step": 436
+    },
+    {
+      "epoch": 0.08615930599369086,
+      "grad_norm": 1.0555900314632443,
+      "learning_rate": 1.997911815524472e-05,
+      "loss": 0.4854,
+      "step": 437
+    },
+    {
+      "epoch": 0.08635646687697161,
+      "grad_norm": 0.8568644478491275,
+      "learning_rate": 1.9979017917505102e-05,
+      "loss": 0.5062,
+      "step": 438
+    },
+    {
+      "epoch": 0.08655362776025237,
+      "grad_norm": 0.899069617822026,
+      "learning_rate": 1.997891744001155e-05,
+      "loss": 0.4572,
+      "step": 439
+    },
+    {
+      "epoch": 0.08675078864353312,
+      "grad_norm": 0.858458600280124,
+      "learning_rate": 1.997881672276646e-05,
+      "loss": 0.4561,
+      "step": 440
+    },
+    {
+      "epoch": 0.08694794952681388,
+      "grad_norm": 0.8872507068731488,
+      "learning_rate": 1.9978715765772266e-05,
+      "loss": 0.4893,
+      "step": 441
+    },
+    {
+      "epoch": 0.08714511041009464,
+      "grad_norm": 0.8001811452184983,
+      "learning_rate": 1.9978614569031388e-05,
+      "loss": 0.4473,
+      "step": 442
+    },
+    {
+      "epoch": 0.08734227129337539,
+      "grad_norm": 0.7719094886129825,
+      "learning_rate": 1.997851313254626e-05,
+      "loss": 0.4162,
+      "step": 443
+    },
+    {
+      "epoch": 0.08753943217665615,
+      "grad_norm": 0.8070187892782263,
+      "learning_rate": 1.997841145631932e-05,
+      "loss": 0.4357,
+      "step": 444
+    },
+    {
+      "epoch": 0.0877365930599369,
+      "grad_norm": 0.7876463515387281,
+      "learning_rate": 1.9978309540353013e-05,
+      "loss": 0.4511,
+      "step": 445
+    },
+    {
+      "epoch": 0.08793375394321766,
+      "grad_norm": 0.888526192109841,
+      "learning_rate": 1.9978207384649778e-05,
+      "loss": 0.482,
+      "step": 446
+    },
+    {
+      "epoch": 0.08813091482649842,
+      "grad_norm": 0.8684918565462284,
+      "learning_rate": 1.9978104989212078e-05,
+      "loss": 0.488,
+      "step": 447
+    },
+    {
+      "epoch": 0.08832807570977919,
+      "grad_norm": 0.9067510910757722,
+      "learning_rate": 1.997800235404237e-05,
+      "loss": 0.4546,
+      "step": 448
+    },
+    {
+      "epoch": 0.08852523659305994,
+      "grad_norm": 0.8445544456418074,
+      "learning_rate": 1.9977899479143117e-05,
+      "loss": 0.4509,
+      "step": 449
+    },
+    {
+      "epoch": 0.0887223974763407,
+      "grad_norm": 0.8441154566325766,
+      "learning_rate": 1.9977796364516796e-05,
+      "loss": 0.4794,
+      "step": 450
+    },
+    {
+      "epoch": 0.08891955835962145,
+      "grad_norm": 0.9219556253568325,
+      "learning_rate": 1.997769301016588e-05,
+      "loss": 0.4876,
+      "step": 451
+    },
+    {
+      "epoch": 0.08911671924290221,
+      "grad_norm": 0.8432925267438068,
+      "learning_rate": 1.997758941609286e-05,
+      "loss": 0.4947,
+      "step": 452
+    },
+    {
+      "epoch": 0.08931388012618297,
+      "grad_norm": 0.8912819991209812,
+      "learning_rate": 1.9977485582300215e-05,
+      "loss": 0.441,
+      "step": 453
+    },
+    {
+      "epoch": 0.08951104100946372,
+      "grad_norm": 0.8123986878321956,
+      "learning_rate": 1.9977381508790446e-05,
+      "loss": 0.461,
+      "step": 454
+    },
+    {
+      "epoch": 0.08970820189274448,
+      "grad_norm": 0.9178026802956625,
+      "learning_rate": 1.997727719556605e-05,
+      "loss": 0.5066,
+      "step": 455
+    },
+    {
+      "epoch": 0.08990536277602523,
+      "grad_norm": 0.8987063196006664,
+      "learning_rate": 1.9977172642629537e-05,
+      "loss": 0.44,
+      "step": 456
+    },
+    {
+      "epoch": 0.09010252365930599,
+      "grad_norm": 0.890159715184649,
+      "learning_rate": 1.9977067849983412e-05,
+      "loss": 0.4605,
+      "step": 457
+    },
+    {
+      "epoch": 0.09029968454258674,
+      "grad_norm": 0.95233992862221,
+      "learning_rate": 1.9976962817630202e-05,
+      "loss": 0.4727,
+      "step": 458
+    },
+    {
+      "epoch": 0.09049684542586751,
+      "grad_norm": 0.9459884973587354,
+      "learning_rate": 1.9976857545572425e-05,
+      "loss": 0.4619,
+      "step": 459
+    },
+    {
+      "epoch": 0.09069400630914827,
+      "grad_norm": 1.2642744480328538,
+      "learning_rate": 1.997675203381261e-05,
+      "loss": 0.5395,
+      "step": 460
+    },
+    {
+      "epoch": 0.09089116719242903,
+      "grad_norm": 1.0028131372893,
+      "learning_rate": 1.997664628235329e-05,
+      "loss": 0.4673,
+      "step": 461
+    },
+    {
+      "epoch": 0.09108832807570978,
+      "grad_norm": 0.9306510247064776,
+      "learning_rate": 1.9976540291197015e-05,
+      "loss": 0.4886,
+      "step": 462
+    },
+    {
+      "epoch": 0.09128548895899054,
+      "grad_norm": 0.905251801766755,
+      "learning_rate": 1.9976434060346324e-05,
+      "loss": 0.5161,
+      "step": 463
+    },
+    {
+      "epoch": 0.0914826498422713,
+      "grad_norm": 0.929602575181201,
+      "learning_rate": 1.9976327589803767e-05,
+      "loss": 0.4745,
+      "step": 464
+    },
+    {
+      "epoch": 0.09167981072555205,
+      "grad_norm": 3.0755405650687653,
+      "learning_rate": 1.997622087957191e-05,
+      "loss": 0.4565,
+      "step": 465
+    },
+    {
+      "epoch": 0.0918769716088328,
+      "grad_norm": 1.3035255174455254,
+      "learning_rate": 1.9976113929653312e-05,
+      "loss": 0.4869,
+      "step": 466
+    },
+    {
+      "epoch": 0.09207413249211356,
+      "grad_norm": 0.863482479091288,
+      "learning_rate": 1.997600674005054e-05,
+      "loss": 0.4533,
+      "step": 467
+    },
+    {
+      "epoch": 0.09227129337539432,
+      "grad_norm": 1.007184735840567,
+      "learning_rate": 1.9975899310766173e-05,
+      "loss": 0.4697,
+      "step": 468
+    },
+    {
+      "epoch": 0.09246845425867507,
+      "grad_norm": 1.1269484430615138,
+      "learning_rate": 1.997579164180279e-05,
+      "loss": 0.5163,
+      "step": 469
+    },
+    {
+      "epoch": 0.09266561514195584,
+      "grad_norm": 1.1397659345471354,
+      "learning_rate": 1.9975683733162987e-05,
+      "loss": 0.4959,
+      "step": 470
+    },
+    {
+      "epoch": 0.0928627760252366,
+      "grad_norm": 0.9860221015384155,
+      "learning_rate": 1.9975575584849346e-05,
+      "loss": 0.4788,
+      "step": 471
+    },
+    {
+      "epoch": 0.09305993690851735,
+      "grad_norm": 0.8672216442193951,
+      "learning_rate": 1.9975467196864465e-05,
+      "loss": 0.4846,
+      "step": 472
+    },
+    {
+      "epoch": 0.09325709779179811,
+      "grad_norm": 1.125980566236637,
+      "learning_rate": 1.9975358569210952e-05,
+      "loss": 0.467,
+      "step": 473
+    },
+    {
+      "epoch": 0.09345425867507887,
+      "grad_norm": 0.9201790577474425,
+      "learning_rate": 1.9975249701891414e-05,
+      "loss": 0.4767,
+      "step": 474
+    },
+    {
+      "epoch": 0.09365141955835962,
+      "grad_norm": 1.0505904231912129,
+      "learning_rate": 1.9975140594908472e-05,
+      "loss": 0.4671,
+      "step": 475
+    },
+    {
+      "epoch": 0.09384858044164038,
+      "grad_norm": 0.8373445019309896,
+      "learning_rate": 1.9975031248264746e-05,
+      "loss": 0.5008,
+      "step": 476
+    },
+    {
+      "epoch": 0.09404574132492113,
+      "grad_norm": 1.016203023349781,
+      "learning_rate": 1.997492166196286e-05,
+      "loss": 0.4614,
+      "step": 477
+    },
+    {
+      "epoch": 0.09424290220820189,
+      "grad_norm": 0.785117485057104,
+      "learning_rate": 1.9974811836005446e-05,
+      "loss": 0.465,
+      "step": 478
+    },
+    {
+      "epoch": 0.09444006309148265,
+      "grad_norm": 1.0204263244954062,
+      "learning_rate": 1.9974701770395147e-05,
+      "loss": 0.461,
+      "step": 479
+    },
+    {
+      "epoch": 0.0946372239747634,
+      "grad_norm": 2.1396373733832785,
+      "learning_rate": 1.9974591465134606e-05,
+      "loss": 0.4751,
+      "step": 480
+    },
+    {
+      "epoch": 0.09483438485804416,
+      "grad_norm": 1.112339191932461,
+      "learning_rate": 1.9974480920226472e-05,
+      "loss": 0.454,
+      "step": 481
+    },
+    {
+      "epoch": 0.09503154574132493,
+      "grad_norm": 1.0016882760903807,
+      "learning_rate": 1.9974370135673398e-05,
+      "loss": 0.4848,
+      "step": 482
+    },
+    {
+      "epoch": 0.09522870662460568,
+      "grad_norm": 1.0366723685296435,
+      "learning_rate": 1.9974259111478054e-05,
+      "loss": 0.4619,
+      "step": 483
+    },
+    {
+      "epoch": 0.09542586750788644,
+      "grad_norm": 0.7816273224252894,
+      "learning_rate": 1.9974147847643103e-05,
+      "loss": 0.4356,
+      "step": 484
+    },
+    {
+      "epoch": 0.0956230283911672,
+      "grad_norm": 2.2595110257069386,
+      "learning_rate": 1.9974036344171215e-05,
+      "loss": 0.528,
+      "step": 485
+    },
+    {
+      "epoch": 0.09582018927444795,
+      "grad_norm": 0.9462755974465362,
+      "learning_rate": 1.997392460106507e-05,
+      "loss": 0.4768,
+      "step": 486
+    },
+    {
+      "epoch": 0.0960173501577287,
+      "grad_norm": 0.9715227579239536,
+      "learning_rate": 1.997381261832736e-05,
+      "loss": 0.4725,
+      "step": 487
+    },
+    {
+      "epoch": 0.09621451104100946,
+      "grad_norm": 0.9626137085486789,
+      "learning_rate": 1.9973700395960765e-05,
+      "loss": 0.4768,
+      "step": 488
+    },
+    {
+      "epoch": 0.09641167192429022,
+      "grad_norm": 0.7957423308586874,
+      "learning_rate": 1.997358793396799e-05,
+      "loss": 0.4333,
+      "step": 489
+    },
+    {
+      "epoch": 0.09660883280757097,
+      "grad_norm": 0.8587376909161047,
+      "learning_rate": 1.9973475232351728e-05,
+      "loss": 0.4677,
+      "step": 490
+    },
+    {
+      "epoch": 0.09680599369085173,
+      "grad_norm": 0.8640406505371612,
+      "learning_rate": 1.9973362291114697e-05,
+      "loss": 0.4915,
+      "step": 491
+    },
+    {
+      "epoch": 0.09700315457413249,
+      "grad_norm": 22.479833252401775,
+      "learning_rate": 1.99732491102596e-05,
+      "loss": 0.5093,
+      "step": 492
+    },
+    {
+      "epoch": 0.09720031545741326,
+      "grad_norm": 1.7104531508622285,
+      "learning_rate": 1.9973135689789167e-05,
+      "loss": 0.4697,
+      "step": 493
+    },
+    {
+      "epoch": 0.09739747634069401,
+      "grad_norm": 0.8723153591836569,
+      "learning_rate": 1.9973022029706117e-05,
+      "loss": 0.4539,
+      "step": 494
+    },
+    {
+      "epoch": 0.09759463722397477,
+      "grad_norm": 1.1539417078528944,
+      "learning_rate": 1.997290813001318e-05,
+      "loss": 0.47,
+      "step": 495
+    },
+    {
+      "epoch": 0.09779179810725552,
+      "grad_norm": 0.8862353677994035,
+      "learning_rate": 1.9972793990713093e-05,
+      "loss": 0.4891,
+      "step": 496
+    },
+    {
+      "epoch": 0.09798895899053628,
+      "grad_norm": 1.1144995212251465,
+      "learning_rate": 1.9972679611808603e-05,
+      "loss": 0.4823,
+      "step": 497
+    },
+    {
+      "epoch": 0.09818611987381703,
+      "grad_norm": 1.4516257465643632,
+      "learning_rate": 1.997256499330245e-05,
+      "loss": 0.5077,
+      "step": 498
+    },
+    {
+      "epoch": 0.09838328075709779,
+      "grad_norm": 0.9004639995601033,
+      "learning_rate": 1.9972450135197397e-05,
+      "loss": 0.4404,
+      "step": 499
+    },
+    {
+      "epoch": 0.09858044164037855,
+      "grad_norm": 0.8035330529361184,
+      "learning_rate": 1.9972335037496195e-05,
+      "loss": 0.456,
+      "step": 500
+    },
+    {
+      "epoch": 0.0987776025236593,
+      "grad_norm": 0.8385329004850893,
+      "learning_rate": 1.9972219700201612e-05,
+      "loss": 0.4332,
+      "step": 501
+    },
+    {
+      "epoch": 0.09897476340694006,
+      "grad_norm": 0.8495426403609305,
+      "learning_rate": 1.9972104123316422e-05,
+      "loss": 0.4541,
+      "step": 502
+    },
+    {
+      "epoch": 0.09917192429022081,
+      "grad_norm": 0.7382887096139712,
+      "learning_rate": 1.9971988306843403e-05,
+      "loss": 0.4095,
+      "step": 503
+    },
+    {
+      "epoch": 0.09936908517350158,
+      "grad_norm": 0.8439026090011481,
+      "learning_rate": 1.997187225078533e-05,
+      "loss": 0.4443,
+      "step": 504
+    },
+    {
+      "epoch": 0.09956624605678234,
+      "grad_norm": 0.8075139072966796,
+      "learning_rate": 1.9971755955144995e-05,
+      "loss": 0.4608,
+      "step": 505
+    },
+    {
+      "epoch": 0.0997634069400631,
+      "grad_norm": 0.818587342922555,
+      "learning_rate": 1.9971639419925197e-05,
+      "loss": 0.4486,
+      "step": 506
+    },
+    {
+      "epoch": 0.09996056782334385,
+      "grad_norm": 0.8577985021487166,
+      "learning_rate": 1.997152264512873e-05,
+      "loss": 0.4494,
+      "step": 507
+    },
+    {
+      "epoch": 0.10015772870662461,
+      "grad_norm": 0.8628896861024081,
+      "learning_rate": 1.9971405630758402e-05,
+      "loss": 0.474,
+      "step": 508
+    },
+    {
+      "epoch": 0.10035488958990536,
+      "grad_norm": 0.799898383721641,
+      "learning_rate": 1.9971288376817023e-05,
+      "loss": 0.4588,
+      "step": 509
+    },
+    {
+      "epoch": 0.10055205047318612,
+      "grad_norm": 0.7095652696234459,
+      "learning_rate": 1.997117088330741e-05,
+      "loss": 0.4359,
+      "step": 510
+    },
+    {
+      "epoch": 0.10074921135646688,
+      "grad_norm": 0.9284886130598523,
+      "learning_rate": 1.9971053150232387e-05,
+      "loss": 0.4517,
+      "step": 511
+    },
+    {
+      "epoch": 0.10094637223974763,
+      "grad_norm": 0.8423011199925328,
+      "learning_rate": 1.9970935177594787e-05,
+      "loss": 0.4511,
+      "step": 512
+    },
+    {
+      "epoch": 0.10114353312302839,
+      "grad_norm": 0.7810718155478004,
+      "learning_rate": 1.9970816965397435e-05,
+      "loss": 0.5072,
+      "step": 513
+    },
+    {
+      "epoch": 0.10134069400630914,
+      "grad_norm": 0.8041359390874309,
+      "learning_rate": 1.9970698513643178e-05,
+      "loss": 0.4703,
+      "step": 514
+    },
+    {
+      "epoch": 0.1015378548895899,
+      "grad_norm": 1.53976548457417,
+      "learning_rate": 1.9970579822334856e-05,
+      "loss": 0.4518,
+      "step": 515
+    },
+    {
+      "epoch": 0.10173501577287067,
+      "grad_norm": 0.9907811278036276,
+      "learning_rate": 1.9970460891475328e-05,
+      "loss": 0.476,
+      "step": 516
+    },
+    {
+      "epoch": 0.10193217665615142,
+      "grad_norm": 0.8081338018995352,
+      "learning_rate": 1.997034172106745e-05,
+      "loss": 0.4691,
+      "step": 517
+    },
+    {
+      "epoch": 0.10212933753943218,
+      "grad_norm": 1.0051078776013984,
+      "learning_rate": 1.9970222311114078e-05,
+      "loss": 0.5074,
+      "step": 518
+    },
+    {
+      "epoch": 0.10232649842271294,
+      "grad_norm": 0.8105369088543491,
+      "learning_rate": 1.9970102661618088e-05,
+      "loss": 0.4502,
+      "step": 519
+    },
+    {
+      "epoch": 0.10252365930599369,
+      "grad_norm": 0.7202414114786588,
+      "learning_rate": 1.9969982772582354e-05,
+      "loss": 0.4452,
+      "step": 520
+    },
+    {
+      "epoch": 0.10272082018927445,
+      "grad_norm": 0.8648373894327899,
+      "learning_rate": 1.996986264400975e-05,
+      "loss": 0.4757,
+      "step": 521
+    },
+    {
+      "epoch": 0.1029179810725552,
+      "grad_norm": 0.7629352205940431,
+      "learning_rate": 1.9969742275903172e-05,
+      "loss": 0.477,
+      "step": 522
+    },
+    {
+      "epoch": 0.10311514195583596,
+      "grad_norm": 0.8526046130077743,
+      "learning_rate": 1.996962166826551e-05,
+      "loss": 0.4341,
+      "step": 523
+    },
+    {
+      "epoch": 0.10331230283911672,
+      "grad_norm": 5.257272773165352,
+      "learning_rate": 1.9969500821099654e-05,
+      "loss": 0.4958,
+      "step": 524
+    },
+    {
+      "epoch": 0.10350946372239747,
+      "grad_norm": 0.8477826266411921,
+      "learning_rate": 1.996937973440851e-05,
+      "loss": 0.43,
+      "step": 525
+    },
+    {
+      "epoch": 0.10370662460567823,
+      "grad_norm": 1.4050180036912192,
+      "learning_rate": 1.9969258408194997e-05,
+      "loss": 0.4446,
+      "step": 526
+    },
+    {
+      "epoch": 0.103903785488959,
+      "grad_norm": 0.976079511985993,
+      "learning_rate": 1.996913684246202e-05,
+      "loss": 0.4718,
+      "step": 527
+    },
+    {
+      "epoch": 0.10410094637223975,
+      "grad_norm": 1.460058823361328,
+      "learning_rate": 1.99690150372125e-05,
+      "loss": 0.4904,
+      "step": 528
+    },
+    {
+      "epoch": 0.10429810725552051,
+      "grad_norm": 1.2260644333639692,
+      "learning_rate": 1.9968892992449364e-05,
+      "loss": 0.4757,
+      "step": 529
+    },
+    {
+      "epoch": 0.10449526813880126,
+      "grad_norm": 0.820353657627711,
+      "learning_rate": 1.9968770708175552e-05,
+      "loss": 0.4823,
+      "step": 530
+    },
+    {
+      "epoch": 0.10469242902208202,
+      "grad_norm": 0.7536246666392749,
+      "learning_rate": 1.996864818439399e-05,
+      "loss": 0.4212,
+      "step": 531
+    },
+    {
+      "epoch": 0.10488958990536278,
+      "grad_norm": 0.8026979740475652,
+      "learning_rate": 1.9968525421107633e-05,
+      "loss": 0.4606,
+      "step": 532
+    },
+    {
+      "epoch": 0.10508675078864353,
+      "grad_norm": 0.8687858454813581,
+      "learning_rate": 1.996840241831942e-05,
+      "loss": 0.4862,
+      "step": 533
+    },
+    {
+      "epoch": 0.10528391167192429,
+      "grad_norm": 3.2595441481445944,
+      "learning_rate": 1.9968279176032314e-05,
+      "loss": 0.4919,
+      "step": 534
+    },
+    {
+      "epoch": 0.10548107255520504,
+      "grad_norm": 0.9637701357273671,
+      "learning_rate": 1.9968155694249274e-05,
+      "loss": 0.4574,
+      "step": 535
+    },
+    {
+      "epoch": 0.1056782334384858,
+      "grad_norm": 0.9015755854427224,
+      "learning_rate": 1.9968031972973266e-05,
+      "loss": 0.4892,
+      "step": 536
+    },
+    {
+      "epoch": 0.10587539432176656,
+      "grad_norm": 0.811289401193313,
+      "learning_rate": 1.996790801220726e-05,
+      "loss": 0.4393,
+      "step": 537
+    },
+    {
+      "epoch": 0.10607255520504733,
+      "grad_norm": 1.15420005303407,
+      "learning_rate": 1.996778381195424e-05,
+      "loss": 0.5066,
+      "step": 538
+    },
+    {
+      "epoch": 0.10626971608832808,
+      "grad_norm": 0.8200798308419749,
+      "learning_rate": 1.9967659372217187e-05,
+      "loss": 0.5189,
+      "step": 539
+    },
+    {
+      "epoch": 0.10646687697160884,
+      "grad_norm": 0.8867823143154604,
+      "learning_rate": 1.9967534692999085e-05,
+      "loss": 0.4365,
+      "step": 540
+    },
+    {
+      "epoch": 0.10666403785488959,
+      "grad_norm": 0.8669753830991764,
+      "learning_rate": 1.996740977430294e-05,
+      "loss": 0.457,
+      "step": 541
+    },
+    {
+      "epoch": 0.10686119873817035,
+      "grad_norm": 0.8954903277161942,
+      "learning_rate": 1.996728461613175e-05,
+      "loss": 0.4597,
+      "step": 542
+    },
+    {
+      "epoch": 0.1070583596214511,
+      "grad_norm": 0.8258256049709248,
+      "learning_rate": 1.9967159218488515e-05,
+      "loss": 0.428,
+      "step": 543
+    },
+    {
+      "epoch": 0.10725552050473186,
+      "grad_norm": 0.811522877066346,
+      "learning_rate": 1.996703358137626e-05,
+      "loss": 0.4708,
+      "step": 544
+    },
+    {
+      "epoch": 0.10745268138801262,
+      "grad_norm": 0.9746081819733365,
+      "learning_rate": 1.996690770479799e-05,
+      "loss": 0.5035,
+      "step": 545
+    },
+    {
+      "epoch": 0.10764984227129337,
+      "grad_norm": 0.8151447726161154,
+      "learning_rate": 1.9966781588756743e-05,
+      "loss": 0.4748,
+      "step": 546
+    },
+    {
+      "epoch": 0.10784700315457413,
+      "grad_norm": 0.9244734430462046,
+      "learning_rate": 1.996665523325554e-05,
+      "loss": 0.4811,
+      "step": 547
+    },
+    {
+      "epoch": 0.10804416403785488,
+      "grad_norm": 0.8967791767272244,
+      "learning_rate": 1.9966528638297415e-05,
+      "loss": 0.4329,
+      "step": 548
+    },
+    {
+      "epoch": 0.10824132492113564,
+      "grad_norm": 1.184825474138752,
+      "learning_rate": 1.9966401803885413e-05,
+      "loss": 0.4622,
+      "step": 549
+    },
+    {
+      "epoch": 0.10843848580441641,
+      "grad_norm": 0.8122484029135365,
+      "learning_rate": 1.9966274730022587e-05,
+      "loss": 0.473,
+      "step": 550
+    },
+    {
+      "epoch": 0.10863564668769717,
+      "grad_norm": 1.1993820808729505,
+      "learning_rate": 1.996614741671198e-05,
+      "loss": 0.4356,
+      "step": 551
+    },
+    {
+      "epoch": 0.10883280757097792,
+      "grad_norm": 0.767742171284427,
+      "learning_rate": 1.996601986395666e-05,
+      "loss": 0.4302,
+      "step": 552
+    },
+    {
+      "epoch": 0.10902996845425868,
+      "grad_norm": 1.1421804231314927,
+      "learning_rate": 1.9965892071759685e-05,
+      "loss": 0.4594,
+      "step": 553
+    },
+    {
+      "epoch": 0.10922712933753943,
+      "grad_norm": 0.8790816673408636,
+      "learning_rate": 1.9965764040124126e-05,
+      "loss": 0.4877,
+      "step": 554
+    },
+    {
+      "epoch": 0.10942429022082019,
+      "grad_norm": 1.2969027619478393,
+      "learning_rate": 1.9965635769053064e-05,
+      "loss": 0.474,
+      "step": 555
+    },
+    {
+      "epoch": 0.10962145110410094,
+      "grad_norm": 0.8359668233842006,
+      "learning_rate": 1.9965507258549573e-05,
+      "loss": 0.4586,
+      "step": 556
+    },
+    {
+      "epoch": 0.1098186119873817,
+      "grad_norm": 0.9318305367356925,
+      "learning_rate": 1.996537850861675e-05,
+      "loss": 0.4439,
+      "step": 557
+    },
+    {
+      "epoch": 0.11001577287066246,
+      "grad_norm": 0.9793179279182511,
+      "learning_rate": 1.9965249519257682e-05,
+      "loss": 0.4953,
+      "step": 558
+    },
+    {
+      "epoch": 0.11021293375394321,
+      "grad_norm": 0.7714402928720476,
+      "learning_rate": 1.9965120290475466e-05,
+      "loss": 0.4635,
+      "step": 559
+    },
+    {
+      "epoch": 0.11041009463722397,
+      "grad_norm": 1.0153326096876347,
+      "learning_rate": 1.996499082227321e-05,
+      "loss": 0.465,
+      "step": 560
+    },
+    {
+      "epoch": 0.11060725552050474,
+      "grad_norm": 0.7664943405741772,
+      "learning_rate": 1.996486111465403e-05,
+      "loss": 0.4849,
+      "step": 561
+    },
+    {
+      "epoch": 0.1108044164037855,
+      "grad_norm": 0.8468927415011162,
+      "learning_rate": 1.996473116762103e-05,
+      "loss": 0.4897,
+      "step": 562
+    },
+    {
+      "epoch": 0.11100157728706625,
+      "grad_norm": 0.7604431065724107,
+      "learning_rate": 1.9964600981177344e-05,
+      "loss": 0.4737,
+      "step": 563
+    },
+    {
+      "epoch": 0.111198738170347,
+      "grad_norm": 0.7359161844087391,
+      "learning_rate": 1.9964470555326096e-05,
+      "loss": 0.4508,
+      "step": 564
+    },
+    {
+      "epoch": 0.11139589905362776,
+      "grad_norm": 1.3795069909449265,
+      "learning_rate": 1.9964339890070415e-05,
+      "loss": 0.4826,
+      "step": 565
+    },
+    {
+      "epoch": 0.11159305993690852,
+      "grad_norm": 0.9524081314639894,
+      "learning_rate": 1.9964208985413448e-05,
+      "loss": 0.4497,
+      "step": 566
+    },
+    {
+      "epoch": 0.11179022082018927,
+      "grad_norm": 0.7435304463746423,
+      "learning_rate": 1.9964077841358333e-05,
+      "loss": 0.4797,
+      "step": 567
+    },
+    {
+      "epoch": 0.11198738170347003,
+      "grad_norm": 2.6303972907019326,
+      "learning_rate": 1.9963946457908223e-05,
+      "loss": 0.4954,
+      "step": 568
+    },
+    {
+      "epoch": 0.11218454258675079,
+      "grad_norm": 0.8886897209525809,
+      "learning_rate": 1.9963814835066274e-05,
+      "loss": 0.4991,
+      "step": 569
+    },
+    {
+      "epoch": 0.11238170347003154,
+      "grad_norm": 16.23439623915258,
+      "learning_rate": 1.9963682972835654e-05,
+      "loss": 0.4808,
+      "step": 570
+    },
+    {
+      "epoch": 0.1125788643533123,
+      "grad_norm": 1.0835312648993332,
+      "learning_rate": 1.9963550871219522e-05,
+      "loss": 0.5025,
+      "step": 571
+    },
+    {
+      "epoch": 0.11277602523659307,
+      "grad_norm": 2.8501049449687157,
+      "learning_rate": 1.996341853022106e-05,
+      "loss": 0.4398,
+      "step": 572
+    },
+    {
+      "epoch": 0.11297318611987382,
+      "grad_norm": 1.4391188096590775,
+      "learning_rate": 1.9963285949843446e-05,
+      "loss": 0.4883,
+      "step": 573
+    },
+    {
+      "epoch": 0.11317034700315458,
+      "grad_norm": 0.8468547462854358,
+      "learning_rate": 1.996315313008986e-05,
+      "loss": 0.4499,
+      "step": 574
+    },
+    {
+      "epoch": 0.11336750788643533,
+      "grad_norm": 1.1620214814742198,
+      "learning_rate": 1.99630200709635e-05,
+      "loss": 0.4489,
+      "step": 575
+    },
+    {
+      "epoch": 0.11356466876971609,
+      "grad_norm": 1.9853935899449358,
+      "learning_rate": 1.996288677246756e-05,
+      "loss": 0.4849,
+      "step": 576
+    },
+    {
+      "epoch": 0.11376182965299685,
+      "grad_norm": 1.8186145471704431,
+      "learning_rate": 1.996275323460524e-05,
+      "loss": 0.5109,
+      "step": 577
+    },
+    {
+      "epoch": 0.1139589905362776,
+      "grad_norm": 0.8478050994792304,
+      "learning_rate": 1.996261945737975e-05,
+      "loss": 0.5038,
+      "step": 578
+    },
+    {
+      "epoch": 0.11415615141955836,
+      "grad_norm": 1.1104246850094956,
+      "learning_rate": 1.9962485440794306e-05,
+      "loss": 0.4551,
+      "step": 579
+    },
+    {
+      "epoch": 0.11435331230283911,
+      "grad_norm": 2.048919052058303,
+      "learning_rate": 1.9962351184852123e-05,
+      "loss": 0.4642,
+      "step": 580
+    },
+    {
+      "epoch": 0.11455047318611987,
+      "grad_norm": 0.9669037613683414,
+      "learning_rate": 1.9962216689556435e-05,
+      "loss": 0.4547,
+      "step": 581
+    },
+    {
+      "epoch": 0.11474763406940063,
+      "grad_norm": 0.7967676382705094,
+      "learning_rate": 1.996208195491047e-05,
+      "loss": 0.4732,
+      "step": 582
+    },
+    {
+      "epoch": 0.11494479495268138,
+      "grad_norm": 0.973744486308741,
+      "learning_rate": 1.9961946980917457e-05,
+      "loss": 0.4612,
+      "step": 583
+    },
+    {
+      "epoch": 0.11514195583596215,
+      "grad_norm": 0.8065923330637983,
+      "learning_rate": 1.9961811767580646e-05,
+      "loss": 0.4467,
+      "step": 584
+    },
+    {
+      "epoch": 0.1153391167192429,
+      "grad_norm": 2.7671844074244127,
+      "learning_rate": 1.996167631490329e-05,
+      "loss": 0.493,
+      "step": 585
+    },
+    {
+      "epoch": 0.11553627760252366,
+      "grad_norm": 0.8616456402628487,
+      "learning_rate": 1.9961540622888637e-05,
+      "loss": 0.4738,
+      "step": 586
+    },
+    {
+      "epoch": 0.11573343848580442,
+      "grad_norm": 4.8009523346043945,
+      "learning_rate": 1.9961404691539947e-05,
+      "loss": 0.4741,
+      "step": 587
+    },
+    {
+      "epoch": 0.11593059936908517,
+      "grad_norm": 1.2289335426336983,
+      "learning_rate": 1.996126852086049e-05,
+      "loss": 0.4626,
+      "step": 588
+    },
+    {
+      "epoch": 0.11612776025236593,
+      "grad_norm": 0.8888193606458007,
+      "learning_rate": 1.996113211085353e-05,
+      "loss": 0.469,
+      "step": 589
+    },
+    {
+      "epoch": 0.11632492113564669,
+      "grad_norm": 1.020387123983096,
+      "learning_rate": 1.996099546152235e-05,
+      "loss": 0.4977,
+      "step": 590
+    },
+    {
+      "epoch": 0.11652208201892744,
+      "grad_norm": 1.5821331097425058,
+      "learning_rate": 1.9960858572870238e-05,
+      "loss": 0.5181,
+      "step": 591
+    },
+    {
+      "epoch": 0.1167192429022082,
+      "grad_norm": 1.406346825248364,
+      "learning_rate": 1.9960721444900475e-05,
+      "loss": 0.5002,
+      "step": 592
+    },
+    {
+      "epoch": 0.11691640378548895,
+      "grad_norm": 0.8333326596917328,
+      "learning_rate": 1.9960584077616356e-05,
+      "loss": 0.4882,
+      "step": 593
+    },
+    {
+      "epoch": 0.11711356466876971,
+      "grad_norm": 0.9769660099184305,
+      "learning_rate": 1.9960446471021187e-05,
+      "loss": 0.4375,
+      "step": 594
+    },
+    {
+      "epoch": 0.11731072555205048,
+      "grad_norm": 0.8908916939407923,
+      "learning_rate": 1.9960308625118265e-05,
+      "loss": 0.475,
+      "step": 595
+    },
+    {
+      "epoch": 0.11750788643533124,
+      "grad_norm": 0.8760104320431786,
+      "learning_rate": 1.996017053991091e-05,
+      "loss": 0.461,
+      "step": 596
+    },
+    {
+      "epoch": 0.11770504731861199,
+      "grad_norm": 1.0317263432759556,
+      "learning_rate": 1.9960032215402436e-05,
+      "loss": 0.4877,
+      "step": 597
+    },
+    {
+      "epoch": 0.11790220820189275,
+      "grad_norm": 1.3949799089229347,
+      "learning_rate": 1.995989365159617e-05,
+      "loss": 0.4721,
+      "step": 598
+    },
+    {
+      "epoch": 0.1180993690851735,
+      "grad_norm": 3.854325351616305,
+      "learning_rate": 1.9959754848495437e-05,
+      "loss": 0.4897,
+      "step": 599
+    },
+    {
+      "epoch": 0.11829652996845426,
+      "grad_norm": 0.9559253651339167,
+      "learning_rate": 1.9959615806103572e-05,
+      "loss": 0.4632,
+      "step": 600
+    },
+    {
+      "epoch": 0.11849369085173501,
+      "grad_norm": 1.864457347756339,
+      "learning_rate": 1.9959476524423917e-05,
+      "loss": 0.4725,
+      "step": 601
+    },
+    {
+      "epoch": 0.11869085173501577,
+      "grad_norm": 1.03878334430699,
+      "learning_rate": 1.9959337003459816e-05,
+      "loss": 0.4476,
+      "step": 602
+    },
+    {
+      "epoch": 0.11888801261829653,
+      "grad_norm": 0.9045258083599287,
+      "learning_rate": 1.995919724321463e-05,
+      "loss": 0.4791,
+      "step": 603
+    },
+    {
+      "epoch": 0.11908517350157728,
+      "grad_norm": 0.9747663255202897,
+      "learning_rate": 1.9959057243691707e-05,
+      "loss": 0.4271,
+      "step": 604
+    },
+    {
+      "epoch": 0.11928233438485804,
+      "grad_norm": 1.2080345292681676,
+      "learning_rate": 1.995891700489441e-05,
+      "loss": 0.4782,
+      "step": 605
+    },
+    {
+      "epoch": 0.11947949526813881,
+      "grad_norm": 0.8131818958865078,
+      "learning_rate": 1.9958776526826115e-05,
+      "loss": 0.4596,
+      "step": 606
+    },
+    {
+      "epoch": 0.11967665615141956,
+      "grad_norm": 0.9375228525284777,
+      "learning_rate": 1.9958635809490195e-05,
+      "loss": 0.4583,
+      "step": 607
+    },
+    {
+      "epoch": 0.11987381703470032,
+      "grad_norm": 0.9475574679425915,
+      "learning_rate": 1.995849485289003e-05,
+      "loss": 0.4766,
+      "step": 608
+    },
+    {
+      "epoch": 0.12007097791798108,
+      "grad_norm": 1.0121761969425598,
+      "learning_rate": 1.9958353657029007e-05,
+      "loss": 0.4816,
+      "step": 609
+    },
+    {
+      "epoch": 0.12026813880126183,
+      "grad_norm": 1.0654177765828532,
+      "learning_rate": 1.9958212221910514e-05,
+      "loss": 0.4907,
+      "step": 610
+    },
+    {
+      "epoch": 0.12046529968454259,
+      "grad_norm": 0.8477027332986309,
+      "learning_rate": 1.9958070547537956e-05,
+      "loss": 0.4706,
+      "step": 611
+    },
+    {
+      "epoch": 0.12066246056782334,
+      "grad_norm": 0.8677643052383288,
+      "learning_rate": 1.9957928633914735e-05,
+      "loss": 0.4858,
+      "step": 612
+    },
+    {
+      "epoch": 0.1208596214511041,
+      "grad_norm": 0.9201595715495308,
+      "learning_rate": 1.9957786481044253e-05,
+      "loss": 0.4733,
+      "step": 613
+    },
+    {
+      "epoch": 0.12105678233438485,
+      "grad_norm": 1.0931868890217196,
+      "learning_rate": 1.995764408892994e-05,
+      "loss": 0.4507,
+      "step": 614
+    },
+    {
+      "epoch": 0.12125394321766561,
+      "grad_norm": 0.9329354196032422,
+      "learning_rate": 1.9957501457575207e-05,
+      "loss": 0.481,
+      "step": 615
+    },
+    {
+      "epoch": 0.12145110410094637,
+      "grad_norm": 0.8796940111635481,
+      "learning_rate": 1.995735858698348e-05,
+      "loss": 0.4416,
+      "step": 616
+    },
+    {
+      "epoch": 0.12164826498422712,
+      "grad_norm": 0.9276302620532716,
+      "learning_rate": 1.9957215477158196e-05,
+      "loss": 0.5071,
+      "step": 617
+    },
+    {
+      "epoch": 0.12184542586750789,
+      "grad_norm": 0.9200272790666039,
+      "learning_rate": 1.9957072128102792e-05,
+      "loss": 0.4609,
+      "step": 618
+    },
+    {
+      "epoch": 0.12204258675078865,
+      "grad_norm": 0.8595424766519794,
+      "learning_rate": 1.995692853982071e-05,
+      "loss": 0.4935,
+      "step": 619
+    },
+    {
+      "epoch": 0.1222397476340694,
+      "grad_norm": 0.8508720329773822,
+      "learning_rate": 1.99567847123154e-05,
+      "loss": 0.4703,
+      "step": 620
+    },
+    {
+      "epoch": 0.12243690851735016,
+      "grad_norm": 0.7662885988170952,
+      "learning_rate": 1.9956640645590326e-05,
+      "loss": 0.4285,
+      "step": 621
+    },
+    {
+      "epoch": 0.12263406940063092,
+      "grad_norm": 0.9173829148052663,
+      "learning_rate": 1.9956496339648936e-05,
+      "loss": 0.4573,
+      "step": 622
+    },
+    {
+      "epoch": 0.12283123028391167,
+      "grad_norm": 0.7616629858014796,
+      "learning_rate": 1.9956351794494706e-05,
+      "loss": 0.5152,
+      "step": 623
+    },
+    {
+      "epoch": 0.12302839116719243,
+      "grad_norm": 0.8272088646379836,
+      "learning_rate": 1.9956207010131107e-05,
+      "loss": 0.477,
+      "step": 624
+    },
+    {
+      "epoch": 0.12322555205047318,
+      "grad_norm": 0.7612317285779934,
+      "learning_rate": 1.9956061986561615e-05,
+      "loss": 0.4668,
+      "step": 625
+    },
+    {
+      "epoch": 0.12342271293375394,
+      "grad_norm": 0.7612795525681632,
+      "learning_rate": 1.9955916723789718e-05,
+      "loss": 0.4562,
+      "step": 626
+    },
+    {
+      "epoch": 0.1236198738170347,
+      "grad_norm": 0.8179624963182074,
+      "learning_rate": 1.99557712218189e-05,
+      "loss": 0.4713,
+      "step": 627
+    },
+    {
+      "epoch": 0.12381703470031545,
+      "grad_norm": 0.8528861188969025,
+      "learning_rate": 1.995562548065267e-05,
+      "loss": 0.4664,
+      "step": 628
+    },
+    {
+      "epoch": 0.12401419558359622,
+      "grad_norm": 0.7690605067619397,
+      "learning_rate": 1.995547950029451e-05,
+      "loss": 0.455,
+      "step": 629
+    },
+    {
+      "epoch": 0.12421135646687698,
+      "grad_norm": 0.7483628212376229,
+      "learning_rate": 1.9955333280747944e-05,
+      "loss": 0.4652,
+      "step": 630
+    },
+    {
+      "epoch": 0.12440851735015773,
+      "grad_norm": 0.9221392473701614,
+      "learning_rate": 1.995518682201648e-05,
+      "loss": 0.4776,
+      "step": 631
+    },
+    {
+      "epoch": 0.12460567823343849,
+      "grad_norm": 0.851528134608556,
+      "learning_rate": 1.995504012410363e-05,
+      "loss": 0.4963,
+      "step": 632
+    },
+    {
+      "epoch": 0.12480283911671924,
+      "grad_norm": 0.7424830012420283,
+      "learning_rate": 1.9954893187012927e-05,
+      "loss": 0.4266,
+      "step": 633
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.8242224350981598,
+      "learning_rate": 1.99547460107479e-05,
+      "loss": 0.4744,
+      "step": 634
+    },
+    {
+      "epoch": 0.12519716088328076,
+      "grad_norm": 0.7923912180491203,
+      "learning_rate": 1.9954598595312084e-05,
+      "loss": 0.4633,
+      "step": 635
+    },
+    {
+      "epoch": 0.1253943217665615,
+      "grad_norm": 0.8875572382982255,
+      "learning_rate": 1.9954450940709018e-05,
+      "loss": 0.4813,
+      "step": 636
+    },
+    {
+      "epoch": 0.12559148264984227,
+      "grad_norm": 0.7926285972835072,
+      "learning_rate": 1.9954303046942255e-05,
+      "loss": 0.4878,
+      "step": 637
+    },
+    {
+      "epoch": 0.12578864353312302,
+      "grad_norm": 0.7499862912600862,
+      "learning_rate": 1.995415491401534e-05,
+      "loss": 0.4734,
+      "step": 638
+    },
+    {
+      "epoch": 0.12598580441640378,
+      "grad_norm": 0.7691964904703829,
+      "learning_rate": 1.9954006541931844e-05,
+      "loss": 0.4509,
+      "step": 639
+    },
+    {
+      "epoch": 0.12618296529968454,
+      "grad_norm": 0.7532129288589031,
+      "learning_rate": 1.9953857930695318e-05,
+      "loss": 0.4733,
+      "step": 640
+    },
+    {
+      "epoch": 0.1263801261829653,
+      "grad_norm": 0.8835688736071808,
+      "learning_rate": 1.995370908030934e-05,
+      "loss": 0.4813,
+      "step": 641
+    },
+    {
+      "epoch": 0.12657728706624605,
+      "grad_norm": 0.8280331392534872,
+      "learning_rate": 1.995355999077749e-05,
+      "loss": 0.5209,
+      "step": 642
+    },
+    {
+      "epoch": 0.1267744479495268,
+      "grad_norm": 0.7790589964453112,
+      "learning_rate": 1.9953410662103346e-05,
+      "loss": 0.468,
+      "step": 643
+    },
+    {
+      "epoch": 0.12697160883280756,
+      "grad_norm": 1.1955818097157254,
+      "learning_rate": 1.995326109429049e-05,
+      "loss": 0.5219,
+      "step": 644
+    },
+    {
+      "epoch": 0.12716876971608831,
+      "grad_norm": 0.760756053231397,
+      "learning_rate": 1.9953111287342524e-05,
+      "loss": 0.4269,
+      "step": 645
+    },
+    {
+      "epoch": 0.1273659305993691,
+      "grad_norm": 0.801104959032305,
+      "learning_rate": 1.9952961241263047e-05,
+      "loss": 0.4947,
+      "step": 646
+    },
+    {
+      "epoch": 0.12756309148264985,
+      "grad_norm": 0.9451949931761754,
+      "learning_rate": 1.9952810956055656e-05,
+      "loss": 0.4904,
+      "step": 647
+    },
+    {
+      "epoch": 0.1277602523659306,
+      "grad_norm": 0.8441360454771364,
+      "learning_rate": 1.995266043172397e-05,
+      "loss": 0.4505,
+      "step": 648
+    },
+    {
+      "epoch": 0.12795741324921137,
+      "grad_norm": 1.975864774694993,
+      "learning_rate": 1.99525096682716e-05,
+      "loss": 0.4824,
+      "step": 649
+    },
+    {
+      "epoch": 0.12815457413249212,
+      "grad_norm": 0.9293646943980033,
+      "learning_rate": 1.995235866570217e-05,
+      "loss": 0.4646,
+      "step": 650
+    },
+    {
+      "epoch": 0.12835173501577288,
+      "grad_norm": 0.7685956450339866,
+      "learning_rate": 1.9952207424019314e-05,
+      "loss": 0.4259,
+      "step": 651
+    },
+    {
+      "epoch": 0.12854889589905363,
+      "grad_norm": 1.0613268056457563,
+      "learning_rate": 1.9952055943226656e-05,
+      "loss": 0.4783,
+      "step": 652
+    },
+    {
+      "epoch": 0.1287460567823344,
+      "grad_norm": 0.7669978348644308,
+      "learning_rate": 1.995190422332784e-05,
+      "loss": 0.4342,
+      "step": 653
+    },
+    {
+      "epoch": 0.12894321766561515,
+      "grad_norm": 1.3019510608457148,
+      "learning_rate": 1.995175226432651e-05,
+      "loss": 0.5192,
+      "step": 654
+    },
+    {
+      "epoch": 0.1291403785488959,
+      "grad_norm": 1.0625567624596923,
+      "learning_rate": 1.995160006622632e-05,
+      "loss": 0.483,
+      "step": 655
+    },
+    {
+      "epoch": 0.12933753943217666,
+      "grad_norm": 0.7982339743597725,
+      "learning_rate": 1.995144762903092e-05,
+      "loss": 0.4548,
+      "step": 656
+    },
+    {
+      "epoch": 0.1295347003154574,
+      "grad_norm": 2.15419728357655,
+      "learning_rate": 1.995129495274398e-05,
+      "loss": 0.4486,
+      "step": 657
+    },
+    {
+      "epoch": 0.12973186119873817,
+      "grad_norm": 0.7953532235642334,
+      "learning_rate": 1.9951142037369163e-05,
+      "loss": 0.5146,
+      "step": 658
+    },
+    {
+      "epoch": 0.12992902208201892,
+      "grad_norm": 0.7733286355003084,
+      "learning_rate": 1.995098888291015e-05,
+      "loss": 0.4509,
+      "step": 659
+    },
+    {
+      "epoch": 0.13012618296529968,
+      "grad_norm": 0.7219639448869223,
+      "learning_rate": 1.995083548937061e-05,
+      "loss": 0.4339,
+      "step": 660
+    },
+    {
+      "epoch": 0.13032334384858044,
+      "grad_norm": 0.6876863470552866,
+      "learning_rate": 1.9950681856754236e-05,
+      "loss": 0.4397,
+      "step": 661
+    },
+    {
+      "epoch": 0.1305205047318612,
+      "grad_norm": 0.7330567511407265,
+      "learning_rate": 1.9950527985064717e-05,
+      "loss": 0.4403,
+      "step": 662
+    },
+    {
+      "epoch": 0.13071766561514195,
+      "grad_norm": 0.6967061391482817,
+      "learning_rate": 1.9950373874305752e-05,
+      "loss": 0.4402,
+      "step": 663
+    },
+    {
+      "epoch": 0.1309148264984227,
+      "grad_norm": 1.2938777572194384,
+      "learning_rate": 1.9950219524481042e-05,
+      "loss": 0.5203,
+      "step": 664
+    },
+    {
+      "epoch": 0.13111198738170346,
+      "grad_norm": 0.9539568996661243,
+      "learning_rate": 1.995006493559429e-05,
+      "loss": 0.4834,
+      "step": 665
+    },
+    {
+      "epoch": 0.13130914826498422,
+      "grad_norm": 0.8042726712344785,
+      "learning_rate": 1.9949910107649218e-05,
+      "loss": 0.4626,
+      "step": 666
+    },
+    {
+      "epoch": 0.13150630914826497,
+      "grad_norm": 1.6472739683316433,
+      "learning_rate": 1.9949755040649545e-05,
+      "loss": 0.479,
+      "step": 667
+    },
+    {
+      "epoch": 0.13170347003154576,
+      "grad_norm": 1.0783243407313121,
+      "learning_rate": 1.9949599734598993e-05,
+      "loss": 0.4656,
+      "step": 668
+    },
+    {
+      "epoch": 0.1319006309148265,
+      "grad_norm": 0.8310407005161574,
+      "learning_rate": 1.9949444189501294e-05,
+      "loss": 0.4254,
+      "step": 669
+    },
+    {
+      "epoch": 0.13209779179810727,
+      "grad_norm": 0.7957824614600361,
+      "learning_rate": 1.9949288405360186e-05,
+      "loss": 0.4683,
+      "step": 670
+    },
+    {
+      "epoch": 0.13229495268138802,
+      "grad_norm": 0.7451707613081128,
+      "learning_rate": 1.9949132382179415e-05,
+      "loss": 0.4632,
+      "step": 671
+    },
+    {
+      "epoch": 0.13249211356466878,
+      "grad_norm": 0.7738134822360583,
+      "learning_rate": 1.9948976119962724e-05,
+      "loss": 0.4252,
+      "step": 672
+    },
+    {
+      "epoch": 0.13268927444794953,
+      "grad_norm": 1.2563956712818225,
+      "learning_rate": 1.9948819618713868e-05,
+      "loss": 0.4297,
+      "step": 673
+    },
+    {
+      "epoch": 0.1328864353312303,
+      "grad_norm": 0.7251684476228085,
+      "learning_rate": 1.994866287843661e-05,
+      "loss": 0.4308,
+      "step": 674
+    },
+    {
+      "epoch": 0.13308359621451105,
+      "grad_norm": 0.8186852540206319,
+      "learning_rate": 1.9948505899134717e-05,
+      "loss": 0.4646,
+      "step": 675
+    },
+    {
+      "epoch": 0.1332807570977918,
+      "grad_norm": 1.7860304743465076,
+      "learning_rate": 1.994834868081196e-05,
+      "loss": 0.4876,
+      "step": 676
+    },
+    {
+      "epoch": 0.13347791798107256,
+      "grad_norm": 0.850580467492657,
+      "learning_rate": 1.9948191223472108e-05,
+      "loss": 0.4828,
+      "step": 677
+    },
+    {
+      "epoch": 0.13367507886435331,
+      "grad_norm": 0.7130834372780571,
+      "learning_rate": 1.9948033527118954e-05,
+      "loss": 0.4874,
+      "step": 678
+    },
+    {
+      "epoch": 0.13387223974763407,
+      "grad_norm": 0.7595923377554168,
+      "learning_rate": 1.9947875591756286e-05,
+      "loss": 0.4508,
+      "step": 679
+    },
+    {
+      "epoch": 0.13406940063091483,
+      "grad_norm": 0.841937797265363,
+      "learning_rate": 1.9947717417387894e-05,
+      "loss": 0.4789,
+      "step": 680
+    },
+    {
+      "epoch": 0.13426656151419558,
+      "grad_norm": 0.7865575362210725,
+      "learning_rate": 1.994755900401758e-05,
+      "loss": 0.439,
+      "step": 681
+    },
+    {
+      "epoch": 0.13446372239747634,
+      "grad_norm": 0.7464667503739694,
+      "learning_rate": 1.9947400351649148e-05,
+      "loss": 0.4574,
+      "step": 682
+    },
+    {
+      "epoch": 0.1346608832807571,
+      "grad_norm": 0.765573758240295,
+      "learning_rate": 1.9947241460286414e-05,
+      "loss": 0.4661,
+      "step": 683
+    },
+    {
+      "epoch": 0.13485804416403785,
+      "grad_norm": 0.8053849370801137,
+      "learning_rate": 1.9947082329933192e-05,
+      "loss": 0.4902,
+      "step": 684
+    },
+    {
+      "epoch": 0.1350552050473186,
+      "grad_norm": 0.9079668954607747,
+      "learning_rate": 1.9946922960593307e-05,
+      "loss": 0.4344,
+      "step": 685
+    },
+    {
+      "epoch": 0.13525236593059936,
+      "grad_norm": 0.7910386751156379,
+      "learning_rate": 1.994676335227059e-05,
+      "loss": 0.4423,
+      "step": 686
+    },
+    {
+      "epoch": 0.13544952681388012,
+      "grad_norm": 1.5549413245532762,
+      "learning_rate": 1.994660350496887e-05,
+      "loss": 0.5308,
+      "step": 687
+    },
+    {
+      "epoch": 0.13564668769716087,
+      "grad_norm": 0.9255915463007742,
+      "learning_rate": 1.9946443418691994e-05,
+      "loss": 0.4618,
+      "step": 688
+    },
+    {
+      "epoch": 0.13584384858044163,
+      "grad_norm": 0.7804360864352426,
+      "learning_rate": 1.9946283093443803e-05,
+      "loss": 0.449,
+      "step": 689
+    },
+    {
+      "epoch": 0.13604100946372238,
+      "grad_norm": 0.7475663046747916,
+      "learning_rate": 1.9946122529228153e-05,
+      "loss": 0.4503,
+      "step": 690
+    },
+    {
+      "epoch": 0.13623817034700317,
+      "grad_norm": 1.2554380433560701,
+      "learning_rate": 1.9945961726048895e-05,
+      "loss": 0.5168,
+      "step": 691
+    },
+    {
+      "epoch": 0.13643533123028392,
+      "grad_norm": 0.7879053546798648,
+      "learning_rate": 1.9945800683909904e-05,
+      "loss": 0.4771,
+      "step": 692
+    },
+    {
+      "epoch": 0.13663249211356468,
+      "grad_norm": 0.7487562573568673,
+      "learning_rate": 1.9945639402815037e-05,
+      "loss": 0.4672,
+      "step": 693
+    },
+    {
+      "epoch": 0.13682965299684544,
+      "grad_norm": 0.7202621138555081,
+      "learning_rate": 1.9945477882768177e-05,
+      "loss": 0.447,
+      "step": 694
+    },
+    {
+      "epoch": 0.1370268138801262,
+      "grad_norm": 0.8717913765163078,
+      "learning_rate": 1.99453161237732e-05,
+      "loss": 0.4599,
+      "step": 695
+    },
+    {
+      "epoch": 0.13722397476340695,
+      "grad_norm": 0.8569531287456702,
+      "learning_rate": 1.9945154125833996e-05,
+      "loss": 0.4604,
+      "step": 696
+    },
+    {
+      "epoch": 0.1374211356466877,
+      "grad_norm": 0.8364641774305389,
+      "learning_rate": 1.9944991888954453e-05,
+      "loss": 0.48,
+      "step": 697
+    },
+    {
+      "epoch": 0.13761829652996846,
+      "grad_norm": 0.7618210867976017,
+      "learning_rate": 1.9944829413138472e-05,
+      "loss": 0.4546,
+      "step": 698
+    },
+    {
+      "epoch": 0.13781545741324921,
+      "grad_norm": 0.745175412220073,
+      "learning_rate": 1.9944666698389957e-05,
+      "loss": 0.465,
+      "step": 699
+    },
+    {
+      "epoch": 0.13801261829652997,
+      "grad_norm": 0.8186541094239488,
+      "learning_rate": 1.9944503744712814e-05,
+      "loss": 0.5081,
+      "step": 700
+    },
+    {
+      "epoch": 0.13820977917981073,
+      "grad_norm": 0.7800566021733093,
+      "learning_rate": 1.994434055211096e-05,
+      "loss": 0.4692,
+      "step": 701
+    },
+    {
+      "epoch": 0.13840694006309148,
+      "grad_norm": 2.337423114442513,
+      "learning_rate": 1.9944177120588318e-05,
+      "loss": 0.4922,
+      "step": 702
+    },
+    {
+      "epoch": 0.13860410094637224,
+      "grad_norm": 0.9603122253059109,
+      "learning_rate": 1.9944013450148812e-05,
+      "loss": 0.5075,
+      "step": 703
+    },
+    {
+      "epoch": 0.138801261829653,
+      "grad_norm": 0.7119216198791689,
+      "learning_rate": 1.9943849540796375e-05,
+      "loss": 0.4626,
+      "step": 704
+    },
+    {
+      "epoch": 0.13899842271293375,
+      "grad_norm": 0.8234426695741333,
+      "learning_rate": 1.9943685392534945e-05,
+      "loss": 0.4274,
+      "step": 705
+    },
+    {
+      "epoch": 0.1391955835962145,
+      "grad_norm": 0.8841914182356732,
+      "learning_rate": 1.9943521005368468e-05,
+      "loss": 0.5169,
+      "step": 706
+    },
+    {
+      "epoch": 0.13939274447949526,
+      "grad_norm": 0.9754273189473419,
+      "learning_rate": 1.994335637930089e-05,
+      "loss": 0.4822,
+      "step": 707
+    },
+    {
+      "epoch": 0.13958990536277602,
+      "grad_norm": 0.8312745222051547,
+      "learning_rate": 1.9943191514336164e-05,
+      "loss": 0.4614,
+      "step": 708
+    },
+    {
+      "epoch": 0.13978706624605677,
+      "grad_norm": 0.8430236639900083,
+      "learning_rate": 1.9943026410478258e-05,
+      "loss": 0.4339,
+      "step": 709
+    },
+    {
+      "epoch": 0.13998422712933753,
+      "grad_norm": 1.0814961430956602,
+      "learning_rate": 1.9942861067731135e-05,
+      "loss": 0.4825,
+      "step": 710
+    },
+    {
+      "epoch": 0.14018138801261829,
+      "grad_norm": 0.77685594526798,
+      "learning_rate": 1.994269548609877e-05,
+      "loss": 0.4801,
+      "step": 711
+    },
+    {
+      "epoch": 0.14037854889589904,
+      "grad_norm": 0.8376210695048821,
+      "learning_rate": 1.9942529665585134e-05,
+      "loss": 0.4742,
+      "step": 712
+    },
+    {
+      "epoch": 0.1405757097791798,
+      "grad_norm": 1.0443709227200295,
+      "learning_rate": 1.994236360619422e-05,
+      "loss": 0.4616,
+      "step": 713
+    },
+    {
+      "epoch": 0.14077287066246058,
+      "grad_norm": 0.7104447806333858,
+      "learning_rate": 1.9942197307930014e-05,
+      "loss": 0.4193,
+      "step": 714
+    },
+    {
+      "epoch": 0.14097003154574134,
+      "grad_norm": 0.7967791879544678,
+      "learning_rate": 1.994203077079651e-05,
+      "loss": 0.4377,
+      "step": 715
+    },
+    {
+      "epoch": 0.1411671924290221,
+      "grad_norm": 0.6852662543141802,
+      "learning_rate": 1.994186399479771e-05,
+      "loss": 0.4558,
+      "step": 716
+    },
+    {
+      "epoch": 0.14136435331230285,
+      "grad_norm": 0.7293696917628472,
+      "learning_rate": 1.9941696979937622e-05,
+      "loss": 0.4695,
+      "step": 717
+    },
+    {
+      "epoch": 0.1415615141955836,
+      "grad_norm": 1.0106248411363608,
+      "learning_rate": 1.994152972622026e-05,
+      "loss": 0.4496,
+      "step": 718
+    },
+    {
+      "epoch": 0.14175867507886436,
+      "grad_norm": 2.069396673415166,
+      "learning_rate": 1.994136223364964e-05,
+      "loss": 0.5479,
+      "step": 719
+    },
+    {
+      "epoch": 0.14195583596214512,
+      "grad_norm": 0.8625581928141414,
+      "learning_rate": 1.994119450222978e-05,
+      "loss": 0.5064,
+      "step": 720
+    },
+    {
+      "epoch": 0.14215299684542587,
+      "grad_norm": 0.9612101953431659,
+      "learning_rate": 1.9941026531964723e-05,
+      "loss": 0.4676,
+      "step": 721
+    },
+    {
+      "epoch": 0.14235015772870663,
+      "grad_norm": 0.775181863734654,
+      "learning_rate": 1.9940858322858493e-05,
+      "loss": 0.4736,
+      "step": 722
+    },
+    {
+      "epoch": 0.14254731861198738,
+      "grad_norm": 1.0608368229238194,
+      "learning_rate": 1.994068987491514e-05,
+      "loss": 0.4761,
+      "step": 723
+    },
+    {
+      "epoch": 0.14274447949526814,
+      "grad_norm": 0.7660677802463719,
+      "learning_rate": 1.9940521188138707e-05,
+      "loss": 0.4652,
+      "step": 724
+    },
+    {
+      "epoch": 0.1429416403785489,
+      "grad_norm": 0.8853653109759493,
+      "learning_rate": 1.9940352262533253e-05,
+      "loss": 0.4351,
+      "step": 725
+    },
+    {
+      "epoch": 0.14313880126182965,
+      "grad_norm": 0.8053400341791106,
+      "learning_rate": 1.9940183098102823e-05,
+      "loss": 0.4688,
+      "step": 726
+    },
+    {
+      "epoch": 0.1433359621451104,
+      "grad_norm": 0.8403515646943758,
+      "learning_rate": 1.9940013694851492e-05,
+      "loss": 0.4709,
+      "step": 727
+    },
+    {
+      "epoch": 0.14353312302839116,
+      "grad_norm": 0.8639620985700686,
+      "learning_rate": 1.9939844052783328e-05,
+      "loss": 0.44,
+      "step": 728
+    },
+    {
+      "epoch": 0.14373028391167192,
+      "grad_norm": 0.7864708085272261,
+      "learning_rate": 1.9939674171902406e-05,
+      "loss": 0.4835,
+      "step": 729
+    },
+    {
+      "epoch": 0.14392744479495267,
+      "grad_norm": 1.5378342164534347,
+      "learning_rate": 1.9939504052212807e-05,
+      "loss": 0.4371,
+      "step": 730
+    },
+    {
+      "epoch": 0.14412460567823343,
+      "grad_norm": 0.8960131433484737,
+      "learning_rate": 1.993933369371862e-05,
+      "loss": 0.4814,
+      "step": 731
+    },
+    {
+      "epoch": 0.1443217665615142,
+      "grad_norm": 1.2989860336198154,
+      "learning_rate": 1.9939163096423936e-05,
+      "loss": 0.4499,
+      "step": 732
+    },
+    {
+      "epoch": 0.14451892744479494,
+      "grad_norm": 0.8858611738667077,
+      "learning_rate": 1.9938992260332854e-05,
+      "loss": 0.4327,
+      "step": 733
+    },
+    {
+      "epoch": 0.1447160883280757,
+      "grad_norm": 1.1182251531253669,
+      "learning_rate": 1.993882118544948e-05,
+      "loss": 0.4575,
+      "step": 734
+    },
+    {
+      "epoch": 0.14491324921135645,
+      "grad_norm": 1.181366460583882,
+      "learning_rate": 1.993864987177792e-05,
+      "loss": 0.4997,
+      "step": 735
+    },
+    {
+      "epoch": 0.14511041009463724,
+      "grad_norm": 0.8730586316239893,
+      "learning_rate": 1.9938478319322296e-05,
+      "loss": 0.4279,
+      "step": 736
+    },
+    {
+      "epoch": 0.145307570977918,
+      "grad_norm": 0.8622249699446333,
+      "learning_rate": 1.9938306528086728e-05,
+      "loss": 0.4689,
+      "step": 737
+    },
+    {
+      "epoch": 0.14550473186119875,
+      "grad_norm": 0.856694039815468,
+      "learning_rate": 1.9938134498075344e-05,
+      "loss": 0.495,
+      "step": 738
+    },
+    {
+      "epoch": 0.1457018927444795,
+      "grad_norm": 1.076447016915175,
+      "learning_rate": 1.993796222929227e-05,
+      "loss": 0.4981,
+      "step": 739
+    },
+    {
+      "epoch": 0.14589905362776026,
+      "grad_norm": 1.1443003392978084,
+      "learning_rate": 1.9937789721741654e-05,
+      "loss": 0.4818,
+      "step": 740
+    },
+    {
+      "epoch": 0.14609621451104102,
+      "grad_norm": 0.7913177157915059,
+      "learning_rate": 1.9937616975427635e-05,
+      "loss": 0.4423,
+      "step": 741
+    },
+    {
+      "epoch": 0.14629337539432177,
+      "grad_norm": 0.8690811074204894,
+      "learning_rate": 1.993744399035437e-05,
+      "loss": 0.4583,
+      "step": 742
+    },
+    {
+      "epoch": 0.14649053627760253,
+      "grad_norm": 0.8018130690999904,
+      "learning_rate": 1.9937270766526007e-05,
+      "loss": 0.4581,
+      "step": 743
+    },
+    {
+      "epoch": 0.14668769716088328,
+      "grad_norm": 0.7068458870399968,
+      "learning_rate": 1.9937097303946712e-05,
+      "loss": 0.4365,
+      "step": 744
+    },
+    {
+      "epoch": 0.14688485804416404,
+      "grad_norm": 1.106160247073608,
+      "learning_rate": 1.993692360262065e-05,
+      "loss": 0.4621,
+      "step": 745
+    },
+    {
+      "epoch": 0.1470820189274448,
+      "grad_norm": 1.2660319824619375,
+      "learning_rate": 1.9936749662552e-05,
+      "loss": 0.4606,
+      "step": 746
+    },
+    {
+      "epoch": 0.14727917981072555,
+      "grad_norm": 1.0379124369378552,
+      "learning_rate": 1.9936575483744934e-05,
+      "loss": 0.5044,
+      "step": 747
+    },
+    {
+      "epoch": 0.1474763406940063,
+      "grad_norm": 2.817773387208584,
+      "learning_rate": 1.993640106620364e-05,
+      "loss": 0.4786,
+      "step": 748
+    },
+    {
+      "epoch": 0.14767350157728706,
+      "grad_norm": 1.2455773315165135,
+      "learning_rate": 1.993622640993231e-05,
+      "loss": 0.4406,
+      "step": 749
+    },
+    {
+      "epoch": 0.14787066246056782,
+      "grad_norm": 0.8090742340724816,
+      "learning_rate": 1.993605151493514e-05,
+      "loss": 0.4581,
+      "step": 750
+    },
+    {
+      "epoch": 0.14806782334384858,
+      "grad_norm": 1.079672829303502,
+      "learning_rate": 1.9935876381216327e-05,
+      "loss": 0.4869,
+      "step": 751
+    },
+    {
+      "epoch": 0.14826498422712933,
+      "grad_norm": 1.1108900962579644,
+      "learning_rate": 1.993570100878009e-05,
+      "loss": 0.4919,
+      "step": 752
+    },
+    {
+      "epoch": 0.1484621451104101,
+      "grad_norm": 0.9754359159817712,
+      "learning_rate": 1.993552539763063e-05,
+      "loss": 0.4823,
+      "step": 753
+    },
+    {
+      "epoch": 0.14865930599369084,
+      "grad_norm": 0.9888856468493855,
+      "learning_rate": 1.9935349547772168e-05,
+      "loss": 0.4685,
+      "step": 754
+    },
+    {
+      "epoch": 0.1488564668769716,
+      "grad_norm": 0.8379931331598425,
+      "learning_rate": 1.993517345920894e-05,
+      "loss": 0.4936,
+      "step": 755
+    },
+    {
+      "epoch": 0.14905362776025236,
+      "grad_norm": 1.0605810276155365,
+      "learning_rate": 1.9934997131945165e-05,
+      "loss": 0.5102,
+      "step": 756
+    },
+    {
+      "epoch": 0.1492507886435331,
+      "grad_norm": 0.7915803452960315,
+      "learning_rate": 1.993482056598508e-05,
+      "loss": 0.4434,
+      "step": 757
+    },
+    {
+      "epoch": 0.14944794952681387,
+      "grad_norm": 0.9745418084289177,
+      "learning_rate": 1.9934643761332933e-05,
+      "loss": 0.493,
+      "step": 758
+    },
+    {
+      "epoch": 0.14964511041009465,
+      "grad_norm": 0.8122670056237417,
+      "learning_rate": 1.993446671799297e-05,
+      "loss": 0.4966,
+      "step": 759
+    },
+    {
+      "epoch": 0.1498422712933754,
+      "grad_norm": 0.8242151071688861,
+      "learning_rate": 1.9934289435969443e-05,
+      "loss": 0.4771,
+      "step": 760
+    },
+    {
+      "epoch": 0.15003943217665616,
+      "grad_norm": 0.8669205650067113,
+      "learning_rate": 1.9934111915266614e-05,
+      "loss": 0.4658,
+      "step": 761
+    },
+    {
+      "epoch": 0.15023659305993692,
+      "grad_norm": 0.8210061083466603,
+      "learning_rate": 1.9933934155888745e-05,
+      "loss": 0.4503,
+      "step": 762
+    },
+    {
+      "epoch": 0.15043375394321767,
+      "grad_norm": 0.8739389286616153,
+      "learning_rate": 1.993375615784011e-05,
+      "loss": 0.4426,
+      "step": 763
+    },
+    {
+      "epoch": 0.15063091482649843,
+      "grad_norm": 0.8857407360915495,
+      "learning_rate": 1.993357792112498e-05,
+      "loss": 0.4337,
+      "step": 764
+    },
+    {
+      "epoch": 0.15082807570977919,
+      "grad_norm": 0.8810531719213002,
+      "learning_rate": 1.9933399445747645e-05,
+      "loss": 0.4614,
+      "step": 765
+    },
+    {
+      "epoch": 0.15102523659305994,
+      "grad_norm": 0.8157853115519594,
+      "learning_rate": 1.9933220731712385e-05,
+      "loss": 0.4531,
+      "step": 766
+    },
+    {
+      "epoch": 0.1512223974763407,
+      "grad_norm": 0.6694213590950058,
+      "learning_rate": 1.9933041779023502e-05,
+      "loss": 0.4429,
+      "step": 767
+    },
+    {
+      "epoch": 0.15141955835962145,
+      "grad_norm": 0.8193309247418029,
+      "learning_rate": 1.993286258768529e-05,
+      "loss": 0.4641,
+      "step": 768
+    },
+    {
+      "epoch": 0.1516167192429022,
+      "grad_norm": 1.13684291197781,
+      "learning_rate": 1.9932683157702054e-05,
+      "loss": 0.4807,
+      "step": 769
+    },
+    {
+      "epoch": 0.15181388012618297,
+      "grad_norm": 0.7712724217918977,
+      "learning_rate": 1.9932503489078105e-05,
+      "loss": 0.496,
+      "step": 770
+    },
+    {
+      "epoch": 0.15201104100946372,
+      "grad_norm": 0.7302957702591549,
+      "learning_rate": 1.993232358181776e-05,
+      "loss": 0.4562,
+      "step": 771
+    },
+    {
+      "epoch": 0.15220820189274448,
+      "grad_norm": 2.0853914712696477,
+      "learning_rate": 1.9932143435925346e-05,
+      "loss": 0.4809,
+      "step": 772
+    },
+    {
+      "epoch": 0.15240536277602523,
+      "grad_norm": 0.7477688111923385,
+      "learning_rate": 1.993196305140519e-05,
+      "loss": 0.4518,
+      "step": 773
+    },
+    {
+      "epoch": 0.152602523659306,
+      "grad_norm": 0.7646481115742062,
+      "learning_rate": 1.993178242826162e-05,
+      "loss": 0.4686,
+      "step": 774
+    },
+    {
+      "epoch": 0.15279968454258674,
+      "grad_norm": 0.7688921576232522,
+      "learning_rate": 1.9931601566498976e-05,
+      "loss": 0.4965,
+      "step": 775
+    },
+    {
+      "epoch": 0.1529968454258675,
+      "grad_norm": 0.749864201600683,
+      "learning_rate": 1.9931420466121613e-05,
+      "loss": 0.435,
+      "step": 776
+    },
+    {
+      "epoch": 0.15319400630914826,
+      "grad_norm": 0.856076678547743,
+      "learning_rate": 1.993123912713387e-05,
+      "loss": 0.5236,
+      "step": 777
+    },
+    {
+      "epoch": 0.153391167192429,
+      "grad_norm": 1.1208704076088636,
+      "learning_rate": 1.9931057549540114e-05,
+      "loss": 0.5461,
+      "step": 778
+    },
+    {
+      "epoch": 0.15358832807570977,
+      "grad_norm": 1.6176725348893128,
+      "learning_rate": 1.9930875733344698e-05,
+      "loss": 0.4927,
+      "step": 779
+    },
+    {
+      "epoch": 0.15378548895899052,
+      "grad_norm": 0.6955630821889409,
+      "learning_rate": 1.9930693678552e-05,
+      "loss": 0.4298,
+      "step": 780
+    },
+    {
+      "epoch": 0.15398264984227128,
+      "grad_norm": 0.7928697755514822,
+      "learning_rate": 1.9930511385166388e-05,
+      "loss": 0.4368,
+      "step": 781
+    },
+    {
+      "epoch": 0.15417981072555206,
+      "grad_norm": 0.9287503152149722,
+      "learning_rate": 1.9930328853192243e-05,
+      "loss": 0.4308,
+      "step": 782
+    },
+    {
+      "epoch": 0.15437697160883282,
+      "grad_norm": 0.9144564030468558,
+      "learning_rate": 1.993014608263395e-05,
+      "loss": 0.4802,
+      "step": 783
+    },
+    {
+      "epoch": 0.15457413249211358,
+      "grad_norm": 0.7843727384732332,
+      "learning_rate": 1.9929963073495896e-05,
+      "loss": 0.4635,
+      "step": 784
+    },
+    {
+      "epoch": 0.15477129337539433,
+      "grad_norm": 0.8191880589199644,
+      "learning_rate": 1.992977982578249e-05,
+      "loss": 0.4816,
+      "step": 785
+    },
+    {
+      "epoch": 0.1549684542586751,
+      "grad_norm": 0.8904807870661935,
+      "learning_rate": 1.9929596339498122e-05,
+      "loss": 0.47,
+      "step": 786
+    },
+    {
+      "epoch": 0.15516561514195584,
+      "grad_norm": 1.8183390384125857,
+      "learning_rate": 1.9929412614647207e-05,
+      "loss": 0.4631,
+      "step": 787
+    },
+    {
+      "epoch": 0.1553627760252366,
+      "grad_norm": 0.8216800794838671,
+      "learning_rate": 1.992922865123416e-05,
+      "loss": 0.4264,
+      "step": 788
+    },
+    {
+      "epoch": 0.15555993690851735,
+      "grad_norm": 0.7165148280953453,
+      "learning_rate": 1.9929044449263397e-05,
+      "loss": 0.4674,
+      "step": 789
+    },
+    {
+      "epoch": 0.1557570977917981,
+      "grad_norm": 1.52791161060643,
+      "learning_rate": 1.9928860008739343e-05,
+      "loss": 0.5036,
+      "step": 790
+    },
+    {
+      "epoch": 0.15595425867507887,
+      "grad_norm": 0.9923711638162429,
+      "learning_rate": 1.9928675329666435e-05,
+      "loss": 0.4778,
+      "step": 791
+    },
+    {
+      "epoch": 0.15615141955835962,
+      "grad_norm": 0.8551576282277356,
+      "learning_rate": 1.9928490412049108e-05,
+      "loss": 0.464,
+      "step": 792
+    },
+    {
+      "epoch": 0.15634858044164038,
+      "grad_norm": 0.7732867427931409,
+      "learning_rate": 1.99283052558918e-05,
+      "loss": 0.4623,
+      "step": 793
+    },
+    {
+      "epoch": 0.15654574132492113,
+      "grad_norm": 0.9096270847699062,
+      "learning_rate": 1.9928119861198962e-05,
+      "loss": 0.5047,
+      "step": 794
+    },
+    {
+      "epoch": 0.1567429022082019,
+      "grad_norm": 0.913028519166887,
+      "learning_rate": 1.9927934227975054e-05,
+      "loss": 0.4832,
+      "step": 795
+    },
+    {
+      "epoch": 0.15694006309148265,
+      "grad_norm": 0.7830739067460275,
+      "learning_rate": 1.9927748356224528e-05,
+      "loss": 0.4744,
+      "step": 796
+    },
+    {
+      "epoch": 0.1571372239747634,
+      "grad_norm": 0.8632625237500243,
+      "learning_rate": 1.9927562245951854e-05,
+      "loss": 0.4458,
+      "step": 797
+    },
+    {
+      "epoch": 0.15733438485804416,
+      "grad_norm": 0.7175002952446166,
+      "learning_rate": 1.9927375897161502e-05,
+      "loss": 0.4602,
+      "step": 798
+    },
+    {
+      "epoch": 0.1575315457413249,
+      "grad_norm": 0.8959680573938172,
+      "learning_rate": 1.9927189309857948e-05,
+      "loss": 0.4611,
+      "step": 799
+    },
+    {
+      "epoch": 0.15772870662460567,
+      "grad_norm": 0.8010317830596815,
+      "learning_rate": 1.9927002484045678e-05,
+      "loss": 0.4761,
+      "step": 800
+    },
+    {
+      "epoch": 0.15792586750788642,
+      "grad_norm": 0.704519354931981,
+      "learning_rate": 1.992681541972918e-05,
+      "loss": 0.4654,
+      "step": 801
+    },
+    {
+      "epoch": 0.15812302839116718,
+      "grad_norm": 1.0431620409267255,
+      "learning_rate": 1.9926628116912946e-05,
+      "loss": 0.519,
+      "step": 802
+    },
+    {
+      "epoch": 0.15832018927444794,
+      "grad_norm": 0.743437599527334,
+      "learning_rate": 1.992644057560148e-05,
+      "loss": 0.4452,
+      "step": 803
+    },
+    {
+      "epoch": 0.15851735015772872,
+      "grad_norm": 2.131396450457408,
+      "learning_rate": 1.992625279579928e-05,
+      "loss": 0.4587,
+      "step": 804
+    },
+    {
+      "epoch": 0.15871451104100948,
+      "grad_norm": 1.6197122809814877,
+      "learning_rate": 1.992606477751087e-05,
+      "loss": 0.4668,
+      "step": 805
+    },
+    {
+      "epoch": 0.15891167192429023,
+      "grad_norm": 0.9266315678990761,
+      "learning_rate": 1.9925876520740758e-05,
+      "loss": 0.5003,
+      "step": 806
+    },
+    {
+      "epoch": 0.159108832807571,
+      "grad_norm": 1.3138014974549306,
+      "learning_rate": 1.9925688025493468e-05,
+      "loss": 0.4816,
+      "step": 807
+    },
+    {
+      "epoch": 0.15930599369085174,
+      "grad_norm": 0.7340535577058777,
+      "learning_rate": 1.9925499291773528e-05,
+      "loss": 0.4804,
+      "step": 808
+    },
+    {
+      "epoch": 0.1595031545741325,
+      "grad_norm": 1.1487683565124704,
+      "learning_rate": 1.9925310319585475e-05,
+      "loss": 0.4804,
+      "step": 809
+    },
+    {
+      "epoch": 0.15970031545741326,
+      "grad_norm": 0.7305026939171207,
+      "learning_rate": 1.9925121108933852e-05,
+      "loss": 0.4774,
+      "step": 810
+    },
+    {
+      "epoch": 0.159897476340694,
+      "grad_norm": 0.9950230997430713,
+      "learning_rate": 1.99249316598232e-05,
+      "loss": 0.4638,
+      "step": 811
+    },
+    {
+      "epoch": 0.16009463722397477,
+      "grad_norm": 0.857287078084844,
+      "learning_rate": 1.9924741972258076e-05,
+      "loss": 0.4742,
+      "step": 812
+    },
+    {
+      "epoch": 0.16029179810725552,
+      "grad_norm": 0.8005250545507363,
+      "learning_rate": 1.9924552046243026e-05,
+      "loss": 0.488,
+      "step": 813
+    },
+    {
+      "epoch": 0.16048895899053628,
+      "grad_norm": 0.90398378601464,
+      "learning_rate": 1.9924361881782625e-05,
+      "loss": 0.4677,
+      "step": 814
+    },
+    {
+      "epoch": 0.16068611987381703,
+      "grad_norm": 0.7372247300869629,
+      "learning_rate": 1.992417147888144e-05,
+      "loss": 0.4529,
+      "step": 815
+    },
+    {
+      "epoch": 0.1608832807570978,
+      "grad_norm": 0.8887941980027961,
+      "learning_rate": 1.992398083754404e-05,
+      "loss": 0.4629,
+      "step": 816
+    },
+    {
+      "epoch": 0.16108044164037855,
+      "grad_norm": 0.6535394579923204,
+      "learning_rate": 1.992378995777501e-05,
+      "loss": 0.4515,
+      "step": 817
+    },
+    {
+      "epoch": 0.1612776025236593,
+      "grad_norm": 0.84734882482852,
+      "learning_rate": 1.9923598839578937e-05,
+      "loss": 0.4263,
+      "step": 818
+    },
+    {
+      "epoch": 0.16147476340694006,
+      "grad_norm": 0.7925717387308715,
+      "learning_rate": 1.9923407482960408e-05,
+      "loss": 0.4775,
+      "step": 819
+    },
+    {
+      "epoch": 0.16167192429022081,
+      "grad_norm": 0.8742951540598778,
+      "learning_rate": 1.9923215887924022e-05,
+      "loss": 0.4637,
+      "step": 820
+    },
+    {
+      "epoch": 0.16186908517350157,
+      "grad_norm": 0.7279557612080059,
+      "learning_rate": 1.9923024054474384e-05,
+      "loss": 0.4636,
+      "step": 821
+    },
+    {
+      "epoch": 0.16206624605678233,
+      "grad_norm": 0.8534513364783904,
+      "learning_rate": 1.99228319826161e-05,
+      "loss": 0.4857,
+      "step": 822
+    },
+    {
+      "epoch": 0.16226340694006308,
+      "grad_norm": 0.7402713942104387,
+      "learning_rate": 1.992263967235379e-05,
+      "loss": 0.481,
+      "step": 823
+    },
+    {
+      "epoch": 0.16246056782334384,
+      "grad_norm": 0.7153338719389193,
+      "learning_rate": 1.992244712369207e-05,
+      "loss": 0.4457,
+      "step": 824
+    },
+    {
+      "epoch": 0.1626577287066246,
+      "grad_norm": 1.4728339073471215,
+      "learning_rate": 1.9922254336635567e-05,
+      "loss": 0.4514,
+      "step": 825
+    },
+    {
+      "epoch": 0.16285488958990535,
+      "grad_norm": 0.8791659930736712,
+      "learning_rate": 1.9922061311188914e-05,
+      "loss": 0.4852,
+      "step": 826
+    },
+    {
+      "epoch": 0.16305205047318613,
+      "grad_norm": 0.7681632672109859,
+      "learning_rate": 1.9921868047356747e-05,
+      "loss": 0.4869,
+      "step": 827
+    },
+    {
+      "epoch": 0.1632492113564669,
+      "grad_norm": 0.8229171295277476,
+      "learning_rate": 1.992167454514371e-05,
+      "loss": 0.4828,
+      "step": 828
+    },
+    {
+      "epoch": 0.16344637223974764,
+      "grad_norm": 0.7529692381939216,
+      "learning_rate": 1.9921480804554453e-05,
+      "loss": 0.46,
+      "step": 829
+    },
+    {
+      "epoch": 0.1636435331230284,
+      "grad_norm": 0.9423934940595431,
+      "learning_rate": 1.9921286825593632e-05,
+      "loss": 0.4418,
+      "step": 830
+    },
+    {
+      "epoch": 0.16384069400630916,
+      "grad_norm": 0.8322161382525551,
+      "learning_rate": 1.9921092608265902e-05,
+      "loss": 0.4489,
+      "step": 831
+    },
+    {
+      "epoch": 0.1640378548895899,
+      "grad_norm": 0.8265863211632591,
+      "learning_rate": 1.9920898152575932e-05,
+      "loss": 0.457,
+      "step": 832
+    },
+    {
+      "epoch": 0.16423501577287067,
+      "grad_norm": 0.8016187387195418,
+      "learning_rate": 1.99207034585284e-05,
+      "loss": 0.4935,
+      "step": 833
+    },
+    {
+      "epoch": 0.16443217665615142,
+      "grad_norm": 0.7562870006268851,
+      "learning_rate": 1.992050852612797e-05,
+      "loss": 0.4901,
+      "step": 834
+    },
+    {
+      "epoch": 0.16462933753943218,
+      "grad_norm": 0.8037484145010387,
+      "learning_rate": 1.992031335537934e-05,
+      "loss": 0.4905,
+      "step": 835
+    },
+    {
+      "epoch": 0.16482649842271294,
+      "grad_norm": 0.8098114037544843,
+      "learning_rate": 1.9920117946287193e-05,
+      "loss": 0.465,
+      "step": 836
+    },
+    {
+      "epoch": 0.1650236593059937,
+      "grad_norm": 0.8269800376599079,
+      "learning_rate": 1.991992229885622e-05,
+      "loss": 0.4735,
+      "step": 837
+    },
+    {
+      "epoch": 0.16522082018927445,
+      "grad_norm": 0.7558534353840859,
+      "learning_rate": 1.9919726413091127e-05,
+      "loss": 0.4824,
+      "step": 838
+    },
+    {
+      "epoch": 0.1654179810725552,
+      "grad_norm": 0.766057843711096,
+      "learning_rate": 1.9919530288996617e-05,
+      "loss": 0.4509,
+      "step": 839
+    },
+    {
+      "epoch": 0.16561514195583596,
+      "grad_norm": 0.8748898217978727,
+      "learning_rate": 1.9919333926577406e-05,
+      "loss": 0.442,
+      "step": 840
+    },
+    {
+      "epoch": 0.16581230283911672,
+      "grad_norm": 0.8934360442693702,
+      "learning_rate": 1.9919137325838208e-05,
+      "loss": 0.5239,
+      "step": 841
+    },
+    {
+      "epoch": 0.16600946372239747,
+      "grad_norm": 0.8552477164679089,
+      "learning_rate": 1.9918940486783752e-05,
+      "loss": 0.4645,
+      "step": 842
+    },
+    {
+      "epoch": 0.16620662460567823,
+      "grad_norm": 0.7957360554655829,
+      "learning_rate": 1.9918743409418756e-05,
+      "loss": 0.4883,
+      "step": 843
+    },
+    {
+      "epoch": 0.16640378548895898,
+      "grad_norm": 0.6874695944020062,
+      "learning_rate": 1.9918546093747965e-05,
+      "loss": 0.4179,
+      "step": 844
+    },
+    {
+      "epoch": 0.16660094637223974,
+      "grad_norm": 0.7762889366838219,
+      "learning_rate": 1.991834853977612e-05,
+      "loss": 0.4905,
+      "step": 845
+    },
+    {
+      "epoch": 0.1667981072555205,
+      "grad_norm": 0.7696965320259346,
+      "learning_rate": 1.9918150747507963e-05,
+      "loss": 0.4404,
+      "step": 846
+    },
+    {
+      "epoch": 0.16699526813880125,
+      "grad_norm": 0.8072545032949946,
+      "learning_rate": 1.9917952716948243e-05,
+      "loss": 0.4681,
+      "step": 847
+    },
+    {
+      "epoch": 0.167192429022082,
+      "grad_norm": 0.8777353512586558,
+      "learning_rate": 1.9917754448101725e-05,
+      "loss": 0.4939,
+      "step": 848
+    },
+    {
+      "epoch": 0.16738958990536276,
+      "grad_norm": 0.7082853265302725,
+      "learning_rate": 1.991755594097317e-05,
+      "loss": 0.4543,
+      "step": 849
+    },
+    {
+      "epoch": 0.16758675078864355,
+      "grad_norm": 0.8604796366799073,
+      "learning_rate": 1.9917357195567347e-05,
+      "loss": 0.4454,
+      "step": 850
+    },
+    {
+      "epoch": 0.1677839116719243,
+      "grad_norm": 0.6898391954794304,
+      "learning_rate": 1.991715821188903e-05,
+      "loss": 0.4238,
+      "step": 851
+    },
+    {
+      "epoch": 0.16798107255520506,
+      "grad_norm": 0.8800493119077631,
+      "learning_rate": 1.9916958989943002e-05,
+      "loss": 0.4569,
+      "step": 852
+    },
+    {
+      "epoch": 0.1681782334384858,
+      "grad_norm": 0.8417073275173806,
+      "learning_rate": 1.9916759529734046e-05,
+      "loss": 0.4777,
+      "step": 853
+    },
+    {
+      "epoch": 0.16837539432176657,
+      "grad_norm": 0.7791244231651858,
+      "learning_rate": 1.9916559831266957e-05,
+      "loss": 0.4544,
+      "step": 854
+    },
+    {
+      "epoch": 0.16857255520504733,
+      "grad_norm": 0.8140536453201463,
+      "learning_rate": 1.9916359894546534e-05,
+      "loss": 0.4634,
+      "step": 855
+    },
+    {
+      "epoch": 0.16876971608832808,
+      "grad_norm": 0.7303022471726613,
+      "learning_rate": 1.9916159719577577e-05,
+      "loss": 0.4519,
+      "step": 856
+    },
+    {
+      "epoch": 0.16896687697160884,
+      "grad_norm": 0.7729780959678046,
+      "learning_rate": 1.9915959306364897e-05,
+      "loss": 0.422,
+      "step": 857
+    },
+    {
+      "epoch": 0.1691640378548896,
+      "grad_norm": 0.8034659806730786,
+      "learning_rate": 1.9915758654913313e-05,
+      "loss": 0.4942,
+      "step": 858
+    },
+    {
+      "epoch": 0.16936119873817035,
+      "grad_norm": 0.7897233162629976,
+      "learning_rate": 1.991555776522764e-05,
+      "loss": 0.4762,
+      "step": 859
+    },
+    {
+      "epoch": 0.1695583596214511,
+      "grad_norm": 0.7803962061080867,
+      "learning_rate": 1.9915356637312704e-05,
+      "loss": 0.4649,
+      "step": 860
+    },
+    {
+      "epoch": 0.16975552050473186,
+      "grad_norm": 0.9547947469437102,
+      "learning_rate": 1.991515527117334e-05,
+      "loss": 0.4652,
+      "step": 861
+    },
+    {
+      "epoch": 0.16995268138801262,
+      "grad_norm": 0.8072759008329805,
+      "learning_rate": 1.9914953666814392e-05,
+      "loss": 0.4665,
+      "step": 862
+    },
+    {
+      "epoch": 0.17014984227129337,
+      "grad_norm": 0.8077676711668577,
+      "learning_rate": 1.9914751824240694e-05,
+      "loss": 0.4617,
+      "step": 863
+    },
+    {
+      "epoch": 0.17034700315457413,
+      "grad_norm": 1.293222408833284,
+      "learning_rate": 1.9914549743457096e-05,
+      "loss": 0.4666,
+      "step": 864
+    },
+    {
+      "epoch": 0.17054416403785488,
+      "grad_norm": 0.7923729068722024,
+      "learning_rate": 1.991434742446846e-05,
+      "loss": 0.4261,
+      "step": 865
+    },
+    {
+      "epoch": 0.17074132492113564,
+      "grad_norm": 14.718649448920804,
+      "learning_rate": 1.9914144867279644e-05,
+      "loss": 0.5072,
+      "step": 866
+    },
+    {
+      "epoch": 0.1709384858044164,
+      "grad_norm": 0.831483953129002,
+      "learning_rate": 1.991394207189551e-05,
+      "loss": 0.4281,
+      "step": 867
+    },
+    {
+      "epoch": 0.17113564668769715,
+      "grad_norm": 0.8157367156824404,
+      "learning_rate": 1.9913739038320935e-05,
+      "loss": 0.4886,
+      "step": 868
+    },
+    {
+      "epoch": 0.1713328075709779,
+      "grad_norm": 0.747934401454802,
+      "learning_rate": 1.99135357665608e-05,
+      "loss": 0.464,
+      "step": 869
+    },
+    {
+      "epoch": 0.17152996845425866,
+      "grad_norm": 0.7940635147564067,
+      "learning_rate": 1.991333225661998e-05,
+      "loss": 0.446,
+      "step": 870
+    },
+    {
+      "epoch": 0.17172712933753942,
+      "grad_norm": 0.7556559635824,
+      "learning_rate": 1.9913128508503373e-05,
+      "loss": 0.4187,
+      "step": 871
+    },
+    {
+      "epoch": 0.1719242902208202,
+      "grad_norm": 0.937585969395132,
+      "learning_rate": 1.991292452221587e-05,
+      "loss": 0.4627,
+      "step": 872
+    },
+    {
+      "epoch": 0.17212145110410096,
+      "grad_norm": 0.7287727570911343,
+      "learning_rate": 1.9912720297762372e-05,
+      "loss": 0.4412,
+      "step": 873
+    },
+    {
+      "epoch": 0.17231861198738171,
+      "grad_norm": 1.4483761847244239,
+      "learning_rate": 1.9912515835147785e-05,
+      "loss": 0.4702,
+      "step": 874
+    },
+    {
+      "epoch": 0.17251577287066247,
+      "grad_norm": 0.7776251644712487,
+      "learning_rate": 1.9912311134377023e-05,
+      "loss": 0.4827,
+      "step": 875
+    },
+    {
+      "epoch": 0.17271293375394323,
+      "grad_norm": 0.811467179063533,
+      "learning_rate": 1.9912106195455002e-05,
+      "loss": 0.4303,
+      "step": 876
+    },
+    {
+      "epoch": 0.17291009463722398,
+      "grad_norm": 0.7347947732597557,
+      "learning_rate": 1.991190101838665e-05,
+      "loss": 0.48,
+      "step": 877
+    },
+    {
+      "epoch": 0.17310725552050474,
+      "grad_norm": 0.7700860751914179,
+      "learning_rate": 1.9911695603176896e-05,
+      "loss": 0.4418,
+      "step": 878
+    },
+    {
+      "epoch": 0.1733044164037855,
+      "grad_norm": 0.7424955632937327,
+      "learning_rate": 1.9911489949830665e-05,
+      "loss": 0.4778,
+      "step": 879
+    },
+    {
+      "epoch": 0.17350157728706625,
+      "grad_norm": 0.6988442966765853,
+      "learning_rate": 1.9911284058352916e-05,
+      "loss": 0.4955,
+      "step": 880
+    },
+    {
+      "epoch": 0.173698738170347,
+      "grad_norm": 1.201129785880802,
+      "learning_rate": 1.9911077928748577e-05,
+      "loss": 0.5094,
+      "step": 881
+    },
+    {
+      "epoch": 0.17389589905362776,
+      "grad_norm": 0.7077413544925765,
+      "learning_rate": 1.9910871561022617e-05,
+      "loss": 0.4489,
+      "step": 882
+    },
+    {
+      "epoch": 0.17409305993690852,
+      "grad_norm": 0.7332451772736706,
+      "learning_rate": 1.9910664955179983e-05,
+      "loss": 0.4479,
+      "step": 883
+    },
+    {
+      "epoch": 0.17429022082018927,
+      "grad_norm": 0.7177585522779164,
+      "learning_rate": 1.9910458111225645e-05,
+      "loss": 0.428,
+      "step": 884
+    },
+    {
+      "epoch": 0.17448738170347003,
+      "grad_norm": 0.6737153923574553,
+      "learning_rate": 1.9910251029164568e-05,
+      "loss": 0.4288,
+      "step": 885
+    },
+    {
+      "epoch": 0.17468454258675079,
+      "grad_norm": 0.8450869761757991,
+      "learning_rate": 1.9910043709001727e-05,
+      "loss": 0.4537,
+      "step": 886
+    },
+    {
+      "epoch": 0.17488170347003154,
+      "grad_norm": 0.6872411051621798,
+      "learning_rate": 1.990983615074211e-05,
+      "loss": 0.4449,
+      "step": 887
+    },
+    {
+      "epoch": 0.1750788643533123,
+      "grad_norm": 0.8721136747808786,
+      "learning_rate": 1.9909628354390697e-05,
+      "loss": 0.4752,
+      "step": 888
+    },
+    {
+      "epoch": 0.17527602523659305,
+      "grad_norm": 0.7655080513181773,
+      "learning_rate": 1.990942031995248e-05,
+      "loss": 0.4868,
+      "step": 889
+    },
+    {
+      "epoch": 0.1754731861198738,
+      "grad_norm": 0.7199528165114193,
+      "learning_rate": 1.9909212047432465e-05,
+      "loss": 0.4528,
+      "step": 890
+    },
+    {
+      "epoch": 0.17567034700315456,
+      "grad_norm": 0.7061497459915089,
+      "learning_rate": 1.990900353683565e-05,
+      "loss": 0.455,
+      "step": 891
+    },
+    {
+      "epoch": 0.17586750788643532,
+      "grad_norm": 2.291375288631725,
+      "learning_rate": 1.990879478816704e-05,
+      "loss": 0.4722,
+      "step": 892
+    },
+    {
+      "epoch": 0.17606466876971608,
+      "grad_norm": 0.8123591618774287,
+      "learning_rate": 1.9908585801431658e-05,
+      "loss": 0.4706,
+      "step": 893
+    },
+    {
+      "epoch": 0.17626182965299683,
+      "grad_norm": 0.785222829757174,
+      "learning_rate": 1.9908376576634526e-05,
+      "loss": 0.4608,
+      "step": 894
+    },
+    {
+      "epoch": 0.17645899053627762,
+      "grad_norm": 0.7633221752691605,
+      "learning_rate": 1.9908167113780665e-05,
+      "loss": 0.4663,
+      "step": 895
+    },
+    {
+      "epoch": 0.17665615141955837,
+      "grad_norm": 0.7931603437941884,
+      "learning_rate": 1.990795741287511e-05,
+      "loss": 0.4979,
+      "step": 896
+    },
+    {
+      "epoch": 0.17685331230283913,
+      "grad_norm": 0.740593821102,
+      "learning_rate": 1.99077474739229e-05,
+      "loss": 0.4595,
+      "step": 897
+    },
+    {
+      "epoch": 0.17705047318611988,
+      "grad_norm": 0.7137127142445181,
+      "learning_rate": 1.9907537296929077e-05,
+      "loss": 0.468,
+      "step": 898
+    },
+    {
+      "epoch": 0.17724763406940064,
+      "grad_norm": 0.7856236565024629,
+      "learning_rate": 1.9907326881898693e-05,
+      "loss": 0.4532,
+      "step": 899
+    },
+    {
+      "epoch": 0.1774447949526814,
+      "grad_norm": 0.8304106213478027,
+      "learning_rate": 1.99071162288368e-05,
+      "loss": 0.4835,
+      "step": 900
+    },
+    {
+      "epoch": 0.17764195583596215,
+      "grad_norm": 0.7075946238835652,
+      "learning_rate": 1.9906905337748466e-05,
+      "loss": 0.4415,
+      "step": 901
+    },
+    {
+      "epoch": 0.1778391167192429,
+      "grad_norm": 0.7255892333472377,
+      "learning_rate": 1.990669420863875e-05,
+      "loss": 0.462,
+      "step": 902
+    },
+    {
+      "epoch": 0.17803627760252366,
+      "grad_norm": 0.7551817312285064,
+      "learning_rate": 1.990648284151273e-05,
+      "loss": 0.4352,
+      "step": 903
+    },
+    {
+      "epoch": 0.17823343848580442,
+      "grad_norm": 0.7462637275359422,
+      "learning_rate": 1.9906271236375478e-05,
+      "loss": 0.4635,
+      "step": 904
+    },
+    {
+      "epoch": 0.17843059936908517,
+      "grad_norm": 0.7285489399185598,
+      "learning_rate": 1.9906059393232088e-05,
+      "loss": 0.4696,
+      "step": 905
+    },
+    {
+      "epoch": 0.17862776025236593,
+      "grad_norm": 0.789466645420734,
+      "learning_rate": 1.990584731208764e-05,
+      "loss": 0.4968,
+      "step": 906
+    },
+    {
+      "epoch": 0.17882492113564669,
+      "grad_norm": 2.968306758137648,
+      "learning_rate": 1.9905634992947235e-05,
+      "loss": 0.491,
+      "step": 907
+    },
+    {
+      "epoch": 0.17902208201892744,
+      "grad_norm": 0.7750070374318527,
+      "learning_rate": 1.990542243581597e-05,
+      "loss": 0.4703,
+      "step": 908
+    },
+    {
+      "epoch": 0.1792192429022082,
+      "grad_norm": 0.695028794150739,
+      "learning_rate": 1.9905209640698952e-05,
+      "loss": 0.4642,
+      "step": 909
+    },
+    {
+      "epoch": 0.17941640378548895,
+      "grad_norm": 0.7695918165395215,
+      "learning_rate": 1.9904996607601303e-05,
+      "loss": 0.4552,
+      "step": 910
+    },
+    {
+      "epoch": 0.1796135646687697,
+      "grad_norm": 0.7202892429327876,
+      "learning_rate": 1.9904783336528128e-05,
+      "loss": 0.4646,
+      "step": 911
+    },
+    {
+      "epoch": 0.17981072555205047,
+      "grad_norm": 0.8135132238328896,
+      "learning_rate": 1.9904569827484556e-05,
+      "loss": 0.4594,
+      "step": 912
+    },
+    {
+      "epoch": 0.18000788643533122,
+      "grad_norm": 0.7554729773822189,
+      "learning_rate": 1.990435608047572e-05,
+      "loss": 0.4008,
+      "step": 913
+    },
+    {
+      "epoch": 0.18020504731861198,
+      "grad_norm": 0.737373759645686,
+      "learning_rate": 1.9904142095506756e-05,
+      "loss": 0.4335,
+      "step": 914
+    },
+    {
+      "epoch": 0.18040220820189273,
+      "grad_norm": 0.8409703840361105,
+      "learning_rate": 1.99039278725828e-05,
+      "loss": 0.4612,
+      "step": 915
+    },
+    {
+      "epoch": 0.1805993690851735,
+      "grad_norm": 0.6825714924533132,
+      "learning_rate": 1.9903713411709003e-05,
+      "loss": 0.4225,
+      "step": 916
+    },
+    {
+      "epoch": 0.18079652996845424,
+      "grad_norm": 0.9821803862497531,
+      "learning_rate": 1.9903498712890516e-05,
+      "loss": 0.4981,
+      "step": 917
+    },
+    {
+      "epoch": 0.18099369085173503,
+      "grad_norm": 0.7273123016078351,
+      "learning_rate": 1.9903283776132495e-05,
+      "loss": 0.4461,
+      "step": 918
+    },
+    {
+      "epoch": 0.18119085173501578,
+      "grad_norm": 0.8665342753995058,
+      "learning_rate": 1.9903068601440106e-05,
+      "loss": 0.4452,
+      "step": 919
+    },
+    {
+      "epoch": 0.18138801261829654,
+      "grad_norm": 0.7822855298339977,
+      "learning_rate": 1.9902853188818518e-05,
+      "loss": 0.4655,
+      "step": 920
+    },
+    {
+      "epoch": 0.1815851735015773,
+      "grad_norm": 0.8357899870781356,
+      "learning_rate": 1.990263753827291e-05,
+      "loss": 0.4416,
+      "step": 921
+    },
+    {
+      "epoch": 0.18178233438485805,
+      "grad_norm": 0.6936838258206445,
+      "learning_rate": 1.990242164980846e-05,
+      "loss": 0.4612,
+      "step": 922
+    },
+    {
+      "epoch": 0.1819794952681388,
+      "grad_norm": 0.8050591566486436,
+      "learning_rate": 1.9902205523430353e-05,
+      "loss": 0.4465,
+      "step": 923
+    },
+    {
+      "epoch": 0.18217665615141956,
+      "grad_norm": 1.0415997120130642,
+      "learning_rate": 1.9901989159143786e-05,
+      "loss": 0.4404,
+      "step": 924
+    },
+    {
+      "epoch": 0.18237381703470032,
+      "grad_norm": 0.7976243404123774,
+      "learning_rate": 1.9901772556953958e-05,
+      "loss": 0.4767,
+      "step": 925
+    },
+    {
+      "epoch": 0.18257097791798108,
+      "grad_norm": 0.7036095691664724,
+      "learning_rate": 1.9901555716866067e-05,
+      "loss": 0.4479,
+      "step": 926
+    },
+    {
+      "epoch": 0.18276813880126183,
+      "grad_norm": 0.7746405790517535,
+      "learning_rate": 1.9901338638885327e-05,
+      "loss": 0.4514,
+      "step": 927
+    },
+    {
+      "epoch": 0.1829652996845426,
+      "grad_norm": 0.7905763277511705,
+      "learning_rate": 1.9901121323016955e-05,
+      "loss": 0.4947,
+      "step": 928
+    },
+    {
+      "epoch": 0.18316246056782334,
+      "grad_norm": 0.8011819612935877,
+      "learning_rate": 1.9900903769266167e-05,
+      "loss": 0.4711,
+      "step": 929
+    },
+    {
+      "epoch": 0.1833596214511041,
+      "grad_norm": 0.7885618953274046,
+      "learning_rate": 1.9900685977638194e-05,
+      "loss": 0.4474,
+      "step": 930
+    },
+    {
+      "epoch": 0.18355678233438485,
+      "grad_norm": 0.8628719762309652,
+      "learning_rate": 1.9900467948138266e-05,
+      "loss": 0.484,
+      "step": 931
+    },
+    {
+      "epoch": 0.1837539432176656,
+      "grad_norm": 0.8426852896992455,
+      "learning_rate": 1.9900249680771622e-05,
+      "loss": 0.4597,
+      "step": 932
+    },
+    {
+      "epoch": 0.18395110410094637,
+      "grad_norm": 0.8844622895009938,
+      "learning_rate": 1.990003117554351e-05,
+      "loss": 0.4837,
+      "step": 933
+    },
+    {
+      "epoch": 0.18414826498422712,
+      "grad_norm": 0.7099824020200362,
+      "learning_rate": 1.9899812432459175e-05,
+      "loss": 0.4425,
+      "step": 934
+    },
+    {
+      "epoch": 0.18434542586750788,
+      "grad_norm": 0.7629633321327819,
+      "learning_rate": 1.9899593451523875e-05,
+      "loss": 0.4526,
+      "step": 935
+    },
+    {
+      "epoch": 0.18454258675078863,
+      "grad_norm": 0.8792149544075311,
+      "learning_rate": 1.989937423274287e-05,
+      "loss": 0.4682,
+      "step": 936
+    },
+    {
+      "epoch": 0.1847397476340694,
+      "grad_norm": 0.6838639839900289,
+      "learning_rate": 1.9899154776121424e-05,
+      "loss": 0.4261,
+      "step": 937
+    },
+    {
+      "epoch": 0.18493690851735015,
+      "grad_norm": 0.706142676214522,
+      "learning_rate": 1.9898935081664814e-05,
+      "loss": 0.4805,
+      "step": 938
+    },
+    {
+      "epoch": 0.1851340694006309,
+      "grad_norm": 0.6859539690000207,
+      "learning_rate": 1.9898715149378317e-05,
+      "loss": 0.4706,
+      "step": 939
+    },
+    {
+      "epoch": 0.18533123028391169,
+      "grad_norm": 0.6812899026189113,
+      "learning_rate": 1.989849497926722e-05,
+      "loss": 0.4256,
+      "step": 940
+    },
+    {
+      "epoch": 0.18552839116719244,
+      "grad_norm": 0.6577900927420254,
+      "learning_rate": 1.989827457133681e-05,
+      "loss": 0.4463,
+      "step": 941
+    },
+    {
+      "epoch": 0.1857255520504732,
+      "grad_norm": 0.7514857045339612,
+      "learning_rate": 1.9898053925592376e-05,
+      "loss": 0.4591,
+      "step": 942
+    },
+    {
+      "epoch": 0.18592271293375395,
+      "grad_norm": 0.6549094044973097,
+      "learning_rate": 1.9897833042039233e-05,
+      "loss": 0.4623,
+      "step": 943
+    },
+    {
+      "epoch": 0.1861198738170347,
+      "grad_norm": 0.7310174851562488,
+      "learning_rate": 1.9897611920682676e-05,
+      "loss": 0.4487,
+      "step": 944
+    },
+    {
+      "epoch": 0.18631703470031546,
+      "grad_norm": 0.6963083495505222,
+      "learning_rate": 1.9897390561528024e-05,
+      "loss": 0.4947,
+      "step": 945
+    },
+    {
+      "epoch": 0.18651419558359622,
+      "grad_norm": 0.6460049210995387,
+      "learning_rate": 1.9897168964580594e-05,
+      "loss": 0.4631,
+      "step": 946
+    },
+    {
+      "epoch": 0.18671135646687698,
+      "grad_norm": 0.7279929742381652,
+      "learning_rate": 1.9896947129845707e-05,
+      "loss": 0.454,
+      "step": 947
+    },
+    {
+      "epoch": 0.18690851735015773,
+      "grad_norm": 0.7314455893730071,
+      "learning_rate": 1.9896725057328695e-05,
+      "loss": 0.468,
+      "step": 948
+    },
+    {
+      "epoch": 0.1871056782334385,
+      "grad_norm": 0.707927540023582,
+      "learning_rate": 1.9896502747034894e-05,
+      "loss": 0.475,
+      "step": 949
+    },
+    {
+      "epoch": 0.18730283911671924,
+      "grad_norm": 0.8331091743901959,
+      "learning_rate": 1.989628019896965e-05,
+      "loss": 0.467,
+      "step": 950
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 0.7225918783951755,
+      "learning_rate": 1.98960574131383e-05,
+      "loss": 0.477,
+      "step": 951
+    },
+    {
+      "epoch": 0.18769716088328076,
+      "grad_norm": 0.6757638681754824,
+      "learning_rate": 1.9895834389546204e-05,
+      "loss": 0.4572,
+      "step": 952
+    },
+    {
+      "epoch": 0.1878943217665615,
+      "grad_norm": 0.7206616057098645,
+      "learning_rate": 1.9895611128198714e-05,
+      "loss": 0.4644,
+      "step": 953
+    },
+    {
+      "epoch": 0.18809148264984227,
+      "grad_norm": 0.6474073350668482,
+      "learning_rate": 1.9895387629101203e-05,
+      "loss": 0.4484,
+      "step": 954
+    },
+    {
+      "epoch": 0.18828864353312302,
+      "grad_norm": 0.7027079831141624,
+      "learning_rate": 1.989516389225903e-05,
+      "loss": 0.4845,
+      "step": 955
+    },
+    {
+      "epoch": 0.18848580441640378,
+      "grad_norm": 0.6725377762031143,
+      "learning_rate": 1.9894939917677577e-05,
+      "loss": 0.4592,
+      "step": 956
+    },
+    {
+      "epoch": 0.18868296529968454,
+      "grad_norm": 0.6747407598105398,
+      "learning_rate": 1.9894715705362227e-05,
+      "loss": 0.4316,
+      "step": 957
+    },
+    {
+      "epoch": 0.1888801261829653,
+      "grad_norm": 0.990350135133828,
+      "learning_rate": 1.9894491255318362e-05,
+      "loss": 0.4676,
+      "step": 958
+    },
+    {
+      "epoch": 0.18907728706624605,
+      "grad_norm": 0.6877906835243988,
+      "learning_rate": 1.9894266567551378e-05,
+      "loss": 0.469,
+      "step": 959
+    },
+    {
+      "epoch": 0.1892744479495268,
+      "grad_norm": 0.6882946457373598,
+      "learning_rate": 1.989404164206667e-05,
+      "loss": 0.4932,
+      "step": 960
+    },
+    {
+      "epoch": 0.18947160883280756,
+      "grad_norm": 0.7061372635020192,
+      "learning_rate": 1.9893816478869646e-05,
+      "loss": 0.484,
+      "step": 961
+    },
+    {
+      "epoch": 0.18966876971608831,
+      "grad_norm": 0.6724470559392719,
+      "learning_rate": 1.989359107796571e-05,
+      "loss": 0.4332,
+      "step": 962
+    },
+    {
+      "epoch": 0.1898659305993691,
+      "grad_norm": 0.7625542937410119,
+      "learning_rate": 1.9893365439360285e-05,
+      "loss": 0.4763,
+      "step": 963
+    },
+    {
+      "epoch": 0.19006309148264985,
+      "grad_norm": 0.6953307451809908,
+      "learning_rate": 1.9893139563058786e-05,
+      "loss": 0.4958,
+      "step": 964
+    },
+    {
+      "epoch": 0.1902602523659306,
+      "grad_norm": 0.7466843034274665,
+      "learning_rate": 1.9892913449066643e-05,
+      "loss": 0.4958,
+      "step": 965
+    },
+    {
+      "epoch": 0.19045741324921137,
+      "grad_norm": 0.6585282349015559,
+      "learning_rate": 1.9892687097389288e-05,
+      "loss": 0.4224,
+      "step": 966
+    },
+    {
+      "epoch": 0.19065457413249212,
+      "grad_norm": 0.6883957756753183,
+      "learning_rate": 1.9892460508032158e-05,
+      "loss": 0.4603,
+      "step": 967
+    },
+    {
+      "epoch": 0.19085173501577288,
+      "grad_norm": 0.7020929418484416,
+      "learning_rate": 1.9892233681000696e-05,
+      "loss": 0.4644,
+      "step": 968
+    },
+    {
+      "epoch": 0.19104889589905363,
+      "grad_norm": 0.659651299875094,
+      "learning_rate": 1.9892006616300358e-05,
+      "loss": 0.4378,
+      "step": 969
+    },
+    {
+      "epoch": 0.1912460567823344,
+      "grad_norm": 0.6871160793712359,
+      "learning_rate": 1.989177931393659e-05,
+      "loss": 0.4432,
+      "step": 970
+    },
+    {
+      "epoch": 0.19144321766561515,
+      "grad_norm": 0.7875222747543351,
+      "learning_rate": 1.989155177391486e-05,
+      "loss": 0.4698,
+      "step": 971
+    },
+    {
+      "epoch": 0.1916403785488959,
+      "grad_norm": 0.6688379722075266,
+      "learning_rate": 1.9891323996240633e-05,
+      "loss": 0.4608,
+      "step": 972
+    },
+    {
+      "epoch": 0.19183753943217666,
+      "grad_norm": 0.7330583737954934,
+      "learning_rate": 1.9891095980919383e-05,
+      "loss": 0.4809,
+      "step": 973
+    },
+    {
+      "epoch": 0.1920347003154574,
+      "grad_norm": 0.6958413705522188,
+      "learning_rate": 1.9890867727956587e-05,
+      "loss": 0.4627,
+      "step": 974
+    },
+    {
+      "epoch": 0.19223186119873817,
+      "grad_norm": 1.5897290586527348,
+      "learning_rate": 1.9890639237357726e-05,
+      "loss": 0.5164,
+      "step": 975
+    },
+    {
+      "epoch": 0.19242902208201892,
+      "grad_norm": 0.7313960505685592,
+      "learning_rate": 1.98904105091283e-05,
+      "loss": 0.4792,
+      "step": 976
+    },
+    {
+      "epoch": 0.19262618296529968,
+      "grad_norm": 0.7063385670757377,
+      "learning_rate": 1.989018154327379e-05,
+      "loss": 0.4605,
+      "step": 977
+    },
+    {
+      "epoch": 0.19282334384858044,
+      "grad_norm": 0.708185416892424,
+      "learning_rate": 1.9889952339799704e-05,
+      "loss": 0.439,
+      "step": 978
+    },
+    {
+      "epoch": 0.1930205047318612,
+      "grad_norm": 0.6540312802061391,
+      "learning_rate": 1.9889722898711546e-05,
+      "loss": 0.4435,
+      "step": 979
+    },
+    {
+      "epoch": 0.19321766561514195,
+      "grad_norm": 0.8450940057465913,
+      "learning_rate": 1.9889493220014837e-05,
+      "loss": 0.4531,
+      "step": 980
+    },
+    {
+      "epoch": 0.1934148264984227,
+      "grad_norm": 0.7708701148620279,
+      "learning_rate": 1.9889263303715086e-05,
+      "loss": 0.4704,
+      "step": 981
+    },
+    {
+      "epoch": 0.19361198738170346,
+      "grad_norm": 0.621772319575249,
+      "learning_rate": 1.9889033149817823e-05,
+      "loss": 0.4514,
+      "step": 982
+    },
+    {
+      "epoch": 0.19380914826498422,
+      "grad_norm": 0.737304053648671,
+      "learning_rate": 1.9888802758328574e-05,
+      "loss": 0.458,
+      "step": 983
+    },
+    {
+      "epoch": 0.19400630914826497,
+      "grad_norm": 0.6741704564219441,
+      "learning_rate": 1.9888572129252875e-05,
+      "loss": 0.4394,
+      "step": 984
+    },
+    {
+      "epoch": 0.19420347003154576,
+      "grad_norm": 0.7931415567269723,
+      "learning_rate": 1.9888341262596266e-05,
+      "loss": 0.4729,
+      "step": 985
+    },
+    {
+      "epoch": 0.1944006309148265,
+      "grad_norm": 0.731130849168293,
+      "learning_rate": 1.9888110158364296e-05,
+      "loss": 0.4546,
+      "step": 986
+    },
+    {
+      "epoch": 0.19459779179810727,
+      "grad_norm": 0.6910048306098391,
+      "learning_rate": 1.988787881656252e-05,
+      "loss": 0.4799,
+      "step": 987
+    },
+    {
+      "epoch": 0.19479495268138802,
+      "grad_norm": 0.6876331628207631,
+      "learning_rate": 1.988764723719649e-05,
+      "loss": 0.4642,
+      "step": 988
+    },
+    {
+      "epoch": 0.19499211356466878,
+      "grad_norm": 0.6926517120324639,
+      "learning_rate": 1.988741542027177e-05,
+      "loss": 0.4482,
+      "step": 989
+    },
+    {
+      "epoch": 0.19518927444794953,
+      "grad_norm": 0.6914878618737181,
+      "learning_rate": 1.9887183365793935e-05,
+      "loss": 0.4126,
+      "step": 990
+    },
+    {
+      "epoch": 0.1953864353312303,
+      "grad_norm": 0.6866079399218992,
+      "learning_rate": 1.9886951073768557e-05,
+      "loss": 0.4373,
+      "step": 991
+    },
+    {
+      "epoch": 0.19558359621451105,
+      "grad_norm": 0.6833596593294384,
+      "learning_rate": 1.988671854420122e-05,
+      "loss": 0.4332,
+      "step": 992
+    },
+    {
+      "epoch": 0.1957807570977918,
+      "grad_norm": 0.7160110640102156,
+      "learning_rate": 1.9886485777097505e-05,
+      "loss": 0.472,
+      "step": 993
+    },
+    {
+      "epoch": 0.19597791798107256,
+      "grad_norm": 0.701325188265206,
+      "learning_rate": 1.9886252772463008e-05,
+      "loss": 0.4854,
+      "step": 994
+    },
+    {
+      "epoch": 0.19617507886435331,
+      "grad_norm": 0.6802161665541124,
+      "learning_rate": 1.9886019530303328e-05,
+      "loss": 0.4466,
+      "step": 995
+    },
+    {
+      "epoch": 0.19637223974763407,
+      "grad_norm": 0.7478910314629341,
+      "learning_rate": 1.9885786050624066e-05,
+      "loss": 0.5023,
+      "step": 996
+    },
+    {
+      "epoch": 0.19656940063091483,
+      "grad_norm": 0.7673069709477471,
+      "learning_rate": 1.9885552333430834e-05,
+      "loss": 0.5205,
+      "step": 997
+    },
+    {
+      "epoch": 0.19676656151419558,
+      "grad_norm": 0.7036229450957218,
+      "learning_rate": 1.9885318378729247e-05,
+      "loss": 0.4693,
+      "step": 998
+    },
+    {
+      "epoch": 0.19696372239747634,
+      "grad_norm": 0.7059304315094523,
+      "learning_rate": 1.9885084186524922e-05,
+      "loss": 0.4687,
+      "step": 999
+    },
+    {
+      "epoch": 0.1971608832807571,
+      "grad_norm": 0.701201649068646,
+      "learning_rate": 1.988484975682349e-05,
+      "loss": 0.4727,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19735804416403785,
+      "grad_norm": 0.6840771386046793,
+      "learning_rate": 1.9884615089630584e-05,
+      "loss": 0.4636,
+      "step": 1001
+    },
+    {
+      "epoch": 0.1975552050473186,
+      "grad_norm": 0.7119449126029943,
+      "learning_rate": 1.988438018495184e-05,
+      "loss": 0.465,
+      "step": 1002
+    },
+    {
+      "epoch": 0.19775236593059936,
+      "grad_norm": 0.7547620511614711,
+      "learning_rate": 1.9884145042792905e-05,
+      "loss": 0.4701,
+      "step": 1003
+    },
+    {
+      "epoch": 0.19794952681388012,
+      "grad_norm": 0.6957906304621474,
+      "learning_rate": 1.9883909663159424e-05,
+      "loss": 0.4531,
+      "step": 1004
+    },
+    {
+      "epoch": 0.19814668769716087,
+      "grad_norm": 0.8261687307427324,
+      "learning_rate": 1.9883674046057054e-05,
+      "loss": 0.4989,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19834384858044163,
+      "grad_norm": 0.6898115598695652,
+      "learning_rate": 1.9883438191491453e-05,
+      "loss": 0.4537,
+      "step": 1006
+    },
+    {
+      "epoch": 0.19854100946372238,
+      "grad_norm": 0.818376757438343,
+      "learning_rate": 1.9883202099468294e-05,
+      "loss": 0.4706,
+      "step": 1007
+    },
+    {
+      "epoch": 0.19873817034700317,
+      "grad_norm": 0.7807033836708956,
+      "learning_rate": 1.988296576999324e-05,
+      "loss": 0.4311,
+      "step": 1008
+    },
+    {
+      "epoch": 0.19893533123028392,
+      "grad_norm": 0.6424414464365055,
+      "learning_rate": 1.988272920307198e-05,
+      "loss": 0.4303,
+      "step": 1009
+    },
+    {
+      "epoch": 0.19913249211356468,
+      "grad_norm": 0.7799809923491292,
+      "learning_rate": 1.9882492398710192e-05,
+      "loss": 0.5081,
+      "step": 1010
+    },
+    {
+      "epoch": 0.19932965299684544,
+      "grad_norm": 0.7784674048585015,
+      "learning_rate": 1.9882255356913563e-05,
+      "loss": 0.5344,
+      "step": 1011
+    },
+    {
+      "epoch": 0.1995268138801262,
+      "grad_norm": 0.7082795210370787,
+      "learning_rate": 1.988201807768779e-05,
+      "loss": 0.4639,
+      "step": 1012
+    },
+    {
+      "epoch": 0.19972397476340695,
+      "grad_norm": 0.6999909023367687,
+      "learning_rate": 1.9881780561038583e-05,
+      "loss": 0.4525,
+      "step": 1013
+    },
+    {
+      "epoch": 0.1999211356466877,
+      "grad_norm": 0.7285326225615099,
+      "learning_rate": 1.988154280697163e-05,
+      "loss": 0.4278,
+      "step": 1014
+    },
+    {
+      "epoch": 0.20011829652996846,
+      "grad_norm": 0.7638833349383104,
+      "learning_rate": 1.988130481549266e-05,
+      "loss": 0.4622,
+      "step": 1015
+    },
+    {
+      "epoch": 0.20031545741324921,
+      "grad_norm": 0.6387623204916645,
+      "learning_rate": 1.9881066586607384e-05,
+      "loss": 0.4361,
+      "step": 1016
+    },
+    {
+      "epoch": 0.20051261829652997,
+      "grad_norm": 0.7537122531684395,
+      "learning_rate": 1.9880828120321523e-05,
+      "loss": 0.5057,
+      "step": 1017
+    },
+    {
+      "epoch": 0.20070977917981073,
+      "grad_norm": 0.6862604234144699,
+      "learning_rate": 1.988058941664081e-05,
+      "loss": 0.4416,
+      "step": 1018
+    },
+    {
+      "epoch": 0.20090694006309148,
+      "grad_norm": 0.7489284571072279,
+      "learning_rate": 1.988035047557098e-05,
+      "loss": 0.4998,
+      "step": 1019
+    },
+    {
+      "epoch": 0.20110410094637224,
+      "grad_norm": 0.7523322609228381,
+      "learning_rate": 1.9880111297117772e-05,
+      "loss": 0.4551,
+      "step": 1020
+    },
+    {
+      "epoch": 0.201301261829653,
+      "grad_norm": 0.7501370176061963,
+      "learning_rate": 1.9879871881286936e-05,
+      "loss": 0.4416,
+      "step": 1021
+    },
+    {
+      "epoch": 0.20149842271293375,
+      "grad_norm": 0.7351622592876997,
+      "learning_rate": 1.9879632228084224e-05,
+      "loss": 0.4639,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2016955835962145,
+      "grad_norm": 0.7423971116335808,
+      "learning_rate": 1.9879392337515385e-05,
+      "loss": 0.488,
+      "step": 1023
+    },
+    {
+      "epoch": 0.20189274447949526,
+      "grad_norm": 0.6751923754084928,
+      "learning_rate": 1.9879152209586193e-05,
+      "loss": 0.3938,
+      "step": 1024
+    },
+    {
+      "epoch": 0.20208990536277602,
+      "grad_norm": 0.722495594329759,
+      "learning_rate": 1.987891184430241e-05,
+      "loss": 0.4837,
+      "step": 1025
+    },
+    {
+      "epoch": 0.20228706624605677,
+      "grad_norm": 0.6782801414805263,
+      "learning_rate": 1.9878671241669824e-05,
+      "loss": 0.4781,
+      "step": 1026
+    },
+    {
+      "epoch": 0.20248422712933753,
+      "grad_norm": 0.7053528722513924,
+      "learning_rate": 1.98784304016942e-05,
+      "loss": 0.4424,
+      "step": 1027
+    },
+    {
+      "epoch": 0.20268138801261829,
+      "grad_norm": 0.6877822028898056,
+      "learning_rate": 1.987818932438133e-05,
+      "loss": 0.4585,
+      "step": 1028
+    },
+    {
+      "epoch": 0.20287854889589904,
+      "grad_norm": 0.7414751335330734,
+      "learning_rate": 1.9877948009737006e-05,
+      "loss": 0.4581,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2030757097791798,
+      "grad_norm": 0.6447511500376699,
+      "learning_rate": 1.9877706457767028e-05,
+      "loss": 0.4246,
+      "step": 1030
+    },
+    {
+      "epoch": 0.20327287066246058,
+      "grad_norm": 0.7084092307640751,
+      "learning_rate": 1.9877464668477195e-05,
+      "loss": 0.4779,
+      "step": 1031
+    },
+    {
+      "epoch": 0.20347003154574134,
+      "grad_norm": 0.753567669793824,
+      "learning_rate": 1.987722264187332e-05,
+      "loss": 0.4658,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2036671924290221,
+      "grad_norm": 0.733202685122525,
+      "learning_rate": 1.987698037796122e-05,
+      "loss": 0.4506,
+      "step": 1033
+    },
+    {
+      "epoch": 0.20386435331230285,
+      "grad_norm": 0.6827619467500508,
+      "learning_rate": 1.987673787674671e-05,
+      "loss": 0.4124,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2040615141955836,
+      "grad_norm": 0.7015654555624304,
+      "learning_rate": 1.987649513823562e-05,
+      "loss": 0.4498,
+      "step": 1035
+    },
+    {
+      "epoch": 0.20425867507886436,
+      "grad_norm": 0.6474610550353069,
+      "learning_rate": 1.987625216243378e-05,
+      "loss": 0.4092,
+      "step": 1036
+    },
+    {
+      "epoch": 0.20445583596214512,
+      "grad_norm": 0.7064160426450053,
+      "learning_rate": 1.987600894934703e-05,
+      "loss": 0.4379,
+      "step": 1037
+    },
+    {
+      "epoch": 0.20465299684542587,
+      "grad_norm": 0.6676749684842929,
+      "learning_rate": 1.987576549898121e-05,
+      "loss": 0.439,
+      "step": 1038
+    },
+    {
+      "epoch": 0.20485015772870663,
+      "grad_norm": 0.7077339162356132,
+      "learning_rate": 1.987552181134217e-05,
+      "loss": 0.4433,
+      "step": 1039
+    },
+    {
+      "epoch": 0.20504731861198738,
+      "grad_norm": 0.7923537698393385,
+      "learning_rate": 1.9875277886435768e-05,
+      "loss": 0.4747,
+      "step": 1040
+    },
+    {
+      "epoch": 0.20524447949526814,
+      "grad_norm": 0.7264062737086919,
+      "learning_rate": 1.9875033724267863e-05,
+      "loss": 0.4479,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2054416403785489,
+      "grad_norm": 0.7252182386797893,
+      "learning_rate": 1.987478932484432e-05,
+      "loss": 0.4701,
+      "step": 1042
+    },
+    {
+      "epoch": 0.20563880126182965,
+      "grad_norm": 0.7474427366745425,
+      "learning_rate": 1.9874544688171008e-05,
+      "loss": 0.4749,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2058359621451104,
+      "grad_norm": 0.8821899810782728,
+      "learning_rate": 1.9874299814253813e-05,
+      "loss": 0.512,
+      "step": 1044
+    },
+    {
+      "epoch": 0.20603312302839116,
+      "grad_norm": 0.7196304240914877,
+      "learning_rate": 1.9874054703098608e-05,
+      "loss": 0.44,
+      "step": 1045
+    },
+    {
+      "epoch": 0.20623028391167192,
+      "grad_norm": 0.6964106812809121,
+      "learning_rate": 1.987380935471129e-05,
+      "loss": 0.4283,
+      "step": 1046
+    },
+    {
+      "epoch": 0.20642744479495267,
+      "grad_norm": 0.7918473693269067,
+      "learning_rate": 1.9873563769097752e-05,
+      "loss": 0.4519,
+      "step": 1047
+    },
+    {
+      "epoch": 0.20662460567823343,
+      "grad_norm": 0.8505419812339252,
+      "learning_rate": 1.9873317946263892e-05,
+      "loss": 0.476,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2068217665615142,
+      "grad_norm": 0.6864258063010205,
+      "learning_rate": 1.9873071886215616e-05,
+      "loss": 0.463,
+      "step": 1049
+    },
+    {
+      "epoch": 0.20701892744479494,
+      "grad_norm": 0.9553769219710571,
+      "learning_rate": 1.987282558895884e-05,
+      "loss": 0.4925,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2072160883280757,
+      "grad_norm": 0.6980654478796005,
+      "learning_rate": 1.9872579054499478e-05,
+      "loss": 0.426,
+      "step": 1051
+    },
+    {
+      "epoch": 0.20741324921135645,
+      "grad_norm": 0.719479062368989,
+      "learning_rate": 1.987233228284345e-05,
+      "loss": 0.4504,
+      "step": 1052
+    },
+    {
+      "epoch": 0.20761041009463724,
+      "grad_norm": 0.7018789823838913,
+      "learning_rate": 1.9872085273996694e-05,
+      "loss": 0.4546,
+      "step": 1053
+    },
+    {
+      "epoch": 0.207807570977918,
+      "grad_norm": 0.7802910640814608,
+      "learning_rate": 1.9871838027965134e-05,
+      "loss": 0.4972,
+      "step": 1054
+    },
+    {
+      "epoch": 0.20800473186119875,
+      "grad_norm": 0.7844445505713344,
+      "learning_rate": 1.987159054475472e-05,
+      "loss": 0.4492,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2082018927444795,
+      "grad_norm": 0.692043038159667,
+      "learning_rate": 1.9871342824371393e-05,
+      "loss": 0.4406,
+      "step": 1056
+    },
+    {
+      "epoch": 0.20839905362776026,
+      "grad_norm": 0.775005815750936,
+      "learning_rate": 1.9871094866821104e-05,
+      "loss": 0.4282,
+      "step": 1057
+    },
+    {
+      "epoch": 0.20859621451104102,
+      "grad_norm": 0.7449051373918187,
+      "learning_rate": 1.987084667210981e-05,
+      "loss": 0.46,
+      "step": 1058
+    },
+    {
+      "epoch": 0.20879337539432177,
+      "grad_norm": 0.7073489451850218,
+      "learning_rate": 1.987059824024348e-05,
+      "loss": 0.429,
+      "step": 1059
+    },
+    {
+      "epoch": 0.20899053627760253,
+      "grad_norm": 0.7309470583978874,
+      "learning_rate": 1.9870349571228075e-05,
+      "loss": 0.4798,
+      "step": 1060
+    },
+    {
+      "epoch": 0.20918769716088328,
+      "grad_norm": 0.7718451336210451,
+      "learning_rate": 1.9870100665069577e-05,
+      "loss": 0.4688,
+      "step": 1061
+    },
+    {
+      "epoch": 0.20938485804416404,
+      "grad_norm": 0.6817754353610815,
+      "learning_rate": 1.9869851521773956e-05,
+      "loss": 0.4495,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2095820189274448,
+      "grad_norm": 0.7665443527171363,
+      "learning_rate": 1.986960214134721e-05,
+      "loss": 0.4802,
+      "step": 1063
+    },
+    {
+      "epoch": 0.20977917981072555,
+      "grad_norm": 0.6633898852299499,
+      "learning_rate": 1.986935252379532e-05,
+      "loss": 0.4538,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2099763406940063,
+      "grad_norm": 0.68853230845626,
+      "learning_rate": 1.9869102669124293e-05,
+      "loss": 0.4882,
+      "step": 1065
+    },
+    {
+      "epoch": 0.21017350157728706,
+      "grad_norm": 0.7017829715484818,
+      "learning_rate": 1.986885257734012e-05,
+      "loss": 0.4855,
+      "step": 1066
+    },
+    {
+      "epoch": 0.21037066246056782,
+      "grad_norm": 0.7815202320893195,
+      "learning_rate": 1.986860224844882e-05,
+      "loss": 0.4763,
+      "step": 1067
+    },
+    {
+      "epoch": 0.21056782334384858,
+      "grad_norm": 0.7066784986762935,
+      "learning_rate": 1.9868351682456408e-05,
+      "loss": 0.481,
+      "step": 1068
+    },
+    {
+      "epoch": 0.21076498422712933,
+      "grad_norm": 0.7041493363838017,
+      "learning_rate": 1.986810087936889e-05,
+      "loss": 0.4637,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2109621451104101,
+      "grad_norm": 0.72824829966811,
+      "learning_rate": 1.9867849839192313e-05,
+      "loss": 0.4651,
+      "step": 1070
+    },
+    {
+      "epoch": 0.21115930599369084,
+      "grad_norm": 0.6680455738505751,
+      "learning_rate": 1.986759856193269e-05,
+      "loss": 0.4665,
+      "step": 1071
+    },
+    {
+      "epoch": 0.2113564668769716,
+      "grad_norm": 0.9281048268306978,
+      "learning_rate": 1.9867347047596066e-05,
+      "loss": 0.471,
+      "step": 1072
+    },
+    {
+      "epoch": 0.21155362776025236,
+      "grad_norm": 0.7395272448325264,
+      "learning_rate": 1.9867095296188483e-05,
+      "loss": 0.4715,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2117507886435331,
+      "grad_norm": 0.7830224178550704,
+      "learning_rate": 1.986684330771599e-05,
+      "loss": 0.4641,
+      "step": 1074
+    },
+    {
+      "epoch": 0.21194794952681387,
+      "grad_norm": 0.7574110213860703,
+      "learning_rate": 1.986659108218464e-05,
+      "loss": 0.4474,
+      "step": 1075
+    },
+    {
+      "epoch": 0.21214511041009465,
+      "grad_norm": 0.6927377615200532,
+      "learning_rate": 1.98663386196005e-05,
+      "loss": 0.4287,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2123422712933754,
+      "grad_norm": 0.7232907585511054,
+      "learning_rate": 1.986608591996962e-05,
+      "loss": 0.471,
+      "step": 1077
+    },
+    {
+      "epoch": 0.21253943217665616,
+      "grad_norm": 0.6491519990201463,
+      "learning_rate": 1.9865832983298085e-05,
+      "loss": 0.4228,
+      "step": 1078
+    },
+    {
+      "epoch": 0.21273659305993692,
+      "grad_norm": 0.6831967457936221,
+      "learning_rate": 1.986557980959197e-05,
+      "loss": 0.4732,
+      "step": 1079
+    },
+    {
+      "epoch": 0.21293375394321767,
+      "grad_norm": 0.6954600579093602,
+      "learning_rate": 1.986532639885735e-05,
+      "loss": 0.4764,
+      "step": 1080
+    },
+    {
+      "epoch": 0.21313091482649843,
+      "grad_norm": 0.6995051067856928,
+      "learning_rate": 1.9865072751100324e-05,
+      "loss": 0.4769,
+      "step": 1081
+    },
+    {
+      "epoch": 0.21332807570977919,
+      "grad_norm": 0.6775668825484679,
+      "learning_rate": 1.9864818866326978e-05,
+      "loss": 0.4461,
+      "step": 1082
+    },
+    {
+      "epoch": 0.21352523659305994,
+      "grad_norm": 0.7519792388180498,
+      "learning_rate": 1.9864564744543412e-05,
+      "loss": 0.4812,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2137223974763407,
+      "grad_norm": 0.7107462899081852,
+      "learning_rate": 1.986431038575574e-05,
+      "loss": 0.5028,
+      "step": 1084
+    },
+    {
+      "epoch": 0.21391955835962145,
+      "grad_norm": 0.6753961183203763,
+      "learning_rate": 1.9864055789970064e-05,
+      "loss": 0.4237,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2141167192429022,
+      "grad_norm": 0.6737584152485184,
+      "learning_rate": 1.9863800957192504e-05,
+      "loss": 0.4461,
+      "step": 1086
+    },
+    {
+      "epoch": 0.21431388012618297,
+      "grad_norm": 0.6556918876340901,
+      "learning_rate": 1.9863545887429185e-05,
+      "loss": 0.4506,
+      "step": 1087
+    },
+    {
+      "epoch": 0.21451104100946372,
+      "grad_norm": 0.6877170697107042,
+      "learning_rate": 1.9863290580686228e-05,
+      "loss": 0.4058,
+      "step": 1088
+    },
+    {
+      "epoch": 0.21470820189274448,
+      "grad_norm": 0.7037870318662259,
+      "learning_rate": 1.9863035036969775e-05,
+      "loss": 0.4576,
+      "step": 1089
+    },
+    {
+      "epoch": 0.21490536277602523,
+      "grad_norm": 0.7399278087323496,
+      "learning_rate": 1.9862779256285964e-05,
+      "loss": 0.4907,
+      "step": 1090
+    },
+    {
+      "epoch": 0.215102523659306,
+      "grad_norm": 0.688883167018551,
+      "learning_rate": 1.986252323864094e-05,
+      "loss": 0.4597,
+      "step": 1091
+    },
+    {
+      "epoch": 0.21529968454258674,
+      "grad_norm": 0.7010288184037431,
+      "learning_rate": 1.9862266984040847e-05,
+      "loss": 0.4519,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2154968454258675,
+      "grad_norm": 0.6241563971147653,
+      "learning_rate": 1.9862010492491852e-05,
+      "loss": 0.3987,
+      "step": 1093
+    },
+    {
+      "epoch": 0.21569400630914826,
+      "grad_norm": 0.6700336161149536,
+      "learning_rate": 1.9861753764000115e-05,
+      "loss": 0.4728,
+      "step": 1094
+    },
+    {
+      "epoch": 0.215891167192429,
+      "grad_norm": 0.7471341036922323,
+      "learning_rate": 1.98614967985718e-05,
+      "loss": 0.446,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21608832807570977,
+      "grad_norm": 0.6506229824272028,
+      "learning_rate": 1.986123959621308e-05,
+      "loss": 0.416,
+      "step": 1096
+    },
+    {
+      "epoch": 0.21628548895899052,
+      "grad_norm": 1.2484822940321072,
+      "learning_rate": 1.986098215693014e-05,
+      "loss": 0.4903,
+      "step": 1097
+    },
+    {
+      "epoch": 0.21648264984227128,
+      "grad_norm": 0.9678245848808749,
+      "learning_rate": 1.986072448072916e-05,
+      "loss": 0.4802,
+      "step": 1098
+    },
+    {
+      "epoch": 0.21667981072555206,
+      "grad_norm": 0.7448588276706483,
+      "learning_rate": 1.9860466567616335e-05,
+      "loss": 0.4231,
+      "step": 1099
+    },
+    {
+      "epoch": 0.21687697160883282,
+      "grad_norm": 0.7225398921155327,
+      "learning_rate": 1.9860208417597863e-05,
+      "loss": 0.4666,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21707413249211358,
+      "grad_norm": 0.8881220700758276,
+      "learning_rate": 1.9859950030679943e-05,
+      "loss": 0.4587,
+      "step": 1101
+    },
+    {
+      "epoch": 0.21727129337539433,
+      "grad_norm": 0.6824253304142237,
+      "learning_rate": 1.985969140686878e-05,
+      "loss": 0.4671,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2174684542586751,
+      "grad_norm": 0.822081525030592,
+      "learning_rate": 1.9859432546170594e-05,
+      "loss": 0.4309,
+      "step": 1103
+    },
+    {
+      "epoch": 0.21766561514195584,
+      "grad_norm": 0.7177176238573195,
+      "learning_rate": 1.98591734485916e-05,
+      "loss": 0.4662,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2178627760252366,
+      "grad_norm": 23.933597214678645,
+      "learning_rate": 1.9858914114138024e-05,
+      "loss": 0.4718,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21805993690851735,
+      "grad_norm": 0.9295012594770319,
+      "learning_rate": 1.9858654542816098e-05,
+      "loss": 0.4374,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2182570977917981,
+      "grad_norm": 0.7110886368537057,
+      "learning_rate": 1.9858394734632054e-05,
+      "loss": 0.4292,
+      "step": 1107
+    },
+    {
+      "epoch": 0.21845425867507887,
+      "grad_norm": 0.745899081526924,
+      "learning_rate": 1.9858134689592143e-05,
+      "loss": 0.4494,
+      "step": 1108
+    },
+    {
+      "epoch": 0.21865141955835962,
+      "grad_norm": 3.2936984686195667,
+      "learning_rate": 1.9857874407702606e-05,
+      "loss": 0.436,
+      "step": 1109
+    },
+    {
+      "epoch": 0.21884858044164038,
+      "grad_norm": 0.9426735384515953,
+      "learning_rate": 1.9857613888969694e-05,
+      "loss": 0.4761,
+      "step": 1110
+    },
+    {
+      "epoch": 0.21904574132492113,
+      "grad_norm": 0.9274354892950216,
+      "learning_rate": 1.9857353133399675e-05,
+      "loss": 0.4885,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2192429022082019,
+      "grad_norm": 0.6753675988750968,
+      "learning_rate": 1.9857092140998807e-05,
+      "loss": 0.4496,
+      "step": 1112
+    },
+    {
+      "epoch": 0.21944006309148265,
+      "grad_norm": 0.9644822228389445,
+      "learning_rate": 1.985683091177336e-05,
+      "loss": 0.4677,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2196372239747634,
+      "grad_norm": 0.7186407032609544,
+      "learning_rate": 1.9856569445729615e-05,
+      "loss": 0.4894,
+      "step": 1114
+    },
+    {
+      "epoch": 0.21983438485804416,
+      "grad_norm": 0.7974174876233592,
+      "learning_rate": 1.9856307742873852e-05,
+      "loss": 0.4593,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2200315457413249,
+      "grad_norm": 0.7199784472630298,
+      "learning_rate": 1.9856045803212356e-05,
+      "loss": 0.4488,
+      "step": 1116
+    },
+    {
+      "epoch": 0.22022870662460567,
+      "grad_norm": 0.7059488613258683,
+      "learning_rate": 1.9855783626751425e-05,
+      "loss": 0.423,
+      "step": 1117
+    },
+    {
+      "epoch": 0.22042586750788642,
+      "grad_norm": 0.6784311746671168,
+      "learning_rate": 1.9855521213497355e-05,
+      "loss": 0.4455,
+      "step": 1118
+    },
+    {
+      "epoch": 0.22062302839116718,
+      "grad_norm": 0.6987440905493829,
+      "learning_rate": 1.9855258563456448e-05,
+      "loss": 0.4458,
+      "step": 1119
+    },
+    {
+      "epoch": 0.22082018927444794,
+      "grad_norm": 0.7453334954062374,
+      "learning_rate": 1.985499567663502e-05,
+      "loss": 0.472,
+      "step": 1120
+    },
+    {
+      "epoch": 0.22101735015772872,
+      "grad_norm": 0.8208236623644202,
+      "learning_rate": 1.9854732553039388e-05,
+      "loss": 0.4327,
+      "step": 1121
+    },
+    {
+      "epoch": 0.22121451104100948,
+      "grad_norm": 0.8076041122842882,
+      "learning_rate": 1.9854469192675868e-05,
+      "loss": 0.4488,
+      "step": 1122
+    },
+    {
+      "epoch": 0.22141167192429023,
+      "grad_norm": 0.6509600595961608,
+      "learning_rate": 1.9854205595550787e-05,
+      "loss": 0.4253,
+      "step": 1123
+    },
+    {
+      "epoch": 0.221608832807571,
+      "grad_norm": 0.8076858203915139,
+      "learning_rate": 1.9853941761670483e-05,
+      "loss": 0.4793,
+      "step": 1124
+    },
+    {
+      "epoch": 0.22180599369085174,
+      "grad_norm": 0.6925472433506414,
+      "learning_rate": 1.9853677691041293e-05,
+      "loss": 0.4703,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2220031545741325,
+      "grad_norm": 0.68014215224375,
+      "learning_rate": 1.985341338366956e-05,
+      "loss": 0.4668,
+      "step": 1126
+    },
+    {
+      "epoch": 0.22220031545741326,
+      "grad_norm": 0.7501920154620356,
+      "learning_rate": 1.9853148839561638e-05,
+      "loss": 0.4611,
+      "step": 1127
+    },
+    {
+      "epoch": 0.222397476340694,
+      "grad_norm": 0.7031267527114912,
+      "learning_rate": 1.985288405872388e-05,
+      "loss": 0.4603,
+      "step": 1128
+    },
+    {
+      "epoch": 0.22259463722397477,
+      "grad_norm": 0.6372019068678325,
+      "learning_rate": 1.9852619041162646e-05,
+      "loss": 0.41,
+      "step": 1129
+    },
+    {
+      "epoch": 0.22279179810725552,
+      "grad_norm": 0.6979715600361694,
+      "learning_rate": 1.9852353786884306e-05,
+      "loss": 0.4666,
+      "step": 1130
+    },
+    {
+      "epoch": 0.22298895899053628,
+      "grad_norm": 0.7323097747430808,
+      "learning_rate": 1.9852088295895232e-05,
+      "loss": 0.4468,
+      "step": 1131
+    },
+    {
+      "epoch": 0.22318611987381703,
+      "grad_norm": 0.6991485532926531,
+      "learning_rate": 1.9851822568201806e-05,
+      "loss": 0.4634,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2233832807570978,
+      "grad_norm": 0.7591783665648759,
+      "learning_rate": 1.9851556603810406e-05,
+      "loss": 0.4419,
+      "step": 1133
+    },
+    {
+      "epoch": 0.22358044164037855,
+      "grad_norm": 1.1746824112571197,
+      "learning_rate": 1.9851290402727426e-05,
+      "loss": 0.4639,
+      "step": 1134
+    },
+    {
+      "epoch": 0.2237776025236593,
+      "grad_norm": 0.7081731495754315,
+      "learning_rate": 1.985102396495926e-05,
+      "loss": 0.4881,
+      "step": 1135
+    },
+    {
+      "epoch": 0.22397476340694006,
+      "grad_norm": 0.6590193639944613,
+      "learning_rate": 1.9850757290512313e-05,
+      "loss": 0.4536,
+      "step": 1136
+    },
+    {
+      "epoch": 0.22417192429022081,
+      "grad_norm": 8.480821491410778,
+      "learning_rate": 1.9850490379392988e-05,
+      "loss": 0.4433,
+      "step": 1137
+    },
+    {
+      "epoch": 0.22436908517350157,
+      "grad_norm": 0.7990748970455287,
+      "learning_rate": 1.9850223231607696e-05,
+      "loss": 0.4704,
+      "step": 1138
+    },
+    {
+      "epoch": 0.22456624605678233,
+      "grad_norm": 0.6257113394097338,
+      "learning_rate": 1.984995584716286e-05,
+      "loss": 0.4377,
+      "step": 1139
+    },
+    {
+      "epoch": 0.22476340694006308,
+      "grad_norm": 4.598690932717382,
+      "learning_rate": 1.9849688226064906e-05,
+      "loss": 0.5025,
+      "step": 1140
+    },
+    {
+      "epoch": 0.22496056782334384,
+      "grad_norm": 0.8210582553644494,
+      "learning_rate": 1.9849420368320254e-05,
+      "loss": 0.4854,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2251577287066246,
+      "grad_norm": 0.7646003921775568,
+      "learning_rate": 1.9849152273935353e-05,
+      "loss": 0.4344,
+      "step": 1142
+    },
+    {
+      "epoch": 0.22535488958990535,
+      "grad_norm": 0.7228017400656176,
+      "learning_rate": 1.9848883942916632e-05,
+      "loss": 0.4785,
+      "step": 1143
+    },
+    {
+      "epoch": 0.22555205047318613,
+      "grad_norm": 0.7985110474365666,
+      "learning_rate": 1.9848615375270547e-05,
+      "loss": 0.4461,
+      "step": 1144
+    },
+    {
+      "epoch": 0.2257492113564669,
+      "grad_norm": 0.8935580670763312,
+      "learning_rate": 1.984834657100354e-05,
+      "loss": 0.4658,
+      "step": 1145
+    },
+    {
+      "epoch": 0.22594637223974764,
+      "grad_norm": 0.8552039095318358,
+      "learning_rate": 1.9848077530122083e-05,
+      "loss": 0.4574,
+      "step": 1146
+    },
+    {
+      "epoch": 0.2261435331230284,
+      "grad_norm": 1.2094865238135273,
+      "learning_rate": 1.984780825263263e-05,
+      "loss": 0.4461,
+      "step": 1147
+    },
+    {
+      "epoch": 0.22634069400630916,
+      "grad_norm": 0.8645616019682253,
+      "learning_rate": 1.984753873854165e-05,
+      "loss": 0.5116,
+      "step": 1148
+    },
+    {
+      "epoch": 0.2265378548895899,
+      "grad_norm": 3.127099963117224,
+      "learning_rate": 1.984726898785563e-05,
+      "loss": 0.4398,
+      "step": 1149
+    },
+    {
+      "epoch": 0.22673501577287067,
+      "grad_norm": 1.1378328088647094,
+      "learning_rate": 1.9846999000581033e-05,
+      "loss": 0.4529,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22693217665615142,
+      "grad_norm": 1.2150716840072067,
+      "learning_rate": 1.9846728776724358e-05,
+      "loss": 0.5032,
+      "step": 1151
+    },
+    {
+      "epoch": 0.22712933753943218,
+      "grad_norm": 1.0286587973333643,
+      "learning_rate": 1.98464583162921e-05,
+      "loss": 0.4525,
+      "step": 1152
+    },
+    {
+      "epoch": 0.22732649842271294,
+      "grad_norm": 1.126989898873081,
+      "learning_rate": 1.9846187619290746e-05,
+      "loss": 0.4292,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2275236593059937,
+      "grad_norm": 1.2199916483600795,
+      "learning_rate": 1.9845916685726808e-05,
+      "loss": 0.4732,
+      "step": 1154
+    },
+    {
+      "epoch": 0.22772082018927445,
+      "grad_norm": 0.9371106577012303,
+      "learning_rate": 1.9845645515606792e-05,
+      "loss": 0.4872,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2279179810725552,
+      "grad_norm": 0.9056546747302068,
+      "learning_rate": 1.9845374108937213e-05,
+      "loss": 0.4975,
+      "step": 1156
+    },
+    {
+      "epoch": 0.22811514195583596,
+      "grad_norm": 1.5812769964252382,
+      "learning_rate": 1.9845102465724593e-05,
+      "loss": 0.4733,
+      "step": 1157
+    },
+    {
+      "epoch": 0.22831230283911672,
+      "grad_norm": 1.1843221653360212,
+      "learning_rate": 1.984483058597546e-05,
+      "loss": 0.4624,
+      "step": 1158
+    },
+    {
+      "epoch": 0.22850946372239747,
+      "grad_norm": 0.9821603418098558,
+      "learning_rate": 1.9844558469696342e-05,
+      "loss": 0.4413,
+      "step": 1159
+    },
+    {
+      "epoch": 0.22870662460567823,
+      "grad_norm": 0.9806072784995279,
+      "learning_rate": 1.984428611689378e-05,
+      "loss": 0.4868,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22890378548895898,
+      "grad_norm": 0.8793984774617989,
+      "learning_rate": 1.9844013527574316e-05,
+      "loss": 0.4267,
+      "step": 1161
+    },
+    {
+      "epoch": 0.22910094637223974,
+      "grad_norm": 0.8495785803021494,
+      "learning_rate": 1.9843740701744497e-05,
+      "loss": 0.4354,
+      "step": 1162
+    },
+    {
+      "epoch": 0.2292981072555205,
+      "grad_norm": 1.7057039922514965,
+      "learning_rate": 1.9843467639410885e-05,
+      "loss": 0.4447,
+      "step": 1163
+    },
+    {
+      "epoch": 0.22949526813880125,
+      "grad_norm": 1.075911659015585,
+      "learning_rate": 1.9843194340580032e-05,
+      "loss": 0.4962,
+      "step": 1164
+    },
+    {
+      "epoch": 0.229692429022082,
+      "grad_norm": 0.8457228409931461,
+      "learning_rate": 1.9842920805258513e-05,
+      "loss": 0.4552,
+      "step": 1165
+    },
+    {
+      "epoch": 0.22988958990536276,
+      "grad_norm": 1.008444001726399,
+      "learning_rate": 1.9842647033452893e-05,
+      "loss": 0.4462,
+      "step": 1166
+    },
+    {
+      "epoch": 0.23008675078864355,
+      "grad_norm": 0.8092829859053057,
+      "learning_rate": 1.984237302516975e-05,
+      "loss": 0.4559,
+      "step": 1167
+    },
+    {
+      "epoch": 0.2302839116719243,
+      "grad_norm": 0.9687384781928995,
+      "learning_rate": 1.984209878041567e-05,
+      "loss": 0.5053,
+      "step": 1168
+    },
+    {
+      "epoch": 0.23048107255520506,
+      "grad_norm": 0.8297737211847974,
+      "learning_rate": 1.984182429919724e-05,
+      "loss": 0.462,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2306782334384858,
+      "grad_norm": 0.7460078380279647,
+      "learning_rate": 1.9841549581521058e-05,
+      "loss": 0.4538,
+      "step": 1170
+    },
+    {
+      "epoch": 0.23087539432176657,
+      "grad_norm": 0.9658528590503032,
+      "learning_rate": 1.984127462739372e-05,
+      "loss": 0.4759,
+      "step": 1171
+    },
+    {
+      "epoch": 0.23107255520504733,
+      "grad_norm": 0.7016223428827921,
+      "learning_rate": 1.9840999436821836e-05,
+      "loss": 0.4328,
+      "step": 1172
+    },
+    {
+      "epoch": 0.23126971608832808,
+      "grad_norm": 0.6671082456539488,
+      "learning_rate": 1.9840724009812013e-05,
+      "loss": 0.4259,
+      "step": 1173
+    },
+    {
+      "epoch": 0.23146687697160884,
+      "grad_norm": 0.7264492795866501,
+      "learning_rate": 1.9840448346370873e-05,
+      "loss": 0.4251,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2316640378548896,
+      "grad_norm": 0.726746284722804,
+      "learning_rate": 1.9840172446505036e-05,
+      "loss": 0.4838,
+      "step": 1175
+    },
+    {
+      "epoch": 0.23186119873817035,
+      "grad_norm": 1.151370972769961,
+      "learning_rate": 1.9839896310221133e-05,
+      "loss": 0.4793,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2320583596214511,
+      "grad_norm": 0.8135672403953714,
+      "learning_rate": 1.9839619937525794e-05,
+      "loss": 0.4496,
+      "step": 1177
+    },
+    {
+      "epoch": 0.23225552050473186,
+      "grad_norm": 0.6927163568862448,
+      "learning_rate": 1.9839343328425668e-05,
+      "loss": 0.4628,
+      "step": 1178
+    },
+    {
+      "epoch": 0.23245268138801262,
+      "grad_norm": 0.6732578753958156,
+      "learning_rate": 1.983906648292739e-05,
+      "loss": 0.43,
+      "step": 1179
+    },
+    {
+      "epoch": 0.23264984227129337,
+      "grad_norm": 0.7086446749133889,
+      "learning_rate": 1.9838789401037616e-05,
+      "loss": 0.4291,
+      "step": 1180
+    },
+    {
+      "epoch": 0.23284700315457413,
+      "grad_norm": 0.7334824055009613,
+      "learning_rate": 1.9838512082763002e-05,
+      "loss": 0.4342,
+      "step": 1181
+    },
+    {
+      "epoch": 0.23304416403785488,
+      "grad_norm": 0.7653457648798104,
+      "learning_rate": 1.983823452811022e-05,
+      "loss": 0.4805,
+      "step": 1182
+    },
+    {
+      "epoch": 0.23324132492113564,
+      "grad_norm": 0.9262252445526374,
+      "learning_rate": 1.9837956737085924e-05,
+      "loss": 0.4788,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2334384858044164,
+      "grad_norm": 0.7504820328186347,
+      "learning_rate": 1.98376787096968e-05,
+      "loss": 0.4555,
+      "step": 1184
+    },
+    {
+      "epoch": 0.23363564668769715,
+      "grad_norm": 0.721933776706364,
+      "learning_rate": 1.983740044594952e-05,
+      "loss": 0.4848,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2338328075709779,
+      "grad_norm": 2.846732366688398,
+      "learning_rate": 1.9837121945850766e-05,
+      "loss": 0.4638,
+      "step": 1186
+    },
+    {
+      "epoch": 0.23402996845425866,
+      "grad_norm": 0.8306542747582979,
+      "learning_rate": 1.9836843209407247e-05,
+      "loss": 0.4082,
+      "step": 1187
+    },
+    {
+      "epoch": 0.23422712933753942,
+      "grad_norm": 0.736142320584185,
+      "learning_rate": 1.983656423662564e-05,
+      "loss": 0.4569,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2344242902208202,
+      "grad_norm": 0.7039751490055907,
+      "learning_rate": 1.983628502751266e-05,
+      "loss": 0.4182,
+      "step": 1189
+    },
+    {
+      "epoch": 0.23462145110410096,
+      "grad_norm": 0.9525090224636836,
+      "learning_rate": 1.983600558207501e-05,
+      "loss": 0.4414,
+      "step": 1190
+    },
+    {
+      "epoch": 0.23481861198738171,
+      "grad_norm": 1.4964168121471113,
+      "learning_rate": 1.9835725900319406e-05,
+      "loss": 0.4468,
+      "step": 1191
+    },
+    {
+      "epoch": 0.23501577287066247,
+      "grad_norm": 0.8670187458357161,
+      "learning_rate": 1.9835445982252565e-05,
+      "loss": 0.4159,
+      "step": 1192
+    },
+    {
+      "epoch": 0.23521293375394323,
+      "grad_norm": 0.6948661600832843,
+      "learning_rate": 1.983516582788121e-05,
+      "loss": 0.4451,
+      "step": 1193
+    },
+    {
+      "epoch": 0.23541009463722398,
+      "grad_norm": 0.7812431346409022,
+      "learning_rate": 1.9834885437212083e-05,
+      "loss": 0.4613,
+      "step": 1194
+    },
+    {
+      "epoch": 0.23560725552050474,
+      "grad_norm": 0.7970118513981926,
+      "learning_rate": 1.983460481025191e-05,
+      "loss": 0.4761,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2358044164037855,
+      "grad_norm": 0.7490679736689834,
+      "learning_rate": 1.9834323947007433e-05,
+      "loss": 0.4931,
+      "step": 1196
+    },
+    {
+      "epoch": 0.23600157728706625,
+      "grad_norm": 0.7282970951897094,
+      "learning_rate": 1.983404284748541e-05,
+      "loss": 0.4184,
+      "step": 1197
+    },
+    {
+      "epoch": 0.236198738170347,
+      "grad_norm": 0.7024789178171861,
+      "learning_rate": 1.9833761511692583e-05,
+      "loss": 0.4604,
+      "step": 1198
+    },
+    {
+      "epoch": 0.23639589905362776,
+      "grad_norm": 0.7562914768111765,
+      "learning_rate": 1.9833479939635724e-05,
+      "loss": 0.4835,
+      "step": 1199
+    },
+    {
+      "epoch": 0.23659305993690852,
+      "grad_norm": 0.6905900737165961,
+      "learning_rate": 1.9833198131321582e-05,
+      "loss": 0.4362,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23679022082018927,
+      "grad_norm": 0.666922032665526,
+      "learning_rate": 1.9832916086756938e-05,
+      "loss": 0.4619,
+      "step": 1201
+    },
+    {
+      "epoch": 0.23698738170347003,
+      "grad_norm": 0.6837662204239521,
+      "learning_rate": 1.983263380594857e-05,
+      "loss": 0.4373,
+      "step": 1202
+    },
+    {
+      "epoch": 0.23718454258675079,
+      "grad_norm": 0.6414003295820169,
+      "learning_rate": 1.9832351288903256e-05,
+      "loss": 0.4317,
+      "step": 1203
+    },
+    {
+      "epoch": 0.23738170347003154,
+      "grad_norm": 0.7460692979898119,
+      "learning_rate": 1.9832068535627785e-05,
+      "loss": 0.479,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2375788643533123,
+      "grad_norm": 0.6428078211747335,
+      "learning_rate": 1.983178554612895e-05,
+      "loss": 0.4469,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23777602523659305,
+      "grad_norm": 0.7517122154912249,
+      "learning_rate": 1.9831502320413543e-05,
+      "loss": 0.4858,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2379731861198738,
+      "grad_norm": 1.3463826441881264,
+      "learning_rate": 1.983121885848838e-05,
+      "loss": 0.4735,
+      "step": 1207
+    },
+    {
+      "epoch": 0.23817034700315456,
+      "grad_norm": 0.7102923990366657,
+      "learning_rate": 1.983093516036027e-05,
+      "loss": 0.487,
+      "step": 1208
+    },
+    {
+      "epoch": 0.23836750788643532,
+      "grad_norm": 2.290103684996036,
+      "learning_rate": 1.9830651226036023e-05,
+      "loss": 0.4727,
+      "step": 1209
+    },
+    {
+      "epoch": 0.23856466876971608,
+      "grad_norm": 0.6811005595914935,
+      "learning_rate": 1.9830367055522463e-05,
+      "loss": 0.4263,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23876182965299683,
+      "grad_norm": 0.799809715831858,
+      "learning_rate": 1.983008264882642e-05,
+      "loss": 0.4097,
+      "step": 1211
+    },
+    {
+      "epoch": 0.23895899053627762,
+      "grad_norm": 0.7354195020597284,
+      "learning_rate": 1.9829798005954725e-05,
+      "loss": 0.4724,
+      "step": 1212
+    },
+    {
+      "epoch": 0.23915615141955837,
+      "grad_norm": 0.8296463030390158,
+      "learning_rate": 1.982951312691422e-05,
+      "loss": 0.4904,
+      "step": 1213
+    },
+    {
+      "epoch": 0.23935331230283913,
+      "grad_norm": 0.6229789713014955,
+      "learning_rate": 1.9829228011711738e-05,
+      "loss": 0.4451,
+      "step": 1214
+    },
+    {
+      "epoch": 0.23955047318611988,
+      "grad_norm": 0.7037507970309669,
+      "learning_rate": 1.9828942660354144e-05,
+      "loss": 0.4382,
+      "step": 1215
+    },
+    {
+      "epoch": 0.23974763406940064,
+      "grad_norm": 1.7862177754634971,
+      "learning_rate": 1.9828657072848284e-05,
+      "loss": 0.4677,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2399447949526814,
+      "grad_norm": 0.7511599646811729,
+      "learning_rate": 1.9828371249201025e-05,
+      "loss": 0.4432,
+      "step": 1217
+    },
+    {
+      "epoch": 0.24014195583596215,
+      "grad_norm": 0.6663595319040087,
+      "learning_rate": 1.982808518941923e-05,
+      "loss": 0.4535,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2403391167192429,
+      "grad_norm": 0.7505465175906688,
+      "learning_rate": 1.982779889350978e-05,
+      "loss": 0.483,
+      "step": 1219
+    },
+    {
+      "epoch": 0.24053627760252366,
+      "grad_norm": 0.7052786101408759,
+      "learning_rate": 1.982751236147954e-05,
+      "loss": 0.4568,
+      "step": 1220
+    },
+    {
+      "epoch": 0.24073343848580442,
+      "grad_norm": 1.0567060689641148,
+      "learning_rate": 1.9827225593335403e-05,
+      "loss": 0.4725,
+      "step": 1221
+    },
+    {
+      "epoch": 0.24093059936908517,
+      "grad_norm": 0.668419640423508,
+      "learning_rate": 1.9826938589084258e-05,
+      "loss": 0.462,
+      "step": 1222
+    },
+    {
+      "epoch": 0.24112776025236593,
+      "grad_norm": 1.2570484391370291,
+      "learning_rate": 1.9826651348733e-05,
+      "loss": 0.5211,
+      "step": 1223
+    },
+    {
+      "epoch": 0.24132492113564669,
+      "grad_norm": 0.7349089292621471,
+      "learning_rate": 1.982636387228853e-05,
+      "loss": 0.4528,
+      "step": 1224
+    },
+    {
+      "epoch": 0.24152208201892744,
+      "grad_norm": 0.8253540404906965,
+      "learning_rate": 1.9826076159757753e-05,
+      "loss": 0.5093,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2417192429022082,
+      "grad_norm": 2.1092492780074745,
+      "learning_rate": 1.9825788211147587e-05,
+      "loss": 0.4845,
+      "step": 1226
+    },
+    {
+      "epoch": 0.24191640378548895,
+      "grad_norm": 0.8850457162579004,
+      "learning_rate": 1.9825500026464947e-05,
+      "loss": 0.4805,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2421135646687697,
+      "grad_norm": 0.6781748488422774,
+      "learning_rate": 1.9825211605716748e-05,
+      "loss": 0.4738,
+      "step": 1228
+    },
+    {
+      "epoch": 0.24231072555205047,
+      "grad_norm": 1.0845607789046796,
+      "learning_rate": 1.9824922948909937e-05,
+      "loss": 0.4449,
+      "step": 1229
+    },
+    {
+      "epoch": 0.24250788643533122,
+      "grad_norm": 1.2815030508313785,
+      "learning_rate": 1.9824634056051436e-05,
+      "loss": 0.473,
+      "step": 1230
+    },
+    {
+      "epoch": 0.24270504731861198,
+      "grad_norm": 0.8080992979642333,
+      "learning_rate": 1.9824344927148193e-05,
+      "loss": 0.4731,
+      "step": 1231
+    },
+    {
+      "epoch": 0.24290220820189273,
+      "grad_norm": 0.7437272267133168,
+      "learning_rate": 1.982405556220715e-05,
+      "loss": 0.4519,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2430993690851735,
+      "grad_norm": 0.7480610069853911,
+      "learning_rate": 1.982376596123526e-05,
+      "loss": 0.4417,
+      "step": 1233
+    },
+    {
+      "epoch": 0.24329652996845424,
+      "grad_norm": 0.7550320130527832,
+      "learning_rate": 1.982347612423948e-05,
+      "loss": 0.439,
+      "step": 1234
+    },
+    {
+      "epoch": 0.24349369085173503,
+      "grad_norm": 0.6749020950952507,
+      "learning_rate": 1.9823186051226783e-05,
+      "loss": 0.4201,
+      "step": 1235
+    },
+    {
+      "epoch": 0.24369085173501578,
+      "grad_norm": 0.8142599734476462,
+      "learning_rate": 1.9822895742204122e-05,
+      "loss": 0.4621,
+      "step": 1236
+    },
+    {
+      "epoch": 0.24388801261829654,
+      "grad_norm": 0.6700713256447526,
+      "learning_rate": 1.9822605197178485e-05,
+      "loss": 0.4387,
+      "step": 1237
+    },
+    {
+      "epoch": 0.2440851735015773,
+      "grad_norm": 0.7745146153654806,
+      "learning_rate": 1.9822314416156848e-05,
+      "loss": 0.4605,
+      "step": 1238
+    },
+    {
+      "epoch": 0.24428233438485805,
+      "grad_norm": 0.9942732385767507,
+      "learning_rate": 1.9822023399146194e-05,
+      "loss": 0.4773,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2444794952681388,
+      "grad_norm": 0.7228782340223426,
+      "learning_rate": 1.9821732146153517e-05,
+      "loss": 0.4754,
+      "step": 1240
+    },
+    {
+      "epoch": 0.24467665615141956,
+      "grad_norm": 0.7905102450931651,
+      "learning_rate": 1.9821440657185822e-05,
+      "loss": 0.4596,
+      "step": 1241
+    },
+    {
+      "epoch": 0.24487381703470032,
+      "grad_norm": 0.8374670342807091,
+      "learning_rate": 1.9821148932250103e-05,
+      "loss": 0.4102,
+      "step": 1242
+    },
+    {
+      "epoch": 0.24507097791798108,
+      "grad_norm": 0.9557468893603367,
+      "learning_rate": 1.9820856971353374e-05,
+      "loss": 0.441,
+      "step": 1243
+    },
+    {
+      "epoch": 0.24526813880126183,
+      "grad_norm": 0.9884255566988156,
+      "learning_rate": 1.9820564774502644e-05,
+      "loss": 0.4866,
+      "step": 1244
+    },
+    {
+      "epoch": 0.2454652996845426,
+      "grad_norm": 0.6985509946646631,
+      "learning_rate": 1.9820272341704937e-05,
+      "loss": 0.4441,
+      "step": 1245
+    },
+    {
+      "epoch": 0.24566246056782334,
+      "grad_norm": 0.7274501226462767,
+      "learning_rate": 1.981997967296728e-05,
+      "loss": 0.4387,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2458596214511041,
+      "grad_norm": 0.7388406161273386,
+      "learning_rate": 1.9819686768296706e-05,
+      "loss": 0.4613,
+      "step": 1247
+    },
+    {
+      "epoch": 0.24605678233438485,
+      "grad_norm": 0.7240588593494661,
+      "learning_rate": 1.9819393627700247e-05,
+      "loss": 0.4782,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2462539432176656,
+      "grad_norm": 0.7023054870349664,
+      "learning_rate": 1.9819100251184945e-05,
+      "loss": 0.4687,
+      "step": 1249
+    },
+    {
+      "epoch": 0.24645110410094637,
+      "grad_norm": 0.7018924187358255,
+      "learning_rate": 1.9818806638757856e-05,
+      "loss": 0.4643,
+      "step": 1250
+    },
+    {
+      "epoch": 0.24664826498422712,
+      "grad_norm": 0.7507947096176073,
+      "learning_rate": 1.981851279042603e-05,
+      "loss": 0.4471,
+      "step": 1251
+    },
+    {
+      "epoch": 0.24684542586750788,
+      "grad_norm": 0.7236171294016701,
+      "learning_rate": 1.9818218706196527e-05,
+      "loss": 0.486,
+      "step": 1252
+    },
+    {
+      "epoch": 0.24704258675078863,
+      "grad_norm": 0.7074399537871575,
+      "learning_rate": 1.9817924386076416e-05,
+      "loss": 0.4696,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2472397476340694,
+      "grad_norm": 0.6990187430370988,
+      "learning_rate": 1.981762983007276e-05,
+      "loss": 0.4539,
+      "step": 1254
+    },
+    {
+      "epoch": 0.24743690851735015,
+      "grad_norm": 0.8201430818139063,
+      "learning_rate": 1.9817335038192644e-05,
+      "loss": 0.4644,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2476340694006309,
+      "grad_norm": 0.6857424544658502,
+      "learning_rate": 1.981704001044315e-05,
+      "loss": 0.4737,
+      "step": 1256
+    },
+    {
+      "epoch": 0.24783123028391169,
+      "grad_norm": 0.6640652585450276,
+      "learning_rate": 1.981674474683136e-05,
+      "loss": 0.4464,
+      "step": 1257
+    },
+    {
+      "epoch": 0.24802839116719244,
+      "grad_norm": 0.7345704399108267,
+      "learning_rate": 1.9816449247364374e-05,
+      "loss": 0.4625,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2482255520504732,
+      "grad_norm": 0.6686587208589408,
+      "learning_rate": 1.981615351204929e-05,
+      "loss": 0.4369,
+      "step": 1259
+    },
+    {
+      "epoch": 0.24842271293375395,
+      "grad_norm": 0.6762802311646812,
+      "learning_rate": 1.9815857540893214e-05,
+      "loss": 0.432,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2486198738170347,
+      "grad_norm": 0.6562041100893717,
+      "learning_rate": 1.9815561333903255e-05,
+      "loss": 0.43,
+      "step": 1261
+    },
+    {
+      "epoch": 0.24881703470031546,
+      "grad_norm": 0.6532239829668289,
+      "learning_rate": 1.981526489108653e-05,
+      "loss": 0.4351,
+      "step": 1262
+    },
+    {
+      "epoch": 0.24901419558359622,
+      "grad_norm": 0.6996616378046743,
+      "learning_rate": 1.981496821245016e-05,
+      "loss": 0.4881,
+      "step": 1263
+    },
+    {
+      "epoch": 0.24921135646687698,
+      "grad_norm": 0.6705242832654518,
+      "learning_rate": 1.9814671298001274e-05,
+      "loss": 0.4536,
+      "step": 1264
+    },
+    {
+      "epoch": 0.24940851735015773,
+      "grad_norm": 0.7366373896084766,
+      "learning_rate": 1.9814374147747012e-05,
+      "loss": 0.4733,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2496056782334385,
+      "grad_norm": 0.6716024398128168,
+      "learning_rate": 1.98140767616945e-05,
+      "loss": 0.4879,
+      "step": 1266
+    },
+    {
+      "epoch": 0.24980283911671924,
+      "grad_norm": 0.7567180353841482,
+      "learning_rate": 1.98137791398509e-05,
+      "loss": 0.4714,
+      "step": 1267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6280236993661752,
+      "learning_rate": 1.981348128222335e-05,
+      "loss": 0.4467,
+      "step": 1268
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.45942220091819763,
+      "eval_runtime": 344.9533,
+      "eval_samples_per_second": 23.568,
+      "eval_steps_per_second": 1.476,
+      "step": 1268
+    },
+    {
+      "epoch": 0.25019716088328076,
+      "grad_norm": 0.7858090304991342,
+      "learning_rate": 1.9813183188819005e-05,
+      "loss": 0.515,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2503943217665615,
+      "grad_norm": 0.7486528699456562,
+      "learning_rate": 1.981288485964503e-05,
+      "loss": 0.482,
+      "step": 1270
+    },
+    {
+      "epoch": 0.25059148264984227,
+      "grad_norm": 0.6374094423910737,
+      "learning_rate": 1.98125862947086e-05,
+      "loss": 0.4309,
+      "step": 1271
+    },
+    {
+      "epoch": 0.250788643533123,
+      "grad_norm": 0.7096340783400006,
+      "learning_rate": 1.981228749401688e-05,
+      "loss": 0.483,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2509858044164038,
+      "grad_norm": 0.7444158649998527,
+      "learning_rate": 1.9811988457577054e-05,
+      "loss": 0.4592,
+      "step": 1273
+    },
+    {
+      "epoch": 0.25118296529968454,
+      "grad_norm": 0.7501735940555235,
+      "learning_rate": 1.98116891853963e-05,
+      "loss": 0.4499,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2513801261829653,
+      "grad_norm": 0.6782176587394123,
+      "learning_rate": 1.9811389677481815e-05,
+      "loss": 0.4538,
+      "step": 1275
+    },
+    {
+      "epoch": 0.25157728706624605,
+      "grad_norm": 0.6415201673231513,
+      "learning_rate": 1.9811089933840788e-05,
+      "loss": 0.3966,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2517744479495268,
+      "grad_norm": 0.7902039683370593,
+      "learning_rate": 1.9810789954480425e-05,
+      "loss": 0.4445,
+      "step": 1277
+    },
+    {
+      "epoch": 0.25197160883280756,
+      "grad_norm": 0.7144540560155214,
+      "learning_rate": 1.9810489739407934e-05,
+      "loss": 0.4711,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2521687697160883,
+      "grad_norm": 0.6516750917925597,
+      "learning_rate": 1.9810189288630524e-05,
+      "loss": 0.4704,
+      "step": 1279
+    },
+    {
+      "epoch": 0.25236593059936907,
+      "grad_norm": 0.6422474397006303,
+      "learning_rate": 1.980988860215542e-05,
+      "loss": 0.4565,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2525630914826498,
+      "grad_norm": 0.6600661011809978,
+      "learning_rate": 1.9809587679989843e-05,
+      "loss": 0.4603,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2527602523659306,
+      "grad_norm": 0.6621495554696686,
+      "learning_rate": 1.980928652214102e-05,
+      "loss": 0.4539,
+      "step": 1282
+    },
+    {
+      "epoch": 0.25295741324921134,
+      "grad_norm": 0.6210665496532219,
+      "learning_rate": 1.980898512861619e-05,
+      "loss": 0.388,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2531545741324921,
+      "grad_norm": 0.6372480467697316,
+      "learning_rate": 1.9808683499422595e-05,
+      "loss": 0.4625,
+      "step": 1284
+    },
+    {
+      "epoch": 0.25335173501577285,
+      "grad_norm": 0.7040049125741846,
+      "learning_rate": 1.9808381634567478e-05,
+      "loss": 0.4644,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2535488958990536,
+      "grad_norm": 0.6788863637030159,
+      "learning_rate": 1.9808079534058092e-05,
+      "loss": 0.4233,
+      "step": 1286
+    },
+    {
+      "epoch": 0.25374605678233436,
+      "grad_norm": 0.7084313524121193,
+      "learning_rate": 1.9807777197901697e-05,
+      "loss": 0.4692,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2539432176656151,
+      "grad_norm": 0.7122952126645031,
+      "learning_rate": 1.9807474626105557e-05,
+      "loss": 0.4687,
+      "step": 1288
+    },
+    {
+      "epoch": 0.2541403785488959,
+      "grad_norm": 0.9877933281913743,
+      "learning_rate": 1.9807171818676944e-05,
+      "loss": 0.4496,
+      "step": 1289
+    },
+    {
+      "epoch": 0.25433753943217663,
+      "grad_norm": 0.6865217783581581,
+      "learning_rate": 1.9806868775623127e-05,
+      "loss": 0.4448,
+      "step": 1290
+    },
+    {
+      "epoch": 0.25453470031545744,
+      "grad_norm": 0.6823194099522876,
+      "learning_rate": 1.980656549695139e-05,
+      "loss": 0.4556,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2547318611987382,
+      "grad_norm": 0.6647345972741581,
+      "learning_rate": 1.9806261982669025e-05,
+      "loss": 0.4491,
+      "step": 1292
+    },
+    {
+      "epoch": 0.25492902208201895,
+      "grad_norm": 0.6302502151295573,
+      "learning_rate": 1.9805958232783314e-05,
+      "loss": 0.4464,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2551261829652997,
+      "grad_norm": 0.6579780881631992,
+      "learning_rate": 1.980565424730156e-05,
+      "loss": 0.4485,
+      "step": 1294
+    },
+    {
+      "epoch": 0.25532334384858046,
+      "grad_norm": 0.7655010402371665,
+      "learning_rate": 1.9805350026231067e-05,
+      "loss": 0.4669,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2555205047318612,
+      "grad_norm": 0.6625750789917858,
+      "learning_rate": 1.9805045569579144e-05,
+      "loss": 0.463,
+      "step": 1296
+    },
+    {
+      "epoch": 0.255717665615142,
+      "grad_norm": 0.6662591258788151,
+      "learning_rate": 1.9804740877353105e-05,
+      "loss": 0.4367,
+      "step": 1297
+    },
+    {
+      "epoch": 0.25591482649842273,
+      "grad_norm": 0.6108948522622889,
+      "learning_rate": 1.980443594956027e-05,
+      "loss": 0.4241,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2561119873817035,
+      "grad_norm": 0.7914240432743016,
+      "learning_rate": 1.9804130786207966e-05,
+      "loss": 0.4388,
+      "step": 1299
+    },
+    {
+      "epoch": 0.25630914826498424,
+      "grad_norm": 0.6627711788576277,
+      "learning_rate": 1.9803825387303525e-05,
+      "loss": 0.4356,
+      "step": 1300
+    },
+    {
+      "epoch": 0.256506309148265,
+      "grad_norm": 0.6421042720478715,
+      "learning_rate": 1.9803519752854284e-05,
+      "loss": 0.4434,
+      "step": 1301
+    },
+    {
+      "epoch": 0.25670347003154576,
+      "grad_norm": 0.9154766392932525,
+      "learning_rate": 1.9803213882867583e-05,
+      "loss": 0.5037,
+      "step": 1302
+    },
+    {
+      "epoch": 0.2569006309148265,
+      "grad_norm": 0.6412059503472081,
+      "learning_rate": 1.9802907777350778e-05,
+      "loss": 0.4555,
+      "step": 1303
+    },
+    {
+      "epoch": 0.25709779179810727,
+      "grad_norm": 0.7328818019918474,
+      "learning_rate": 1.9802601436311223e-05,
+      "loss": 0.4795,
+      "step": 1304
+    },
+    {
+      "epoch": 0.257294952681388,
+      "grad_norm": 1.290582323709545,
+      "learning_rate": 1.980229485975627e-05,
+      "loss": 0.4574,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2574921135646688,
+      "grad_norm": 0.7047549042877954,
+      "learning_rate": 1.980198804769329e-05,
+      "loss": 0.4441,
+      "step": 1306
+    },
+    {
+      "epoch": 0.25768927444794953,
+      "grad_norm": 0.6450143398070649,
+      "learning_rate": 1.9801681000129652e-05,
+      "loss": 0.4523,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2578864353312303,
+      "grad_norm": 0.7236989507032727,
+      "learning_rate": 1.9801373717072732e-05,
+      "loss": 0.4811,
+      "step": 1308
+    },
+    {
+      "epoch": 0.25808359621451105,
+      "grad_norm": 0.5982776267715486,
+      "learning_rate": 1.980106619852992e-05,
+      "loss": 0.4257,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2582807570977918,
+      "grad_norm": 1.5437331705738089,
+      "learning_rate": 1.98007584445086e-05,
+      "loss": 0.4755,
+      "step": 1310
+    },
+    {
+      "epoch": 0.25847791798107256,
+      "grad_norm": 0.665819086891947,
+      "learning_rate": 1.9800450455016163e-05,
+      "loss": 0.426,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2586750788643533,
+      "grad_norm": 0.6986032017043294,
+      "learning_rate": 1.9800142230060012e-05,
+      "loss": 0.4677,
+      "step": 1312
+    },
+    {
+      "epoch": 0.25887223974763407,
+      "grad_norm": 0.7076440147684866,
+      "learning_rate": 1.9799833769647553e-05,
+      "loss": 0.4849,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2590694006309148,
+      "grad_norm": 0.6993982811871999,
+      "learning_rate": 1.9799525073786196e-05,
+      "loss": 0.4636,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2592665615141956,
+      "grad_norm": 0.6663223299060342,
+      "learning_rate": 1.9799216142483358e-05,
+      "loss": 0.4374,
+      "step": 1315
+    },
+    {
+      "epoch": 0.25946372239747634,
+      "grad_norm": 0.7322544020861116,
+      "learning_rate": 1.9798906975746462e-05,
+      "loss": 0.4616,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2596608832807571,
+      "grad_norm": 4.868764964094287,
+      "learning_rate": 1.9798597573582935e-05,
+      "loss": 0.5068,
+      "step": 1317
+    },
+    {
+      "epoch": 0.25985804416403785,
+      "grad_norm": 1.6121049926148328,
+      "learning_rate": 1.9798287936000203e-05,
+      "loss": 0.4858,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2600552050473186,
+      "grad_norm": 0.7881657434369355,
+      "learning_rate": 1.979797806300572e-05,
+      "loss": 0.4707,
+      "step": 1319
+    },
+    {
+      "epoch": 0.26025236593059936,
+      "grad_norm": 0.6719396543796176,
+      "learning_rate": 1.9797667954606923e-05,
+      "loss": 0.4375,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2604495268138801,
+      "grad_norm": 0.677646758521938,
+      "learning_rate": 1.9797357610811264e-05,
+      "loss": 0.3941,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2606466876971609,
+      "grad_norm": 0.6629789255172303,
+      "learning_rate": 1.9797047031626197e-05,
+      "loss": 0.4666,
+      "step": 1322
+    },
+    {
+      "epoch": 0.26084384858044163,
+      "grad_norm": 0.7016369371591297,
+      "learning_rate": 1.9796736217059184e-05,
+      "loss": 0.4176,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2610410094637224,
+      "grad_norm": 0.7392651241817684,
+      "learning_rate": 1.9796425167117697e-05,
+      "loss": 0.4569,
+      "step": 1324
+    },
+    {
+      "epoch": 0.26123817034700314,
+      "grad_norm": 3.4278364452599943,
+      "learning_rate": 1.9796113881809207e-05,
+      "loss": 0.4897,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2614353312302839,
+      "grad_norm": 0.7430113958710807,
+      "learning_rate": 1.979580236114119e-05,
+      "loss": 0.4281,
+      "step": 1326
+    },
+    {
+      "epoch": 0.26163249211356465,
+      "grad_norm": 0.7574764251542814,
+      "learning_rate": 1.9795490605121133e-05,
+      "loss": 0.4285,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2618296529968454,
+      "grad_norm": 0.7573150663369252,
+      "learning_rate": 1.9795178613756526e-05,
+      "loss": 0.4214,
+      "step": 1328
+    },
+    {
+      "epoch": 0.26202681388012616,
+      "grad_norm": 0.7233219537322484,
+      "learning_rate": 1.9794866387054866e-05,
+      "loss": 0.4333,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2622239747634069,
+      "grad_norm": 0.6536227209487654,
+      "learning_rate": 1.9794553925023648e-05,
+      "loss": 0.508,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2624211356466877,
+      "grad_norm": 0.697150695998406,
+      "learning_rate": 1.979424122767039e-05,
+      "loss": 0.4402,
+      "step": 1331
+    },
+    {
+      "epoch": 0.26261829652996843,
+      "grad_norm": 1.1510688800982456,
+      "learning_rate": 1.97939282950026e-05,
+      "loss": 0.4935,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2628154574132492,
+      "grad_norm": 0.6936122849738883,
+      "learning_rate": 1.979361512702779e-05,
+      "loss": 0.4369,
+      "step": 1333
+    },
+    {
+      "epoch": 0.26301261829652994,
+      "grad_norm": 0.7316037732825653,
+      "learning_rate": 1.9793301723753494e-05,
+      "loss": 0.4774,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2632097791798107,
+      "grad_norm": 0.7046667225142733,
+      "learning_rate": 1.9792988085187237e-05,
+      "loss": 0.4586,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2634069400630915,
+      "grad_norm": 0.7576252182556432,
+      "learning_rate": 1.9792674211336557e-05,
+      "loss": 0.4444,
+      "step": 1336
+    },
+    {
+      "epoch": 0.26360410094637227,
+      "grad_norm": 0.702436953338292,
+      "learning_rate": 1.9792360102208987e-05,
+      "loss": 0.4508,
+      "step": 1337
+    },
+    {
+      "epoch": 0.263801261829653,
+      "grad_norm": 0.8581168449077242,
+      "learning_rate": 1.9792045757812083e-05,
+      "loss": 0.4536,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2639984227129338,
+      "grad_norm": 0.7421141473102278,
+      "learning_rate": 1.9791731178153398e-05,
+      "loss": 0.4884,
+      "step": 1339
+    },
+    {
+      "epoch": 0.26419558359621453,
+      "grad_norm": 0.8029593819817289,
+      "learning_rate": 1.979141636324048e-05,
+      "loss": 0.5005,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2643927444794953,
+      "grad_norm": 0.7626034173756989,
+      "learning_rate": 1.97911013130809e-05,
+      "loss": 0.4701,
+      "step": 1341
+    },
+    {
+      "epoch": 0.26458990536277605,
+      "grad_norm": 0.6947755523676065,
+      "learning_rate": 1.979078602768223e-05,
+      "loss": 0.4649,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2647870662460568,
+      "grad_norm": 0.7173606924037355,
+      "learning_rate": 1.9790470507052043e-05,
+      "loss": 0.4636,
+      "step": 1343
+    },
+    {
+      "epoch": 0.26498422712933756,
+      "grad_norm": 0.7174536980291077,
+      "learning_rate": 1.979015475119791e-05,
+      "loss": 0.4722,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2651813880126183,
+      "grad_norm": 0.7091313571434286,
+      "learning_rate": 1.978983876012743e-05,
+      "loss": 0.4604,
+      "step": 1345
+    },
+    {
+      "epoch": 0.26537854889589907,
+      "grad_norm": 0.7938427813870562,
+      "learning_rate": 1.978952253384819e-05,
+      "loss": 0.4683,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2655757097791798,
+      "grad_norm": 0.8665000592475282,
+      "learning_rate": 1.9789206072367788e-05,
+      "loss": 0.4651,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2657728706624606,
+      "grad_norm": 0.7476112206662554,
+      "learning_rate": 1.9788889375693826e-05,
+      "loss": 0.4501,
+      "step": 1348
+    },
+    {
+      "epoch": 0.26597003154574134,
+      "grad_norm": 0.6411119915064069,
+      "learning_rate": 1.978857244383391e-05,
+      "loss": 0.4332,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2661671924290221,
+      "grad_norm": 0.7501426692435633,
+      "learning_rate": 1.9788255276795665e-05,
+      "loss": 0.4482,
+      "step": 1350
+    },
+    {
+      "epoch": 0.26636435331230285,
+      "grad_norm": 0.6924050647879918,
+      "learning_rate": 1.97879378745867e-05,
+      "loss": 0.4781,
+      "step": 1351
+    },
+    {
+      "epoch": 0.2665615141955836,
+      "grad_norm": 0.7692951834407704,
+      "learning_rate": 1.9787620237214648e-05,
+      "loss": 0.4492,
+      "step": 1352
+    },
+    {
+      "epoch": 0.26675867507886436,
+      "grad_norm": 0.7874785369772107,
+      "learning_rate": 1.9787302364687137e-05,
+      "loss": 0.4625,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2669558359621451,
+      "grad_norm": 0.7706538665841983,
+      "learning_rate": 1.9786984257011804e-05,
+      "loss": 0.5065,
+      "step": 1354
+    },
+    {
+      "epoch": 0.26715299684542587,
+      "grad_norm": 0.9023943582130787,
+      "learning_rate": 1.9786665914196293e-05,
+      "loss": 0.4269,
+      "step": 1355
+    },
+    {
+      "epoch": 0.26735015772870663,
+      "grad_norm": 0.7444077480799488,
+      "learning_rate": 1.978634733624825e-05,
+      "loss": 0.4632,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2675473186119874,
+      "grad_norm": 0.750725217959198,
+      "learning_rate": 1.9786028523175334e-05,
+      "loss": 0.4813,
+      "step": 1357
+    },
+    {
+      "epoch": 0.26774447949526814,
+      "grad_norm": 0.6159139836796179,
+      "learning_rate": 1.9785709474985205e-05,
+      "loss": 0.4647,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2679416403785489,
+      "grad_norm": 0.7129947513385421,
+      "learning_rate": 1.978539019168552e-05,
+      "loss": 0.4177,
+      "step": 1359
+    },
+    {
+      "epoch": 0.26813880126182965,
+      "grad_norm": 0.6516268614502504,
+      "learning_rate": 1.9785070673283958e-05,
+      "loss": 0.4365,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2683359621451104,
+      "grad_norm": 0.7106387839347237,
+      "learning_rate": 1.9784750919788192e-05,
+      "loss": 0.4908,
+      "step": 1361
+    },
+    {
+      "epoch": 0.26853312302839116,
+      "grad_norm": 0.6645867370230059,
+      "learning_rate": 1.978443093120591e-05,
+      "loss": 0.4646,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2687302839116719,
+      "grad_norm": 0.6609654446353767,
+      "learning_rate": 1.978411070754479e-05,
+      "loss": 0.4464,
+      "step": 1363
+    },
+    {
+      "epoch": 0.2689274447949527,
+      "grad_norm": 0.6509684679600655,
+      "learning_rate": 1.9783790248812535e-05,
+      "loss": 0.4378,
+      "step": 1364
+    },
+    {
+      "epoch": 0.26912460567823343,
+      "grad_norm": 0.7410255193953944,
+      "learning_rate": 1.9783469555016838e-05,
+      "loss": 0.4883,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2693217665615142,
+      "grad_norm": 0.6410457849255176,
+      "learning_rate": 1.9783148626165408e-05,
+      "loss": 0.4566,
+      "step": 1366
+    },
+    {
+      "epoch": 0.26951892744479494,
+      "grad_norm": 0.7212189325599111,
+      "learning_rate": 1.978282746226595e-05,
+      "loss": 0.4748,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2697160883280757,
+      "grad_norm": 0.6386044383277297,
+      "learning_rate": 1.9782506063326188e-05,
+      "loss": 0.4525,
+      "step": 1368
+    },
+    {
+      "epoch": 0.26991324921135645,
+      "grad_norm": 0.6183024303790638,
+      "learning_rate": 1.978218442935384e-05,
+      "loss": 0.4419,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2701104100946372,
+      "grad_norm": 1.0311747282997497,
+      "learning_rate": 1.9781862560356632e-05,
+      "loss": 0.4705,
+      "step": 1370
+    },
+    {
+      "epoch": 0.27030757097791797,
+      "grad_norm": 0.8876601008055358,
+      "learning_rate": 1.97815404563423e-05,
+      "loss": 0.4952,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2705047318611987,
+      "grad_norm": 0.6707553837334197,
+      "learning_rate": 1.978121811731858e-05,
+      "loss": 0.4727,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2707018927444795,
+      "grad_norm": 0.6324861998325033,
+      "learning_rate": 1.978089554329322e-05,
+      "loss": 0.4512,
+      "step": 1373
+    },
+    {
+      "epoch": 0.27089905362776023,
+      "grad_norm": 0.7208718808928518,
+      "learning_rate": 1.9780572734273965e-05,
+      "loss": 0.4708,
+      "step": 1374
+    },
+    {
+      "epoch": 0.271096214511041,
+      "grad_norm": 0.6289916939191427,
+      "learning_rate": 1.9780249690268577e-05,
+      "loss": 0.46,
+      "step": 1375
+    },
+    {
+      "epoch": 0.27129337539432175,
+      "grad_norm": 0.6255312002347381,
+      "learning_rate": 1.977992641128481e-05,
+      "loss": 0.4113,
+      "step": 1376
+    },
+    {
+      "epoch": 0.2714905362776025,
+      "grad_norm": 0.7538032993514353,
+      "learning_rate": 1.977960289733044e-05,
+      "loss": 0.5171,
+      "step": 1377
+    },
+    {
+      "epoch": 0.27168769716088326,
+      "grad_norm": 0.6617762757068099,
+      "learning_rate": 1.977927914841323e-05,
+      "loss": 0.4574,
+      "step": 1378
+    },
+    {
+      "epoch": 0.271884858044164,
+      "grad_norm": 0.6317670145204217,
+      "learning_rate": 1.9778955164540966e-05,
+      "loss": 0.4416,
+      "step": 1379
+    },
+    {
+      "epoch": 0.27208201892744477,
+      "grad_norm": 0.6625339760031199,
+      "learning_rate": 1.977863094572143e-05,
+      "loss": 0.4695,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2722791798107255,
+      "grad_norm": 0.8851653534578081,
+      "learning_rate": 1.977830649196241e-05,
+      "loss": 0.469,
+      "step": 1381
+    },
+    {
+      "epoch": 0.27247634069400634,
+      "grad_norm": 0.6525212576179507,
+      "learning_rate": 1.9777981803271702e-05,
+      "loss": 0.4607,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2726735015772871,
+      "grad_norm": 0.687794207764544,
+      "learning_rate": 1.9777656879657104e-05,
+      "loss": 0.4329,
+      "step": 1383
+    },
+    {
+      "epoch": 0.27287066246056785,
+      "grad_norm": 0.6477780040394759,
+      "learning_rate": 1.9777331721126432e-05,
+      "loss": 0.455,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2730678233438486,
+      "grad_norm": 0.6704764509430108,
+      "learning_rate": 1.9777006327687486e-05,
+      "loss": 0.4756,
+      "step": 1385
+    },
+    {
+      "epoch": 0.27326498422712936,
+      "grad_norm": 0.6673187768283187,
+      "learning_rate": 1.9776680699348093e-05,
+      "loss": 0.4525,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2734621451104101,
+      "grad_norm": 0.6725761328570758,
+      "learning_rate": 1.977635483611607e-05,
+      "loss": 0.4465,
+      "step": 1387
+    },
+    {
+      "epoch": 0.27365930599369087,
+      "grad_norm": 0.6036295529404659,
+      "learning_rate": 1.9776028737999256e-05,
+      "loss": 0.4371,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2738564668769716,
+      "grad_norm": 0.649936023702026,
+      "learning_rate": 1.9775702405005473e-05,
+      "loss": 0.4419,
+      "step": 1389
+    },
+    {
+      "epoch": 0.2740536277602524,
+      "grad_norm": 0.6259329731865518,
+      "learning_rate": 1.977537583714257e-05,
+      "loss": 0.4607,
+      "step": 1390
+    },
+    {
+      "epoch": 0.27425078864353314,
+      "grad_norm": 0.9337258788421762,
+      "learning_rate": 1.9775049034418384e-05,
+      "loss": 0.4314,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2744479495268139,
+      "grad_norm": 0.7553643495249596,
+      "learning_rate": 1.977472199684078e-05,
+      "loss": 0.4019,
+      "step": 1392
+    },
+    {
+      "epoch": 0.27464511041009465,
+      "grad_norm": 0.6681682048717218,
+      "learning_rate": 1.9774394724417608e-05,
+      "loss": 0.4669,
+      "step": 1393
+    },
+    {
+      "epoch": 0.2748422712933754,
+      "grad_norm": 0.6660846989527527,
+      "learning_rate": 1.977406721715673e-05,
+      "loss": 0.4592,
+      "step": 1394
+    },
+    {
+      "epoch": 0.27503943217665616,
+      "grad_norm": 1.030125307014766,
+      "learning_rate": 1.9773739475066015e-05,
+      "loss": 0.4519,
+      "step": 1395
+    },
+    {
+      "epoch": 0.2752365930599369,
+      "grad_norm": 0.6165906390246885,
+      "learning_rate": 1.977341149815334e-05,
+      "loss": 0.4611,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2754337539432177,
+      "grad_norm": 0.6301071810608176,
+      "learning_rate": 1.977308328642658e-05,
+      "loss": 0.4728,
+      "step": 1397
+    },
+    {
+      "epoch": 0.27563091482649843,
+      "grad_norm": 0.603328270231981,
+      "learning_rate": 1.9772754839893627e-05,
+      "loss": 0.434,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2758280757097792,
+      "grad_norm": 0.6648768357451612,
+      "learning_rate": 1.9772426158562367e-05,
+      "loss": 0.4612,
+      "step": 1399
+    },
+    {
+      "epoch": 0.27602523659305994,
+      "grad_norm": 0.6209334691515703,
+      "learning_rate": 1.9772097242440703e-05,
+      "loss": 0.4629,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2762223974763407,
+      "grad_norm": 0.6615598121420424,
+      "learning_rate": 1.9771768091536528e-05,
+      "loss": 0.4546,
+      "step": 1401
+    },
+    {
+      "epoch": 0.27641955835962145,
+      "grad_norm": 0.6824317459563148,
+      "learning_rate": 1.977143870585776e-05,
+      "loss": 0.4428,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2766167192429022,
+      "grad_norm": 0.6670787550837416,
+      "learning_rate": 1.9771109085412304e-05,
+      "loss": 0.4997,
+      "step": 1403
+    },
+    {
+      "epoch": 0.27681388012618297,
+      "grad_norm": 0.8229581570264787,
+      "learning_rate": 1.9770779230208088e-05,
+      "loss": 0.4375,
+      "step": 1404
+    },
+    {
+      "epoch": 0.2770110410094637,
+      "grad_norm": 0.637951151429009,
+      "learning_rate": 1.977044914025303e-05,
+      "loss": 0.4322,
+      "step": 1405
+    },
+    {
+      "epoch": 0.2772082018927445,
+      "grad_norm": 0.607442608155513,
+      "learning_rate": 1.9770118815555063e-05,
+      "loss": 0.4242,
+      "step": 1406
+    },
+    {
+      "epoch": 0.27740536277602523,
+      "grad_norm": 0.6993192647998349,
+      "learning_rate": 1.9769788256122125e-05,
+      "loss": 0.4741,
+      "step": 1407
+    },
+    {
+      "epoch": 0.277602523659306,
+      "grad_norm": 0.6691537581034921,
+      "learning_rate": 1.9769457461962154e-05,
+      "loss": 0.4504,
+      "step": 1408
+    },
+    {
+      "epoch": 0.27779968454258674,
+      "grad_norm": 0.7485422674380534,
+      "learning_rate": 1.9769126433083102e-05,
+      "loss": 0.4825,
+      "step": 1409
+    },
+    {
+      "epoch": 0.2779968454258675,
+      "grad_norm": 0.6486895377627935,
+      "learning_rate": 1.976879516949292e-05,
+      "loss": 0.459,
+      "step": 1410
+    },
+    {
+      "epoch": 0.27819400630914826,
+      "grad_norm": 0.669976702587913,
+      "learning_rate": 1.976846367119957e-05,
+      "loss": 0.4618,
+      "step": 1411
+    },
+    {
+      "epoch": 0.278391167192429,
+      "grad_norm": 0.6947541489605775,
+      "learning_rate": 1.976813193821101e-05,
+      "loss": 0.4782,
+      "step": 1412
+    },
+    {
+      "epoch": 0.27858832807570977,
+      "grad_norm": 0.7279085002000397,
+      "learning_rate": 1.9767799970535214e-05,
+      "loss": 0.4584,
+      "step": 1413
+    },
+    {
+      "epoch": 0.2787854889589905,
+      "grad_norm": 0.6324612775847488,
+      "learning_rate": 1.9767467768180163e-05,
+      "loss": 0.4429,
+      "step": 1414
+    },
+    {
+      "epoch": 0.2789826498422713,
+      "grad_norm": 0.7075686828831383,
+      "learning_rate": 1.9767135331153827e-05,
+      "loss": 0.497,
+      "step": 1415
+    },
+    {
+      "epoch": 0.27917981072555204,
+      "grad_norm": 0.677366770732698,
+      "learning_rate": 1.97668026594642e-05,
+      "loss": 0.4739,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2793769716088328,
+      "grad_norm": 0.6413070384790073,
+      "learning_rate": 1.9766469753119274e-05,
+      "loss": 0.4609,
+      "step": 1417
+    },
+    {
+      "epoch": 0.27957413249211355,
+      "grad_norm": 0.6875558936981296,
+      "learning_rate": 1.976613661212705e-05,
+      "loss": 0.5027,
+      "step": 1418
+    },
+    {
+      "epoch": 0.2797712933753943,
+      "grad_norm": 0.7069012738884121,
+      "learning_rate": 1.976580323649553e-05,
+      "loss": 0.4675,
+      "step": 1419
+    },
+    {
+      "epoch": 0.27996845425867506,
+      "grad_norm": 0.6999152312103506,
+      "learning_rate": 1.976546962623272e-05,
+      "loss": 0.479,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2801656151419558,
+      "grad_norm": 0.632966968386772,
+      "learning_rate": 1.9765135781346637e-05,
+      "loss": 0.4458,
+      "step": 1421
+    },
+    {
+      "epoch": 0.28036277602523657,
+      "grad_norm": 0.680032576952019,
+      "learning_rate": 1.9764801701845307e-05,
+      "loss": 0.4904,
+      "step": 1422
+    },
+    {
+      "epoch": 0.2805599369085173,
+      "grad_norm": 0.6793474835680345,
+      "learning_rate": 1.9764467387736748e-05,
+      "loss": 0.4594,
+      "step": 1423
+    },
+    {
+      "epoch": 0.2807570977917981,
+      "grad_norm": 0.6990046169084455,
+      "learning_rate": 1.9764132839029e-05,
+      "loss": 0.4825,
+      "step": 1424
+    },
+    {
+      "epoch": 0.28095425867507884,
+      "grad_norm": 0.7012958434364689,
+      "learning_rate": 1.9763798055730096e-05,
+      "loss": 0.4394,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2811514195583596,
+      "grad_norm": 0.649381673277496,
+      "learning_rate": 1.9763463037848082e-05,
+      "loss": 0.4832,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2813485804416404,
+      "grad_norm": 0.7051628624005888,
+      "learning_rate": 1.9763127785391007e-05,
+      "loss": 0.4962,
+      "step": 1427
+    },
+    {
+      "epoch": 0.28154574132492116,
+      "grad_norm": 0.6757095316378556,
+      "learning_rate": 1.976279229836692e-05,
+      "loss": 0.4448,
+      "step": 1428
+    },
+    {
+      "epoch": 0.2817429022082019,
+      "grad_norm": 0.7505169408393606,
+      "learning_rate": 1.976245657678389e-05,
+      "loss": 0.4875,
+      "step": 1429
+    },
+    {
+      "epoch": 0.2819400630914827,
+      "grad_norm": 0.6482256189057635,
+      "learning_rate": 1.9762120620649978e-05,
+      "loss": 0.4457,
+      "step": 1430
+    },
+    {
+      "epoch": 0.28213722397476343,
+      "grad_norm": 0.6341748454996773,
+      "learning_rate": 1.9761784429973257e-05,
+      "loss": 0.4298,
+      "step": 1431
+    },
+    {
+      "epoch": 0.2823343848580442,
+      "grad_norm": 0.6426740598337791,
+      "learning_rate": 1.9761448004761804e-05,
+      "loss": 0.4351,
+      "step": 1432
+    },
+    {
+      "epoch": 0.28253154574132494,
+      "grad_norm": 0.6893390849187452,
+      "learning_rate": 1.9761111345023702e-05,
+      "loss": 0.4528,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2827287066246057,
+      "grad_norm": 0.6931202899611968,
+      "learning_rate": 1.976077445076704e-05,
+      "loss": 0.4504,
+      "step": 1434
+    },
+    {
+      "epoch": 0.28292586750788645,
+      "grad_norm": 0.6396492394646054,
+      "learning_rate": 1.976043732199991e-05,
+      "loss": 0.5051,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2831230283911672,
+      "grad_norm": 0.6068638024331832,
+      "learning_rate": 1.9760099958730414e-05,
+      "loss": 0.3965,
+      "step": 1436
+    },
+    {
+      "epoch": 0.28332018927444796,
+      "grad_norm": 0.6793441643898733,
+      "learning_rate": 1.9759762360966658e-05,
+      "loss": 0.4565,
+      "step": 1437
+    },
+    {
+      "epoch": 0.2835173501577287,
+      "grad_norm": 0.6338152033705464,
+      "learning_rate": 1.9759424528716748e-05,
+      "loss": 0.3908,
+      "step": 1438
+    },
+    {
+      "epoch": 0.2837145110410095,
+      "grad_norm": 0.6375149590618007,
+      "learning_rate": 1.975908646198881e-05,
+      "loss": 0.4457,
+      "step": 1439
+    },
+    {
+      "epoch": 0.28391167192429023,
+      "grad_norm": 0.6912444261554432,
+      "learning_rate": 1.9758748160790956e-05,
+      "loss": 0.4556,
+      "step": 1440
+    },
+    {
+      "epoch": 0.284108832807571,
+      "grad_norm": 0.7308297162675815,
+      "learning_rate": 1.975840962513132e-05,
+      "loss": 0.484,
+      "step": 1441
+    },
+    {
+      "epoch": 0.28430599369085174,
+      "grad_norm": 0.7358253544207752,
+      "learning_rate": 1.9758070855018033e-05,
+      "loss": 0.4667,
+      "step": 1442
+    },
+    {
+      "epoch": 0.2845031545741325,
+      "grad_norm": 0.6240361195624918,
+      "learning_rate": 1.975773185045924e-05,
+      "loss": 0.4549,
+      "step": 1443
+    },
+    {
+      "epoch": 0.28470031545741326,
+      "grad_norm": 0.6895402014641637,
+      "learning_rate": 1.975739261146308e-05,
+      "loss": 0.4234,
+      "step": 1444
+    },
+    {
+      "epoch": 0.284897476340694,
+      "grad_norm": 2.677764220478418,
+      "learning_rate": 1.97570531380377e-05,
+      "loss": 0.4986,
+      "step": 1445
+    },
+    {
+      "epoch": 0.28509463722397477,
+      "grad_norm": 0.7134750571138294,
+      "learning_rate": 1.975671343019126e-05,
+      "loss": 0.4143,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2852917981072555,
+      "grad_norm": 0.7374402505710781,
+      "learning_rate": 1.9756373487931932e-05,
+      "loss": 0.478,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2854889589905363,
+      "grad_norm": 0.7841802091710707,
+      "learning_rate": 1.975603331126787e-05,
+      "loss": 0.4791,
+      "step": 1448
+    },
+    {
+      "epoch": 0.28568611987381703,
+      "grad_norm": 0.7851659427951097,
+      "learning_rate": 1.975569290020725e-05,
+      "loss": 0.4445,
+      "step": 1449
+    },
+    {
+      "epoch": 0.2858832807570978,
+      "grad_norm": 0.7016574816261865,
+      "learning_rate": 1.9755352254758253e-05,
+      "loss": 0.4718,
+      "step": 1450
+    },
+    {
+      "epoch": 0.28608044164037855,
+      "grad_norm": 1.360154947822406,
+      "learning_rate": 1.975501137492906e-05,
+      "loss": 0.4778,
+      "step": 1451
+    },
+    {
+      "epoch": 0.2862776025236593,
+      "grad_norm": 0.7088540961818571,
+      "learning_rate": 1.9754670260727865e-05,
+      "loss": 0.4956,
+      "step": 1452
+    },
+    {
+      "epoch": 0.28647476340694006,
+      "grad_norm": 0.6877943073595973,
+      "learning_rate": 1.9754328912162864e-05,
+      "loss": 0.4666,
+      "step": 1453
+    },
+    {
+      "epoch": 0.2866719242902208,
+      "grad_norm": 0.7632821008022346,
+      "learning_rate": 1.975398732924225e-05,
+      "loss": 0.4624,
+      "step": 1454
+    },
+    {
+      "epoch": 0.28686908517350157,
+      "grad_norm": 0.649955168958308,
+      "learning_rate": 1.975364551197424e-05,
+      "loss": 0.4327,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2870662460567823,
+      "grad_norm": 1.5748278853806321,
+      "learning_rate": 1.975330346036704e-05,
+      "loss": 0.4463,
+      "step": 1456
+    },
+    {
+      "epoch": 0.2872634069400631,
+      "grad_norm": 0.6418235523968513,
+      "learning_rate": 1.975296117442887e-05,
+      "loss": 0.4426,
+      "step": 1457
+    },
+    {
+      "epoch": 0.28746056782334384,
+      "grad_norm": 0.7529789256896934,
+      "learning_rate": 1.9752618654167954e-05,
+      "loss": 0.4525,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2876577287066246,
+      "grad_norm": 0.6837140757349681,
+      "learning_rate": 1.975227589959252e-05,
+      "loss": 0.4664,
+      "step": 1459
+    },
+    {
+      "epoch": 0.28785488958990535,
+      "grad_norm": 0.7302417926946705,
+      "learning_rate": 1.9751932910710808e-05,
+      "loss": 0.4344,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2880520504731861,
+      "grad_norm": 0.6549023096371199,
+      "learning_rate": 1.9751589687531052e-05,
+      "loss": 0.4637,
+      "step": 1461
+    },
+    {
+      "epoch": 0.28824921135646686,
+      "grad_norm": 0.7572816037486845,
+      "learning_rate": 1.97512462300615e-05,
+      "loss": 0.4808,
+      "step": 1462
+    },
+    {
+      "epoch": 0.2884463722397476,
+      "grad_norm": 0.7438685296075004,
+      "learning_rate": 1.9750902538310407e-05,
+      "loss": 0.4554,
+      "step": 1463
+    },
+    {
+      "epoch": 0.2886435331230284,
+      "grad_norm": 0.7921043989775165,
+      "learning_rate": 1.9750558612286025e-05,
+      "loss": 0.4669,
+      "step": 1464
+    },
+    {
+      "epoch": 0.28884069400630913,
+      "grad_norm": 1.7620430759198427,
+      "learning_rate": 1.9750214451996623e-05,
+      "loss": 0.4356,
+      "step": 1465
+    },
+    {
+      "epoch": 0.2890378548895899,
+      "grad_norm": 0.7869312505011672,
+      "learning_rate": 1.9749870057450464e-05,
+      "loss": 0.4502,
+      "step": 1466
+    },
+    {
+      "epoch": 0.28923501577287064,
+      "grad_norm": 0.7845532647336225,
+      "learning_rate": 1.974952542865583e-05,
+      "loss": 0.4835,
+      "step": 1467
+    },
+    {
+      "epoch": 0.2894321766561514,
+      "grad_norm": 0.7515938063415738,
+      "learning_rate": 1.9749180565620995e-05,
+      "loss": 0.4507,
+      "step": 1468
+    },
+    {
+      "epoch": 0.28962933753943215,
+      "grad_norm": 0.8913235246679625,
+      "learning_rate": 1.9748835468354243e-05,
+      "loss": 0.4729,
+      "step": 1469
+    },
+    {
+      "epoch": 0.2898264984227129,
+      "grad_norm": 0.8870979739457769,
+      "learning_rate": 1.9748490136863867e-05,
+      "loss": 0.4391,
+      "step": 1470
+    },
+    {
+      "epoch": 0.29002365930599366,
+      "grad_norm": 0.7727887896133023,
+      "learning_rate": 1.9748144571158167e-05,
+      "loss": 0.501,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2902208201892745,
+      "grad_norm": 0.6725824982109148,
+      "learning_rate": 1.9747798771245446e-05,
+      "loss": 0.4284,
+      "step": 1472
+    },
+    {
+      "epoch": 0.29041798107255523,
+      "grad_norm": 0.7636840941258795,
+      "learning_rate": 1.9747452737134006e-05,
+      "loss": 0.4975,
+      "step": 1473
+    },
+    {
+      "epoch": 0.290615141955836,
+      "grad_norm": 0.695534933075417,
+      "learning_rate": 1.974710646883217e-05,
+      "loss": 0.4367,
+      "step": 1474
+    },
+    {
+      "epoch": 0.29081230283911674,
+      "grad_norm": 0.7808275437629759,
+      "learning_rate": 1.9746759966348244e-05,
+      "loss": 0.4821,
+      "step": 1475
+    },
+    {
+      "epoch": 0.2910094637223975,
+      "grad_norm": 0.6839691016507893,
+      "learning_rate": 1.9746413229690565e-05,
+      "loss": 0.4496,
+      "step": 1476
+    },
+    {
+      "epoch": 0.29120662460567825,
+      "grad_norm": 0.7291630955736097,
+      "learning_rate": 1.974606625886746e-05,
+      "loss": 0.467,
+      "step": 1477
+    },
+    {
+      "epoch": 0.291403785488959,
+      "grad_norm": 0.7415071271998195,
+      "learning_rate": 1.9745719053887265e-05,
+      "loss": 0.4563,
+      "step": 1478
+    },
+    {
+      "epoch": 0.29160094637223977,
+      "grad_norm": 0.721134342027533,
+      "learning_rate": 1.974537161475832e-05,
+      "loss": 0.4157,
+      "step": 1479
+    },
+    {
+      "epoch": 0.2917981072555205,
+      "grad_norm": 0.6815829464296383,
+      "learning_rate": 1.9745023941488974e-05,
+      "loss": 0.4959,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2919952681388013,
+      "grad_norm": 0.6455560207981099,
+      "learning_rate": 1.974467603408758e-05,
+      "loss": 0.4131,
+      "step": 1481
+    },
+    {
+      "epoch": 0.29219242902208203,
+      "grad_norm": 0.7495965931748421,
+      "learning_rate": 1.9744327892562497e-05,
+      "loss": 0.4536,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2923895899053628,
+      "grad_norm": 4.622108646120691,
+      "learning_rate": 1.974397951692209e-05,
+      "loss": 0.4472,
+      "step": 1483
+    },
+    {
+      "epoch": 0.29258675078864355,
+      "grad_norm": 0.7671628175372462,
+      "learning_rate": 1.9743630907174727e-05,
+      "loss": 0.4463,
+      "step": 1484
+    },
+    {
+      "epoch": 0.2927839116719243,
+      "grad_norm": 0.6306822829326908,
+      "learning_rate": 1.9743282063328786e-05,
+      "loss": 0.4321,
+      "step": 1485
+    },
+    {
+      "epoch": 0.29298107255520506,
+      "grad_norm": 0.8026977169526986,
+      "learning_rate": 1.9742932985392646e-05,
+      "loss": 0.44,
+      "step": 1486
+    },
+    {
+      "epoch": 0.2931782334384858,
+      "grad_norm": 0.6458962336741704,
+      "learning_rate": 1.97425836733747e-05,
+      "loss": 0.415,
+      "step": 1487
+    },
+    {
+      "epoch": 0.29337539432176657,
+      "grad_norm": 0.8366441003492903,
+      "learning_rate": 1.9742234127283328e-05,
+      "loss": 0.5109,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2935725552050473,
+      "grad_norm": 3.206384511793217,
+      "learning_rate": 1.9741884347126937e-05,
+      "loss": 0.4654,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2937697160883281,
+      "grad_norm": 1.0905491570369534,
+      "learning_rate": 1.9741534332913934e-05,
+      "loss": 0.4734,
+      "step": 1490
+    },
+    {
+      "epoch": 0.29396687697160884,
+      "grad_norm": 0.7932149212181082,
+      "learning_rate": 1.9741184084652723e-05,
+      "loss": 0.4809,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2941640378548896,
+      "grad_norm": 0.907352399489384,
+      "learning_rate": 1.9740833602351718e-05,
+      "loss": 0.4856,
+      "step": 1492
+    },
+    {
+      "epoch": 0.29436119873817035,
+      "grad_norm": 0.8172339427555916,
+      "learning_rate": 1.9740482886019342e-05,
+      "loss": 0.4348,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2945583596214511,
+      "grad_norm": 0.7379386874605652,
+      "learning_rate": 1.974013193566402e-05,
+      "loss": 0.4315,
+      "step": 1494
+    },
+    {
+      "epoch": 0.29475552050473186,
+      "grad_norm": 0.7315832037117764,
+      "learning_rate": 1.9739780751294188e-05,
+      "loss": 0.4302,
+      "step": 1495
+    },
+    {
+      "epoch": 0.2949526813880126,
+      "grad_norm": 0.7363421312105899,
+      "learning_rate": 1.9739429332918276e-05,
+      "loss": 0.4428,
+      "step": 1496
+    },
+    {
+      "epoch": 0.29514984227129337,
+      "grad_norm": 0.7954453286705275,
+      "learning_rate": 1.973907768054473e-05,
+      "loss": 0.4826,
+      "step": 1497
+    },
+    {
+      "epoch": 0.29534700315457413,
+      "grad_norm": 0.6920665293293689,
+      "learning_rate": 1.9738725794182004e-05,
+      "loss": 0.45,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2955441640378549,
+      "grad_norm": 0.7476711485594634,
+      "learning_rate": 1.9738373673838545e-05,
+      "loss": 0.47,
+      "step": 1499
+    },
+    {
+      "epoch": 0.29574132492113564,
+      "grad_norm": 0.6378986468629754,
+      "learning_rate": 1.9738021319522817e-05,
+      "loss": 0.446,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2959384858044164,
+      "grad_norm": 0.7449299773188399,
+      "learning_rate": 1.9737668731243284e-05,
+      "loss": 0.4622,
+      "step": 1501
+    },
+    {
+      "epoch": 0.29613564668769715,
+      "grad_norm": 0.6648614812967949,
+      "learning_rate": 1.973731590900842e-05,
+      "loss": 0.4512,
+      "step": 1502
+    },
+    {
+      "epoch": 0.2963328075709779,
+      "grad_norm": 0.7321775927414682,
+      "learning_rate": 1.97369628528267e-05,
+      "loss": 0.423,
+      "step": 1503
+    },
+    {
+      "epoch": 0.29652996845425866,
+      "grad_norm": 0.6801575733467247,
+      "learning_rate": 1.9736609562706604e-05,
+      "loss": 0.4424,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2967271293375394,
+      "grad_norm": 0.6497325650317418,
+      "learning_rate": 1.9736256038656624e-05,
+      "loss": 0.4494,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2969242902208202,
+      "grad_norm": 0.7063493502541413,
+      "learning_rate": 1.9735902280685252e-05,
+      "loss": 0.4375,
+      "step": 1506
+    },
+    {
+      "epoch": 0.29712145110410093,
+      "grad_norm": 0.6907055254378035,
+      "learning_rate": 1.9735548288800988e-05,
+      "loss": 0.4723,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2973186119873817,
+      "grad_norm": 0.6955235821677667,
+      "learning_rate": 1.9735194063012337e-05,
+      "loss": 0.4377,
+      "step": 1508
+    },
+    {
+      "epoch": 0.29751577287066244,
+      "grad_norm": 0.6505153342650509,
+      "learning_rate": 1.9734839603327805e-05,
+      "loss": 0.4741,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2977129337539432,
+      "grad_norm": 1.21645699588291,
+      "learning_rate": 1.973448490975592e-05,
+      "loss": 0.488,
+      "step": 1510
+    },
+    {
+      "epoch": 0.29791009463722395,
+      "grad_norm": 5.633561734216787,
+      "learning_rate": 1.9734129982305187e-05,
+      "loss": 0.5188,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2981072555205047,
+      "grad_norm": 5.960949833229553,
+      "learning_rate": 1.9733774820984146e-05,
+      "loss": 0.4864,
+      "step": 1512
+    },
+    {
+      "epoch": 0.29830441640378547,
+      "grad_norm": 0.8936987496926002,
+      "learning_rate": 1.9733419425801326e-05,
+      "loss": 0.4904,
+      "step": 1513
+    },
+    {
+      "epoch": 0.2985015772870662,
+      "grad_norm": 0.8736184927683982,
+      "learning_rate": 1.9733063796765267e-05,
+      "loss": 0.4807,
+      "step": 1514
+    },
+    {
+      "epoch": 0.298698738170347,
+      "grad_norm": 0.8389206829886541,
+      "learning_rate": 1.9732707933884508e-05,
+      "loss": 0.4708,
+      "step": 1515
+    },
+    {
+      "epoch": 0.29889589905362773,
+      "grad_norm": 0.877518939540052,
+      "learning_rate": 1.973235183716761e-05,
+      "loss": 0.4064,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2990930599369085,
+      "grad_norm": 1.0406389624738088,
+      "learning_rate": 1.9731995506623118e-05,
+      "loss": 0.4824,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2992902208201893,
+      "grad_norm": 0.8112544551340672,
+      "learning_rate": 1.9731638942259596e-05,
+      "loss": 0.4748,
+      "step": 1518
+    },
+    {
+      "epoch": 0.29948738170347006,
+      "grad_norm": 0.7560476486010564,
+      "learning_rate": 1.9731282144085613e-05,
+      "loss": 0.4459,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2996845425867508,
+      "grad_norm": 0.8150605495041554,
+      "learning_rate": 1.973092511210974e-05,
+      "loss": 0.4697,
+      "step": 1520
+    },
+    {
+      "epoch": 0.29988170347003157,
+      "grad_norm": 0.7564808190893473,
+      "learning_rate": 1.9730567846340552e-05,
+      "loss": 0.4705,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3000788643533123,
+      "grad_norm": 0.7262145297310585,
+      "learning_rate": 1.973021034678664e-05,
+      "loss": 0.3978,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3002760252365931,
+      "grad_norm": 0.7038980359439974,
+      "learning_rate": 1.9729852613456586e-05,
+      "loss": 0.4584,
+      "step": 1523
+    },
+    {
+      "epoch": 0.30047318611987384,
+      "grad_norm": 0.7588921782146265,
+      "learning_rate": 1.972949464635899e-05,
+      "loss": 0.4819,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3006703470031546,
+      "grad_norm": 0.7168024227685293,
+      "learning_rate": 1.9729136445502446e-05,
+      "loss": 0.4864,
+      "step": 1525
+    },
+    {
+      "epoch": 0.30086750788643535,
+      "grad_norm": 0.6493407202654529,
+      "learning_rate": 1.9728778010895567e-05,
+      "loss": 0.4107,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3010646687697161,
+      "grad_norm": 0.6952700541408848,
+      "learning_rate": 1.9728419342546962e-05,
+      "loss": 0.495,
+      "step": 1527
+    },
+    {
+      "epoch": 0.30126182965299686,
+      "grad_norm": 1.1339617708695726,
+      "learning_rate": 1.9728060440465246e-05,
+      "loss": 0.4262,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3014589905362776,
+      "grad_norm": 0.6443436669507463,
+      "learning_rate": 1.9727701304659046e-05,
+      "loss": 0.4323,
+      "step": 1529
+    },
+    {
+      "epoch": 0.30165615141955837,
+      "grad_norm": 0.7193540103567199,
+      "learning_rate": 1.9727341935136987e-05,
+      "loss": 0.4688,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3018533123028391,
+      "grad_norm": 0.6808229865935738,
+      "learning_rate": 1.9726982331907706e-05,
+      "loss": 0.4637,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3020504731861199,
+      "grad_norm": 0.6469885340908015,
+      "learning_rate": 1.972662249497984e-05,
+      "loss": 0.4533,
+      "step": 1532
+    },
+    {
+      "epoch": 0.30224763406940064,
+      "grad_norm": 0.8050241386096714,
+      "learning_rate": 1.9726262424362033e-05,
+      "loss": 0.4629,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3024447949526814,
+      "grad_norm": 1.1279481880937756,
+      "learning_rate": 1.9725902120062942e-05,
+      "loss": 0.4769,
+      "step": 1534
+    },
+    {
+      "epoch": 0.30264195583596215,
+      "grad_norm": 0.9362607518901818,
+      "learning_rate": 1.972554158209122e-05,
+      "loss": 0.4499,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3028391167192429,
+      "grad_norm": 0.6311429250558253,
+      "learning_rate": 1.972518081045553e-05,
+      "loss": 0.4707,
+      "step": 1536
+    },
+    {
+      "epoch": 0.30303627760252366,
+      "grad_norm": 0.6908119659769627,
+      "learning_rate": 1.9724819805164542e-05,
+      "loss": 0.459,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3032334384858044,
+      "grad_norm": 1.3140084577134628,
+      "learning_rate": 1.972445856622692e-05,
+      "loss": 0.4604,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3034305993690852,
+      "grad_norm": 1.4336266646144273,
+      "learning_rate": 1.9724097093651356e-05,
+      "loss": 0.5054,
+      "step": 1539
+    },
+    {
+      "epoch": 0.30362776025236593,
+      "grad_norm": 0.6266235358734742,
+      "learning_rate": 1.9723735387446526e-05,
+      "loss": 0.4195,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3038249211356467,
+      "grad_norm": 0.664253082557445,
+      "learning_rate": 1.9723373447621125e-05,
+      "loss": 0.4466,
+      "step": 1541
+    },
+    {
+      "epoch": 0.30402208201892744,
+      "grad_norm": 0.6631894672531906,
+      "learning_rate": 1.9723011274183844e-05,
+      "loss": 0.4613,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3042192429022082,
+      "grad_norm": 0.9552950397058307,
+      "learning_rate": 1.9722648867143384e-05,
+      "loss": 0.4328,
+      "step": 1543
+    },
+    {
+      "epoch": 0.30441640378548895,
+      "grad_norm": 0.6534330380880412,
+      "learning_rate": 1.972228622650846e-05,
+      "loss": 0.472,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3046135646687697,
+      "grad_norm": 0.6792854850313309,
+      "learning_rate": 1.972192335228778e-05,
+      "loss": 0.4639,
+      "step": 1545
+    },
+    {
+      "epoch": 0.30481072555205047,
+      "grad_norm": 0.8246065548514672,
+      "learning_rate": 1.972156024449006e-05,
+      "loss": 0.4678,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3050078864353312,
+      "grad_norm": 0.6805673028086667,
+      "learning_rate": 1.972119690312403e-05,
+      "loss": 0.4392,
+      "step": 1547
+    },
+    {
+      "epoch": 0.305205047318612,
+      "grad_norm": 1.1098363616208942,
+      "learning_rate": 1.9720833328198416e-05,
+      "loss": 0.457,
+      "step": 1548
+    },
+    {
+      "epoch": 0.30540220820189273,
+      "grad_norm": 1.9882754170201284,
+      "learning_rate": 1.972046951972195e-05,
+      "loss": 0.4782,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3055993690851735,
+      "grad_norm": 0.6655712983680275,
+      "learning_rate": 1.972010547770338e-05,
+      "loss": 0.4654,
+      "step": 1550
+    },
+    {
+      "epoch": 0.30579652996845424,
+      "grad_norm": 0.6357244449306213,
+      "learning_rate": 1.9719741202151442e-05,
+      "loss": 0.4337,
+      "step": 1551
+    },
+    {
+      "epoch": 0.305993690851735,
+      "grad_norm": 0.6432848249907509,
+      "learning_rate": 1.9719376693074898e-05,
+      "loss": 0.4656,
+      "step": 1552
+    },
+    {
+      "epoch": 0.30619085173501576,
+      "grad_norm": 0.6490834027479839,
+      "learning_rate": 1.97190119504825e-05,
+      "loss": 0.4442,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3063880126182965,
+      "grad_norm": 0.9593162347229904,
+      "learning_rate": 1.9718646974383016e-05,
+      "loss": 0.417,
+      "step": 1554
+    },
+    {
+      "epoch": 0.30658517350157727,
+      "grad_norm": 0.8582061855411773,
+      "learning_rate": 1.9718281764785213e-05,
+      "loss": 0.4533,
+      "step": 1555
+    },
+    {
+      "epoch": 0.306782334384858,
+      "grad_norm": 0.6829232907025286,
+      "learning_rate": 1.9717916321697862e-05,
+      "loss": 0.4531,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3069794952681388,
+      "grad_norm": 0.6946751006171858,
+      "learning_rate": 1.9717550645129745e-05,
+      "loss": 0.4332,
+      "step": 1557
+    },
+    {
+      "epoch": 0.30717665615141954,
+      "grad_norm": 0.7343732521374964,
+      "learning_rate": 1.971718473508965e-05,
+      "loss": 0.4563,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3073738170347003,
+      "grad_norm": 0.7092985413488407,
+      "learning_rate": 1.9716818591586367e-05,
+      "loss": 0.4427,
+      "step": 1559
+    },
+    {
+      "epoch": 0.30757097791798105,
+      "grad_norm": 0.6920199000103088,
+      "learning_rate": 1.9716452214628688e-05,
+      "loss": 0.4867,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3077681388012618,
+      "grad_norm": 0.7518768098411801,
+      "learning_rate": 1.9716085604225425e-05,
+      "loss": 0.4288,
+      "step": 1561
+    },
+    {
+      "epoch": 0.30796529968454256,
+      "grad_norm": 0.9921646912510204,
+      "learning_rate": 1.9715718760385377e-05,
+      "loss": 0.458,
+      "step": 1562
+    },
+    {
+      "epoch": 0.30816246056782337,
+      "grad_norm": 0.723626350748604,
+      "learning_rate": 1.9715351683117364e-05,
+      "loss": 0.4865,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3083596214511041,
+      "grad_norm": 0.6699114054096069,
+      "learning_rate": 1.9714984372430205e-05,
+      "loss": 0.4719,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3085567823343849,
+      "grad_norm": 0.7208356036921162,
+      "learning_rate": 1.971461682833272e-05,
+      "loss": 0.4499,
+      "step": 1565
+    },
+    {
+      "epoch": 0.30875394321766564,
+      "grad_norm": 0.696266504957194,
+      "learning_rate": 1.9714249050833743e-05,
+      "loss": 0.4685,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3089511041009464,
+      "grad_norm": 0.7045591753323159,
+      "learning_rate": 1.971388103994211e-05,
+      "loss": 0.4693,
+      "step": 1567
+    },
+    {
+      "epoch": 0.30914826498422715,
+      "grad_norm": 0.9972061656036422,
+      "learning_rate": 1.9713512795666663e-05,
+      "loss": 0.4655,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3093454258675079,
+      "grad_norm": 0.6343439339945319,
+      "learning_rate": 1.971314431801625e-05,
+      "loss": 0.4607,
+      "step": 1569
+    },
+    {
+      "epoch": 0.30954258675078866,
+      "grad_norm": 1.3214688571967796,
+      "learning_rate": 1.9712775606999718e-05,
+      "loss": 0.5253,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3097397476340694,
+      "grad_norm": 1.1643733839819645,
+      "learning_rate": 1.9712406662625934e-05,
+      "loss": 0.4845,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3099369085173502,
+      "grad_norm": 0.6395778049163567,
+      "learning_rate": 1.9712037484903758e-05,
+      "loss": 0.4646,
+      "step": 1572
+    },
+    {
+      "epoch": 0.31013406940063093,
+      "grad_norm": 0.728804625037493,
+      "learning_rate": 1.971166807384206e-05,
+      "loss": 0.4829,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3103312302839117,
+      "grad_norm": 0.6692642203500485,
+      "learning_rate": 1.9711298429449716e-05,
+      "loss": 0.4628,
+      "step": 1574
+    },
+    {
+      "epoch": 0.31052839116719244,
+      "grad_norm": 0.6801411829078509,
+      "learning_rate": 1.9710928551735606e-05,
+      "loss": 0.4806,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3107255520504732,
+      "grad_norm": 0.9147951787329351,
+      "learning_rate": 1.971055844070862e-05,
+      "loss": 0.4823,
+      "step": 1576
+    },
+    {
+      "epoch": 0.31092271293375395,
+      "grad_norm": 0.6704142482301383,
+      "learning_rate": 1.9710188096377645e-05,
+      "loss": 0.493,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3111198738170347,
+      "grad_norm": 0.7934181252723825,
+      "learning_rate": 1.9709817518751585e-05,
+      "loss": 0.4711,
+      "step": 1578
+    },
+    {
+      "epoch": 0.31131703470031546,
+      "grad_norm": 0.9638101618156172,
+      "learning_rate": 1.9709446707839336e-05,
+      "loss": 0.4734,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3115141955835962,
+      "grad_norm": 0.8058930264480674,
+      "learning_rate": 1.9709075663649812e-05,
+      "loss": 0.4672,
+      "step": 1580
+    },
+    {
+      "epoch": 0.311711356466877,
+      "grad_norm": 0.648347607286343,
+      "learning_rate": 1.9708704386191924e-05,
+      "loss": 0.4387,
+      "step": 1581
+    },
+    {
+      "epoch": 0.31190851735015773,
+      "grad_norm": 0.677310939518794,
+      "learning_rate": 1.97083328754746e-05,
+      "loss": 0.4441,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3121056782334385,
+      "grad_norm": 1.102580749565265,
+      "learning_rate": 1.9707961131506756e-05,
+      "loss": 0.4659,
+      "step": 1583
+    },
+    {
+      "epoch": 0.31230283911671924,
+      "grad_norm": 0.7198499814401207,
+      "learning_rate": 1.9707589154297328e-05,
+      "loss": 0.4805,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 0.6444043482783797,
+      "learning_rate": 1.9707216943855258e-05,
+      "loss": 0.4437,
+      "step": 1585
+    },
+    {
+      "epoch": 0.31269716088328076,
+      "grad_norm": 0.8244715296197911,
+      "learning_rate": 1.970684450018948e-05,
+      "loss": 0.4363,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3128943217665615,
+      "grad_norm": 0.6445385325086617,
+      "learning_rate": 1.9706471823308946e-05,
+      "loss": 0.4599,
+      "step": 1587
+    },
+    {
+      "epoch": 0.31309148264984227,
+      "grad_norm": 0.699849167744654,
+      "learning_rate": 1.9706098913222608e-05,
+      "loss": 0.5236,
+      "step": 1588
+    },
+    {
+      "epoch": 0.313288643533123,
+      "grad_norm": 0.7229780587068287,
+      "learning_rate": 1.970572576993943e-05,
+      "loss": 0.412,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3134858044164038,
+      "grad_norm": 0.6479915833004943,
+      "learning_rate": 1.9705352393468374e-05,
+      "loss": 0.4697,
+      "step": 1590
+    },
+    {
+      "epoch": 0.31368296529968454,
+      "grad_norm": 0.650021857370611,
+      "learning_rate": 1.9704978783818413e-05,
+      "loss": 0.4326,
+      "step": 1591
+    },
+    {
+      "epoch": 0.3138801261829653,
+      "grad_norm": 0.6861324174548465,
+      "learning_rate": 1.970460494099852e-05,
+      "loss": 0.4125,
+      "step": 1592
+    },
+    {
+      "epoch": 0.31407728706624605,
+      "grad_norm": 0.6801449769884733,
+      "learning_rate": 1.9704230865017675e-05,
+      "loss": 0.5069,
+      "step": 1593
+    },
+    {
+      "epoch": 0.3142744479495268,
+      "grad_norm": 0.7205207674630276,
+      "learning_rate": 1.970385655588487e-05,
+      "loss": 0.4757,
+      "step": 1594
+    },
+    {
+      "epoch": 0.31447160883280756,
+      "grad_norm": 0.6629052999763434,
+      "learning_rate": 1.9703482013609098e-05,
+      "loss": 0.4351,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3146687697160883,
+      "grad_norm": 0.7088009223433047,
+      "learning_rate": 1.9703107238199356e-05,
+      "loss": 0.4652,
+      "step": 1596
+    },
+    {
+      "epoch": 0.31486593059936907,
+      "grad_norm": 0.9099735140613966,
+      "learning_rate": 1.9702732229664653e-05,
+      "loss": 0.4856,
+      "step": 1597
+    },
+    {
+      "epoch": 0.3150630914826498,
+      "grad_norm": 0.744041311387149,
+      "learning_rate": 1.9702356988013988e-05,
+      "loss": 0.4675,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3152602523659306,
+      "grad_norm": 0.6721850270536037,
+      "learning_rate": 1.970198151325639e-05,
+      "loss": 0.4876,
+      "step": 1599
+    },
+    {
+      "epoch": 0.31545741324921134,
+      "grad_norm": 0.7183831817138312,
+      "learning_rate": 1.9701605805400866e-05,
+      "loss": 0.4715,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3156545741324921,
+      "grad_norm": 0.6393423636617389,
+      "learning_rate": 1.9701229864456452e-05,
+      "loss": 0.4591,
+      "step": 1601
+    },
+    {
+      "epoch": 0.31585173501577285,
+      "grad_norm": 0.6363184919995188,
+      "learning_rate": 1.970085369043218e-05,
+      "loss": 0.4134,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3160488958990536,
+      "grad_norm": 0.703734451223413,
+      "learning_rate": 1.9700477283337084e-05,
+      "loss": 0.4589,
+      "step": 1603
+    },
+    {
+      "epoch": 0.31624605678233436,
+      "grad_norm": 0.6584994125713037,
+      "learning_rate": 1.9700100643180213e-05,
+      "loss": 0.4482,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3164432176656151,
+      "grad_norm": 0.6078730627425877,
+      "learning_rate": 1.9699723769970608e-05,
+      "loss": 0.4399,
+      "step": 1605
+    },
+    {
+      "epoch": 0.3166403785488959,
+      "grad_norm": 0.6961873263732445,
+      "learning_rate": 1.969934666371733e-05,
+      "loss": 0.4775,
+      "step": 1606
+    },
+    {
+      "epoch": 0.31683753943217663,
+      "grad_norm": 0.724397941037564,
+      "learning_rate": 1.969896932442944e-05,
+      "loss": 0.4981,
+      "step": 1607
+    },
+    {
+      "epoch": 0.31703470031545744,
+      "grad_norm": 0.7312028182555175,
+      "learning_rate": 1.9698591752115997e-05,
+      "loss": 0.4688,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3172318611987382,
+      "grad_norm": 0.6505489803123736,
+      "learning_rate": 1.969821394678608e-05,
+      "loss": 0.4298,
+      "step": 1609
+    },
+    {
+      "epoch": 0.31742902208201895,
+      "grad_norm": 0.6779628156309334,
+      "learning_rate": 1.969783590844876e-05,
+      "loss": 0.4806,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3176261829652997,
+      "grad_norm": 0.7180529086461475,
+      "learning_rate": 1.9697457637113126e-05,
+      "loss": 0.5021,
+      "step": 1611
+    },
+    {
+      "epoch": 0.31782334384858046,
+      "grad_norm": 0.6009368475283028,
+      "learning_rate": 1.969707913278826e-05,
+      "loss": 0.4411,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3180205047318612,
+      "grad_norm": 0.6337105416643031,
+      "learning_rate": 1.969670039548326e-05,
+      "loss": 0.4554,
+      "step": 1613
+    },
+    {
+      "epoch": 0.318217665615142,
+      "grad_norm": 0.7204205483322568,
+      "learning_rate": 1.9696321425207227e-05,
+      "loss": 0.4912,
+      "step": 1614
+    },
+    {
+      "epoch": 0.31841482649842273,
+      "grad_norm": 0.6061996997155742,
+      "learning_rate": 1.969594222196926e-05,
+      "loss": 0.4442,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3186119873817035,
+      "grad_norm": 0.6433833110520849,
+      "learning_rate": 1.9695562785778473e-05,
+      "loss": 0.4289,
+      "step": 1616
+    },
+    {
+      "epoch": 0.31880914826498424,
+      "grad_norm": 0.5906452185455755,
+      "learning_rate": 1.9695183116643983e-05,
+      "loss": 0.408,
+      "step": 1617
+    },
+    {
+      "epoch": 0.319006309148265,
+      "grad_norm": 0.6795767414105931,
+      "learning_rate": 1.9694803214574914e-05,
+      "loss": 0.4325,
+      "step": 1618
+    },
+    {
+      "epoch": 0.31920347003154576,
+      "grad_norm": 0.6337536961202738,
+      "learning_rate": 1.9694423079580387e-05,
+      "loss": 0.4426,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3194006309148265,
+      "grad_norm": 0.6793514535697415,
+      "learning_rate": 1.969404271166954e-05,
+      "loss": 0.4756,
+      "step": 1620
+    },
+    {
+      "epoch": 0.31959779179810727,
+      "grad_norm": 0.5997185524987442,
+      "learning_rate": 1.9693662110851507e-05,
+      "loss": 0.4417,
+      "step": 1621
+    },
+    {
+      "epoch": 0.319794952681388,
+      "grad_norm": 0.6223038049361463,
+      "learning_rate": 1.969328127713544e-05,
+      "loss": 0.4317,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3199921135646688,
+      "grad_norm": 0.6472194848635398,
+      "learning_rate": 1.9692900210530482e-05,
+      "loss": 0.4702,
+      "step": 1623
+    },
+    {
+      "epoch": 0.32018927444794953,
+      "grad_norm": 0.5962256875505306,
+      "learning_rate": 1.9692518911045793e-05,
+      "loss": 0.4499,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3203864353312303,
+      "grad_norm": 0.6998497786192203,
+      "learning_rate": 1.969213737869053e-05,
+      "loss": 0.4395,
+      "step": 1625
+    },
+    {
+      "epoch": 0.32058359621451105,
+      "grad_norm": 0.6753566663287628,
+      "learning_rate": 1.969175561347386e-05,
+      "loss": 0.4741,
+      "step": 1626
+    },
+    {
+      "epoch": 0.3207807570977918,
+      "grad_norm": 0.6364586995100593,
+      "learning_rate": 1.969137361540496e-05,
+      "loss": 0.425,
+      "step": 1627
+    },
+    {
+      "epoch": 0.32097791798107256,
+      "grad_norm": 0.6316173890013236,
+      "learning_rate": 1.9690991384493002e-05,
+      "loss": 0.426,
+      "step": 1628
+    },
+    {
+      "epoch": 0.3211750788643533,
+      "grad_norm": 0.6592976432535022,
+      "learning_rate": 1.969060892074717e-05,
+      "loss": 0.4435,
+      "step": 1629
+    },
+    {
+      "epoch": 0.32137223974763407,
+      "grad_norm": 0.6057087728738385,
+      "learning_rate": 1.969022622417666e-05,
+      "loss": 0.4434,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3215694006309148,
+      "grad_norm": 0.6341926570691082,
+      "learning_rate": 1.968984329479066e-05,
+      "loss": 0.4479,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3217665615141956,
+      "grad_norm": 0.6729916605171744,
+      "learning_rate": 1.9689460132598372e-05,
+      "loss": 0.4832,
+      "step": 1632
+    },
+    {
+      "epoch": 0.32196372239747634,
+      "grad_norm": 0.8010790117508052,
+      "learning_rate": 1.9689076737608998e-05,
+      "loss": 0.4815,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3221608832807571,
+      "grad_norm": 0.6342908894048237,
+      "learning_rate": 1.9688693109831755e-05,
+      "loss": 0.4412,
+      "step": 1634
+    },
+    {
+      "epoch": 0.32235804416403785,
+      "grad_norm": 0.6574543494629974,
+      "learning_rate": 1.9688309249275857e-05,
+      "loss": 0.4337,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3225552050473186,
+      "grad_norm": 0.6700412322589491,
+      "learning_rate": 1.9687925155950526e-05,
+      "loss": 0.4612,
+      "step": 1636
+    },
+    {
+      "epoch": 0.32275236593059936,
+      "grad_norm": 0.6880798043825762,
+      "learning_rate": 1.9687540829864996e-05,
+      "loss": 0.4633,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3229495268138801,
+      "grad_norm": 0.5858751037584682,
+      "learning_rate": 1.9687156271028493e-05,
+      "loss": 0.4147,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3231466876971609,
+      "grad_norm": 1.1648261296219227,
+      "learning_rate": 1.968677147945026e-05,
+      "loss": 0.4663,
+      "step": 1639
+    },
+    {
+      "epoch": 0.32334384858044163,
+      "grad_norm": 0.6393378112398456,
+      "learning_rate": 1.9686386455139544e-05,
+      "loss": 0.4392,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3235410094637224,
+      "grad_norm": 0.6736562294968578,
+      "learning_rate": 1.9686001198105587e-05,
+      "loss": 0.459,
+      "step": 1641
+    },
+    {
+      "epoch": 0.32373817034700314,
+      "grad_norm": 0.6291265298426209,
+      "learning_rate": 1.9685615708357656e-05,
+      "loss": 0.455,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3239353312302839,
+      "grad_norm": 0.6956696368636405,
+      "learning_rate": 1.9685229985905007e-05,
+      "loss": 0.4735,
+      "step": 1643
+    },
+    {
+      "epoch": 0.32413249211356465,
+      "grad_norm": 0.7248600768104795,
+      "learning_rate": 1.9684844030756907e-05,
+      "loss": 0.4599,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3243296529968454,
+      "grad_norm": 0.8868152265197408,
+      "learning_rate": 1.9684457842922632e-05,
+      "loss": 0.4353,
+      "step": 1645
+    },
+    {
+      "epoch": 0.32452681388012616,
+      "grad_norm": 0.7520432959600722,
+      "learning_rate": 1.9684071422411456e-05,
+      "loss": 0.4873,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3247239747634069,
+      "grad_norm": 0.7377406807353308,
+      "learning_rate": 1.968368476923267e-05,
+      "loss": 0.461,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3249211356466877,
+      "grad_norm": 0.708141024519448,
+      "learning_rate": 1.968329788339555e-05,
+      "loss": 0.467,
+      "step": 1648
+    },
+    {
+      "epoch": 0.32511829652996843,
+      "grad_norm": 0.6848367537784837,
+      "learning_rate": 1.9682910764909405e-05,
+      "loss": 0.496,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3253154574132492,
+      "grad_norm": 0.6665467412273421,
+      "learning_rate": 1.9682523413783533e-05,
+      "loss": 0.4732,
+      "step": 1650
+    },
+    {
+      "epoch": 0.32551261829652994,
+      "grad_norm": 0.6953139108715455,
+      "learning_rate": 1.968213583002724e-05,
+      "loss": 0.4518,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3257097791798107,
+      "grad_norm": 0.6427842151833715,
+      "learning_rate": 1.9681748013649834e-05,
+      "loss": 0.4153,
+      "step": 1652
+    },
+    {
+      "epoch": 0.3259069400630915,
+      "grad_norm": 1.183472131884011,
+      "learning_rate": 1.968135996466064e-05,
+      "loss": 0.4764,
+      "step": 1653
+    },
+    {
+      "epoch": 0.32610410094637227,
+      "grad_norm": 0.617392531814384,
+      "learning_rate": 1.968097168306897e-05,
+      "loss": 0.4081,
+      "step": 1654
+    },
+    {
+      "epoch": 0.326301261829653,
+      "grad_norm": 0.6869167388782502,
+      "learning_rate": 1.9680583168884163e-05,
+      "loss": 0.4505,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3264984227129338,
+      "grad_norm": 0.8700641935343537,
+      "learning_rate": 1.9680194422115548e-05,
+      "loss": 0.5009,
+      "step": 1656
+    },
+    {
+      "epoch": 0.32669558359621453,
+      "grad_norm": 0.696926074465654,
+      "learning_rate": 1.9679805442772464e-05,
+      "loss": 0.4871,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3268927444794953,
+      "grad_norm": 0.7205797312738156,
+      "learning_rate": 1.9679416230864265e-05,
+      "loss": 0.487,
+      "step": 1658
+    },
+    {
+      "epoch": 0.32708990536277605,
+      "grad_norm": 0.7034418967452446,
+      "learning_rate": 1.967902678640029e-05,
+      "loss": 0.4442,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3272870662460568,
+      "grad_norm": 0.6205957767275062,
+      "learning_rate": 1.9678637109389903e-05,
+      "loss": 0.4606,
+      "step": 1660
+    },
+    {
+      "epoch": 0.32748422712933756,
+      "grad_norm": 0.6767913682774107,
+      "learning_rate": 1.967824719984247e-05,
+      "loss": 0.4449,
+      "step": 1661
+    },
+    {
+      "epoch": 0.3276813880126183,
+      "grad_norm": 0.7628685405323068,
+      "learning_rate": 1.967785705776735e-05,
+      "loss": 0.4912,
+      "step": 1662
+    },
+    {
+      "epoch": 0.32787854889589907,
+      "grad_norm": 0.6449573923061704,
+      "learning_rate": 1.9677466683173922e-05,
+      "loss": 0.4424,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3280757097791798,
+      "grad_norm": 0.6596295374884681,
+      "learning_rate": 1.9677076076071568e-05,
+      "loss": 0.4491,
+      "step": 1664
+    },
+    {
+      "epoch": 0.3282728706624606,
+      "grad_norm": 0.6027707481490366,
+      "learning_rate": 1.967668523646966e-05,
+      "loss": 0.4272,
+      "step": 1665
+    },
+    {
+      "epoch": 0.32847003154574134,
+      "grad_norm": 0.6919043992802008,
+      "learning_rate": 1.9676294164377603e-05,
+      "loss": 0.4597,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3286671924290221,
+      "grad_norm": 0.6975499115349002,
+      "learning_rate": 1.9675902859804786e-05,
+      "loss": 0.4978,
+      "step": 1667
+    },
+    {
+      "epoch": 0.32886435331230285,
+      "grad_norm": 0.6526383582936955,
+      "learning_rate": 1.967551132276061e-05,
+      "loss": 0.4506,
+      "step": 1668
+    },
+    {
+      "epoch": 0.3290615141955836,
+      "grad_norm": 0.6951437907766501,
+      "learning_rate": 1.9675119553254477e-05,
+      "loss": 0.4678,
+      "step": 1669
+    },
+    {
+      "epoch": 0.32925867507886436,
+      "grad_norm": 0.6405156580206511,
+      "learning_rate": 1.9674727551295812e-05,
+      "loss": 0.4657,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3294558359621451,
+      "grad_norm": 0.6686701711260229,
+      "learning_rate": 1.9674335316894024e-05,
+      "loss": 0.4424,
+      "step": 1671
+    },
+    {
+      "epoch": 0.32965299684542587,
+      "grad_norm": 0.6364715264193335,
+      "learning_rate": 1.9673942850058542e-05,
+      "loss": 0.4423,
+      "step": 1672
+    },
+    {
+      "epoch": 0.32985015772870663,
+      "grad_norm": 0.6132748220879023,
+      "learning_rate": 1.9673550150798787e-05,
+      "loss": 0.4247,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3300473186119874,
+      "grad_norm": 0.6645141377088348,
+      "learning_rate": 1.9673157219124207e-05,
+      "loss": 0.466,
+      "step": 1674
+    },
+    {
+      "epoch": 0.33024447949526814,
+      "grad_norm": 0.6213535676703986,
+      "learning_rate": 1.967276405504423e-05,
+      "loss": 0.4326,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3304416403785489,
+      "grad_norm": 1.0248391227718814,
+      "learning_rate": 1.9672370658568306e-05,
+      "loss": 0.5144,
+      "step": 1676
+    },
+    {
+      "epoch": 0.33063880126182965,
+      "grad_norm": 0.6705592825340083,
+      "learning_rate": 1.967197702970589e-05,
+      "loss": 0.4504,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3308359621451104,
+      "grad_norm": 0.6159547925544582,
+      "learning_rate": 1.967158316846644e-05,
+      "loss": 0.4262,
+      "step": 1678
+    },
+    {
+      "epoch": 0.33103312302839116,
+      "grad_norm": 0.6625745295364192,
+      "learning_rate": 1.9671189074859412e-05,
+      "loss": 0.463,
+      "step": 1679
+    },
+    {
+      "epoch": 0.3312302839116719,
+      "grad_norm": 0.6192642001745661,
+      "learning_rate": 1.967079474889428e-05,
+      "loss": 0.4283,
+      "step": 1680
+    },
+    {
+      "epoch": 0.3314274447949527,
+      "grad_norm": 0.659809255678557,
+      "learning_rate": 1.9670400190580516e-05,
+      "loss": 0.4443,
+      "step": 1681
+    },
+    {
+      "epoch": 0.33162460567823343,
+      "grad_norm": 0.60127325810791,
+      "learning_rate": 1.9670005399927602e-05,
+      "loss": 0.4579,
+      "step": 1682
+    },
+    {
+      "epoch": 0.3318217665615142,
+      "grad_norm": 0.7413899493606506,
+      "learning_rate": 1.9669610376945013e-05,
+      "loss": 0.487,
+      "step": 1683
+    },
+    {
+      "epoch": 0.33201892744479494,
+      "grad_norm": 0.6582626993097654,
+      "learning_rate": 1.9669215121642255e-05,
+      "loss": 0.4734,
+      "step": 1684
+    },
+    {
+      "epoch": 0.3322160883280757,
+      "grad_norm": 0.7151538898984898,
+      "learning_rate": 1.9668819634028816e-05,
+      "loss": 0.4446,
+      "step": 1685
+    },
+    {
+      "epoch": 0.33241324921135645,
+      "grad_norm": 0.6913048824848471,
+      "learning_rate": 1.96684239141142e-05,
+      "loss": 0.4253,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3326104100946372,
+      "grad_norm": 0.6414540359989238,
+      "learning_rate": 1.966802796190791e-05,
+      "loss": 0.4317,
+      "step": 1687
+    },
+    {
+      "epoch": 0.33280757097791797,
+      "grad_norm": 0.9207986171363428,
+      "learning_rate": 1.9667631777419466e-05,
+      "loss": 0.4627,
+      "step": 1688
+    },
+    {
+      "epoch": 0.3330047318611987,
+      "grad_norm": 0.616029959208574,
+      "learning_rate": 1.966723536065838e-05,
+      "loss": 0.4655,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3332018927444795,
+      "grad_norm": 0.6883300541317802,
+      "learning_rate": 1.9666838711634182e-05,
+      "loss": 0.4618,
+      "step": 1690
+    },
+    {
+      "epoch": 0.33339905362776023,
+      "grad_norm": 0.650342941791099,
+      "learning_rate": 1.9666441830356397e-05,
+      "loss": 0.444,
+      "step": 1691
+    },
+    {
+      "epoch": 0.333596214511041,
+      "grad_norm": 0.6491057879424494,
+      "learning_rate": 1.9666044716834566e-05,
+      "loss": 0.4624,
+      "step": 1692
+    },
+    {
+      "epoch": 0.33379337539432175,
+      "grad_norm": 0.6311071726696637,
+      "learning_rate": 1.9665647371078225e-05,
+      "loss": 0.4329,
+      "step": 1693
+    },
+    {
+      "epoch": 0.3339905362776025,
+      "grad_norm": 0.7246225471538845,
+      "learning_rate": 1.966524979309692e-05,
+      "loss": 0.5341,
+      "step": 1694
+    },
+    {
+      "epoch": 0.33418769716088326,
+      "grad_norm": 0.6987766778208062,
+      "learning_rate": 1.966485198290021e-05,
+      "loss": 0.48,
+      "step": 1695
+    },
+    {
+      "epoch": 0.334384858044164,
+      "grad_norm": 0.6138348348118343,
+      "learning_rate": 1.9664453940497642e-05,
+      "loss": 0.4447,
+      "step": 1696
+    },
+    {
+      "epoch": 0.33458201892744477,
+      "grad_norm": 0.6567138637243869,
+      "learning_rate": 1.966405566589879e-05,
+      "loss": 0.4991,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3347791798107255,
+      "grad_norm": 0.6352662846456245,
+      "learning_rate": 1.9663657159113217e-05,
+      "loss": 0.4652,
+      "step": 1698
+    },
+    {
+      "epoch": 0.33497634069400634,
+      "grad_norm": 0.8883699236623674,
+      "learning_rate": 1.96632584201505e-05,
+      "loss": 0.4797,
+      "step": 1699
+    },
+    {
+      "epoch": 0.3351735015772871,
+      "grad_norm": 0.6256688722873787,
+      "learning_rate": 1.9662859449020214e-05,
+      "loss": 0.4771,
+      "step": 1700
+    },
+    {
+      "epoch": 0.33537066246056785,
+      "grad_norm": 0.6792106596589087,
+      "learning_rate": 1.966246024573195e-05,
+      "loss": 0.4561,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3355678233438486,
+      "grad_norm": 0.5975526642006157,
+      "learning_rate": 1.96620608102953e-05,
+      "loss": 0.4372,
+      "step": 1702
+    },
+    {
+      "epoch": 0.33576498422712936,
+      "grad_norm": 0.6481968588062909,
+      "learning_rate": 1.9661661142719856e-05,
+      "loss": 0.4679,
+      "step": 1703
+    },
+    {
+      "epoch": 0.3359621451104101,
+      "grad_norm": 0.625117310912806,
+      "learning_rate": 1.966126124301522e-05,
+      "loss": 0.4722,
+      "step": 1704
+    },
+    {
+      "epoch": 0.33615930599369087,
+      "grad_norm": 0.6779932286491652,
+      "learning_rate": 1.9660861111191004e-05,
+      "loss": 0.4548,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3363564668769716,
+      "grad_norm": 0.5897827532389534,
+      "learning_rate": 1.9660460747256823e-05,
+      "loss": 0.4221,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3365536277602524,
+      "grad_norm": 0.600745494776992,
+      "learning_rate": 1.9660060151222292e-05,
+      "loss": 0.4359,
+      "step": 1707
+    },
+    {
+      "epoch": 0.33675078864353314,
+      "grad_norm": 0.6263245381562469,
+      "learning_rate": 1.9659659323097037e-05,
+      "loss": 0.4537,
+      "step": 1708
+    },
+    {
+      "epoch": 0.3369479495268139,
+      "grad_norm": 0.6059447411600158,
+      "learning_rate": 1.9659258262890683e-05,
+      "loss": 0.4557,
+      "step": 1709
+    },
+    {
+      "epoch": 0.33714511041009465,
+      "grad_norm": 0.6276333332545005,
+      "learning_rate": 1.9658856970612878e-05,
+      "loss": 0.4539,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3373422712933754,
+      "grad_norm": 0.61826217784891,
+      "learning_rate": 1.965845544627325e-05,
+      "loss": 0.441,
+      "step": 1711
+    },
+    {
+      "epoch": 0.33753943217665616,
+      "grad_norm": 0.6053223097223316,
+      "learning_rate": 1.9658053689881453e-05,
+      "loss": 0.4813,
+      "step": 1712
+    },
+    {
+      "epoch": 0.3377365930599369,
+      "grad_norm": 0.5730530402368444,
+      "learning_rate": 1.965765170144714e-05,
+      "loss": 0.4452,
+      "step": 1713
+    },
+    {
+      "epoch": 0.3379337539432177,
+      "grad_norm": 4.5489083531739265,
+      "learning_rate": 1.9657249480979968e-05,
+      "loss": 0.4672,
+      "step": 1714
+    },
+    {
+      "epoch": 0.33813091482649843,
+      "grad_norm": 0.6468829198177741,
+      "learning_rate": 1.9656847028489597e-05,
+      "loss": 0.447,
+      "step": 1715
+    },
+    {
+      "epoch": 0.3383280757097792,
+      "grad_norm": 0.6197378264065176,
+      "learning_rate": 1.9656444343985705e-05,
+      "loss": 0.4681,
+      "step": 1716
+    },
+    {
+      "epoch": 0.33852523659305994,
+      "grad_norm": 0.7104123212898853,
+      "learning_rate": 1.9656041427477957e-05,
+      "loss": 0.4457,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3387223974763407,
+      "grad_norm": 0.7070319729266388,
+      "learning_rate": 1.965563827897604e-05,
+      "loss": 0.4517,
+      "step": 1718
+    },
+    {
+      "epoch": 0.33891955835962145,
+      "grad_norm": 0.6119590288529166,
+      "learning_rate": 1.9655234898489634e-05,
+      "loss": 0.4346,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3391167192429022,
+      "grad_norm": 0.6245842477692922,
+      "learning_rate": 1.965483128602844e-05,
+      "loss": 0.4604,
+      "step": 1720
+    },
+    {
+      "epoch": 0.33931388012618297,
+      "grad_norm": 1.2443841217690506,
+      "learning_rate": 1.9654427441602145e-05,
+      "loss": 0.479,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3395110410094637,
+      "grad_norm": 0.6732229801004805,
+      "learning_rate": 1.9654023365220456e-05,
+      "loss": 0.4513,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3397082018927445,
+      "grad_norm": 0.6568708942903064,
+      "learning_rate": 1.9653619056893082e-05,
+      "loss": 0.4406,
+      "step": 1723
+    },
+    {
+      "epoch": 0.33990536277602523,
+      "grad_norm": 0.663524577710516,
+      "learning_rate": 1.9653214516629737e-05,
+      "loss": 0.4763,
+      "step": 1724
+    },
+    {
+      "epoch": 0.340102523659306,
+      "grad_norm": 0.6675810755775939,
+      "learning_rate": 1.965280974444014e-05,
+      "loss": 0.4489,
+      "step": 1725
+    },
+    {
+      "epoch": 0.34029968454258674,
+      "grad_norm": 0.6563303589975222,
+      "learning_rate": 1.9652404740334015e-05,
+      "loss": 0.4309,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3404968454258675,
+      "grad_norm": 0.7598373649184584,
+      "learning_rate": 1.9651999504321094e-05,
+      "loss": 0.4272,
+      "step": 1727
+    },
+    {
+      "epoch": 0.34069400630914826,
+      "grad_norm": 0.651084326470726,
+      "learning_rate": 1.9651594036411107e-05,
+      "loss": 0.4475,
+      "step": 1728
+    },
+    {
+      "epoch": 0.340891167192429,
+      "grad_norm": 0.6799855930682017,
+      "learning_rate": 1.9651188336613807e-05,
+      "loss": 0.4524,
+      "step": 1729
+    },
+    {
+      "epoch": 0.34108832807570977,
+      "grad_norm": 0.8147708964689288,
+      "learning_rate": 1.9650782404938933e-05,
+      "loss": 0.4748,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3412854889589905,
+      "grad_norm": 1.9599798652041918,
+      "learning_rate": 1.965037624139624e-05,
+      "loss": 0.4634,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3414826498422713,
+      "grad_norm": 0.9750946996661596,
+      "learning_rate": 1.9649969845995486e-05,
+      "loss": 0.4522,
+      "step": 1732
+    },
+    {
+      "epoch": 0.34167981072555204,
+      "grad_norm": 0.7830583556096765,
+      "learning_rate": 1.9649563218746436e-05,
+      "loss": 0.4837,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3418769716088328,
+      "grad_norm": 1.8522800709343359,
+      "learning_rate": 1.9649156359658857e-05,
+      "loss": 0.4907,
+      "step": 1734
+    },
+    {
+      "epoch": 0.34207413249211355,
+      "grad_norm": 0.7087784601927121,
+      "learning_rate": 1.964874926874253e-05,
+      "loss": 0.4537,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3422712933753943,
+      "grad_norm": 0.9263867760746262,
+      "learning_rate": 1.9648341946007228e-05,
+      "loss": 0.4481,
+      "step": 1736
+    },
+    {
+      "epoch": 0.34246845425867506,
+      "grad_norm": 0.7131592632953431,
+      "learning_rate": 1.9647934391462743e-05,
+      "loss": 0.4859,
+      "step": 1737
+    },
+    {
+      "epoch": 0.3426656151419558,
+      "grad_norm": 0.6812545398255945,
+      "learning_rate": 1.9647526605118863e-05,
+      "loss": 0.4419,
+      "step": 1738
+    },
+    {
+      "epoch": 0.34286277602523657,
+      "grad_norm": 0.8324328718435976,
+      "learning_rate": 1.964711858698539e-05,
+      "loss": 0.4495,
+      "step": 1739
+    },
+    {
+      "epoch": 0.3430599369085173,
+      "grad_norm": 0.7569149729504874,
+      "learning_rate": 1.964671033707212e-05,
+      "loss": 0.4779,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3432570977917981,
+      "grad_norm": 0.7011727969185421,
+      "learning_rate": 1.9646301855388868e-05,
+      "loss": 0.486,
+      "step": 1741
+    },
+    {
+      "epoch": 0.34345425867507884,
+      "grad_norm": 0.6487547062067612,
+      "learning_rate": 1.9645893141945444e-05,
+      "loss": 0.4265,
+      "step": 1742
+    },
+    {
+      "epoch": 0.3436514195583596,
+      "grad_norm": 0.7657920846629185,
+      "learning_rate": 1.9645484196751676e-05,
+      "loss": 0.455,
+      "step": 1743
+    },
+    {
+      "epoch": 0.3438485804416404,
+      "grad_norm": 0.696631675500191,
+      "learning_rate": 1.9645075019817374e-05,
+      "loss": 0.4743,
+      "step": 1744
+    },
+    {
+      "epoch": 0.34404574132492116,
+      "grad_norm": 0.6735160176131795,
+      "learning_rate": 1.9644665611152384e-05,
+      "loss": 0.458,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3442429022082019,
+      "grad_norm": 0.7721263917052835,
+      "learning_rate": 1.964425597076653e-05,
+      "loss": 0.4765,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3444400630914827,
+      "grad_norm": 0.7497548232948051,
+      "learning_rate": 1.9643846098669664e-05,
+      "loss": 0.4772,
+      "step": 1747
+    },
+    {
+      "epoch": 0.34463722397476343,
+      "grad_norm": 0.6847650811355418,
+      "learning_rate": 1.9643435994871626e-05,
+      "loss": 0.4417,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3448343848580442,
+      "grad_norm": 0.6433657002066264,
+      "learning_rate": 1.9643025659382274e-05,
+      "loss": 0.415,
+      "step": 1749
+    },
+    {
+      "epoch": 0.34503154574132494,
+      "grad_norm": 0.6855538495003168,
+      "learning_rate": 1.9642615092211468e-05,
+      "loss": 0.4665,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3452287066246057,
+      "grad_norm": 0.6890543003455781,
+      "learning_rate": 1.9642204293369066e-05,
+      "loss": 0.4571,
+      "step": 1751
+    },
+    {
+      "epoch": 0.34542586750788645,
+      "grad_norm": 0.8006291058144541,
+      "learning_rate": 1.9641793262864942e-05,
+      "loss": 0.4617,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3456230283911672,
+      "grad_norm": 0.7454470217038145,
+      "learning_rate": 1.9641382000708972e-05,
+      "loss": 0.4697,
+      "step": 1753
+    },
+    {
+      "epoch": 0.34582018927444796,
+      "grad_norm": 0.6090095712411363,
+      "learning_rate": 1.9640970506911033e-05,
+      "loss": 0.4375,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3460173501577287,
+      "grad_norm": 0.7015624149655699,
+      "learning_rate": 1.9640558781481015e-05,
+      "loss": 0.4219,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3462145110410095,
+      "grad_norm": 0.7670637934340552,
+      "learning_rate": 1.9640146824428807e-05,
+      "loss": 0.4899,
+      "step": 1756
+    },
+    {
+      "epoch": 0.34641167192429023,
+      "grad_norm": 0.7363863325209535,
+      "learning_rate": 1.963973463576431e-05,
+      "loss": 0.4664,
+      "step": 1757
+    },
+    {
+      "epoch": 0.346608832807571,
+      "grad_norm": 0.6131339077035906,
+      "learning_rate": 1.9639322215497423e-05,
+      "loss": 0.4072,
+      "step": 1758
+    },
+    {
+      "epoch": 0.34680599369085174,
+      "grad_norm": 0.6707877768295646,
+      "learning_rate": 1.963890956363806e-05,
+      "loss": 0.4662,
+      "step": 1759
+    },
+    {
+      "epoch": 0.3470031545741325,
+      "grad_norm": 0.677317205988406,
+      "learning_rate": 1.9638496680196135e-05,
+      "loss": 0.4106,
+      "step": 1760
+    },
+    {
+      "epoch": 0.34720031545741326,
+      "grad_norm": 0.6492787706072033,
+      "learning_rate": 1.963808356518156e-05,
+      "loss": 0.4077,
+      "step": 1761
+    },
+    {
+      "epoch": 0.347397476340694,
+      "grad_norm": 0.7823981215932518,
+      "learning_rate": 1.9637670218604267e-05,
+      "loss": 0.4573,
+      "step": 1762
+    },
+    {
+      "epoch": 0.34759463722397477,
+      "grad_norm": 0.5953310114048408,
+      "learning_rate": 1.9637256640474187e-05,
+      "loss": 0.4539,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3477917981072555,
+      "grad_norm": 0.759527295176903,
+      "learning_rate": 1.9636842830801255e-05,
+      "loss": 0.4354,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3479889589905363,
+      "grad_norm": 0.8180906973487583,
+      "learning_rate": 1.9636428789595413e-05,
+      "loss": 0.4312,
+      "step": 1765
+    },
+    {
+      "epoch": 0.34818611987381703,
+      "grad_norm": 4.382079962402747,
+      "learning_rate": 1.963601451686661e-05,
+      "loss": 0.4406,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3483832807570978,
+      "grad_norm": 0.7560610305482834,
+      "learning_rate": 1.9635600012624798e-05,
+      "loss": 0.4313,
+      "step": 1767
+    },
+    {
+      "epoch": 0.34858044164037855,
+      "grad_norm": 0.7418288958717549,
+      "learning_rate": 1.9635185276879936e-05,
+      "loss": 0.4793,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3487776025236593,
+      "grad_norm": 1.072851573671909,
+      "learning_rate": 1.963477030964199e-05,
+      "loss": 0.485,
+      "step": 1769
+    },
+    {
+      "epoch": 0.34897476340694006,
+      "grad_norm": 0.7920038346678471,
+      "learning_rate": 1.963435511092093e-05,
+      "loss": 0.4696,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3491719242902208,
+      "grad_norm": 0.7093452794943754,
+      "learning_rate": 1.9633939680726724e-05,
+      "loss": 0.4727,
+      "step": 1771
+    },
+    {
+      "epoch": 0.34936908517350157,
+      "grad_norm": 0.8112163451470932,
+      "learning_rate": 1.9633524019069365e-05,
+      "loss": 0.5013,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3495662460567823,
+      "grad_norm": 3.917584519790698,
+      "learning_rate": 1.963310812595883e-05,
+      "loss": 0.5098,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3497634069400631,
+      "grad_norm": 1.000199323692884,
+      "learning_rate": 1.9632692001405113e-05,
+      "loss": 0.4594,
+      "step": 1774
+    },
+    {
+      "epoch": 0.34996056782334384,
+      "grad_norm": 0.8133085348010322,
+      "learning_rate": 1.9632275645418218e-05,
+      "loss": 0.4881,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3501577287066246,
+      "grad_norm": 0.9472822270595948,
+      "learning_rate": 1.963185905800814e-05,
+      "loss": 0.4973,
+      "step": 1776
+    },
+    {
+      "epoch": 0.35035488958990535,
+      "grad_norm": 0.9453081319239314,
+      "learning_rate": 1.9631442239184894e-05,
+      "loss": 0.4266,
+      "step": 1777
+    },
+    {
+      "epoch": 0.3505520504731861,
+      "grad_norm": 0.7602206983614878,
+      "learning_rate": 1.9631025188958492e-05,
+      "loss": 0.4783,
+      "step": 1778
+    },
+    {
+      "epoch": 0.35074921135646686,
+      "grad_norm": 0.8604246184240586,
+      "learning_rate": 1.9630607907338953e-05,
+      "loss": 0.4293,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3509463722397476,
+      "grad_norm": 0.7616354505596147,
+      "learning_rate": 1.9630190394336304e-05,
+      "loss": 0.5098,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3511435331230284,
+      "grad_norm": 0.819713708870105,
+      "learning_rate": 1.9629772649960574e-05,
+      "loss": 0.4656,
+      "step": 1781
+    },
+    {
+      "epoch": 0.35134069400630913,
+      "grad_norm": 0.6981497101426539,
+      "learning_rate": 1.9629354674221803e-05,
+      "loss": 0.4555,
+      "step": 1782
+    },
+    {
+      "epoch": 0.3515378548895899,
+      "grad_norm": 0.8239076401594878,
+      "learning_rate": 1.962893646713003e-05,
+      "loss": 0.4619,
+      "step": 1783
+    },
+    {
+      "epoch": 0.35173501577287064,
+      "grad_norm": 0.6349622015636447,
+      "learning_rate": 1.9628518028695307e-05,
+      "loss": 0.4508,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3519321766561514,
+      "grad_norm": 0.7787614675612314,
+      "learning_rate": 1.9628099358927684e-05,
+      "loss": 0.4399,
+      "step": 1785
+    },
+    {
+      "epoch": 0.35212933753943215,
+      "grad_norm": 0.6213692400849354,
+      "learning_rate": 1.962768045783722e-05,
+      "loss": 0.4302,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3523264984227129,
+      "grad_norm": 0.8089918611577429,
+      "learning_rate": 1.9627261325433976e-05,
+      "loss": 0.4447,
+      "step": 1787
+    },
+    {
+      "epoch": 0.35252365930599366,
+      "grad_norm": 0.8637850568343813,
+      "learning_rate": 1.962684196172803e-05,
+      "loss": 0.4683,
+      "step": 1788
+    },
+    {
+      "epoch": 0.3527208201892745,
+      "grad_norm": 0.7190171498413642,
+      "learning_rate": 1.9626422366729453e-05,
+      "loss": 0.4591,
+      "step": 1789
+    },
+    {
+      "epoch": 0.35291798107255523,
+      "grad_norm": 0.7746352347399187,
+      "learning_rate": 1.9626002540448325e-05,
+      "loss": 0.4545,
+      "step": 1790
+    },
+    {
+      "epoch": 0.353115141955836,
+      "grad_norm": 0.7705463288801266,
+      "learning_rate": 1.9625582482894735e-05,
+      "loss": 0.4505,
+      "step": 1791
+    },
+    {
+      "epoch": 0.35331230283911674,
+      "grad_norm": 0.6701387506672304,
+      "learning_rate": 1.9625162194078775e-05,
+      "loss": 0.4823,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3535094637223975,
+      "grad_norm": 0.7068914753075993,
+      "learning_rate": 1.9624741674010544e-05,
+      "loss": 0.4654,
+      "step": 1793
+    },
+    {
+      "epoch": 0.35370662460567825,
+      "grad_norm": 0.6064047478157758,
+      "learning_rate": 1.9624320922700138e-05,
+      "loss": 0.4057,
+      "step": 1794
+    },
+    {
+      "epoch": 0.353903785488959,
+      "grad_norm": 0.6898360183497596,
+      "learning_rate": 1.9623899940157675e-05,
+      "loss": 0.4221,
+      "step": 1795
+    },
+    {
+      "epoch": 0.35410094637223977,
+      "grad_norm": 0.6629057087304656,
+      "learning_rate": 1.9623478726393266e-05,
+      "loss": 0.5155,
+      "step": 1796
+    },
+    {
+      "epoch": 0.3542981072555205,
+      "grad_norm": 0.6794292374990497,
+      "learning_rate": 1.9623057281417028e-05,
+      "loss": 0.4402,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3544952681388013,
+      "grad_norm": 0.6122789190306797,
+      "learning_rate": 1.9622635605239095e-05,
+      "loss": 0.4431,
+      "step": 1798
+    },
+    {
+      "epoch": 0.35469242902208203,
+      "grad_norm": 0.6459796432517372,
+      "learning_rate": 1.9622213697869587e-05,
+      "loss": 0.4388,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3548895899053628,
+      "grad_norm": 0.6623695426508792,
+      "learning_rate": 1.9621791559318648e-05,
+      "loss": 0.4812,
+      "step": 1800
+    },
+    {
+      "epoch": 0.35508675078864355,
+      "grad_norm": 0.6324064996658313,
+      "learning_rate": 1.962136918959642e-05,
+      "loss": 0.4675,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3552839116719243,
+      "grad_norm": 0.5963028662353506,
+      "learning_rate": 1.9620946588713048e-05,
+      "loss": 0.4412,
+      "step": 1802
+    },
+    {
+      "epoch": 0.35548107255520506,
+      "grad_norm": 0.6744345317892008,
+      "learning_rate": 1.9620523756678685e-05,
+      "loss": 0.4177,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3556782334384858,
+      "grad_norm": 0.645630963745389,
+      "learning_rate": 1.9620100693503494e-05,
+      "loss": 0.4511,
+      "step": 1804
+    },
+    {
+      "epoch": 0.35587539432176657,
+      "grad_norm": 0.6672682741456414,
+      "learning_rate": 1.9619677399197634e-05,
+      "loss": 0.4382,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3560725552050473,
+      "grad_norm": 0.7740606862475723,
+      "learning_rate": 1.961925387377128e-05,
+      "loss": 0.4462,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3562697160883281,
+      "grad_norm": 0.6497305621170518,
+      "learning_rate": 1.9618830117234603e-05,
+      "loss": 0.4725,
+      "step": 1807
+    },
+    {
+      "epoch": 0.35646687697160884,
+      "grad_norm": 0.640815305534194,
+      "learning_rate": 1.9618406129597787e-05,
+      "loss": 0.4651,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3566640378548896,
+      "grad_norm": 0.6155816175110397,
+      "learning_rate": 1.961798191087102e-05,
+      "loss": 0.4281,
+      "step": 1809
+    },
+    {
+      "epoch": 0.35686119873817035,
+      "grad_norm": 0.633839320877963,
+      "learning_rate": 1.9617557461064495e-05,
+      "loss": 0.4938,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3570583596214511,
+      "grad_norm": 0.635922394462134,
+      "learning_rate": 1.96171327801884e-05,
+      "loss": 0.4393,
+      "step": 1811
+    },
+    {
+      "epoch": 0.35725552050473186,
+      "grad_norm": 0.6410247655705709,
+      "learning_rate": 1.961670786825295e-05,
+      "loss": 0.4629,
+      "step": 1812
+    },
+    {
+      "epoch": 0.3574526813880126,
+      "grad_norm": 1.9279852657649488,
+      "learning_rate": 1.9616282725268347e-05,
+      "loss": 0.4694,
+      "step": 1813
+    },
+    {
+      "epoch": 0.35764984227129337,
+      "grad_norm": 0.6710119594137577,
+      "learning_rate": 1.961585735124481e-05,
+      "loss": 0.4594,
+      "step": 1814
+    },
+    {
+      "epoch": 0.35784700315457413,
+      "grad_norm": 0.7397595757858347,
+      "learning_rate": 1.9615431746192553e-05,
+      "loss": 0.4534,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3580441640378549,
+      "grad_norm": 0.8225787141793098,
+      "learning_rate": 1.9615005910121806e-05,
+      "loss": 0.4832,
+      "step": 1816
+    },
+    {
+      "epoch": 0.35824132492113564,
+      "grad_norm": 0.6983577011027828,
+      "learning_rate": 1.96145798430428e-05,
+      "loss": 0.4377,
+      "step": 1817
+    },
+    {
+      "epoch": 0.3584384858044164,
+      "grad_norm": 0.7361587716765654,
+      "learning_rate": 1.9614153544965773e-05,
+      "loss": 0.4893,
+      "step": 1818
+    },
+    {
+      "epoch": 0.35863564668769715,
+      "grad_norm": 0.6203265764684508,
+      "learning_rate": 1.9613727015900962e-05,
+      "loss": 0.3938,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3588328075709779,
+      "grad_norm": 0.6980283936814929,
+      "learning_rate": 1.9613300255858615e-05,
+      "loss": 0.4953,
+      "step": 1820
+    },
+    {
+      "epoch": 0.35902996845425866,
+      "grad_norm": 0.7026401510766678,
+      "learning_rate": 1.9612873264848994e-05,
+      "loss": 0.4594,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3592271293375394,
+      "grad_norm": 0.6405607709544402,
+      "learning_rate": 1.9612446042882345e-05,
+      "loss": 0.4146,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3594242902208202,
+      "grad_norm": 0.8137624725348499,
+      "learning_rate": 1.961201858996894e-05,
+      "loss": 0.4603,
+      "step": 1823
+    },
+    {
+      "epoch": 0.35962145110410093,
+      "grad_norm": 0.6238197035171101,
+      "learning_rate": 1.9611590906119055e-05,
+      "loss": 0.4301,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3598186119873817,
+      "grad_norm": 0.7765705822707861,
+      "learning_rate": 1.9611162991342952e-05,
+      "loss": 0.4622,
+      "step": 1825
+    },
+    {
+      "epoch": 0.36001577287066244,
+      "grad_norm": 0.8129569752866966,
+      "learning_rate": 1.961073484565092e-05,
+      "loss": 0.4634,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3602129337539432,
+      "grad_norm": 0.7674729263580945,
+      "learning_rate": 1.9610306469053243e-05,
+      "loss": 0.4797,
+      "step": 1827
+    },
+    {
+      "epoch": 0.36041009463722395,
+      "grad_norm": 0.6433782160365153,
+      "learning_rate": 1.9609877861560213e-05,
+      "loss": 0.4427,
+      "step": 1828
+    },
+    {
+      "epoch": 0.3606072555205047,
+      "grad_norm": 0.8189751748622848,
+      "learning_rate": 1.9609449023182133e-05,
+      "loss": 0.4925,
+      "step": 1829
+    },
+    {
+      "epoch": 0.36080441640378547,
+      "grad_norm": 0.6603921192162837,
+      "learning_rate": 1.9609019953929298e-05,
+      "loss": 0.4328,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3610015772870662,
+      "grad_norm": 0.7020104873363077,
+      "learning_rate": 1.960859065381202e-05,
+      "loss": 0.4532,
+      "step": 1831
+    },
+    {
+      "epoch": 0.361198738170347,
+      "grad_norm": 0.6514968871084654,
+      "learning_rate": 1.9608161122840614e-05,
+      "loss": 0.4262,
+      "step": 1832
+    },
+    {
+      "epoch": 0.36139589905362773,
+      "grad_norm": 1.3363500449543768,
+      "learning_rate": 1.9607731361025402e-05,
+      "loss": 0.4738,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3615930599369085,
+      "grad_norm": 0.6353691592363815,
+      "learning_rate": 1.9607301368376706e-05,
+      "loss": 0.4918,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3617902208201893,
+      "grad_norm": 0.6784040176469287,
+      "learning_rate": 1.9606871144904855e-05,
+      "loss": 0.4827,
+      "step": 1835
+    },
+    {
+      "epoch": 0.36198738170347006,
+      "grad_norm": 0.6264342538936625,
+      "learning_rate": 1.960644069062019e-05,
+      "loss": 0.4172,
+      "step": 1836
+    },
+    {
+      "epoch": 0.3621845425867508,
+      "grad_norm": 3.5120658505114277,
+      "learning_rate": 1.9606010005533055e-05,
+      "loss": 0.4948,
+      "step": 1837
+    },
+    {
+      "epoch": 0.36238170347003157,
+      "grad_norm": 0.9363587178765608,
+      "learning_rate": 1.960557908965379e-05,
+      "loss": 0.468,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3625788643533123,
+      "grad_norm": 0.6948665441777858,
+      "learning_rate": 1.9605147942992752e-05,
+      "loss": 0.4493,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3627760252365931,
+      "grad_norm": 0.7506323131912949,
+      "learning_rate": 1.9604716565560303e-05,
+      "loss": 0.4361,
+      "step": 1840
+    },
+    {
+      "epoch": 0.36297318611987384,
+      "grad_norm": 0.6646663396446123,
+      "learning_rate": 1.96042849573668e-05,
+      "loss": 0.4565,
+      "step": 1841
+    },
+    {
+      "epoch": 0.3631703470031546,
+      "grad_norm": 0.7021081874099678,
+      "learning_rate": 1.9603853118422618e-05,
+      "loss": 0.4791,
+      "step": 1842
+    },
+    {
+      "epoch": 0.36336750788643535,
+      "grad_norm": 0.7405667369095288,
+      "learning_rate": 1.960342104873813e-05,
+      "loss": 0.4507,
+      "step": 1843
+    },
+    {
+      "epoch": 0.3635646687697161,
+      "grad_norm": 0.8937542644763167,
+      "learning_rate": 1.9602988748323718e-05,
+      "loss": 0.4739,
+      "step": 1844
+    },
+    {
+      "epoch": 0.36376182965299686,
+      "grad_norm": 0.647962965489675,
+      "learning_rate": 1.960255621718977e-05,
+      "loss": 0.4505,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3639589905362776,
+      "grad_norm": 0.6679476483113629,
+      "learning_rate": 1.9602123455346677e-05,
+      "loss": 0.4594,
+      "step": 1846
+    },
+    {
+      "epoch": 0.36415615141955837,
+      "grad_norm": 0.6248714424291023,
+      "learning_rate": 1.960169046280483e-05,
+      "loss": 0.4452,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3643533123028391,
+      "grad_norm": 0.5629821941055176,
+      "learning_rate": 1.960125723957464e-05,
+      "loss": 0.4021,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3645504731861199,
+      "grad_norm": 0.6196696074939825,
+      "learning_rate": 1.9600823785666515e-05,
+      "loss": 0.4365,
+      "step": 1849
+    },
+    {
+      "epoch": 0.36474763406940064,
+      "grad_norm": 0.6115029014398423,
+      "learning_rate": 1.9600390101090867e-05,
+      "loss": 0.4279,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3649447949526814,
+      "grad_norm": 0.6120747583019098,
+      "learning_rate": 1.9599956185858112e-05,
+      "loss": 0.4548,
+      "step": 1851
+    },
+    {
+      "epoch": 0.36514195583596215,
+      "grad_norm": 0.7496535911475231,
+      "learning_rate": 1.959952203997868e-05,
+      "loss": 0.4646,
+      "step": 1852
+    },
+    {
+      "epoch": 0.3653391167192429,
+      "grad_norm": 0.6382300755564924,
+      "learning_rate": 1.9599087663463003e-05,
+      "loss": 0.4355,
+      "step": 1853
+    },
+    {
+      "epoch": 0.36553627760252366,
+      "grad_norm": 0.6602319601851913,
+      "learning_rate": 1.9598653056321512e-05,
+      "loss": 0.4632,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3657334384858044,
+      "grad_norm": 0.6191066463069663,
+      "learning_rate": 1.9598218218564656e-05,
+      "loss": 0.4238,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3659305993690852,
+      "grad_norm": 0.6072202775912791,
+      "learning_rate": 1.9597783150202873e-05,
+      "loss": 0.4683,
+      "step": 1856
+    },
+    {
+      "epoch": 0.36612776025236593,
+      "grad_norm": 0.5689905034550738,
+      "learning_rate": 1.9597347851246623e-05,
+      "loss": 0.4102,
+      "step": 1857
+    },
+    {
+      "epoch": 0.3663249211356467,
+      "grad_norm": 0.6023153592091132,
+      "learning_rate": 1.959691232170636e-05,
+      "loss": 0.4748,
+      "step": 1858
+    },
+    {
+      "epoch": 0.36652208201892744,
+      "grad_norm": 0.6035103505300697,
+      "learning_rate": 1.9596476561592553e-05,
+      "loss": 0.429,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3667192429022082,
+      "grad_norm": 0.6246219996358824,
+      "learning_rate": 1.9596040570915666e-05,
+      "loss": 0.4179,
+      "step": 1860
+    },
+    {
+      "epoch": 0.36691640378548895,
+      "grad_norm": 0.6764471802280777,
+      "learning_rate": 1.959560434968618e-05,
+      "loss": 0.5002,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3671135646687697,
+      "grad_norm": 0.6218434897962343,
+      "learning_rate": 1.959516789791457e-05,
+      "loss": 0.4436,
+      "step": 1862
+    },
+    {
+      "epoch": 0.36731072555205047,
+      "grad_norm": 0.6206678178233864,
+      "learning_rate": 1.959473121561132e-05,
+      "loss": 0.4553,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3675078864353312,
+      "grad_norm": 0.5901550517207009,
+      "learning_rate": 1.9594294302786933e-05,
+      "loss": 0.4377,
+      "step": 1864
+    },
+    {
+      "epoch": 0.367705047318612,
+      "grad_norm": 0.6112016781481284,
+      "learning_rate": 1.9593857159451897e-05,
+      "loss": 0.4622,
+      "step": 1865
+    },
+    {
+      "epoch": 0.36790220820189273,
+      "grad_norm": 0.6154064090788753,
+      "learning_rate": 1.9593419785616716e-05,
+      "loss": 0.4587,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3680993690851735,
+      "grad_norm": 0.6350370151618049,
+      "learning_rate": 1.95929821812919e-05,
+      "loss": 0.4683,
+      "step": 1867
+    },
+    {
+      "epoch": 0.36829652996845424,
+      "grad_norm": 1.6840479608127483,
+      "learning_rate": 1.9592544346487958e-05,
+      "loss": 0.4761,
+      "step": 1868
+    },
+    {
+      "epoch": 0.368493690851735,
+      "grad_norm": 0.586315544529278,
+      "learning_rate": 1.9592106281215418e-05,
+      "loss": 0.4171,
+      "step": 1869
+    },
+    {
+      "epoch": 0.36869085173501576,
+      "grad_norm": 0.6098943093456206,
+      "learning_rate": 1.95916679854848e-05,
+      "loss": 0.4483,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3688880126182965,
+      "grad_norm": 0.6030006559111836,
+      "learning_rate": 1.959122945930663e-05,
+      "loss": 0.4379,
+      "step": 1871
+    },
+    {
+      "epoch": 0.36908517350157727,
+      "grad_norm": 0.6788148490316216,
+      "learning_rate": 1.9590790702691453e-05,
+      "loss": 0.4691,
+      "step": 1872
+    },
+    {
+      "epoch": 0.369282334384858,
+      "grad_norm": 0.6629668903573882,
+      "learning_rate": 1.9590351715649803e-05,
+      "loss": 0.4551,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3694794952681388,
+      "grad_norm": 0.7978043884977309,
+      "learning_rate": 1.9589912498192233e-05,
+      "loss": 0.4405,
+      "step": 1874
+    },
+    {
+      "epoch": 0.36967665615141954,
+      "grad_norm": 0.7049104610666904,
+      "learning_rate": 1.958947305032929e-05,
+      "loss": 0.4673,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3698738170347003,
+      "grad_norm": 0.6949512497820435,
+      "learning_rate": 1.9589033372071537e-05,
+      "loss": 0.4147,
+      "step": 1876
+    },
+    {
+      "epoch": 0.37007097791798105,
+      "grad_norm": 0.7972703080682879,
+      "learning_rate": 1.9588593463429532e-05,
+      "loss": 0.4636,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3702681388012618,
+      "grad_norm": 0.6733154115167292,
+      "learning_rate": 1.958815332441385e-05,
+      "loss": 0.4556,
+      "step": 1878
+    },
+    {
+      "epoch": 0.37046529968454256,
+      "grad_norm": 0.633932691284626,
+      "learning_rate": 1.9587712955035064e-05,
+      "loss": 0.457,
+      "step": 1879
+    },
+    {
+      "epoch": 0.37066246056782337,
+      "grad_norm": 0.6626177934715735,
+      "learning_rate": 1.958727235530375e-05,
+      "loss": 0.4682,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3708596214511041,
+      "grad_norm": 0.6359733697611749,
+      "learning_rate": 1.9586831525230496e-05,
+      "loss": 0.4482,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3710567823343849,
+      "grad_norm": 0.8674782473950938,
+      "learning_rate": 1.9586390464825896e-05,
+      "loss": 0.447,
+      "step": 1882
+    },
+    {
+      "epoch": 0.37125394321766564,
+      "grad_norm": 0.7212842297017934,
+      "learning_rate": 1.958594917410055e-05,
+      "loss": 0.4928,
+      "step": 1883
+    },
+    {
+      "epoch": 0.3714511041009464,
+      "grad_norm": 0.5994454432815358,
+      "learning_rate": 1.958550765306505e-05,
+      "loss": 0.4354,
+      "step": 1884
+    },
+    {
+      "epoch": 0.37164826498422715,
+      "grad_norm": 0.5881488736842267,
+      "learning_rate": 1.9585065901730013e-05,
+      "loss": 0.4114,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3718454258675079,
+      "grad_norm": 0.5729776216743723,
+      "learning_rate": 1.9584623920106044e-05,
+      "loss": 0.3996,
+      "step": 1886
+    },
+    {
+      "epoch": 0.37204258675078866,
+      "grad_norm": 0.6292024150236367,
+      "learning_rate": 1.9584181708203772e-05,
+      "loss": 0.4452,
+      "step": 1887
+    },
+    {
+      "epoch": 0.3722397476340694,
+      "grad_norm": 0.6434967873079545,
+      "learning_rate": 1.958373926603381e-05,
+      "loss": 0.4864,
+      "step": 1888
+    },
+    {
+      "epoch": 0.3724369085173502,
+      "grad_norm": 0.6189683570278633,
+      "learning_rate": 1.95832965936068e-05,
+      "loss": 0.4288,
+      "step": 1889
+    },
+    {
+      "epoch": 0.37263406940063093,
+      "grad_norm": 0.6925481427276183,
+      "learning_rate": 1.958285369093337e-05,
+      "loss": 0.4982,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3728312302839117,
+      "grad_norm": 0.5954823776943323,
+      "learning_rate": 1.9582410558024162e-05,
+      "loss": 0.4159,
+      "step": 1891
+    },
+    {
+      "epoch": 0.37302839116719244,
+      "grad_norm": 0.6671992939469052,
+      "learning_rate": 1.9581967194889826e-05,
+      "loss": 0.4524,
+      "step": 1892
+    },
+    {
+      "epoch": 0.3732255520504732,
+      "grad_norm": 0.6322337439769842,
+      "learning_rate": 1.9581523601541012e-05,
+      "loss": 0.4909,
+      "step": 1893
+    },
+    {
+      "epoch": 0.37342271293375395,
+      "grad_norm": 0.6712642730299693,
+      "learning_rate": 1.9581079777988375e-05,
+      "loss": 0.4483,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3736198738170347,
+      "grad_norm": 0.6673822790768886,
+      "learning_rate": 1.958063572424258e-05,
+      "loss": 0.4431,
+      "step": 1895
+    },
+    {
+      "epoch": 0.37381703470031546,
+      "grad_norm": 2.28916156009429,
+      "learning_rate": 1.9580191440314304e-05,
+      "loss": 0.5002,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3740141955835962,
+      "grad_norm": 0.6439890915303407,
+      "learning_rate": 1.9579746926214205e-05,
+      "loss": 0.4325,
+      "step": 1897
+    },
+    {
+      "epoch": 0.374211356466877,
+      "grad_norm": 0.7309784046021603,
+      "learning_rate": 1.9579302181952977e-05,
+      "loss": 0.4615,
+      "step": 1898
+    },
+    {
+      "epoch": 0.37440851735015773,
+      "grad_norm": 0.5998628834963219,
+      "learning_rate": 1.9578857207541296e-05,
+      "loss": 0.4563,
+      "step": 1899
+    },
+    {
+      "epoch": 0.3746056782334385,
+      "grad_norm": 0.6925977755644835,
+      "learning_rate": 1.957841200298986e-05,
+      "loss": 0.4705,
+      "step": 1900
+    },
+    {
+      "epoch": 0.37480283911671924,
+      "grad_norm": 0.7297338025869237,
+      "learning_rate": 1.9577966568309358e-05,
+      "loss": 0.4749,
+      "step": 1901
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.7547412508318946,
+      "learning_rate": 1.9577520903510497e-05,
+      "loss": 0.4826,
+      "step": 1902
+    },
+    {
+      "epoch": 0.37519716088328076,
+      "grad_norm": 0.6220978689967857,
+      "learning_rate": 1.957707500860399e-05,
+      "loss": 0.4348,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3753943217665615,
+      "grad_norm": 0.6472282954783044,
+      "learning_rate": 1.9576628883600533e-05,
+      "loss": 0.4102,
+      "step": 1904
+    },
+    {
+      "epoch": 0.37559148264984227,
+      "grad_norm": 0.6841423764886676,
+      "learning_rate": 1.9576182528510864e-05,
+      "loss": 0.4776,
+      "step": 1905
+    },
+    {
+      "epoch": 0.375788643533123,
+      "grad_norm": 0.5900859903843134,
+      "learning_rate": 1.957573594334569e-05,
+      "loss": 0.4334,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3759858044164038,
+      "grad_norm": 0.6428081199582281,
+      "learning_rate": 1.9575289128115758e-05,
+      "loss": 0.4459,
+      "step": 1907
+    },
+    {
+      "epoch": 0.37618296529968454,
+      "grad_norm": 0.6123822425477046,
+      "learning_rate": 1.9574842082831788e-05,
+      "loss": 0.4674,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3763801261829653,
+      "grad_norm": 0.7460401880955564,
+      "learning_rate": 1.957439480750453e-05,
+      "loss": 0.4364,
+      "step": 1909
+    },
+    {
+      "epoch": 0.37657728706624605,
+      "grad_norm": 0.6613935489626954,
+      "learning_rate": 1.957394730214472e-05,
+      "loss": 0.4688,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3767744479495268,
+      "grad_norm": 1.1706846880413952,
+      "learning_rate": 1.9573499566763124e-05,
+      "loss": 0.4627,
+      "step": 1911
+    },
+    {
+      "epoch": 0.37697160883280756,
+      "grad_norm": 1.4249063300767362,
+      "learning_rate": 1.9573051601370485e-05,
+      "loss": 0.4546,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3771687697160883,
+      "grad_norm": 0.7269862868607329,
+      "learning_rate": 1.9572603405977573e-05,
+      "loss": 0.4929,
+      "step": 1913
+    },
+    {
+      "epoch": 0.37736593059936907,
+      "grad_norm": 0.7060411924408715,
+      "learning_rate": 1.957215498059516e-05,
+      "loss": 0.4406,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3775630914826498,
+      "grad_norm": 0.6494093122802824,
+      "learning_rate": 1.957170632523401e-05,
+      "loss": 0.4622,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3777602523659306,
+      "grad_norm": 0.6622590367838171,
+      "learning_rate": 1.957125743990491e-05,
+      "loss": 0.5108,
+      "step": 1916
+    },
+    {
+      "epoch": 0.37795741324921134,
+      "grad_norm": 0.6744814131396353,
+      "learning_rate": 1.9570808324618646e-05,
+      "loss": 0.4583,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3781545741324921,
+      "grad_norm": 3.197758463117182,
+      "learning_rate": 1.9570358979386e-05,
+      "loss": 0.4479,
+      "step": 1918
+    },
+    {
+      "epoch": 0.37835173501577285,
+      "grad_norm": 0.6725127235656607,
+      "learning_rate": 1.956990940421777e-05,
+      "loss": 0.4218,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3785488958990536,
+      "grad_norm": 0.6774677552967868,
+      "learning_rate": 1.9569459599124765e-05,
+      "loss": 0.4654,
+      "step": 1920
+    },
+    {
+      "epoch": 0.37874605678233436,
+      "grad_norm": 2.1023918912871715,
+      "learning_rate": 1.9569009564117783e-05,
+      "loss": 0.4601,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3789432176656151,
+      "grad_norm": 0.8729529815113425,
+      "learning_rate": 1.9568559299207645e-05,
+      "loss": 0.4609,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3791403785488959,
+      "grad_norm": 0.6586715530887697,
+      "learning_rate": 1.9568108804405162e-05,
+      "loss": 0.454,
+      "step": 1923
+    },
+    {
+      "epoch": 0.37933753943217663,
+      "grad_norm": 0.6950591232290906,
+      "learning_rate": 1.956765807972116e-05,
+      "loss": 0.4587,
+      "step": 1924
+    },
+    {
+      "epoch": 0.37953470031545744,
+      "grad_norm": 0.5859553428507086,
+      "learning_rate": 1.9567207125166466e-05,
+      "loss": 0.4559,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3797318611987382,
+      "grad_norm": 0.654061925363833,
+      "learning_rate": 1.9566755940751916e-05,
+      "loss": 0.4539,
+      "step": 1926
+    },
+    {
+      "epoch": 0.37992902208201895,
+      "grad_norm": 0.7394858768473541,
+      "learning_rate": 1.956630452648835e-05,
+      "loss": 0.4135,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3801261829652997,
+      "grad_norm": 0.7096797120999595,
+      "learning_rate": 1.956585288238662e-05,
+      "loss": 0.4421,
+      "step": 1928
+    },
+    {
+      "epoch": 0.38032334384858046,
+      "grad_norm": 0.8603308022382222,
+      "learning_rate": 1.9565401008457567e-05,
+      "loss": 0.4694,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3805205047318612,
+      "grad_norm": 0.6288376088275459,
+      "learning_rate": 1.956494890471205e-05,
+      "loss": 0.4178,
+      "step": 1930
+    },
+    {
+      "epoch": 0.380717665615142,
+      "grad_norm": 0.6656485291036316,
+      "learning_rate": 1.9564496571160935e-05,
+      "loss": 0.4571,
+      "step": 1931
+    },
+    {
+      "epoch": 0.38091482649842273,
+      "grad_norm": 0.6168171412436806,
+      "learning_rate": 1.9564044007815087e-05,
+      "loss": 0.4142,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3811119873817035,
+      "grad_norm": 0.7052282174762466,
+      "learning_rate": 1.956359121468538e-05,
+      "loss": 0.4445,
+      "step": 1933
+    },
+    {
+      "epoch": 0.38130914826498424,
+      "grad_norm": 0.7120124227298402,
+      "learning_rate": 1.9563138191782692e-05,
+      "loss": 0.4557,
+      "step": 1934
+    },
+    {
+      "epoch": 0.381506309148265,
+      "grad_norm": 0.6317595911299724,
+      "learning_rate": 1.956268493911791e-05,
+      "loss": 0.4587,
+      "step": 1935
+    },
+    {
+      "epoch": 0.38170347003154576,
+      "grad_norm": 0.6798856792123635,
+      "learning_rate": 1.9562231456701922e-05,
+      "loss": 0.4612,
+      "step": 1936
+    },
+    {
+      "epoch": 0.3819006309148265,
+      "grad_norm": 1.1002750501181588,
+      "learning_rate": 1.9561777744545616e-05,
+      "loss": 0.4516,
+      "step": 1937
+    },
+    {
+      "epoch": 0.38209779179810727,
+      "grad_norm": 0.7801889450244034,
+      "learning_rate": 1.9561323802659908e-05,
+      "loss": 0.4882,
+      "step": 1938
+    },
+    {
+      "epoch": 0.382294952681388,
+      "grad_norm": 0.6312650051830359,
+      "learning_rate": 1.9560869631055693e-05,
+      "loss": 0.4616,
+      "step": 1939
+    },
+    {
+      "epoch": 0.3824921135646688,
+      "grad_norm": 0.802737212887057,
+      "learning_rate": 1.9560415229743885e-05,
+      "loss": 0.4417,
+      "step": 1940
+    },
+    {
+      "epoch": 0.38268927444794953,
+      "grad_norm": 0.7357740131308222,
+      "learning_rate": 1.9559960598735403e-05,
+      "loss": 0.4701,
+      "step": 1941
+    },
+    {
+      "epoch": 0.3828864353312303,
+      "grad_norm": 0.6284042286654438,
+      "learning_rate": 1.9559505738041167e-05,
+      "loss": 0.4459,
+      "step": 1942
+    },
+    {
+      "epoch": 0.38308359621451105,
+      "grad_norm": 0.6451867717261572,
+      "learning_rate": 1.955905064767211e-05,
+      "loss": 0.4818,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3832807570977918,
+      "grad_norm": 0.9841866252584824,
+      "learning_rate": 1.955859532763916e-05,
+      "loss": 0.412,
+      "step": 1944
+    },
+    {
+      "epoch": 0.38347791798107256,
+      "grad_norm": 0.7692893239650029,
+      "learning_rate": 1.955813977795326e-05,
+      "loss": 0.4328,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3836750788643533,
+      "grad_norm": 0.740511429503883,
+      "learning_rate": 1.955768399862536e-05,
+      "loss": 0.4491,
+      "step": 1946
+    },
+    {
+      "epoch": 0.38387223974763407,
+      "grad_norm": 0.6674798746464747,
+      "learning_rate": 1.95572279896664e-05,
+      "loss": 0.499,
+      "step": 1947
+    },
+    {
+      "epoch": 0.3840694006309148,
+      "grad_norm": 0.8324473358303699,
+      "learning_rate": 1.9556771751087343e-05,
+      "loss": 0.4653,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3842665615141956,
+      "grad_norm": 0.7000802793104866,
+      "learning_rate": 1.955631528289915e-05,
+      "loss": 0.4386,
+      "step": 1949
+    },
+    {
+      "epoch": 0.38446372239747634,
+      "grad_norm": 0.7005116186670807,
+      "learning_rate": 1.9555858585112784e-05,
+      "loss": 0.4615,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3846608832807571,
+      "grad_norm": 0.7023205716896679,
+      "learning_rate": 1.9555401657739222e-05,
+      "loss": 0.4507,
+      "step": 1951
+    },
+    {
+      "epoch": 0.38485804416403785,
+      "grad_norm": 0.6267525065077723,
+      "learning_rate": 1.9554944500789438e-05,
+      "loss": 0.4291,
+      "step": 1952
+    },
+    {
+      "epoch": 0.3850552050473186,
+      "grad_norm": 0.6719159244755133,
+      "learning_rate": 1.955448711427442e-05,
+      "loss": 0.4891,
+      "step": 1953
+    },
+    {
+      "epoch": 0.38525236593059936,
+      "grad_norm": 0.6908923759078575,
+      "learning_rate": 1.9554029498205154e-05,
+      "loss": 0.4753,
+      "step": 1954
+    },
+    {
+      "epoch": 0.3854495268138801,
+      "grad_norm": 0.6560268963325017,
+      "learning_rate": 1.9553571652592637e-05,
+      "loss": 0.4271,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3856466876971609,
+      "grad_norm": 0.8242232954813055,
+      "learning_rate": 1.9553113577447866e-05,
+      "loss": 0.4337,
+      "step": 1956
+    },
+    {
+      "epoch": 0.38584384858044163,
+      "grad_norm": 0.5786503721027892,
+      "learning_rate": 1.9552655272781848e-05,
+      "loss": 0.4491,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3860410094637224,
+      "grad_norm": 0.6178851746400039,
+      "learning_rate": 1.9552196738605596e-05,
+      "loss": 0.4676,
+      "step": 1958
+    },
+    {
+      "epoch": 0.38623817034700314,
+      "grad_norm": 0.6433261446448528,
+      "learning_rate": 1.9551737974930124e-05,
+      "loss": 0.4405,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3864353312302839,
+      "grad_norm": 0.657702346180845,
+      "learning_rate": 1.9551278981766453e-05,
+      "loss": 0.4697,
+      "step": 1960
+    },
+    {
+      "epoch": 0.38663249211356465,
+      "grad_norm": 0.5617264106547595,
+      "learning_rate": 1.9550819759125613e-05,
+      "loss": 0.4263,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3868296529968454,
+      "grad_norm": 0.6078848131857936,
+      "learning_rate": 1.955036030701864e-05,
+      "loss": 0.4693,
+      "step": 1962
+    },
+    {
+      "epoch": 0.38702681388012616,
+      "grad_norm": 0.6098840021530041,
+      "learning_rate": 1.954990062545657e-05,
+      "loss": 0.4535,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3872239747634069,
+      "grad_norm": 0.6666148228118189,
+      "learning_rate": 1.9549440714450447e-05,
+      "loss": 0.4376,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3874211356466877,
+      "grad_norm": 0.6007383801169028,
+      "learning_rate": 1.9548980574011315e-05,
+      "loss": 0.4488,
+      "step": 1965
+    },
+    {
+      "epoch": 0.38761829652996843,
+      "grad_norm": 1.1165576178172691,
+      "learning_rate": 1.954852020415024e-05,
+      "loss": 0.4622,
+      "step": 1966
+    },
+    {
+      "epoch": 0.3878154574132492,
+      "grad_norm": 0.6069174688430108,
+      "learning_rate": 1.9548059604878277e-05,
+      "loss": 0.414,
+      "step": 1967
+    },
+    {
+      "epoch": 0.38801261829652994,
+      "grad_norm": 0.6426555376380005,
+      "learning_rate": 1.9547598776206492e-05,
+      "loss": 0.4859,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3882097791798107,
+      "grad_norm": 0.643323906059335,
+      "learning_rate": 1.954713771814596e-05,
+      "loss": 0.476,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3884069400630915,
+      "grad_norm": 0.6214422402840049,
+      "learning_rate": 1.9546676430707758e-05,
+      "loss": 0.456,
+      "step": 1970
+    },
+    {
+      "epoch": 0.38860410094637227,
+      "grad_norm": 0.5933166205495725,
+      "learning_rate": 1.954621491390296e-05,
+      "loss": 0.4282,
+      "step": 1971
+    },
+    {
+      "epoch": 0.388801261829653,
+      "grad_norm": 0.6469075563555098,
+      "learning_rate": 1.9545753167742664e-05,
+      "loss": 0.4395,
+      "step": 1972
+    },
+    {
+      "epoch": 0.3889984227129338,
+      "grad_norm": 0.575746232416297,
+      "learning_rate": 1.9545291192237962e-05,
+      "loss": 0.4426,
+      "step": 1973
+    },
+    {
+      "epoch": 0.38919558359621453,
+      "grad_norm": 0.8936881259787512,
+      "learning_rate": 1.954482898739995e-05,
+      "loss": 0.4218,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3893927444794953,
+      "grad_norm": 0.5907560428830476,
+      "learning_rate": 1.9544366553239738e-05,
+      "loss": 0.4408,
+      "step": 1975
+    },
+    {
+      "epoch": 0.38958990536277605,
+      "grad_norm": 0.6531849605246557,
+      "learning_rate": 1.9543903889768435e-05,
+      "loss": 0.4718,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3897870662460568,
+      "grad_norm": 0.5847696666259158,
+      "learning_rate": 1.9543440996997152e-05,
+      "loss": 0.4284,
+      "step": 1977
+    },
+    {
+      "epoch": 0.38998422712933756,
+      "grad_norm": 0.6685285347892534,
+      "learning_rate": 1.9542977874937014e-05,
+      "loss": 0.3962,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3901813880126183,
+      "grad_norm": 0.5988664018630944,
+      "learning_rate": 1.954251452359915e-05,
+      "loss": 0.4522,
+      "step": 1979
+    },
+    {
+      "epoch": 0.39037854889589907,
+      "grad_norm": 0.683915321776376,
+      "learning_rate": 1.9542050942994686e-05,
+      "loss": 0.4443,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3905757097791798,
+      "grad_norm": 0.5965994237087414,
+      "learning_rate": 1.9541587133134766e-05,
+      "loss": 0.4836,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3907728706624606,
+      "grad_norm": 0.6874247628508368,
+      "learning_rate": 1.9541123094030528e-05,
+      "loss": 0.4478,
+      "step": 1982
+    },
+    {
+      "epoch": 0.39097003154574134,
+      "grad_norm": 0.6164585865243617,
+      "learning_rate": 1.954065882569313e-05,
+      "loss": 0.4458,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3911671924290221,
+      "grad_norm": 0.6370965104931681,
+      "learning_rate": 1.954019432813372e-05,
+      "loss": 0.421,
+      "step": 1984
+    },
+    {
+      "epoch": 0.39136435331230285,
+      "grad_norm": 0.6849623127670889,
+      "learning_rate": 1.9539729601363456e-05,
+      "loss": 0.4799,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3915615141955836,
+      "grad_norm": 0.6885520124409992,
+      "learning_rate": 1.9539264645393508e-05,
+      "loss": 0.4901,
+      "step": 1986
+    },
+    {
+      "epoch": 0.39175867507886436,
+      "grad_norm": 0.6261317061235863,
+      "learning_rate": 1.9538799460235044e-05,
+      "loss": 0.4266,
+      "step": 1987
+    },
+    {
+      "epoch": 0.3919558359621451,
+      "grad_norm": 0.6441025202938847,
+      "learning_rate": 1.953833404589924e-05,
+      "loss": 0.4428,
+      "step": 1988
+    },
+    {
+      "epoch": 0.39215299684542587,
+      "grad_norm": 0.7945693849241009,
+      "learning_rate": 1.953786840239728e-05,
+      "loss": 0.4498,
+      "step": 1989
+    },
+    {
+      "epoch": 0.39235015772870663,
+      "grad_norm": 0.6236426936646374,
+      "learning_rate": 1.953740252974035e-05,
+      "loss": 0.4433,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3925473186119874,
+      "grad_norm": 0.7354383225024252,
+      "learning_rate": 1.9536936427939647e-05,
+      "loss": 0.4586,
+      "step": 1991
+    },
+    {
+      "epoch": 0.39274447949526814,
+      "grad_norm": 0.7831532504579899,
+      "learning_rate": 1.9536470097006363e-05,
+      "loss": 0.4887,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3929416403785489,
+      "grad_norm": 0.6872639246256627,
+      "learning_rate": 1.9536003536951708e-05,
+      "loss": 0.4405,
+      "step": 1993
+    },
+    {
+      "epoch": 0.39313880126182965,
+      "grad_norm": 0.6736696437372744,
+      "learning_rate": 1.9535536747786884e-05,
+      "loss": 0.441,
+      "step": 1994
+    },
+    {
+      "epoch": 0.3933359621451104,
+      "grad_norm": 0.7082789400644977,
+      "learning_rate": 1.953506972952312e-05,
+      "loss": 0.4596,
+      "step": 1995
+    },
+    {
+      "epoch": 0.39353312302839116,
+      "grad_norm": 0.7181262186981547,
+      "learning_rate": 1.9534602482171618e-05,
+      "loss": 0.4626,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3937302839116719,
+      "grad_norm": 0.7063551366210826,
+      "learning_rate": 1.9534135005743614e-05,
+      "loss": 0.4472,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3939274447949527,
+      "grad_norm": 0.7562123798608619,
+      "learning_rate": 1.9533667300250343e-05,
+      "loss": 0.4832,
+      "step": 1998
+    },
+    {
+      "epoch": 0.39412460567823343,
+      "grad_norm": 0.6940975477174752,
+      "learning_rate": 1.9533199365703035e-05,
+      "loss": 0.4811,
+      "step": 1999
+    },
+    {
+      "epoch": 0.3943217665615142,
+      "grad_norm": 0.6260579790621669,
+      "learning_rate": 1.9532731202112935e-05,
+      "loss": 0.4616,
+      "step": 2000
+    },
+    {
+      "epoch": 0.39451892744479494,
+      "grad_norm": 0.6557298785672231,
+      "learning_rate": 1.9532262809491294e-05,
+      "loss": 0.4785,
+      "step": 2001
+    },
+    {
+      "epoch": 0.3947160883280757,
+      "grad_norm": 0.6533941335015038,
+      "learning_rate": 1.953179418784936e-05,
+      "loss": 0.4448,
+      "step": 2002
+    },
+    {
+      "epoch": 0.39491324921135645,
+      "grad_norm": 0.6709234409900684,
+      "learning_rate": 1.9531325337198394e-05,
+      "loss": 0.4481,
+      "step": 2003
+    },
+    {
+      "epoch": 0.3951104100946372,
+      "grad_norm": 1.528459654964046,
+      "learning_rate": 1.9530856257549664e-05,
+      "loss": 0.4861,
+      "step": 2004
+    },
+    {
+      "epoch": 0.39530757097791797,
+      "grad_norm": 0.6914635119888252,
+      "learning_rate": 1.9530386948914436e-05,
+      "loss": 0.4359,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3955047318611987,
+      "grad_norm": 0.8898433190590056,
+      "learning_rate": 1.9529917411303984e-05,
+      "loss": 0.4643,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3957018927444795,
+      "grad_norm": 0.6810967882315899,
+      "learning_rate": 1.95294476447296e-05,
+      "loss": 0.458,
+      "step": 2007
+    },
+    {
+      "epoch": 0.39589905362776023,
+      "grad_norm": 0.6805859515758788,
+      "learning_rate": 1.9528977649202554e-05,
+      "loss": 0.4199,
+      "step": 2008
+    },
+    {
+      "epoch": 0.396096214511041,
+      "grad_norm": 0.5998994373884956,
+      "learning_rate": 1.9528507424734148e-05,
+      "loss": 0.4118,
+      "step": 2009
+    },
+    {
+      "epoch": 0.39629337539432175,
+      "grad_norm": 0.8444604178654211,
+      "learning_rate": 1.9528036971335678e-05,
+      "loss": 0.4933,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3964905362776025,
+      "grad_norm": 1.2768307772665541,
+      "learning_rate": 1.952756628901845e-05,
+      "loss": 0.4256,
+      "step": 2011
+    },
+    {
+      "epoch": 0.39668769716088326,
+      "grad_norm": 0.7512293164398522,
+      "learning_rate": 1.9527095377793765e-05,
+      "loss": 0.4721,
+      "step": 2012
+    },
+    {
+      "epoch": 0.396884858044164,
+      "grad_norm": 0.9372345492820653,
+      "learning_rate": 1.9526624237672945e-05,
+      "loss": 0.4319,
+      "step": 2013
+    },
+    {
+      "epoch": 0.39708201892744477,
+      "grad_norm": 3.610388601557778,
+      "learning_rate": 1.9526152868667302e-05,
+      "loss": 0.511,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3972791798107255,
+      "grad_norm": 0.9133527465332176,
+      "learning_rate": 1.952568127078817e-05,
+      "loss": 0.4852,
+      "step": 2015
+    },
+    {
+      "epoch": 0.39747634069400634,
+      "grad_norm": 0.801460794349205,
+      "learning_rate": 1.952520944404687e-05,
+      "loss": 0.4679,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3976735015772871,
+      "grad_norm": 0.6184192490263251,
+      "learning_rate": 1.9524737388454745e-05,
+      "loss": 0.4306,
+      "step": 2017
+    },
+    {
+      "epoch": 0.39787066246056785,
+      "grad_norm": 0.7074190008986967,
+      "learning_rate": 1.9524265104023133e-05,
+      "loss": 0.4233,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3980678233438486,
+      "grad_norm": 0.7029283154800328,
+      "learning_rate": 1.9523792590763382e-05,
+      "loss": 0.4627,
+      "step": 2019
+    },
+    {
+      "epoch": 0.39826498422712936,
+      "grad_norm": 0.851532590870496,
+      "learning_rate": 1.9523319848686845e-05,
+      "loss": 0.4506,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3984621451104101,
+      "grad_norm": 0.7525221362585885,
+      "learning_rate": 1.952284687780488e-05,
+      "loss": 0.4434,
+      "step": 2021
+    },
+    {
+      "epoch": 0.39865930599369087,
+      "grad_norm": 0.8241310038169775,
+      "learning_rate": 1.952237367812885e-05,
+      "loss": 0.4552,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3988564668769716,
+      "grad_norm": 0.6291466115452881,
+      "learning_rate": 1.952190024967012e-05,
+      "loss": 0.441,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3990536277602524,
+      "grad_norm": 0.8316147796097879,
+      "learning_rate": 1.9521426592440075e-05,
+      "loss": 0.4394,
+      "step": 2024
+    },
+    {
+      "epoch": 0.39925078864353314,
+      "grad_norm": 0.6195605062293554,
+      "learning_rate": 1.9520952706450083e-05,
+      "loss": 0.4547,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3994479495268139,
+      "grad_norm": 1.0728222016893565,
+      "learning_rate": 1.952047859171154e-05,
+      "loss": 0.4847,
+      "step": 2026
+    },
+    {
+      "epoch": 0.39964511041009465,
+      "grad_norm": 0.6543029079987064,
+      "learning_rate": 1.9520004248235826e-05,
+      "loss": 0.4412,
+      "step": 2027
+    },
+    {
+      "epoch": 0.3998422712933754,
+      "grad_norm": 0.8215362022698259,
+      "learning_rate": 1.9519529676034347e-05,
+      "loss": 0.4413,
+      "step": 2028
+    },
+    {
+      "epoch": 0.40003943217665616,
+      "grad_norm": 0.6394436878246803,
+      "learning_rate": 1.95190548751185e-05,
+      "loss": 0.4091,
+      "step": 2029
+    },
+    {
+      "epoch": 0.4002365930599369,
+      "grad_norm": 0.8545402719858461,
+      "learning_rate": 1.9518579845499698e-05,
+      "loss": 0.4713,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4004337539432177,
+      "grad_norm": 0.6930745549889441,
+      "learning_rate": 1.9518104587189348e-05,
+      "loss": 0.4474,
+      "step": 2031
+    },
+    {
+      "epoch": 0.40063091482649843,
+      "grad_norm": 0.8730350351232401,
+      "learning_rate": 1.951762910019887e-05,
+      "loss": 0.4678,
+      "step": 2032
+    },
+    {
+      "epoch": 0.4008280757097792,
+      "grad_norm": 0.7083655229046756,
+      "learning_rate": 1.9517153384539685e-05,
+      "loss": 0.4452,
+      "step": 2033
+    },
+    {
+      "epoch": 0.40102523659305994,
+      "grad_norm": 0.9617195558090388,
+      "learning_rate": 1.9516677440223232e-05,
+      "loss": 0.4631,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4012223974763407,
+      "grad_norm": 0.7580517534109538,
+      "learning_rate": 1.9516201267260935e-05,
+      "loss": 0.4548,
+      "step": 2035
+    },
+    {
+      "epoch": 0.40141955835962145,
+      "grad_norm": 1.0525747613425824,
+      "learning_rate": 1.9515724865664242e-05,
+      "loss": 0.4953,
+      "step": 2036
+    },
+    {
+      "epoch": 0.4016167192429022,
+      "grad_norm": 0.7820613688472805,
+      "learning_rate": 1.9515248235444595e-05,
+      "loss": 0.4339,
+      "step": 2037
+    },
+    {
+      "epoch": 0.40181388012618297,
+      "grad_norm": 0.7403224002007617,
+      "learning_rate": 1.9514771376613446e-05,
+      "loss": 0.4766,
+      "step": 2038
+    },
+    {
+      "epoch": 0.4020110410094637,
+      "grad_norm": 0.8359871396836571,
+      "learning_rate": 1.9514294289182253e-05,
+      "loss": 0.5007,
+      "step": 2039
+    },
+    {
+      "epoch": 0.4022082018927445,
+      "grad_norm": 51.06976499844091,
+      "learning_rate": 1.9513816973162475e-05,
+      "loss": 0.7096,
+      "step": 2040
+    },
+    {
+      "epoch": 0.40240536277602523,
+      "grad_norm": 0.9110749969850523,
+      "learning_rate": 1.9513339428565588e-05,
+      "loss": 0.4644,
+      "step": 2041
+    },
+    {
+      "epoch": 0.402602523659306,
+      "grad_norm": 0.7178947368074214,
+      "learning_rate": 1.9512861655403057e-05,
+      "loss": 0.4586,
+      "step": 2042
+    },
+    {
+      "epoch": 0.40279968454258674,
+      "grad_norm": 0.6762975622616101,
+      "learning_rate": 1.9512383653686364e-05,
+      "loss": 0.4217,
+      "step": 2043
+    },
+    {
+      "epoch": 0.4029968454258675,
+      "grad_norm": 0.7412728040347322,
+      "learning_rate": 1.9511905423426992e-05,
+      "loss": 0.5072,
+      "step": 2044
+    },
+    {
+      "epoch": 0.40319400630914826,
+      "grad_norm": 0.8922820307079298,
+      "learning_rate": 1.9511426964636437e-05,
+      "loss": 0.4553,
+      "step": 2045
+    },
+    {
+      "epoch": 0.403391167192429,
+      "grad_norm": 0.6604569080074368,
+      "learning_rate": 1.9510948277326188e-05,
+      "loss": 0.4388,
+      "step": 2046
+    },
+    {
+      "epoch": 0.40358832807570977,
+      "grad_norm": 0.7657116520818048,
+      "learning_rate": 1.9510469361507747e-05,
+      "loss": 0.4703,
+      "step": 2047
+    },
+    {
+      "epoch": 0.4037854889589905,
+      "grad_norm": 0.6403472748402534,
+      "learning_rate": 1.950999021719262e-05,
+      "loss": 0.4426,
+      "step": 2048
+    },
+    {
+      "epoch": 0.4039826498422713,
+      "grad_norm": 0.7942729480613527,
+      "learning_rate": 1.950951084439232e-05,
+      "loss": 0.4722,
+      "step": 2049
+    },
+    {
+      "epoch": 0.40417981072555204,
+      "grad_norm": 0.6777646106474043,
+      "learning_rate": 1.9509031243118365e-05,
+      "loss": 0.4596,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4043769716088328,
+      "grad_norm": 0.6657983977712271,
+      "learning_rate": 1.9508551413382274e-05,
+      "loss": 0.4649,
+      "step": 2051
+    },
+    {
+      "epoch": 0.40457413249211355,
+      "grad_norm": 0.7409227648041087,
+      "learning_rate": 1.950807135519558e-05,
+      "loss": 0.4949,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4047712933753943,
+      "grad_norm": 0.6979782066122284,
+      "learning_rate": 1.9507591068569812e-05,
+      "loss": 0.519,
+      "step": 2053
+    },
+    {
+      "epoch": 0.40496845425867506,
+      "grad_norm": 0.6505454029015588,
+      "learning_rate": 1.9507110553516518e-05,
+      "loss": 0.4148,
+      "step": 2054
+    },
+    {
+      "epoch": 0.4051656151419558,
+      "grad_norm": 0.6905829191499583,
+      "learning_rate": 1.9506629810047233e-05,
+      "loss": 0.4269,
+      "step": 2055
+    },
+    {
+      "epoch": 0.40536277602523657,
+      "grad_norm": 0.6726973855018206,
+      "learning_rate": 1.9506148838173512e-05,
+      "loss": 0.4714,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4055599369085173,
+      "grad_norm": 0.6449181697582981,
+      "learning_rate": 1.950566763790691e-05,
+      "loss": 0.4611,
+      "step": 2057
+    },
+    {
+      "epoch": 0.4057570977917981,
+      "grad_norm": 0.6541125973886486,
+      "learning_rate": 1.9505186209258987e-05,
+      "loss": 0.4657,
+      "step": 2058
+    },
+    {
+      "epoch": 0.40595425867507884,
+      "grad_norm": 0.6492031501355185,
+      "learning_rate": 1.950470455224131e-05,
+      "loss": 0.419,
+      "step": 2059
+    },
+    {
+      "epoch": 0.4061514195583596,
+      "grad_norm": 0.6958217286443721,
+      "learning_rate": 1.9504222666865457e-05,
+      "loss": 0.4559,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4063485804416404,
+      "grad_norm": 0.6660544919270931,
+      "learning_rate": 1.9503740553142995e-05,
+      "loss": 0.4614,
+      "step": 2061
+    },
+    {
+      "epoch": 0.40654574132492116,
+      "grad_norm": 0.6441549325271921,
+      "learning_rate": 1.9503258211085515e-05,
+      "loss": 0.4832,
+      "step": 2062
+    },
+    {
+      "epoch": 0.4067429022082019,
+      "grad_norm": 1.5150959012964547,
+      "learning_rate": 1.9502775640704606e-05,
+      "loss": 0.434,
+      "step": 2063
+    },
+    {
+      "epoch": 0.4069400630914827,
+      "grad_norm": 0.6244719601196084,
+      "learning_rate": 1.9502292842011857e-05,
+      "loss": 0.4173,
+      "step": 2064
+    },
+    {
+      "epoch": 0.40713722397476343,
+      "grad_norm": 0.6628533283416533,
+      "learning_rate": 1.950180981501887e-05,
+      "loss": 0.4414,
+      "step": 2065
+    },
+    {
+      "epoch": 0.4073343848580442,
+      "grad_norm": 0.7233298009650888,
+      "learning_rate": 1.950132655973725e-05,
+      "loss": 0.4296,
+      "step": 2066
+    },
+    {
+      "epoch": 0.40753154574132494,
+      "grad_norm": 0.6648377046224346,
+      "learning_rate": 1.9500843076178612e-05,
+      "loss": 0.4455,
+      "step": 2067
+    },
+    {
+      "epoch": 0.4077287066246057,
+      "grad_norm": 0.6564892733023274,
+      "learning_rate": 1.9500359364354565e-05,
+      "loss": 0.4545,
+      "step": 2068
+    },
+    {
+      "epoch": 0.40792586750788645,
+      "grad_norm": 0.6435233243344421,
+      "learning_rate": 1.9499875424276734e-05,
+      "loss": 0.4551,
+      "step": 2069
+    },
+    {
+      "epoch": 0.4081230283911672,
+      "grad_norm": 0.6478000940884499,
+      "learning_rate": 1.9499391255956745e-05,
+      "loss": 0.4262,
+      "step": 2070
+    },
+    {
+      "epoch": 0.40832018927444796,
+      "grad_norm": 2.6194886693717487,
+      "learning_rate": 1.949890685940623e-05,
+      "loss": 0.4949,
+      "step": 2071
+    },
+    {
+      "epoch": 0.4085173501577287,
+      "grad_norm": 0.7796501553012304,
+      "learning_rate": 1.949842223463683e-05,
+      "loss": 0.4521,
+      "step": 2072
+    },
+    {
+      "epoch": 0.4087145110410095,
+      "grad_norm": 1.1718471052149675,
+      "learning_rate": 1.9497937381660188e-05,
+      "loss": 0.43,
+      "step": 2073
+    },
+    {
+      "epoch": 0.40891167192429023,
+      "grad_norm": 0.7069733785890165,
+      "learning_rate": 1.949745230048795e-05,
+      "loss": 0.4567,
+      "step": 2074
+    },
+    {
+      "epoch": 0.409108832807571,
+      "grad_norm": 0.7134975969386784,
+      "learning_rate": 1.9496966991131775e-05,
+      "loss": 0.3937,
+      "step": 2075
+    },
+    {
+      "epoch": 0.40930599369085174,
+      "grad_norm": 0.6204131343159649,
+      "learning_rate": 1.9496481453603318e-05,
+      "loss": 0.4501,
+      "step": 2076
+    },
+    {
+      "epoch": 0.4095031545741325,
+      "grad_norm": 0.6233092334560004,
+      "learning_rate": 1.9495995687914244e-05,
+      "loss": 0.4122,
+      "step": 2077
+    },
+    {
+      "epoch": 0.40970031545741326,
+      "grad_norm": 0.6614510142665343,
+      "learning_rate": 1.949550969407623e-05,
+      "loss": 0.4269,
+      "step": 2078
+    },
+    {
+      "epoch": 0.409897476340694,
+      "grad_norm": 0.6647109624994366,
+      "learning_rate": 1.949502347210095e-05,
+      "loss": 0.464,
+      "step": 2079
+    },
+    {
+      "epoch": 0.41009463722397477,
+      "grad_norm": 0.6704844205652758,
+      "learning_rate": 1.949453702200008e-05,
+      "loss": 0.471,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4102917981072555,
+      "grad_norm": 0.6946952276485558,
+      "learning_rate": 1.9494050343785317e-05,
+      "loss": 0.461,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4104889589905363,
+      "grad_norm": 0.6724334955144293,
+      "learning_rate": 1.9493563437468344e-05,
+      "loss": 0.4693,
+      "step": 2082
+    },
+    {
+      "epoch": 0.41068611987381703,
+      "grad_norm": 0.7089092005187811,
+      "learning_rate": 1.9493076303060866e-05,
+      "loss": 0.4482,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4108832807570978,
+      "grad_norm": 0.5996155537534351,
+      "learning_rate": 1.9492588940574588e-05,
+      "loss": 0.445,
+      "step": 2084
+    },
+    {
+      "epoch": 0.41108044164037855,
+      "grad_norm": 0.6632658160746489,
+      "learning_rate": 1.9492101350021216e-05,
+      "loss": 0.42,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4112776025236593,
+      "grad_norm": 0.6093300916331452,
+      "learning_rate": 1.9491613531412463e-05,
+      "loss": 0.4507,
+      "step": 2086
+    },
+    {
+      "epoch": 0.41147476340694006,
+      "grad_norm": 0.6358547202310659,
+      "learning_rate": 1.949112548476005e-05,
+      "loss": 0.4588,
+      "step": 2087
+    },
+    {
+      "epoch": 0.4116719242902208,
+      "grad_norm": 0.6371219707058599,
+      "learning_rate": 1.9490637210075708e-05,
+      "loss": 0.4772,
+      "step": 2088
+    },
+    {
+      "epoch": 0.41186908517350157,
+      "grad_norm": 0.5999352611847947,
+      "learning_rate": 1.9490148707371163e-05,
+      "loss": 0.4604,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4120662460567823,
+      "grad_norm": 0.653380052430745,
+      "learning_rate": 1.9489659976658152e-05,
+      "loss": 0.4377,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4122634069400631,
+      "grad_norm": 0.6174059523962631,
+      "learning_rate": 1.948917101794842e-05,
+      "loss": 0.4545,
+      "step": 2091
+    },
+    {
+      "epoch": 0.41246056782334384,
+      "grad_norm": 0.6989529903953323,
+      "learning_rate": 1.9488681831253706e-05,
+      "loss": 0.4615,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4126577287066246,
+      "grad_norm": 0.6751206895361119,
+      "learning_rate": 1.9488192416585775e-05,
+      "loss": 0.4749,
+      "step": 2093
+    },
+    {
+      "epoch": 0.41285488958990535,
+      "grad_norm": 0.6067234928335872,
+      "learning_rate": 1.948770277395638e-05,
+      "loss": 0.4495,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4130520504731861,
+      "grad_norm": 0.6038861471266463,
+      "learning_rate": 1.9487212903377286e-05,
+      "loss": 0.4318,
+      "step": 2095
+    },
+    {
+      "epoch": 0.41324921135646686,
+      "grad_norm": 0.6099951703516203,
+      "learning_rate": 1.9486722804860262e-05,
+      "loss": 0.4232,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4134463722397476,
+      "grad_norm": 0.6018481400930096,
+      "learning_rate": 1.948623247841708e-05,
+      "loss": 0.4583,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4136435331230284,
+      "grad_norm": 0.6300586347763767,
+      "learning_rate": 1.948574192405953e-05,
+      "loss": 0.4805,
+      "step": 2098
+    },
+    {
+      "epoch": 0.41384069400630913,
+      "grad_norm": 0.6261753743075679,
+      "learning_rate": 1.9485251141799387e-05,
+      "loss": 0.4929,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4140378548895899,
+      "grad_norm": 0.5502783264260944,
+      "learning_rate": 1.9484760131648447e-05,
+      "loss": 0.4201,
+      "step": 2100
+    },
+    {
+      "epoch": 0.41423501577287064,
+      "grad_norm": 0.6373152594526875,
+      "learning_rate": 1.9484268893618504e-05,
+      "loss": 0.5034,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4144321766561514,
+      "grad_norm": 0.550511658521269,
+      "learning_rate": 1.9483777427721367e-05,
+      "loss": 0.4215,
+      "step": 2102
+    },
+    {
+      "epoch": 0.41462933753943215,
+      "grad_norm": 0.600033811832785,
+      "learning_rate": 1.948328573396884e-05,
+      "loss": 0.4536,
+      "step": 2103
+    },
+    {
+      "epoch": 0.4148264984227129,
+      "grad_norm": 0.591267981029353,
+      "learning_rate": 1.9482793812372732e-05,
+      "loss": 0.4195,
+      "step": 2104
+    },
+    {
+      "epoch": 0.41502365930599366,
+      "grad_norm": 0.6355093625177733,
+      "learning_rate": 1.9482301662944872e-05,
+      "loss": 0.45,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4152208201892745,
+      "grad_norm": 0.6066083667212502,
+      "learning_rate": 1.9481809285697076e-05,
+      "loss": 0.438,
+      "step": 2106
+    },
+    {
+      "epoch": 0.41541798107255523,
+      "grad_norm": 0.6268489081451546,
+      "learning_rate": 1.9481316680641175e-05,
+      "loss": 0.4482,
+      "step": 2107
+    },
+    {
+      "epoch": 0.415615141955836,
+      "grad_norm": 0.6292327520694636,
+      "learning_rate": 1.9480823847789007e-05,
+      "loss": 0.4657,
+      "step": 2108
+    },
+    {
+      "epoch": 0.41581230283911674,
+      "grad_norm": 0.5829052402848848,
+      "learning_rate": 1.9480330787152413e-05,
+      "loss": 0.429,
+      "step": 2109
+    },
+    {
+      "epoch": 0.4160094637223975,
+      "grad_norm": 1.3748165275525082,
+      "learning_rate": 1.9479837498743236e-05,
+      "loss": 0.4193,
+      "step": 2110
+    },
+    {
+      "epoch": 0.41620662460567825,
+      "grad_norm": 0.5682917102110292,
+      "learning_rate": 1.9479343982573326e-05,
+      "loss": 0.4422,
+      "step": 2111
+    },
+    {
+      "epoch": 0.416403785488959,
+      "grad_norm": 0.5947136497114629,
+      "learning_rate": 1.9478850238654546e-05,
+      "loss": 0.4614,
+      "step": 2112
+    },
+    {
+      "epoch": 0.41660094637223977,
+      "grad_norm": 0.7037292516298015,
+      "learning_rate": 1.9478356266998757e-05,
+      "loss": 0.5048,
+      "step": 2113
+    },
+    {
+      "epoch": 0.4167981072555205,
+      "grad_norm": 0.6478893055908316,
+      "learning_rate": 1.947786206761782e-05,
+      "loss": 0.4463,
+      "step": 2114
+    },
+    {
+      "epoch": 0.4169952681388013,
+      "grad_norm": 0.6013780035939855,
+      "learning_rate": 1.9477367640523622e-05,
+      "loss": 0.4595,
+      "step": 2115
+    },
+    {
+      "epoch": 0.41719242902208203,
+      "grad_norm": 0.5984366322194706,
+      "learning_rate": 1.947687298572803e-05,
+      "loss": 0.4312,
+      "step": 2116
+    },
+    {
+      "epoch": 0.4173895899053628,
+      "grad_norm": 0.6057400198881939,
+      "learning_rate": 1.9476378103242934e-05,
+      "loss": 0.4423,
+      "step": 2117
+    },
+    {
+      "epoch": 0.41758675078864355,
+      "grad_norm": 0.5985571921288283,
+      "learning_rate": 1.9475882993080223e-05,
+      "loss": 0.4192,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4177839116719243,
+      "grad_norm": 0.6313104451272632,
+      "learning_rate": 1.947538765525179e-05,
+      "loss": 0.4382,
+      "step": 2119
+    },
+    {
+      "epoch": 0.41798107255520506,
+      "grad_norm": 0.6434996318524585,
+      "learning_rate": 1.9474892089769538e-05,
+      "loss": 0.4643,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4181782334384858,
+      "grad_norm": 0.5983507556937149,
+      "learning_rate": 1.947439629664538e-05,
+      "loss": 0.4306,
+      "step": 2121
+    },
+    {
+      "epoch": 0.41837539432176657,
+      "grad_norm": 0.6227894910009829,
+      "learning_rate": 1.9473900275891214e-05,
+      "loss": 0.4185,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4185725552050473,
+      "grad_norm": 0.6249459358371175,
+      "learning_rate": 1.9473404027518965e-05,
+      "loss": 0.4422,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4187697160883281,
+      "grad_norm": 0.5581857659991755,
+      "learning_rate": 1.9472907551540557e-05,
+      "loss": 0.3957,
+      "step": 2124
+    },
+    {
+      "epoch": 0.41896687697160884,
+      "grad_norm": 0.6086539772510696,
+      "learning_rate": 1.9472410847967917e-05,
+      "loss": 0.4538,
+      "step": 2125
+    },
+    {
+      "epoch": 0.4191640378548896,
+      "grad_norm": 0.6001107475689484,
+      "learning_rate": 1.947191391681298e-05,
+      "loss": 0.4105,
+      "step": 2126
+    },
+    {
+      "epoch": 0.41936119873817035,
+      "grad_norm": 0.66517888101447,
+      "learning_rate": 1.947141675808768e-05,
+      "loss": 0.4399,
+      "step": 2127
+    },
+    {
+      "epoch": 0.4195583596214511,
+      "grad_norm": 0.5752990303617271,
+      "learning_rate": 1.9470919371803966e-05,
+      "loss": 0.4208,
+      "step": 2128
+    },
+    {
+      "epoch": 0.41975552050473186,
+      "grad_norm": 0.6156099256870735,
+      "learning_rate": 1.947042175797379e-05,
+      "loss": 0.4578,
+      "step": 2129
+    },
+    {
+      "epoch": 0.4199526813880126,
+      "grad_norm": 0.5825517100633804,
+      "learning_rate": 1.94699239166091e-05,
+      "loss": 0.45,
+      "step": 2130
+    },
+    {
+      "epoch": 0.42014984227129337,
+      "grad_norm": 0.6011744479891673,
+      "learning_rate": 1.9469425847721865e-05,
+      "loss": 0.4734,
+      "step": 2131
+    },
+    {
+      "epoch": 0.42034700315457413,
+      "grad_norm": 0.5606446063107278,
+      "learning_rate": 1.9468927551324045e-05,
+      "loss": 0.4242,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4205441640378549,
+      "grad_norm": 0.5730880385070863,
+      "learning_rate": 1.946842902742762e-05,
+      "loss": 0.4439,
+      "step": 2133
+    },
+    {
+      "epoch": 0.42074132492113564,
+      "grad_norm": 0.6125765323790818,
+      "learning_rate": 1.9467930276044557e-05,
+      "loss": 0.4655,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4209384858044164,
+      "grad_norm": 0.5806851297859154,
+      "learning_rate": 1.946743129718685e-05,
+      "loss": 0.441,
+      "step": 2135
+    },
+    {
+      "epoch": 0.42113564668769715,
+      "grad_norm": 0.6077365774621822,
+      "learning_rate": 1.946693209086648e-05,
+      "loss": 0.4479,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4213328075709779,
+      "grad_norm": 0.6092599482658368,
+      "learning_rate": 1.9466432657095443e-05,
+      "loss": 0.4439,
+      "step": 2137
+    },
+    {
+      "epoch": 0.42152996845425866,
+      "grad_norm": 0.5972879209489058,
+      "learning_rate": 1.9465932995885737e-05,
+      "loss": 0.4174,
+      "step": 2138
+    },
+    {
+      "epoch": 0.4217271293375394,
+      "grad_norm": 0.6448412899637194,
+      "learning_rate": 1.946543310724937e-05,
+      "loss": 0.4802,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4219242902208202,
+      "grad_norm": 0.5919021815947381,
+      "learning_rate": 1.946493299119835e-05,
+      "loss": 0.428,
+      "step": 2140
+    },
+    {
+      "epoch": 0.42212145110410093,
+      "grad_norm": 0.5832897545671972,
+      "learning_rate": 1.9464432647744693e-05,
+      "loss": 0.4152,
+      "step": 2141
+    },
+    {
+      "epoch": 0.4223186119873817,
+      "grad_norm": 0.7594410044159404,
+      "learning_rate": 1.9463932076900416e-05,
+      "loss": 0.4921,
+      "step": 2142
+    },
+    {
+      "epoch": 0.42251577287066244,
+      "grad_norm": 0.5750713155262214,
+      "learning_rate": 1.9463431278677552e-05,
+      "loss": 0.4064,
+      "step": 2143
+    },
+    {
+      "epoch": 0.4227129337539432,
+      "grad_norm": 0.6628109681634553,
+      "learning_rate": 1.946293025308813e-05,
+      "loss": 0.4372,
+      "step": 2144
+    },
+    {
+      "epoch": 0.42291009463722395,
+      "grad_norm": 0.6309603811899096,
+      "learning_rate": 1.946242900014419e-05,
+      "loss": 0.4862,
+      "step": 2145
+    },
+    {
+      "epoch": 0.4231072555205047,
+      "grad_norm": 0.6540676115612354,
+      "learning_rate": 1.9461927519857772e-05,
+      "loss": 0.4509,
+      "step": 2146
+    },
+    {
+      "epoch": 0.42330441640378547,
+      "grad_norm": 0.6395341421094025,
+      "learning_rate": 1.9461425812240925e-05,
+      "loss": 0.4526,
+      "step": 2147
+    },
+    {
+      "epoch": 0.4235015772870662,
+      "grad_norm": 0.6394713150271438,
+      "learning_rate": 1.9460923877305706e-05,
+      "loss": 0.4616,
+      "step": 2148
+    },
+    {
+      "epoch": 0.423698738170347,
+      "grad_norm": 0.6069065769872102,
+      "learning_rate": 1.9460421715064172e-05,
+      "loss": 0.4289,
+      "step": 2149
+    },
+    {
+      "epoch": 0.42389589905362773,
+      "grad_norm": 0.6064632740648996,
+      "learning_rate": 1.9459919325528384e-05,
+      "loss": 0.4635,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4240930599369085,
+      "grad_norm": 0.616873552338082,
+      "learning_rate": 1.945941670871042e-05,
+      "loss": 0.4648,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4242902208201893,
+      "grad_norm": 0.5805094134522223,
+      "learning_rate": 1.945891386462235e-05,
+      "loss": 0.4254,
+      "step": 2152
+    },
+    {
+      "epoch": 0.42448738170347006,
+      "grad_norm": 0.6016630618010061,
+      "learning_rate": 1.9458410793276256e-05,
+      "loss": 0.4537,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4246845425867508,
+      "grad_norm": 0.6221143855785904,
+      "learning_rate": 1.9457907494684227e-05,
+      "loss": 0.4689,
+      "step": 2154
+    },
+    {
+      "epoch": 0.42488170347003157,
+      "grad_norm": 0.786382722019145,
+      "learning_rate": 1.9457403968858358e-05,
+      "loss": 0.4614,
+      "step": 2155
+    },
+    {
+      "epoch": 0.4250788643533123,
+      "grad_norm": 0.5847320244340704,
+      "learning_rate": 1.9456900215810737e-05,
+      "loss": 0.4293,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4252760252365931,
+      "grad_norm": 0.695033418240993,
+      "learning_rate": 1.9456396235553474e-05,
+      "loss": 0.4215,
+      "step": 2157
+    },
+    {
+      "epoch": 0.42547318611987384,
+      "grad_norm": 0.6844045999840678,
+      "learning_rate": 1.9455892028098677e-05,
+      "loss": 0.5193,
+      "step": 2158
+    },
+    {
+      "epoch": 0.4256703470031546,
+      "grad_norm": 0.6184226453391258,
+      "learning_rate": 1.945538759345846e-05,
+      "loss": 0.486,
+      "step": 2159
+    },
+    {
+      "epoch": 0.42586750788643535,
+      "grad_norm": 0.6324228885355624,
+      "learning_rate": 1.9454882931644942e-05,
+      "loss": 0.4585,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4260646687697161,
+      "grad_norm": 0.6165127471239128,
+      "learning_rate": 1.9454378042670245e-05,
+      "loss": 0.46,
+      "step": 2161
+    },
+    {
+      "epoch": 0.42626182965299686,
+      "grad_norm": 0.6347931854232857,
+      "learning_rate": 1.9453872926546505e-05,
+      "loss": 0.4598,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4264589905362776,
+      "grad_norm": 0.6539072568560251,
+      "learning_rate": 1.9453367583285853e-05,
+      "loss": 0.4291,
+      "step": 2163
+    },
+    {
+      "epoch": 0.42665615141955837,
+      "grad_norm": 0.6184889881753938,
+      "learning_rate": 1.945286201290043e-05,
+      "loss": 0.4624,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4268533123028391,
+      "grad_norm": 0.619366970782029,
+      "learning_rate": 1.945235621540239e-05,
+      "loss": 0.4838,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4270504731861199,
+      "grad_norm": 0.5940273163643314,
+      "learning_rate": 1.9451850190803877e-05,
+      "loss": 0.4219,
+      "step": 2166
+    },
+    {
+      "epoch": 0.42724763406940064,
+      "grad_norm": 0.6055311252553117,
+      "learning_rate": 1.9451343939117052e-05,
+      "loss": 0.4413,
+      "step": 2167
+    },
+    {
+      "epoch": 0.4274447949526814,
+      "grad_norm": 0.6326779232777279,
+      "learning_rate": 1.9450837460354073e-05,
+      "loss": 0.4506,
+      "step": 2168
+    },
+    {
+      "epoch": 0.42764195583596215,
+      "grad_norm": 0.6835756822811451,
+      "learning_rate": 1.9450330754527118e-05,
+      "loss": 0.4916,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4278391167192429,
+      "grad_norm": 0.6561532078291141,
+      "learning_rate": 1.9449823821648357e-05,
+      "loss": 0.461,
+      "step": 2170
+    },
+    {
+      "epoch": 0.42803627760252366,
+      "grad_norm": 0.7237386411398435,
+      "learning_rate": 1.944931666172997e-05,
+      "loss": 0.4694,
+      "step": 2171
+    },
+    {
+      "epoch": 0.4282334384858044,
+      "grad_norm": 0.615056095691249,
+      "learning_rate": 1.9448809274784136e-05,
+      "loss": 0.4374,
+      "step": 2172
+    },
+    {
+      "epoch": 0.4284305993690852,
+      "grad_norm": 0.5965296460751116,
+      "learning_rate": 1.944830166082305e-05,
+      "loss": 0.4651,
+      "step": 2173
+    },
+    {
+      "epoch": 0.42862776025236593,
+      "grad_norm": 0.6166157125058764,
+      "learning_rate": 1.9447793819858912e-05,
+      "loss": 0.4427,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4288249211356467,
+      "grad_norm": 0.5827660473699446,
+      "learning_rate": 1.944728575190392e-05,
+      "loss": 0.4428,
+      "step": 2175
+    },
+    {
+      "epoch": 0.42902208201892744,
+      "grad_norm": 0.5867427682120541,
+      "learning_rate": 1.9446777456970276e-05,
+      "loss": 0.4346,
+      "step": 2176
+    },
+    {
+      "epoch": 0.4292192429022082,
+      "grad_norm": 0.5549950519191269,
+      "learning_rate": 1.9446268935070197e-05,
+      "loss": 0.4119,
+      "step": 2177
+    },
+    {
+      "epoch": 0.42941640378548895,
+      "grad_norm": 0.6125305517287657,
+      "learning_rate": 1.94457601862159e-05,
+      "loss": 0.4829,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4296135646687697,
+      "grad_norm": 0.6921490966397458,
+      "learning_rate": 1.944525121041961e-05,
+      "loss": 0.4758,
+      "step": 2179
+    },
+    {
+      "epoch": 0.42981072555205047,
+      "grad_norm": 0.585352142286126,
+      "learning_rate": 1.944474200769355e-05,
+      "loss": 0.4497,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4300078864353312,
+      "grad_norm": 0.6266546137858079,
+      "learning_rate": 1.944423257804996e-05,
+      "loss": 0.4481,
+      "step": 2181
+    },
+    {
+      "epoch": 0.430205047318612,
+      "grad_norm": 0.6516582163069267,
+      "learning_rate": 1.9443722921501074e-05,
+      "loss": 0.4473,
+      "step": 2182
+    },
+    {
+      "epoch": 0.43040220820189273,
+      "grad_norm": 0.6099356219217796,
+      "learning_rate": 1.9443213038059145e-05,
+      "loss": 0.4546,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4305993690851735,
+      "grad_norm": 0.6526605435968623,
+      "learning_rate": 1.944270292773641e-05,
+      "loss": 0.4225,
+      "step": 2184
+    },
+    {
+      "epoch": 0.43079652996845424,
+      "grad_norm": 0.7044967095269495,
+      "learning_rate": 1.944219259054514e-05,
+      "loss": 0.4483,
+      "step": 2185
+    },
+    {
+      "epoch": 0.430993690851735,
+      "grad_norm": 0.6635399522882317,
+      "learning_rate": 1.9441682026497587e-05,
+      "loss": 0.4396,
+      "step": 2186
+    },
+    {
+      "epoch": 0.43119085173501576,
+      "grad_norm": 0.7462506165933283,
+      "learning_rate": 1.944117123560602e-05,
+      "loss": 0.4184,
+      "step": 2187
+    },
+    {
+      "epoch": 0.4313880126182965,
+      "grad_norm": 0.630701248717489,
+      "learning_rate": 1.944066021788271e-05,
+      "loss": 0.4723,
+      "step": 2188
+    },
+    {
+      "epoch": 0.43158517350157727,
+      "grad_norm": 0.6815590189822918,
+      "learning_rate": 1.9440148973339937e-05,
+      "loss": 0.461,
+      "step": 2189
+    },
+    {
+      "epoch": 0.431782334384858,
+      "grad_norm": 0.5608877290281826,
+      "learning_rate": 1.9439637501989984e-05,
+      "loss": 0.4261,
+      "step": 2190
+    },
+    {
+      "epoch": 0.4319794952681388,
+      "grad_norm": 0.5678482544734272,
+      "learning_rate": 1.9439125803845136e-05,
+      "loss": 0.4343,
+      "step": 2191
+    },
+    {
+      "epoch": 0.43217665615141954,
+      "grad_norm": 0.6142946467397039,
+      "learning_rate": 1.9438613878917693e-05,
+      "loss": 0.4446,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4323738170347003,
+      "grad_norm": 0.5554804879141013,
+      "learning_rate": 1.9438101727219946e-05,
+      "loss": 0.4212,
+      "step": 2193
+    },
+    {
+      "epoch": 0.43257097791798105,
+      "grad_norm": 0.7084789316007144,
+      "learning_rate": 1.943758934876421e-05,
+      "loss": 0.4591,
+      "step": 2194
+    },
+    {
+      "epoch": 0.4327681388012618,
+      "grad_norm": 0.615339701668569,
+      "learning_rate": 1.9437076743562785e-05,
+      "loss": 0.4361,
+      "step": 2195
+    },
+    {
+      "epoch": 0.43296529968454256,
+      "grad_norm": 0.790061368854975,
+      "learning_rate": 1.943656391162799e-05,
+      "loss": 0.4531,
+      "step": 2196
+    },
+    {
+      "epoch": 0.43316246056782337,
+      "grad_norm": 0.6198525493844337,
+      "learning_rate": 1.943605085297215e-05,
+      "loss": 0.4438,
+      "step": 2197
+    },
+    {
+      "epoch": 0.4333596214511041,
+      "grad_norm": 0.6215188641437128,
+      "learning_rate": 1.943553756760759e-05,
+      "loss": 0.4106,
+      "step": 2198
+    },
+    {
+      "epoch": 0.4335567823343849,
+      "grad_norm": 0.6051622234791362,
+      "learning_rate": 1.9435024055546644e-05,
+      "loss": 0.4059,
+      "step": 2199
+    },
+    {
+      "epoch": 0.43375394321766564,
+      "grad_norm": 0.6070081944728062,
+      "learning_rate": 1.9434510316801644e-05,
+      "loss": 0.4448,
+      "step": 2200
+    },
+    {
+      "epoch": 0.4339511041009464,
+      "grad_norm": 0.6052231890529391,
+      "learning_rate": 1.9433996351384936e-05,
+      "loss": 0.4059,
+      "step": 2201
+    },
+    {
+      "epoch": 0.43414826498422715,
+      "grad_norm": 0.654268310381098,
+      "learning_rate": 1.943348215930887e-05,
+      "loss": 0.4578,
+      "step": 2202
+    },
+    {
+      "epoch": 0.4343454258675079,
+      "grad_norm": 0.5896345157357851,
+      "learning_rate": 1.9432967740585797e-05,
+      "loss": 0.4017,
+      "step": 2203
+    },
+    {
+      "epoch": 0.43454258675078866,
+      "grad_norm": 0.7026359807455905,
+      "learning_rate": 1.9432453095228078e-05,
+      "loss": 0.4789,
+      "step": 2204
+    },
+    {
+      "epoch": 0.4347397476340694,
+      "grad_norm": 0.9860127509953628,
+      "learning_rate": 1.9431938223248076e-05,
+      "loss": 0.4495,
+      "step": 2205
+    },
+    {
+      "epoch": 0.4349369085173502,
+      "grad_norm": 0.5609344606036918,
+      "learning_rate": 1.9431423124658165e-05,
+      "loss": 0.4384,
+      "step": 2206
+    },
+    {
+      "epoch": 0.43513406940063093,
+      "grad_norm": 0.6369183512758136,
+      "learning_rate": 1.943090779947072e-05,
+      "loss": 0.4943,
+      "step": 2207
+    },
+    {
+      "epoch": 0.4353312302839117,
+      "grad_norm": 0.5765316628970028,
+      "learning_rate": 1.9430392247698117e-05,
+      "loss": 0.4622,
+      "step": 2208
+    },
+    {
+      "epoch": 0.43552839116719244,
+      "grad_norm": 0.5960810357622368,
+      "learning_rate": 1.9429876469352746e-05,
+      "loss": 0.4399,
+      "step": 2209
+    },
+    {
+      "epoch": 0.4357255520504732,
+      "grad_norm": 0.6096812488671478,
+      "learning_rate": 1.9429360464447e-05,
+      "loss": 0.4627,
+      "step": 2210
+    },
+    {
+      "epoch": 0.43592271293375395,
+      "grad_norm": 0.5776557605282492,
+      "learning_rate": 1.9428844232993275e-05,
+      "loss": 0.411,
+      "step": 2211
+    },
+    {
+      "epoch": 0.4361198738170347,
+      "grad_norm": 0.6052522052120832,
+      "learning_rate": 1.9428327775003978e-05,
+      "loss": 0.4448,
+      "step": 2212
+    },
+    {
+      "epoch": 0.43631703470031546,
+      "grad_norm": 0.5994361538832661,
+      "learning_rate": 1.942781109049151e-05,
+      "loss": 0.4331,
+      "step": 2213
+    },
+    {
+      "epoch": 0.4365141955835962,
+      "grad_norm": 0.6941586070053137,
+      "learning_rate": 1.9427294179468287e-05,
+      "loss": 0.457,
+      "step": 2214
+    },
+    {
+      "epoch": 0.436711356466877,
+      "grad_norm": 0.6424583665064718,
+      "learning_rate": 1.942677704194673e-05,
+      "loss": 0.4651,
+      "step": 2215
+    },
+    {
+      "epoch": 0.43690851735015773,
+      "grad_norm": 0.6074306424288837,
+      "learning_rate": 1.9426259677939264e-05,
+      "loss": 0.4518,
+      "step": 2216
+    },
+    {
+      "epoch": 0.4371056782334385,
+      "grad_norm": 0.6404681725403938,
+      "learning_rate": 1.9425742087458318e-05,
+      "loss": 0.4514,
+      "step": 2217
+    },
+    {
+      "epoch": 0.43730283911671924,
+      "grad_norm": 0.656741496330048,
+      "learning_rate": 1.942522427051633e-05,
+      "loss": 0.4355,
+      "step": 2218
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 0.6260232241706599,
+      "learning_rate": 1.942470622712574e-05,
+      "loss": 0.4314,
+      "step": 2219
+    },
+    {
+      "epoch": 0.43769716088328076,
+      "grad_norm": 0.6238147702128097,
+      "learning_rate": 1.942418795729899e-05,
+      "loss": 0.3956,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4378943217665615,
+      "grad_norm": 0.6274596830509961,
+      "learning_rate": 1.9423669461048534e-05,
+      "loss": 0.4483,
+      "step": 2221
+    },
+    {
+      "epoch": 0.43809148264984227,
+      "grad_norm": 0.6495347166026656,
+      "learning_rate": 1.942315073838683e-05,
+      "loss": 0.4945,
+      "step": 2222
+    },
+    {
+      "epoch": 0.438288643533123,
+      "grad_norm": 0.6005997995972219,
+      "learning_rate": 1.942263178932634e-05,
+      "loss": 0.4541,
+      "step": 2223
+    },
+    {
+      "epoch": 0.4384858044164038,
+      "grad_norm": 0.5741164876920474,
+      "learning_rate": 1.942211261387954e-05,
+      "loss": 0.4174,
+      "step": 2224
+    },
+    {
+      "epoch": 0.43868296529968454,
+      "grad_norm": 0.6438713859666909,
+      "learning_rate": 1.9421593212058894e-05,
+      "loss": 0.4801,
+      "step": 2225
+    },
+    {
+      "epoch": 0.4388801261829653,
+      "grad_norm": 0.6248222261459098,
+      "learning_rate": 1.9421073583876882e-05,
+      "loss": 0.4509,
+      "step": 2226
+    },
+    {
+      "epoch": 0.43907728706624605,
+      "grad_norm": 0.6012676799031479,
+      "learning_rate": 1.9420553729345993e-05,
+      "loss": 0.4304,
+      "step": 2227
+    },
+    {
+      "epoch": 0.4392744479495268,
+      "grad_norm": 0.6083239790788654,
+      "learning_rate": 1.942003364847871e-05,
+      "loss": 0.4619,
+      "step": 2228
+    },
+    {
+      "epoch": 0.43947160883280756,
+      "grad_norm": 0.624739792282654,
+      "learning_rate": 1.9419513341287537e-05,
+      "loss": 0.4327,
+      "step": 2229
+    },
+    {
+      "epoch": 0.4396687697160883,
+      "grad_norm": 0.6099088599325223,
+      "learning_rate": 1.9418992807784967e-05,
+      "loss": 0.4361,
+      "step": 2230
+    },
+    {
+      "epoch": 0.43986593059936907,
+      "grad_norm": 0.6184689759512897,
+      "learning_rate": 1.9418472047983512e-05,
+      "loss": 0.4254,
+      "step": 2231
+    },
+    {
+      "epoch": 0.4400630914826498,
+      "grad_norm": 0.6336984491426801,
+      "learning_rate": 1.941795106189568e-05,
+      "loss": 0.4799,
+      "step": 2232
+    },
+    {
+      "epoch": 0.4402602523659306,
+      "grad_norm": 0.6363183171959824,
+      "learning_rate": 1.9417429849533992e-05,
+      "loss": 0.4786,
+      "step": 2233
+    },
+    {
+      "epoch": 0.44045741324921134,
+      "grad_norm": 0.634185491973065,
+      "learning_rate": 1.9416908410910965e-05,
+      "loss": 0.4645,
+      "step": 2234
+    },
+    {
+      "epoch": 0.4406545741324921,
+      "grad_norm": 0.5890796293233364,
+      "learning_rate": 1.941638674603913e-05,
+      "loss": 0.4377,
+      "step": 2235
+    },
+    {
+      "epoch": 0.44085173501577285,
+      "grad_norm": 0.7416948900678109,
+      "learning_rate": 1.9415864854931024e-05,
+      "loss": 0.4953,
+      "step": 2236
+    },
+    {
+      "epoch": 0.4410488958990536,
+      "grad_norm": 0.6100278936853679,
+      "learning_rate": 1.941534273759918e-05,
+      "loss": 0.4839,
+      "step": 2237
+    },
+    {
+      "epoch": 0.44124605678233436,
+      "grad_norm": 0.679877854057984,
+      "learning_rate": 1.9414820394056143e-05,
+      "loss": 0.4574,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4414432176656151,
+      "grad_norm": 0.6337313987110674,
+      "learning_rate": 1.9414297824314466e-05,
+      "loss": 0.4798,
+      "step": 2239
+    },
+    {
+      "epoch": 0.4416403785488959,
+      "grad_norm": 0.602381577533317,
+      "learning_rate": 1.9413775028386702e-05,
+      "loss": 0.4421,
+      "step": 2240
+    },
+    {
+      "epoch": 0.44183753943217663,
+      "grad_norm": 0.7468962543654029,
+      "learning_rate": 1.9413252006285416e-05,
+      "loss": 0.4569,
+      "step": 2241
+    },
+    {
+      "epoch": 0.44203470031545744,
+      "grad_norm": 0.6131581613338967,
+      "learning_rate": 1.9412728758023166e-05,
+      "loss": 0.469,
+      "step": 2242
+    },
+    {
+      "epoch": 0.4422318611987382,
+      "grad_norm": 0.6247638637761316,
+      "learning_rate": 1.9412205283612527e-05,
+      "loss": 0.4668,
+      "step": 2243
+    },
+    {
+      "epoch": 0.44242902208201895,
+      "grad_norm": 0.6074735973961677,
+      "learning_rate": 1.9411681583066077e-05,
+      "loss": 0.4293,
+      "step": 2244
+    },
+    {
+      "epoch": 0.4426261829652997,
+      "grad_norm": 0.5960980998564303,
+      "learning_rate": 1.94111576563964e-05,
+      "loss": 0.3827,
+      "step": 2245
+    },
+    {
+      "epoch": 0.44282334384858046,
+      "grad_norm": 0.5711582764920223,
+      "learning_rate": 1.9410633503616077e-05,
+      "loss": 0.4133,
+      "step": 2246
+    },
+    {
+      "epoch": 0.4430205047318612,
+      "grad_norm": 0.572834729047988,
+      "learning_rate": 1.9410109124737708e-05,
+      "loss": 0.4487,
+      "step": 2247
+    },
+    {
+      "epoch": 0.443217665615142,
+      "grad_norm": 0.5941534079486432,
+      "learning_rate": 1.940958451977389e-05,
+      "loss": 0.439,
+      "step": 2248
+    },
+    {
+      "epoch": 0.44341482649842273,
+      "grad_norm": 0.6057683342246314,
+      "learning_rate": 1.9409059688737226e-05,
+      "loss": 0.466,
+      "step": 2249
+    },
+    {
+      "epoch": 0.4436119873817035,
+      "grad_norm": 0.6128435575114525,
+      "learning_rate": 1.9408534631640328e-05,
+      "loss": 0.4392,
+      "step": 2250
+    },
+    {
+      "epoch": 0.44380914826498424,
+      "grad_norm": 0.593633225129246,
+      "learning_rate": 1.9408009348495808e-05,
+      "loss": 0.4084,
+      "step": 2251
+    },
+    {
+      "epoch": 0.444006309148265,
+      "grad_norm": 0.656165917461389,
+      "learning_rate": 1.9407483839316284e-05,
+      "loss": 0.4554,
+      "step": 2252
+    },
+    {
+      "epoch": 0.44420347003154576,
+      "grad_norm": 0.5874000065946977,
+      "learning_rate": 1.9406958104114387e-05,
+      "loss": 0.4118,
+      "step": 2253
+    },
+    {
+      "epoch": 0.4444006309148265,
+      "grad_norm": 0.6161464173752417,
+      "learning_rate": 1.9406432142902748e-05,
+      "loss": 0.4291,
+      "step": 2254
+    },
+    {
+      "epoch": 0.44459779179810727,
+      "grad_norm": 0.6113342594946223,
+      "learning_rate": 1.9405905955694e-05,
+      "loss": 0.4582,
+      "step": 2255
+    },
+    {
+      "epoch": 0.444794952681388,
+      "grad_norm": 0.6825339340881406,
+      "learning_rate": 1.9405379542500786e-05,
+      "loss": 0.4354,
+      "step": 2256
+    },
+    {
+      "epoch": 0.4449921135646688,
+      "grad_norm": 0.6663414389872298,
+      "learning_rate": 1.9404852903335752e-05,
+      "loss": 0.4717,
+      "step": 2257
+    },
+    {
+      "epoch": 0.44518927444794953,
+      "grad_norm": 0.5510169867193097,
+      "learning_rate": 1.9404326038211558e-05,
+      "loss": 0.397,
+      "step": 2258
+    },
+    {
+      "epoch": 0.4453864353312303,
+      "grad_norm": 0.6117720610593416,
+      "learning_rate": 1.9403798947140857e-05,
+      "loss": 0.4518,
+      "step": 2259
+    },
+    {
+      "epoch": 0.44558359621451105,
+      "grad_norm": 0.6131687458644133,
+      "learning_rate": 1.9403271630136312e-05,
+      "loss": 0.4612,
+      "step": 2260
+    },
+    {
+      "epoch": 0.4457807570977918,
+      "grad_norm": 0.5958608772380412,
+      "learning_rate": 1.9402744087210594e-05,
+      "loss": 0.4574,
+      "step": 2261
+    },
+    {
+      "epoch": 0.44597791798107256,
+      "grad_norm": 0.6397300220383113,
+      "learning_rate": 1.9402216318376377e-05,
+      "loss": 0.4359,
+      "step": 2262
+    },
+    {
+      "epoch": 0.4461750788643533,
+      "grad_norm": 0.6132315582223438,
+      "learning_rate": 1.940168832364634e-05,
+      "loss": 0.4525,
+      "step": 2263
+    },
+    {
+      "epoch": 0.44637223974763407,
+      "grad_norm": 0.6126178978812497,
+      "learning_rate": 1.9401160103033173e-05,
+      "loss": 0.449,
+      "step": 2264
+    },
+    {
+      "epoch": 0.4465694006309148,
+      "grad_norm": 0.6024472555100853,
+      "learning_rate": 1.9400631656549566e-05,
+      "loss": 0.4552,
+      "step": 2265
+    },
+    {
+      "epoch": 0.4467665615141956,
+      "grad_norm": 0.6325409183391781,
+      "learning_rate": 1.9400102984208208e-05,
+      "loss": 0.4723,
+      "step": 2266
+    },
+    {
+      "epoch": 0.44696372239747634,
+      "grad_norm": 0.5774023003574291,
+      "learning_rate": 1.939957408602181e-05,
+      "loss": 0.43,
+      "step": 2267
+    },
+    {
+      "epoch": 0.4471608832807571,
+      "grad_norm": 0.5773695869045778,
+      "learning_rate": 1.939904496200307e-05,
+      "loss": 0.4507,
+      "step": 2268
+    },
+    {
+      "epoch": 0.44735804416403785,
+      "grad_norm": 0.5873923629131955,
+      "learning_rate": 1.939851561216471e-05,
+      "loss": 0.426,
+      "step": 2269
+    },
+    {
+      "epoch": 0.4475552050473186,
+      "grad_norm": 0.6124825928799854,
+      "learning_rate": 1.939798603651944e-05,
+      "loss": 0.4621,
+      "step": 2270
+    },
+    {
+      "epoch": 0.44775236593059936,
+      "grad_norm": 0.6004472428898958,
+      "learning_rate": 1.939745623507999e-05,
+      "loss": 0.4213,
+      "step": 2271
+    },
+    {
+      "epoch": 0.4479495268138801,
+      "grad_norm": 0.584595466954028,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 0.4066,
+      "step": 2272
+    },
+    {
+      "epoch": 0.4481466876971609,
+      "grad_norm": 0.6100782446987316,
+      "learning_rate": 1.9396395954869463e-05,
+      "loss": 0.4527,
+      "step": 2273
+    },
+    {
+      "epoch": 0.44834384858044163,
+      "grad_norm": 0.5988478613742948,
+      "learning_rate": 1.939586547612386e-05,
+      "loss": 0.4301,
+      "step": 2274
+    },
+    {
+      "epoch": 0.4485410094637224,
+      "grad_norm": 0.5892383164078985,
+      "learning_rate": 1.939533477163502e-05,
+      "loss": 0.4637,
+      "step": 2275
+    },
+    {
+      "epoch": 0.44873817034700314,
+      "grad_norm": 0.6188258893089192,
+      "learning_rate": 1.93948038414157e-05,
+      "loss": 0.4286,
+      "step": 2276
+    },
+    {
+      "epoch": 0.4489353312302839,
+      "grad_norm": 0.5859833886607909,
+      "learning_rate": 1.9394272685478646e-05,
+      "loss": 0.4393,
+      "step": 2277
+    },
+    {
+      "epoch": 0.44913249211356465,
+      "grad_norm": 0.6243646580192782,
+      "learning_rate": 1.9393741303836633e-05,
+      "loss": 0.4472,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4493296529968454,
+      "grad_norm": 0.597880669331921,
+      "learning_rate": 1.9393209696502416e-05,
+      "loss": 0.4429,
+      "step": 2279
+    },
+    {
+      "epoch": 0.44952681388012616,
+      "grad_norm": 0.6092038363303693,
+      "learning_rate": 1.9392677863488773e-05,
+      "loss": 0.4427,
+      "step": 2280
+    },
+    {
+      "epoch": 0.4497239747634069,
+      "grad_norm": 0.5945045969064071,
+      "learning_rate": 1.9392145804808484e-05,
+      "loss": 0.4521,
+      "step": 2281
+    },
+    {
+      "epoch": 0.4499211356466877,
+      "grad_norm": 0.5341262843103938,
+      "learning_rate": 1.939161352047432e-05,
+      "loss": 0.4044,
+      "step": 2282
+    },
+    {
+      "epoch": 0.45011829652996843,
+      "grad_norm": 0.6003095082379514,
+      "learning_rate": 1.9391081010499085e-05,
+      "loss": 0.4407,
+      "step": 2283
+    },
+    {
+      "epoch": 0.4503154574132492,
+      "grad_norm": 0.6015907908122791,
+      "learning_rate": 1.9390548274895563e-05,
+      "loss": 0.4399,
+      "step": 2284
+    },
+    {
+      "epoch": 0.45051261829652994,
+      "grad_norm": 0.5889743466466599,
+      "learning_rate": 1.9390015313676558e-05,
+      "loss": 0.4748,
+      "step": 2285
+    },
+    {
+      "epoch": 0.4507097791798107,
+      "grad_norm": 0.589434443034765,
+      "learning_rate": 1.938948212685487e-05,
+      "loss": 0.4502,
+      "step": 2286
+    },
+    {
+      "epoch": 0.4509069400630915,
+      "grad_norm": 0.5794383012877807,
+      "learning_rate": 1.9388948714443317e-05,
+      "loss": 0.4249,
+      "step": 2287
+    },
+    {
+      "epoch": 0.45110410094637227,
+      "grad_norm": 0.5975560713887517,
+      "learning_rate": 1.938841507645471e-05,
+      "loss": 0.463,
+      "step": 2288
+    },
+    {
+      "epoch": 0.451301261829653,
+      "grad_norm": 0.5692154145280981,
+      "learning_rate": 1.938788121290187e-05,
+      "loss": 0.4282,
+      "step": 2289
+    },
+    {
+      "epoch": 0.4514984227129338,
+      "grad_norm": 0.602494745609716,
+      "learning_rate": 1.938734712379762e-05,
+      "loss": 0.4488,
+      "step": 2290
+    },
+    {
+      "epoch": 0.45169558359621453,
+      "grad_norm": 0.587913158355413,
+      "learning_rate": 1.93868128091548e-05,
+      "loss": 0.4679,
+      "step": 2291
+    },
+    {
+      "epoch": 0.4518927444794953,
+      "grad_norm": 0.5718028537813307,
+      "learning_rate": 1.9386278268986243e-05,
+      "loss": 0.4573,
+      "step": 2292
+    },
+    {
+      "epoch": 0.45208990536277605,
+      "grad_norm": 0.6226913772837778,
+      "learning_rate": 1.938574350330479e-05,
+      "loss": 0.4422,
+      "step": 2293
+    },
+    {
+      "epoch": 0.4522870662460568,
+      "grad_norm": 0.6822157964341871,
+      "learning_rate": 1.9385208512123293e-05,
+      "loss": 0.4359,
+      "step": 2294
+    },
+    {
+      "epoch": 0.45248422712933756,
+      "grad_norm": 0.631137044281597,
+      "learning_rate": 1.9384673295454603e-05,
+      "loss": 0.446,
+      "step": 2295
+    },
+    {
+      "epoch": 0.4526813880126183,
+      "grad_norm": 0.6264948809573254,
+      "learning_rate": 1.9384137853311576e-05,
+      "loss": 0.4355,
+      "step": 2296
+    },
+    {
+      "epoch": 0.45287854889589907,
+      "grad_norm": 0.566650305099432,
+      "learning_rate": 1.9383602185707082e-05,
+      "loss": 0.4404,
+      "step": 2297
+    },
+    {
+      "epoch": 0.4530757097791798,
+      "grad_norm": 0.6365696893318133,
+      "learning_rate": 1.938306629265399e-05,
+      "loss": 0.4407,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4532728706624606,
+      "grad_norm": 0.6109514828404591,
+      "learning_rate": 1.9382530174165175e-05,
+      "loss": 0.4282,
+      "step": 2299
+    },
+    {
+      "epoch": 0.45347003154574134,
+      "grad_norm": 0.5636586001477206,
+      "learning_rate": 1.9381993830253515e-05,
+      "loss": 0.4267,
+      "step": 2300
+    },
+    {
+      "epoch": 0.4536671924290221,
+      "grad_norm": 0.705135133601523,
+      "learning_rate": 1.93814572609319e-05,
+      "loss": 0.4815,
+      "step": 2301
+    },
+    {
+      "epoch": 0.45386435331230285,
+      "grad_norm": 0.6238025886975589,
+      "learning_rate": 1.9380920466213217e-05,
+      "loss": 0.4439,
+      "step": 2302
+    },
+    {
+      "epoch": 0.4540615141955836,
+      "grad_norm": 0.7081859082809666,
+      "learning_rate": 1.9380383446110368e-05,
+      "loss": 0.4483,
+      "step": 2303
+    },
+    {
+      "epoch": 0.45425867507886436,
+      "grad_norm": 0.6531999149924236,
+      "learning_rate": 1.937984620063625e-05,
+      "loss": 0.4331,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4544558359621451,
+      "grad_norm": 0.578550340908979,
+      "learning_rate": 1.9379308729803775e-05,
+      "loss": 0.3904,
+      "step": 2305
+    },
+    {
+      "epoch": 0.45465299684542587,
+      "grad_norm": 0.638465715516517,
+      "learning_rate": 1.9378771033625855e-05,
+      "loss": 0.4311,
+      "step": 2306
+    },
+    {
+      "epoch": 0.45485015772870663,
+      "grad_norm": 1.0881560912923343,
+      "learning_rate": 1.9378233112115406e-05,
+      "loss": 0.4634,
+      "step": 2307
+    },
+    {
+      "epoch": 0.4550473186119874,
+      "grad_norm": 0.6713906883533458,
+      "learning_rate": 1.9377694965285356e-05,
+      "loss": 0.4347,
+      "step": 2308
+    },
+    {
+      "epoch": 0.45524447949526814,
+      "grad_norm": 0.6120327107112462,
+      "learning_rate": 1.9377156593148632e-05,
+      "loss": 0.4135,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4554416403785489,
+      "grad_norm": 0.6652885718648691,
+      "learning_rate": 1.937661799571817e-05,
+      "loss": 0.4182,
+      "step": 2310
+    },
+    {
+      "epoch": 0.45563880126182965,
+      "grad_norm": 0.8090629623799119,
+      "learning_rate": 1.937607917300691e-05,
+      "loss": 0.4534,
+      "step": 2311
+    },
+    {
+      "epoch": 0.4558359621451104,
+      "grad_norm": 1.1834526081440842,
+      "learning_rate": 1.9375540125027796e-05,
+      "loss": 0.4866,
+      "step": 2312
+    },
+    {
+      "epoch": 0.45603312302839116,
+      "grad_norm": 0.6151631247816741,
+      "learning_rate": 1.937500085179378e-05,
+      "loss": 0.4517,
+      "step": 2313
+    },
+    {
+      "epoch": 0.4562302839116719,
+      "grad_norm": 1.0542335058904364,
+      "learning_rate": 1.937446135331782e-05,
+      "loss": 0.4423,
+      "step": 2314
+    },
+    {
+      "epoch": 0.4564274447949527,
+      "grad_norm": 0.6255363893355489,
+      "learning_rate": 1.9373921629612876e-05,
+      "loss": 0.46,
+      "step": 2315
+    },
+    {
+      "epoch": 0.45662460567823343,
+      "grad_norm": 0.5392418471066178,
+      "learning_rate": 1.937338168069192e-05,
+      "loss": 0.3992,
+      "step": 2316
+    },
+    {
+      "epoch": 0.4568217665615142,
+      "grad_norm": 0.7140904386654281,
+      "learning_rate": 1.9372841506567916e-05,
+      "loss": 0.4168,
+      "step": 2317
+    },
+    {
+      "epoch": 0.45701892744479494,
+      "grad_norm": 0.8124956257464673,
+      "learning_rate": 1.937230110725385e-05,
+      "loss": 0.4444,
+      "step": 2318
+    },
+    {
+      "epoch": 0.4572160883280757,
+      "grad_norm": 0.5825397168265147,
+      "learning_rate": 1.93717604827627e-05,
+      "loss": 0.4125,
+      "step": 2319
+    },
+    {
+      "epoch": 0.45741324921135645,
+      "grad_norm": 0.5926863567356023,
+      "learning_rate": 1.937121963310746e-05,
+      "loss": 0.4414,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4576104100946372,
+      "grad_norm": 0.649677750333782,
+      "learning_rate": 1.9370678558301117e-05,
+      "loss": 0.4315,
+      "step": 2321
+    },
+    {
+      "epoch": 0.45780757097791797,
+      "grad_norm": 1.1050583010405093,
+      "learning_rate": 1.937013725835668e-05,
+      "loss": 0.4948,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4580047318611987,
+      "grad_norm": 0.6313289986933639,
+      "learning_rate": 1.9369595733287147e-05,
+      "loss": 0.4265,
+      "step": 2323
+    },
+    {
+      "epoch": 0.4582018927444795,
+      "grad_norm": 0.6683692439342528,
+      "learning_rate": 1.9369053983105533e-05,
+      "loss": 0.4648,
+      "step": 2324
+    },
+    {
+      "epoch": 0.45839905362776023,
+      "grad_norm": 0.7637595774268595,
+      "learning_rate": 1.9368512007824852e-05,
+      "loss": 0.4634,
+      "step": 2325
+    },
+    {
+      "epoch": 0.458596214511041,
+      "grad_norm": 0.634965367240232,
+      "learning_rate": 1.9367969807458125e-05,
+      "loss": 0.4925,
+      "step": 2326
+    },
+    {
+      "epoch": 0.45879337539432175,
+      "grad_norm": 0.6164240403541263,
+      "learning_rate": 1.936742738201838e-05,
+      "loss": 0.4655,
+      "step": 2327
+    },
+    {
+      "epoch": 0.4589905362776025,
+      "grad_norm": 0.6481028520735399,
+      "learning_rate": 1.9366884731518648e-05,
+      "loss": 0.4842,
+      "step": 2328
+    },
+    {
+      "epoch": 0.45918769716088326,
+      "grad_norm": 0.6211874377041051,
+      "learning_rate": 1.9366341855971967e-05,
+      "loss": 0.468,
+      "step": 2329
+    },
+    {
+      "epoch": 0.459384858044164,
+      "grad_norm": 0.7567018049090279,
+      "learning_rate": 1.936579875539138e-05,
+      "loss": 0.4688,
+      "step": 2330
+    },
+    {
+      "epoch": 0.45958201892744477,
+      "grad_norm": 0.6196167687850394,
+      "learning_rate": 1.9365255429789934e-05,
+      "loss": 0.4041,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4597791798107255,
+      "grad_norm": 0.5842148695471889,
+      "learning_rate": 1.9364711879180688e-05,
+      "loss": 0.4289,
+      "step": 2332
+    },
+    {
+      "epoch": 0.45997634069400634,
+      "grad_norm": 0.6604436365980313,
+      "learning_rate": 1.9364168103576696e-05,
+      "loss": 0.4489,
+      "step": 2333
+    },
+    {
+      "epoch": 0.4601735015772871,
+      "grad_norm": 1.3308649937601518,
+      "learning_rate": 1.9363624102991022e-05,
+      "loss": 0.451,
+      "step": 2334
+    },
+    {
+      "epoch": 0.46037066246056785,
+      "grad_norm": 0.6093584018980908,
+      "learning_rate": 1.9363079877436744e-05,
+      "loss": 0.4565,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4605678233438486,
+      "grad_norm": 0.7358912383290898,
+      "learning_rate": 1.936253542692693e-05,
+      "loss": 0.4456,
+      "step": 2336
+    },
+    {
+      "epoch": 0.46076498422712936,
+      "grad_norm": 0.5582505407627746,
+      "learning_rate": 1.936199075147466e-05,
+      "loss": 0.3969,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4609621451104101,
+      "grad_norm": 0.6033721253583829,
+      "learning_rate": 1.936144585109302e-05,
+      "loss": 0.4284,
+      "step": 2338
+    },
+    {
+      "epoch": 0.46115930599369087,
+      "grad_norm": 0.6175274199097704,
+      "learning_rate": 1.9360900725795112e-05,
+      "loss": 0.4613,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4613564668769716,
+      "grad_norm": 0.5765237098211964,
+      "learning_rate": 1.9360355375594025e-05,
+      "loss": 0.4307,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4615536277602524,
+      "grad_norm": 0.7141205911743096,
+      "learning_rate": 1.9359809800502858e-05,
+      "loss": 0.4762,
+      "step": 2341
+    },
+    {
+      "epoch": 0.46175078864353314,
+      "grad_norm": 0.8234992518594635,
+      "learning_rate": 1.9359264000534726e-05,
+      "loss": 0.4912,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4619479495268139,
+      "grad_norm": 0.5994202989618423,
+      "learning_rate": 1.9358717975702735e-05,
+      "loss": 0.4382,
+      "step": 2343
+    },
+    {
+      "epoch": 0.46214511041009465,
+      "grad_norm": 0.6761671021504989,
+      "learning_rate": 1.9358171726020014e-05,
+      "loss": 0.447,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4623422712933754,
+      "grad_norm": 0.6365526106317112,
+      "learning_rate": 1.9357625251499682e-05,
+      "loss": 0.4319,
+      "step": 2345
+    },
+    {
+      "epoch": 0.46253943217665616,
+      "grad_norm": 0.7520248964390419,
+      "learning_rate": 1.9357078552154864e-05,
+      "loss": 0.4967,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4627365930599369,
+      "grad_norm": 0.6558895684983119,
+      "learning_rate": 1.9356531627998696e-05,
+      "loss": 0.4489,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4629337539432177,
+      "grad_norm": 0.6828662741868297,
+      "learning_rate": 1.9355984479044324e-05,
+      "loss": 0.4477,
+      "step": 2348
+    },
+    {
+      "epoch": 0.46313091482649843,
+      "grad_norm": 0.6163218102500873,
+      "learning_rate": 1.9355437105304893e-05,
+      "loss": 0.4666,
+      "step": 2349
+    },
+    {
+      "epoch": 0.4633280757097792,
+      "grad_norm": 0.629723944986669,
+      "learning_rate": 1.9354889506793548e-05,
+      "loss": 0.4734,
+      "step": 2350
+    },
+    {
+      "epoch": 0.46352523659305994,
+      "grad_norm": 0.5904354659114018,
+      "learning_rate": 1.935434168352345e-05,
+      "loss": 0.4522,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4637223974763407,
+      "grad_norm": 0.6172193362795846,
+      "learning_rate": 1.935379363550776e-05,
+      "loss": 0.4478,
+      "step": 2352
+    },
+    {
+      "epoch": 0.46391955835962145,
+      "grad_norm": 0.6097296169978487,
+      "learning_rate": 1.9353245362759647e-05,
+      "loss": 0.4407,
+      "step": 2353
+    },
+    {
+      "epoch": 0.4641167192429022,
+      "grad_norm": 0.6073643122176055,
+      "learning_rate": 1.9352696865292278e-05,
+      "loss": 0.4116,
+      "step": 2354
+    },
+    {
+      "epoch": 0.46431388012618297,
+      "grad_norm": 0.6723779390449313,
+      "learning_rate": 1.935214814311884e-05,
+      "loss": 0.4183,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4645110410094637,
+      "grad_norm": 0.5416360235400502,
+      "learning_rate": 1.935159919625251e-05,
+      "loss": 0.4128,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4647082018927445,
+      "grad_norm": 0.6434910428180349,
+      "learning_rate": 1.9351050024706476e-05,
+      "loss": 0.3999,
+      "step": 2357
+    },
+    {
+      "epoch": 0.46490536277602523,
+      "grad_norm": 0.5545962685011621,
+      "learning_rate": 1.9350500628493938e-05,
+      "loss": 0.4059,
+      "step": 2358
+    },
+    {
+      "epoch": 0.465102523659306,
+      "grad_norm": 0.6598887221595904,
+      "learning_rate": 1.9349951007628093e-05,
+      "loss": 0.449,
+      "step": 2359
+    },
+    {
+      "epoch": 0.46529968454258674,
+      "grad_norm": 0.6254882672594428,
+      "learning_rate": 1.934940116212214e-05,
+      "loss": 0.4422,
+      "step": 2360
+    },
+    {
+      "epoch": 0.4654968454258675,
+      "grad_norm": 0.6643297973480146,
+      "learning_rate": 1.93488510919893e-05,
+      "loss": 0.4693,
+      "step": 2361
+    },
+    {
+      "epoch": 0.46569400630914826,
+      "grad_norm": 0.6671960843874735,
+      "learning_rate": 1.9348300797242784e-05,
+      "loss": 0.4674,
+      "step": 2362
+    },
+    {
+      "epoch": 0.465891167192429,
+      "grad_norm": 0.6583457703373533,
+      "learning_rate": 1.934775027789581e-05,
+      "loss": 0.4454,
+      "step": 2363
+    },
+    {
+      "epoch": 0.46608832807570977,
+      "grad_norm": 0.6019895419204817,
+      "learning_rate": 1.934719953396161e-05,
+      "loss": 0.4585,
+      "step": 2364
+    },
+    {
+      "epoch": 0.4662854889589905,
+      "grad_norm": 0.5539573629879283,
+      "learning_rate": 1.9346648565453412e-05,
+      "loss": 0.3792,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4664826498422713,
+      "grad_norm": 0.5851973307214345,
+      "learning_rate": 1.934609737238446e-05,
+      "loss": 0.4116,
+      "step": 2366
+    },
+    {
+      "epoch": 0.46667981072555204,
+      "grad_norm": 0.5995849117001575,
+      "learning_rate": 1.9345545954767985e-05,
+      "loss": 0.4468,
+      "step": 2367
+    },
+    {
+      "epoch": 0.4668769716088328,
+      "grad_norm": 0.713328117155134,
+      "learning_rate": 1.934499431261725e-05,
+      "loss": 0.4863,
+      "step": 2368
+    },
+    {
+      "epoch": 0.46707413249211355,
+      "grad_norm": 0.6211541889706627,
+      "learning_rate": 1.93444424459455e-05,
+      "loss": 0.4355,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4672712933753943,
+      "grad_norm": 0.5811288166838467,
+      "learning_rate": 1.934389035476599e-05,
+      "loss": 0.4487,
+      "step": 2370
+    },
+    {
+      "epoch": 0.46746845425867506,
+      "grad_norm": 0.7101811484585888,
+      "learning_rate": 1.9343338039091992e-05,
+      "loss": 0.4968,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4676656151419558,
+      "grad_norm": 0.6800579406115673,
+      "learning_rate": 1.9342785498936775e-05,
+      "loss": 0.4606,
+      "step": 2372
+    },
+    {
+      "epoch": 0.46786277602523657,
+      "grad_norm": 0.6222849178571329,
+      "learning_rate": 1.934223273431361e-05,
+      "loss": 0.4408,
+      "step": 2373
+    },
+    {
+      "epoch": 0.4680599369085173,
+      "grad_norm": 0.6463617017560804,
+      "learning_rate": 1.9341679745235783e-05,
+      "loss": 0.3811,
+      "step": 2374
+    },
+    {
+      "epoch": 0.4682570977917981,
+      "grad_norm": 0.6104669769368634,
+      "learning_rate": 1.9341126531716575e-05,
+      "loss": 0.4963,
+      "step": 2375
+    },
+    {
+      "epoch": 0.46845425867507884,
+      "grad_norm": 0.6755247672185276,
+      "learning_rate": 1.934057309376928e-05,
+      "loss": 0.4448,
+      "step": 2376
+    },
+    {
+      "epoch": 0.4686514195583596,
+      "grad_norm": 0.5842414289182047,
+      "learning_rate": 1.93400194314072e-05,
+      "loss": 0.4339,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4688485804416404,
+      "grad_norm": 0.6069017142653811,
+      "learning_rate": 1.9339465544643623e-05,
+      "loss": 0.4667,
+      "step": 2378
+    },
+    {
+      "epoch": 0.46904574132492116,
+      "grad_norm": 0.7521849096494124,
+      "learning_rate": 1.9338911433491868e-05,
+      "loss": 0.4821,
+      "step": 2379
+    },
+    {
+      "epoch": 0.4692429022082019,
+      "grad_norm": 0.6450195680499727,
+      "learning_rate": 1.933835709796525e-05,
+      "loss": 0.451,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4694400630914827,
+      "grad_norm": 0.6039688998164772,
+      "learning_rate": 1.933780253807708e-05,
+      "loss": 0.4661,
+      "step": 2381
+    },
+    {
+      "epoch": 0.46963722397476343,
+      "grad_norm": 0.5834276251343435,
+      "learning_rate": 1.933724775384068e-05,
+      "loss": 0.4327,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4698343848580442,
+      "grad_norm": 0.6142679272929852,
+      "learning_rate": 1.9336692745269388e-05,
+      "loss": 0.4644,
+      "step": 2383
+    },
+    {
+      "epoch": 0.47003154574132494,
+      "grad_norm": 0.6124895653388481,
+      "learning_rate": 1.9336137512376532e-05,
+      "loss": 0.4099,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4702287066246057,
+      "grad_norm": 0.6384760441175886,
+      "learning_rate": 1.9335582055175454e-05,
+      "loss": 0.4585,
+      "step": 2385
+    },
+    {
+      "epoch": 0.47042586750788645,
+      "grad_norm": 0.6605064367442711,
+      "learning_rate": 1.9335026373679503e-05,
+      "loss": 0.4624,
+      "step": 2386
+    },
+    {
+      "epoch": 0.4706230283911672,
+      "grad_norm": 0.9621836476711323,
+      "learning_rate": 1.9334470467902024e-05,
+      "loss": 0.4828,
+      "step": 2387
+    },
+    {
+      "epoch": 0.47082018927444796,
+      "grad_norm": 0.6671978180371507,
+      "learning_rate": 1.9333914337856373e-05,
+      "loss": 0.4651,
+      "step": 2388
+    },
+    {
+      "epoch": 0.4710173501577287,
+      "grad_norm": 0.6234859549259715,
+      "learning_rate": 1.933335798355591e-05,
+      "loss": 0.4176,
+      "step": 2389
+    },
+    {
+      "epoch": 0.4712145110410095,
+      "grad_norm": 0.7746006898875549,
+      "learning_rate": 1.9332801405014013e-05,
+      "loss": 0.4175,
+      "step": 2390
+    },
+    {
+      "epoch": 0.47141167192429023,
+      "grad_norm": 0.6524991860281778,
+      "learning_rate": 1.9332244602244042e-05,
+      "loss": 0.4798,
+      "step": 2391
+    },
+    {
+      "epoch": 0.471608832807571,
+      "grad_norm": 0.6671756189730206,
+      "learning_rate": 1.9331687575259378e-05,
+      "loss": 0.4623,
+      "step": 2392
+    },
+    {
+      "epoch": 0.47180599369085174,
+      "grad_norm": 0.5980869066369106,
+      "learning_rate": 1.933113032407341e-05,
+      "loss": 0.3908,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4720031545741325,
+      "grad_norm": 0.6523513564819309,
+      "learning_rate": 1.933057284869952e-05,
+      "loss": 0.4516,
+      "step": 2394
+    },
+    {
+      "epoch": 0.47220031545741326,
+      "grad_norm": 0.6761092970782154,
+      "learning_rate": 1.93300151491511e-05,
+      "loss": 0.4672,
+      "step": 2395
+    },
+    {
+      "epoch": 0.472397476340694,
+      "grad_norm": 0.6145122564628309,
+      "learning_rate": 1.9329457225441554e-05,
+      "loss": 0.4365,
+      "step": 2396
+    },
+    {
+      "epoch": 0.47259463722397477,
+      "grad_norm": 0.9246573499880514,
+      "learning_rate": 1.932889907758429e-05,
+      "loss": 0.4831,
+      "step": 2397
+    },
+    {
+      "epoch": 0.4727917981072555,
+      "grad_norm": 3.0455979525914487,
+      "learning_rate": 1.9328340705592708e-05,
+      "loss": 0.4655,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4729889589905363,
+      "grad_norm": 3.2802273681352085,
+      "learning_rate": 1.932778210948023e-05,
+      "loss": 0.4443,
+      "step": 2399
+    },
+    {
+      "epoch": 0.47318611987381703,
+      "grad_norm": 0.715855291735479,
+      "learning_rate": 1.9327223289260274e-05,
+      "loss": 0.4375,
+      "step": 2400
+    },
+    {
+      "epoch": 0.4733832807570978,
+      "grad_norm": 0.7740268211710657,
+      "learning_rate": 1.932666424494627e-05,
+      "loss": 0.4653,
+      "step": 2401
+    },
+    {
+      "epoch": 0.47358044164037855,
+      "grad_norm": 0.6204311682591072,
+      "learning_rate": 1.9326104976551643e-05,
+      "loss": 0.4134,
+      "step": 2402
+    },
+    {
+      "epoch": 0.4737776025236593,
+      "grad_norm": 0.7091391867960442,
+      "learning_rate": 1.932554548408984e-05,
+      "loss": 0.449,
+      "step": 2403
+    },
+    {
+      "epoch": 0.47397476340694006,
+      "grad_norm": 0.7803886372499086,
+      "learning_rate": 1.932498576757429e-05,
+      "loss": 0.4777,
+      "step": 2404
+    },
+    {
+      "epoch": 0.4741719242902208,
+      "grad_norm": 0.716866961037204,
+      "learning_rate": 1.9324425827018452e-05,
+      "loss": 0.4242,
+      "step": 2405
+    },
+    {
+      "epoch": 0.47436908517350157,
+      "grad_norm": 0.6489682185779229,
+      "learning_rate": 1.932386566243577e-05,
+      "loss": 0.4261,
+      "step": 2406
+    },
+    {
+      "epoch": 0.4745662460567823,
+      "grad_norm": 0.6254026998694395,
+      "learning_rate": 1.9323305273839713e-05,
+      "loss": 0.4012,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4747634069400631,
+      "grad_norm": 0.664068211708124,
+      "learning_rate": 1.9322744661243732e-05,
+      "loss": 0.4401,
+      "step": 2408
+    },
+    {
+      "epoch": 0.47496056782334384,
+      "grad_norm": 0.859745025842503,
+      "learning_rate": 1.9322183824661306e-05,
+      "loss": 0.4634,
+      "step": 2409
+    },
+    {
+      "epoch": 0.4751577287066246,
+      "grad_norm": 0.7510084681382883,
+      "learning_rate": 1.932162276410591e-05,
+      "loss": 0.4403,
+      "step": 2410
+    },
+    {
+      "epoch": 0.47535488958990535,
+      "grad_norm": 0.62038420709768,
+      "learning_rate": 1.9321061479591017e-05,
+      "loss": 0.4503,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4755520504731861,
+      "grad_norm": 0.6669051292027413,
+      "learning_rate": 1.9320499971130114e-05,
+      "loss": 0.4232,
+      "step": 2412
+    },
+    {
+      "epoch": 0.47574921135646686,
+      "grad_norm": 0.6402332048996934,
+      "learning_rate": 1.93199382387367e-05,
+      "loss": 0.4647,
+      "step": 2413
+    },
+    {
+      "epoch": 0.4759463722397476,
+      "grad_norm": 0.6407606660148513,
+      "learning_rate": 1.9319376282424255e-05,
+      "loss": 0.415,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4761435331230284,
+      "grad_norm": 0.6662928458835895,
+      "learning_rate": 1.9318814102206296e-05,
+      "loss": 0.4525,
+      "step": 2415
+    },
+    {
+      "epoch": 0.47634069400630913,
+      "grad_norm": 0.66510230765238,
+      "learning_rate": 1.9318251698096322e-05,
+      "loss": 0.4325,
+      "step": 2416
+    },
+    {
+      "epoch": 0.4765378548895899,
+      "grad_norm": 0.8178454623617458,
+      "learning_rate": 1.931768907010785e-05,
+      "loss": 0.448,
+      "step": 2417
+    },
+    {
+      "epoch": 0.47673501577287064,
+      "grad_norm": 0.6008989268966388,
+      "learning_rate": 1.931712621825439e-05,
+      "loss": 0.4561,
+      "step": 2418
+    },
+    {
+      "epoch": 0.4769321766561514,
+      "grad_norm": 0.6058146707232347,
+      "learning_rate": 1.9316563142549475e-05,
+      "loss": 0.4263,
+      "step": 2419
+    },
+    {
+      "epoch": 0.47712933753943215,
+      "grad_norm": 0.6620728520314264,
+      "learning_rate": 1.9315999843006624e-05,
+      "loss": 0.4697,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4773264984227129,
+      "grad_norm": 0.6180305422847591,
+      "learning_rate": 1.9315436319639375e-05,
+      "loss": 0.4495,
+      "step": 2421
+    },
+    {
+      "epoch": 0.47752365930599366,
+      "grad_norm": 0.7048061474007149,
+      "learning_rate": 1.9314872572461265e-05,
+      "loss": 0.4593,
+      "step": 2422
+    },
+    {
+      "epoch": 0.4777208201892745,
+      "grad_norm": 1.0945355318730723,
+      "learning_rate": 1.9314308601485842e-05,
+      "loss": 0.4426,
+      "step": 2423
+    },
+    {
+      "epoch": 0.47791798107255523,
+      "grad_norm": 0.5490282310707831,
+      "learning_rate": 1.9313744406726656e-05,
+      "loss": 0.4061,
+      "step": 2424
+    },
+    {
+      "epoch": 0.478115141955836,
+      "grad_norm": 0.6400659060403348,
+      "learning_rate": 1.931317998819726e-05,
+      "loss": 0.4378,
+      "step": 2425
+    },
+    {
+      "epoch": 0.47831230283911674,
+      "grad_norm": 0.6364254959448777,
+      "learning_rate": 1.931261534591121e-05,
+      "loss": 0.4796,
+      "step": 2426
+    },
+    {
+      "epoch": 0.4785094637223975,
+      "grad_norm": 0.5901543416183237,
+      "learning_rate": 1.9312050479882082e-05,
+      "loss": 0.4326,
+      "step": 2427
+    },
+    {
+      "epoch": 0.47870662460567825,
+      "grad_norm": 0.7359527860806958,
+      "learning_rate": 1.9311485390123442e-05,
+      "loss": 0.5009,
+      "step": 2428
+    },
+    {
+      "epoch": 0.478903785488959,
+      "grad_norm": 0.9323925635649081,
+      "learning_rate": 1.931092007664886e-05,
+      "loss": 0.4826,
+      "step": 2429
+    },
+    {
+      "epoch": 0.47910094637223977,
+      "grad_norm": 0.5787415894460493,
+      "learning_rate": 1.9310354539471935e-05,
+      "loss": 0.4639,
+      "step": 2430
+    },
+    {
+      "epoch": 0.4792981072555205,
+      "grad_norm": 0.6479424198694343,
+      "learning_rate": 1.930978877860624e-05,
+      "loss": 0.4823,
+      "step": 2431
+    },
+    {
+      "epoch": 0.4794952681388013,
+      "grad_norm": 0.5894770732012385,
+      "learning_rate": 1.9309222794065373e-05,
+      "loss": 0.4604,
+      "step": 2432
+    },
+    {
+      "epoch": 0.47969242902208203,
+      "grad_norm": 0.639884540685977,
+      "learning_rate": 1.930865658586293e-05,
+      "loss": 0.4675,
+      "step": 2433
+    },
+    {
+      "epoch": 0.4798895899053628,
+      "grad_norm": 0.6684635586126257,
+      "learning_rate": 1.930809015401252e-05,
+      "loss": 0.4585,
+      "step": 2434
+    },
+    {
+      "epoch": 0.48008675078864355,
+      "grad_norm": 0.6336667692984291,
+      "learning_rate": 1.9307523498527744e-05,
+      "loss": 0.471,
+      "step": 2435
+    },
+    {
+      "epoch": 0.4802839116719243,
+      "grad_norm": 0.6876919208132971,
+      "learning_rate": 1.930695661942222e-05,
+      "loss": 0.4582,
+      "step": 2436
+    },
+    {
+      "epoch": 0.48048107255520506,
+      "grad_norm": 0.5960087219303766,
+      "learning_rate": 1.9306389516709575e-05,
+      "loss": 0.4476,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4806782334384858,
+      "grad_norm": 0.6171728207106327,
+      "learning_rate": 1.9305822190403422e-05,
+      "loss": 0.4362,
+      "step": 2438
+    },
+    {
+      "epoch": 0.48087539432176657,
+      "grad_norm": 0.6355089188560651,
+      "learning_rate": 1.9305254640517398e-05,
+      "loss": 0.4392,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4810725552050473,
+      "grad_norm": 1.1033461138258283,
+      "learning_rate": 1.9304686867065138e-05,
+      "loss": 0.4564,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4812697160883281,
+      "grad_norm": 0.6067878466977591,
+      "learning_rate": 1.9304118870060283e-05,
+      "loss": 0.4234,
+      "step": 2441
+    },
+    {
+      "epoch": 0.48146687697160884,
+      "grad_norm": 0.6684348483494865,
+      "learning_rate": 1.930355064951648e-05,
+      "loss": 0.4558,
+      "step": 2442
+    },
+    {
+      "epoch": 0.4816640378548896,
+      "grad_norm": 0.62445972264211,
+      "learning_rate": 1.930298220544738e-05,
+      "loss": 0.4464,
+      "step": 2443
+    },
+    {
+      "epoch": 0.48186119873817035,
+      "grad_norm": 0.6151015537788406,
+      "learning_rate": 1.9302413537866642e-05,
+      "loss": 0.4514,
+      "step": 2444
+    },
+    {
+      "epoch": 0.4820583596214511,
+      "grad_norm": 0.6241079129743627,
+      "learning_rate": 1.9301844646787927e-05,
+      "loss": 0.4433,
+      "step": 2445
+    },
+    {
+      "epoch": 0.48225552050473186,
+      "grad_norm": 0.7891240381146803,
+      "learning_rate": 1.93012755322249e-05,
+      "loss": 0.4893,
+      "step": 2446
+    },
+    {
+      "epoch": 0.4824526813880126,
+      "grad_norm": 0.6404678342701038,
+      "learning_rate": 1.9300706194191244e-05,
+      "loss": 0.4532,
+      "step": 2447
+    },
+    {
+      "epoch": 0.48264984227129337,
+      "grad_norm": 0.6493424413450147,
+      "learning_rate": 1.930013663270063e-05,
+      "loss": 0.4498,
+      "step": 2448
+    },
+    {
+      "epoch": 0.48284700315457413,
+      "grad_norm": 0.6508616343699691,
+      "learning_rate": 1.929956684776674e-05,
+      "loss": 0.487,
+      "step": 2449
+    },
+    {
+      "epoch": 0.4830441640378549,
+      "grad_norm": 0.6434094284010808,
+      "learning_rate": 1.929899683940327e-05,
+      "loss": 0.4437,
+      "step": 2450
+    },
+    {
+      "epoch": 0.48324132492113564,
+      "grad_norm": 0.6735775372022499,
+      "learning_rate": 1.9298426607623915e-05,
+      "loss": 0.4301,
+      "step": 2451
+    },
+    {
+      "epoch": 0.4834384858044164,
+      "grad_norm": 0.6116434104938371,
+      "learning_rate": 1.929785615244237e-05,
+      "loss": 0.433,
+      "step": 2452
+    },
+    {
+      "epoch": 0.48363564668769715,
+      "grad_norm": 0.6272151580269819,
+      "learning_rate": 1.9297285473872343e-05,
+      "loss": 0.4447,
+      "step": 2453
+    },
+    {
+      "epoch": 0.4838328075709779,
+      "grad_norm": 0.6438884702800719,
+      "learning_rate": 1.929671457192755e-05,
+      "loss": 0.475,
+      "step": 2454
+    },
+    {
+      "epoch": 0.48402996845425866,
+      "grad_norm": 0.7535169667410828,
+      "learning_rate": 1.9296143446621697e-05,
+      "loss": 0.4784,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4842271293375394,
+      "grad_norm": 44.84808687603809,
+      "learning_rate": 1.9295572097968514e-05,
+      "loss": 0.7183,
+      "step": 2456
+    },
+    {
+      "epoch": 0.4844242902208202,
+      "grad_norm": 0.6854267912996111,
+      "learning_rate": 1.9295000525981725e-05,
+      "loss": 0.4581,
+      "step": 2457
+    },
+    {
+      "epoch": 0.48462145110410093,
+      "grad_norm": 0.7188232334712162,
+      "learning_rate": 1.929442873067506e-05,
+      "loss": 0.4354,
+      "step": 2458
+    },
+    {
+      "epoch": 0.4848186119873817,
+      "grad_norm": 9.85290231143863,
+      "learning_rate": 1.9293856712062267e-05,
+      "loss": 0.5323,
+      "step": 2459
+    },
+    {
+      "epoch": 0.48501577287066244,
+      "grad_norm": 0.6382321261076438,
+      "learning_rate": 1.9293284470157082e-05,
+      "loss": 0.4332,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4852129337539432,
+      "grad_norm": 0.8581690728691596,
+      "learning_rate": 1.9292712004973248e-05,
+      "loss": 0.4314,
+      "step": 2461
+    },
+    {
+      "epoch": 0.48541009463722395,
+      "grad_norm": 0.6494937725982591,
+      "learning_rate": 1.9292139316524528e-05,
+      "loss": 0.4891,
+      "step": 2462
+    },
+    {
+      "epoch": 0.4856072555205047,
+      "grad_norm": 0.6193320454689136,
+      "learning_rate": 1.9291566404824676e-05,
+      "loss": 0.4077,
+      "step": 2463
+    },
+    {
+      "epoch": 0.48580441640378547,
+      "grad_norm": 0.6374169775943455,
+      "learning_rate": 1.9290993269887458e-05,
+      "loss": 0.4282,
+      "step": 2464
+    },
+    {
+      "epoch": 0.4860015772870662,
+      "grad_norm": 0.6787679196471768,
+      "learning_rate": 1.9290419911726647e-05,
+      "loss": 0.4902,
+      "step": 2465
+    },
+    {
+      "epoch": 0.486198738170347,
+      "grad_norm": 0.6598399921444862,
+      "learning_rate": 1.9289846330356018e-05,
+      "loss": 0.4991,
+      "step": 2466
+    },
+    {
+      "epoch": 0.48639589905362773,
+      "grad_norm": 0.6206359700753303,
+      "learning_rate": 1.9289272525789348e-05,
+      "loss": 0.4457,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4865930599369085,
+      "grad_norm": 0.5930790685776189,
+      "learning_rate": 1.9288698498040423e-05,
+      "loss": 0.4066,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4867902208201893,
+      "grad_norm": 0.651999060628564,
+      "learning_rate": 1.928812424712304e-05,
+      "loss": 0.4191,
+      "step": 2469
+    },
+    {
+      "epoch": 0.48698738170347006,
+      "grad_norm": 0.615287126559333,
+      "learning_rate": 1.9287549773050988e-05,
+      "loss": 0.4301,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4871845425867508,
+      "grad_norm": 0.7055975170195661,
+      "learning_rate": 1.9286975075838077e-05,
+      "loss": 0.4719,
+      "step": 2471
+    },
+    {
+      "epoch": 0.48738170347003157,
+      "grad_norm": 0.5856890230925806,
+      "learning_rate": 1.9286400155498107e-05,
+      "loss": 0.3913,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4875788643533123,
+      "grad_norm": 0.8762126171240702,
+      "learning_rate": 1.92858250120449e-05,
+      "loss": 0.4059,
+      "step": 2473
+    },
+    {
+      "epoch": 0.4877760252365931,
+      "grad_norm": 0.6128580611588724,
+      "learning_rate": 1.9285249645492266e-05,
+      "loss": 0.4451,
+      "step": 2474
+    },
+    {
+      "epoch": 0.48797318611987384,
+      "grad_norm": 0.803850920215572,
+      "learning_rate": 1.928467405585403e-05,
+      "loss": 0.4162,
+      "step": 2475
+    },
+    {
+      "epoch": 0.4881703470031546,
+      "grad_norm": 0.684760361276952,
+      "learning_rate": 1.9284098243144028e-05,
+      "loss": 0.4731,
+      "step": 2476
+    },
+    {
+      "epoch": 0.48836750788643535,
+      "grad_norm": 0.6490585074155004,
+      "learning_rate": 1.9283522207376088e-05,
+      "loss": 0.4238,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4885646687697161,
+      "grad_norm": 0.6065001605718633,
+      "learning_rate": 1.9282945948564047e-05,
+      "loss": 0.435,
+      "step": 2478
+    },
+    {
+      "epoch": 0.48876182965299686,
+      "grad_norm": 0.6557206451679547,
+      "learning_rate": 1.9282369466721756e-05,
+      "loss": 0.4611,
+      "step": 2479
+    },
+    {
+      "epoch": 0.4889589905362776,
+      "grad_norm": 0.5885897744446494,
+      "learning_rate": 1.9281792761863067e-05,
+      "loss": 0.475,
+      "step": 2480
+    },
+    {
+      "epoch": 0.48915615141955837,
+      "grad_norm": 0.6645075192181713,
+      "learning_rate": 1.928121583400183e-05,
+      "loss": 0.4432,
+      "step": 2481
+    },
+    {
+      "epoch": 0.4893533123028391,
+      "grad_norm": 0.854625778831215,
+      "learning_rate": 1.9280638683151903e-05,
+      "loss": 0.4287,
+      "step": 2482
+    },
+    {
+      "epoch": 0.4895504731861199,
+      "grad_norm": 10.207500341633537,
+      "learning_rate": 1.9280061309327164e-05,
+      "loss": 0.4656,
+      "step": 2483
+    },
+    {
+      "epoch": 0.48974763406940064,
+      "grad_norm": 0.8188261610562382,
+      "learning_rate": 1.9279483712541477e-05,
+      "loss": 0.4747,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4899447949526814,
+      "grad_norm": 0.6708338710083666,
+      "learning_rate": 1.9278905892808725e-05,
+      "loss": 0.466,
+      "step": 2485
+    },
+    {
+      "epoch": 0.49014195583596215,
+      "grad_norm": 0.827156096180289,
+      "learning_rate": 1.9278327850142783e-05,
+      "loss": 0.4859,
+      "step": 2486
+    },
+    {
+      "epoch": 0.4903391167192429,
+      "grad_norm": 0.7039638642215147,
+      "learning_rate": 1.9277749584557543e-05,
+      "loss": 0.457,
+      "step": 2487
+    },
+    {
+      "epoch": 0.49053627760252366,
+      "grad_norm": 0.6238491942723505,
+      "learning_rate": 1.9277171096066895e-05,
+      "loss": 0.4593,
+      "step": 2488
+    },
+    {
+      "epoch": 0.4907334384858044,
+      "grad_norm": 0.6466067939563221,
+      "learning_rate": 1.9276592384684745e-05,
+      "loss": 0.4694,
+      "step": 2489
+    },
+    {
+      "epoch": 0.4909305993690852,
+      "grad_norm": 0.6199822833188464,
+      "learning_rate": 1.9276013450424995e-05,
+      "loss": 0.4572,
+      "step": 2490
+    },
+    {
+      "epoch": 0.49112776025236593,
+      "grad_norm": 1.053077823726099,
+      "learning_rate": 1.9275434293301544e-05,
+      "loss": 0.4559,
+      "step": 2491
+    },
+    {
+      "epoch": 0.4913249211356467,
+      "grad_norm": 0.6349781493349106,
+      "learning_rate": 1.9274854913328317e-05,
+      "loss": 0.5007,
+      "step": 2492
+    },
+    {
+      "epoch": 0.49152208201892744,
+      "grad_norm": 0.616100989000789,
+      "learning_rate": 1.9274275310519234e-05,
+      "loss": 0.417,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4917192429022082,
+      "grad_norm": 0.6027522466238368,
+      "learning_rate": 1.9273695484888216e-05,
+      "loss": 0.454,
+      "step": 2494
+    },
+    {
+      "epoch": 0.49191640378548895,
+      "grad_norm": 0.6388698511914507,
+      "learning_rate": 1.9273115436449198e-05,
+      "loss": 0.4398,
+      "step": 2495
+    },
+    {
+      "epoch": 0.4921135646687697,
+      "grad_norm": 1.088727821676831,
+      "learning_rate": 1.9272535165216112e-05,
+      "loss": 0.4679,
+      "step": 2496
+    },
+    {
+      "epoch": 0.49231072555205047,
+      "grad_norm": 0.6152883070859394,
+      "learning_rate": 1.9271954671202902e-05,
+      "loss": 0.4152,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4925078864353312,
+      "grad_norm": 0.5214691152255505,
+      "learning_rate": 1.9271373954423517e-05,
+      "loss": 0.4056,
+      "step": 2498
+    },
+    {
+      "epoch": 0.492705047318612,
+      "grad_norm": 0.641068964137141,
+      "learning_rate": 1.9270793014891906e-05,
+      "loss": 0.4716,
+      "step": 2499
+    },
+    {
+      "epoch": 0.49290220820189273,
+      "grad_norm": 0.5667713763054852,
+      "learning_rate": 1.9270211852622024e-05,
+      "loss": 0.4235,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4930993690851735,
+      "grad_norm": 0.7075480774876152,
+      "learning_rate": 1.926963046762784e-05,
+      "loss": 0.4775,
+      "step": 2501
+    },
+    {
+      "epoch": 0.49329652996845424,
+      "grad_norm": 0.5892770191488615,
+      "learning_rate": 1.9269048859923318e-05,
+      "loss": 0.4885,
+      "step": 2502
+    },
+    {
+      "epoch": 0.493493690851735,
+      "grad_norm": 0.6792628358438927,
+      "learning_rate": 1.9268467029522432e-05,
+      "loss": 0.4683,
+      "step": 2503
+    },
+    {
+      "epoch": 0.49369085173501576,
+      "grad_norm": 0.637559207810441,
+      "learning_rate": 1.9267884976439163e-05,
+      "loss": 0.4298,
+      "step": 2504
+    },
+    {
+      "epoch": 0.4938880126182965,
+      "grad_norm": 0.5698260135323511,
+      "learning_rate": 1.9267302700687494e-05,
+      "loss": 0.4026,
+      "step": 2505
+    },
+    {
+      "epoch": 0.49408517350157727,
+      "grad_norm": 0.5949689949186419,
+      "learning_rate": 1.9266720202281413e-05,
+      "loss": 0.4076,
+      "step": 2506
+    },
+    {
+      "epoch": 0.494282334384858,
+      "grad_norm": 0.6042856982676498,
+      "learning_rate": 1.9266137481234918e-05,
+      "loss": 0.4727,
+      "step": 2507
+    },
+    {
+      "epoch": 0.4944794952681388,
+      "grad_norm": 0.5738882939815504,
+      "learning_rate": 1.9265554537562008e-05,
+      "loss": 0.4585,
+      "step": 2508
+    },
+    {
+      "epoch": 0.49467665615141954,
+      "grad_norm": 1.6832821361365304,
+      "learning_rate": 1.926497137127669e-05,
+      "loss": 0.435,
+      "step": 2509
+    },
+    {
+      "epoch": 0.4948738170347003,
+      "grad_norm": 0.7344488196925597,
+      "learning_rate": 1.9264387982392972e-05,
+      "loss": 0.4663,
+      "step": 2510
+    },
+    {
+      "epoch": 0.49507097791798105,
+      "grad_norm": 0.8273300571453689,
+      "learning_rate": 1.926380437092487e-05,
+      "loss": 0.4682,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4952681388012618,
+      "grad_norm": 0.6195232818493738,
+      "learning_rate": 1.9263220536886413e-05,
+      "loss": 0.4296,
+      "step": 2512
+    },
+    {
+      "epoch": 0.49546529968454256,
+      "grad_norm": 0.6274418197133689,
+      "learning_rate": 1.9262636480291618e-05,
+      "loss": 0.4445,
+      "step": 2513
+    },
+    {
+      "epoch": 0.49566246056782337,
+      "grad_norm": 0.6220175805229026,
+      "learning_rate": 1.9262052201154525e-05,
+      "loss": 0.4491,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4958596214511041,
+      "grad_norm": 1.2792193280662592,
+      "learning_rate": 1.926146769948917e-05,
+      "loss": 0.441,
+      "step": 2515
+    },
+    {
+      "epoch": 0.4960567823343849,
+      "grad_norm": 0.6309907361680024,
+      "learning_rate": 1.926088297530959e-05,
+      "loss": 0.4213,
+      "step": 2516
+    },
+    {
+      "epoch": 0.49625394321766564,
+      "grad_norm": 0.6327839276931332,
+      "learning_rate": 1.9260298028629846e-05,
+      "loss": 0.4576,
+      "step": 2517
+    },
+    {
+      "epoch": 0.4964511041009464,
+      "grad_norm": 0.6462083871492283,
+      "learning_rate": 1.925971285946398e-05,
+      "loss": 0.4825,
+      "step": 2518
+    },
+    {
+      "epoch": 0.49664826498422715,
+      "grad_norm": 0.659095376571992,
+      "learning_rate": 1.9259127467826055e-05,
+      "loss": 0.4919,
+      "step": 2519
+    },
+    {
+      "epoch": 0.4968454258675079,
+      "grad_norm": 0.7047541361963703,
+      "learning_rate": 1.925854185373014e-05,
+      "loss": 0.4622,
+      "step": 2520
+    },
+    {
+      "epoch": 0.49704258675078866,
+      "grad_norm": 0.6556353060348127,
+      "learning_rate": 1.9257956017190297e-05,
+      "loss": 0.4269,
+      "step": 2521
+    },
+    {
+      "epoch": 0.4972397476340694,
+      "grad_norm": 1.249097804465773,
+      "learning_rate": 1.9257369958220612e-05,
+      "loss": 0.4644,
+      "step": 2522
+    },
+    {
+      "epoch": 0.4974369085173502,
+      "grad_norm": 0.5862685793054593,
+      "learning_rate": 1.9256783676835153e-05,
+      "loss": 0.4157,
+      "step": 2523
+    },
+    {
+      "epoch": 0.49763406940063093,
+      "grad_norm": 0.5843281300414458,
+      "learning_rate": 1.9256197173048013e-05,
+      "loss": 0.4691,
+      "step": 2524
+    },
+    {
+      "epoch": 0.4978312302839117,
+      "grad_norm": 0.642053300811986,
+      "learning_rate": 1.925561044687328e-05,
+      "loss": 0.4671,
+      "step": 2525
+    },
+    {
+      "epoch": 0.49802839116719244,
+      "grad_norm": 2.457307700484372,
+      "learning_rate": 1.9255023498325055e-05,
+      "loss": 0.4732,
+      "step": 2526
+    },
+    {
+      "epoch": 0.4982255520504732,
+      "grad_norm": 1.0179133708857775,
+      "learning_rate": 1.9254436327417436e-05,
+      "loss": 0.4766,
+      "step": 2527
+    },
+    {
+      "epoch": 0.49842271293375395,
+      "grad_norm": 0.5737583222016323,
+      "learning_rate": 1.9253848934164533e-05,
+      "loss": 0.4076,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4986198738170347,
+      "grad_norm": 1.222029502495448,
+      "learning_rate": 1.9253261318580456e-05,
+      "loss": 0.482,
+      "step": 2529
+    },
+    {
+      "epoch": 0.49881703470031546,
+      "grad_norm": 0.7053057865997824,
+      "learning_rate": 1.9252673480679328e-05,
+      "loss": 0.4837,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4990141955835962,
+      "grad_norm": 0.6259696508511272,
+      "learning_rate": 1.9252085420475263e-05,
+      "loss": 0.4232,
+      "step": 2531
+    },
+    {
+      "epoch": 0.499211356466877,
+      "grad_norm": 0.6853061232406765,
+      "learning_rate": 1.92514971379824e-05,
+      "loss": 0.443,
+      "step": 2532
+    },
+    {
+      "epoch": 0.49940851735015773,
+      "grad_norm": 0.8584950321451298,
+      "learning_rate": 1.9250908633214863e-05,
+      "loss": 0.4483,
+      "step": 2533
+    },
+    {
+      "epoch": 0.4996056782334385,
+      "grad_norm": 0.6391676971123401,
+      "learning_rate": 1.92503199061868e-05,
+      "loss": 0.4797,
+      "step": 2534
+    },
+    {
+      "epoch": 0.49980283911671924,
+      "grad_norm": 0.638850981357959,
+      "learning_rate": 1.924973095691235e-05,
+      "loss": 0.4311,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6527695472648593,
+      "learning_rate": 1.9249141785405666e-05,
+      "loss": 0.4149,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.45128634572029114,
+      "eval_runtime": 344.9468,
+      "eval_samples_per_second": 23.569,
+      "eval_steps_per_second": 1.476,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5001971608832808,
+      "grad_norm": 0.6134815871569128,
+      "learning_rate": 1.9248552391680902e-05,
+      "loss": 0.4394,
+      "step": 2537
+    },
+    {
+      "epoch": 0.5003943217665615,
+      "grad_norm": 0.7190198346074168,
+      "learning_rate": 1.924796277575222e-05,
+      "loss": 0.4758,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5005914826498423,
+      "grad_norm": 0.6176497275780272,
+      "learning_rate": 1.9247372937633785e-05,
+      "loss": 0.4417,
+      "step": 2539
+    },
+    {
+      "epoch": 0.500788643533123,
+      "grad_norm": 0.6368942778751133,
+      "learning_rate": 1.9246782877339767e-05,
+      "loss": 0.4418,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5009858044164038,
+      "grad_norm": 0.6136453962550936,
+      "learning_rate": 1.9246192594884344e-05,
+      "loss": 0.4454,
+      "step": 2541
+    },
+    {
+      "epoch": 0.5011829652996845,
+      "grad_norm": 0.6118109954623908,
+      "learning_rate": 1.9245602090281698e-05,
+      "loss": 0.4704,
+      "step": 2542
+    },
+    {
+      "epoch": 0.5013801261829653,
+      "grad_norm": 0.5608720572816418,
+      "learning_rate": 1.924501136354602e-05,
+      "loss": 0.4331,
+      "step": 2543
+    },
+    {
+      "epoch": 0.501577287066246,
+      "grad_norm": 0.818907256932753,
+      "learning_rate": 1.924442041469149e-05,
+      "loss": 0.4582,
+      "step": 2544
+    },
+    {
+      "epoch": 0.5017744479495269,
+      "grad_norm": 0.5786413688970435,
+      "learning_rate": 1.9243829243732324e-05,
+      "loss": 0.4307,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5019716088328076,
+      "grad_norm": 0.7567124449798754,
+      "learning_rate": 1.924323785068271e-05,
+      "loss": 0.4212,
+      "step": 2546
+    },
+    {
+      "epoch": 0.5021687697160884,
+      "grad_norm": 0.7087458808657718,
+      "learning_rate": 1.9242646235556868e-05,
+      "loss": 0.4734,
+      "step": 2547
+    },
+    {
+      "epoch": 0.5023659305993691,
+      "grad_norm": 0.8128756129153467,
+      "learning_rate": 1.9242054398369005e-05,
+      "loss": 0.4908,
+      "step": 2548
+    },
+    {
+      "epoch": 0.5025630914826499,
+      "grad_norm": 0.7036392962632404,
+      "learning_rate": 1.9241462339133342e-05,
+      "loss": 0.429,
+      "step": 2549
+    },
+    {
+      "epoch": 0.5027602523659306,
+      "grad_norm": 2.7528973749927608,
+      "learning_rate": 1.9240870057864106e-05,
+      "loss": 0.4387,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5029574132492114,
+      "grad_norm": 0.6836200522610701,
+      "learning_rate": 1.9240277554575523e-05,
+      "loss": 0.434,
+      "step": 2551
+    },
+    {
+      "epoch": 0.5031545741324921,
+      "grad_norm": 0.6838558197223942,
+      "learning_rate": 1.923968482928183e-05,
+      "loss": 0.4878,
+      "step": 2552
+    },
+    {
+      "epoch": 0.5033517350157729,
+      "grad_norm": 0.6771965904434417,
+      "learning_rate": 1.9239091881997274e-05,
+      "loss": 0.4582,
+      "step": 2553
+    },
+    {
+      "epoch": 0.5035488958990536,
+      "grad_norm": 0.9137036161581693,
+      "learning_rate": 1.923849871273609e-05,
+      "loss": 0.4471,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5037460567823344,
+      "grad_norm": 0.5838279107533122,
+      "learning_rate": 1.923790532151254e-05,
+      "loss": 0.4392,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5039432176656151,
+      "grad_norm": 0.751287528093079,
+      "learning_rate": 1.9237311708340867e-05,
+      "loss": 0.4608,
+      "step": 2556
+    },
+    {
+      "epoch": 0.5041403785488959,
+      "grad_norm": 0.6438474116366795,
+      "learning_rate": 1.9236717873235347e-05,
+      "loss": 0.4483,
+      "step": 2557
+    },
+    {
+      "epoch": 0.5043375394321766,
+      "grad_norm": 0.6691961023553488,
+      "learning_rate": 1.923612381621024e-05,
+      "loss": 0.4729,
+      "step": 2558
+    },
+    {
+      "epoch": 0.5045347003154574,
+      "grad_norm": 0.8212757423894174,
+      "learning_rate": 1.923552953727982e-05,
+      "loss": 0.3985,
+      "step": 2559
+    },
+    {
+      "epoch": 0.5047318611987381,
+      "grad_norm": 0.8451322507658569,
+      "learning_rate": 1.923493503645837e-05,
+      "loss": 0.4875,
+      "step": 2560
+    },
+    {
+      "epoch": 0.504929022082019,
+      "grad_norm": 1.3003193055502538,
+      "learning_rate": 1.9234340313760163e-05,
+      "loss": 0.4569,
+      "step": 2561
+    },
+    {
+      "epoch": 0.5051261829652997,
+      "grad_norm": 0.6906807412142325,
+      "learning_rate": 1.9233745369199495e-05,
+      "loss": 0.4659,
+      "step": 2562
+    },
+    {
+      "epoch": 0.5053233438485805,
+      "grad_norm": 0.7693812343976348,
+      "learning_rate": 1.923315020279066e-05,
+      "loss": 0.4422,
+      "step": 2563
+    },
+    {
+      "epoch": 0.5055205047318612,
+      "grad_norm": 1.128657931598075,
+      "learning_rate": 1.9232554814547953e-05,
+      "loss": 0.4615,
+      "step": 2564
+    },
+    {
+      "epoch": 0.505717665615142,
+      "grad_norm": 1.1572958591539255,
+      "learning_rate": 1.923195920448569e-05,
+      "loss": 0.4783,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5059148264984227,
+      "grad_norm": 0.698446780450944,
+      "learning_rate": 1.9231363372618165e-05,
+      "loss": 0.4343,
+      "step": 2566
+    },
+    {
+      "epoch": 0.5061119873817035,
+      "grad_norm": 0.5921597967019708,
+      "learning_rate": 1.92307673189597e-05,
+      "loss": 0.3959,
+      "step": 2567
+    },
+    {
+      "epoch": 0.5063091482649842,
+      "grad_norm": 0.7489688974414478,
+      "learning_rate": 1.923017104352462e-05,
+      "loss": 0.4237,
+      "step": 2568
+    },
+    {
+      "epoch": 0.506506309148265,
+      "grad_norm": 0.6464100318912962,
+      "learning_rate": 1.9229574546327247e-05,
+      "loss": 0.4587,
+      "step": 2569
+    },
+    {
+      "epoch": 0.5067034700315457,
+      "grad_norm": 0.5894037282136724,
+      "learning_rate": 1.9228977827381914e-05,
+      "loss": 0.412,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5069006309148265,
+      "grad_norm": 0.5896415150715301,
+      "learning_rate": 1.922838088670296e-05,
+      "loss": 0.4415,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5070977917981072,
+      "grad_norm": 0.6093444816516489,
+      "learning_rate": 1.9227783724304716e-05,
+      "loss": 0.4246,
+      "step": 2572
+    },
+    {
+      "epoch": 0.507294952681388,
+      "grad_norm": 0.6133472401577025,
+      "learning_rate": 1.922718634020154e-05,
+      "loss": 0.4265,
+      "step": 2573
+    },
+    {
+      "epoch": 0.5074921135646687,
+      "grad_norm": 0.6834888429313623,
+      "learning_rate": 1.922658873440778e-05,
+      "loss": 0.4595,
+      "step": 2574
+    },
+    {
+      "epoch": 0.5076892744479495,
+      "grad_norm": 0.6948580579520518,
+      "learning_rate": 1.92259909069378e-05,
+      "loss": 0.448,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5078864353312302,
+      "grad_norm": 0.6374166697259309,
+      "learning_rate": 1.9225392857805955e-05,
+      "loss": 0.436,
+      "step": 2576
+    },
+    {
+      "epoch": 0.508083596214511,
+      "grad_norm": 0.6666234831625328,
+      "learning_rate": 1.922479458702662e-05,
+      "loss": 0.4592,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5082807570977917,
+      "grad_norm": 0.6537577476123761,
+      "learning_rate": 1.9224196094614163e-05,
+      "loss": 0.4444,
+      "step": 2578
+    },
+    {
+      "epoch": 0.5084779179810726,
+      "grad_norm": 0.608893272413261,
+      "learning_rate": 1.9223597380582967e-05,
+      "loss": 0.4611,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5086750788643533,
+      "grad_norm": 0.6235476792381212,
+      "learning_rate": 1.9222998444947417e-05,
+      "loss": 0.4138,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5088722397476341,
+      "grad_norm": 0.6352774634847786,
+      "learning_rate": 1.92223992877219e-05,
+      "loss": 0.4596,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5090694006309149,
+      "grad_norm": 0.7096357346244977,
+      "learning_rate": 1.922179990892082e-05,
+      "loss": 0.4706,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5092665615141956,
+      "grad_norm": 0.9207989672506494,
+      "learning_rate": 1.9221200308558566e-05,
+      "loss": 0.4623,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5094637223974764,
+      "grad_norm": 0.6470580276117848,
+      "learning_rate": 1.922060048664955e-05,
+      "loss": 0.4691,
+      "step": 2584
+    },
+    {
+      "epoch": 0.5096608832807571,
+      "grad_norm": 0.6435644649688526,
+      "learning_rate": 1.9220000443208183e-05,
+      "loss": 0.4633,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5098580441640379,
+      "grad_norm": 0.7356132379703066,
+      "learning_rate": 1.9219400178248876e-05,
+      "loss": 0.4805,
+      "step": 2586
+    },
+    {
+      "epoch": 0.5100552050473186,
+      "grad_norm": 0.7125001237201177,
+      "learning_rate": 1.9218799691786062e-05,
+      "loss": 0.4589,
+      "step": 2587
+    },
+    {
+      "epoch": 0.5102523659305994,
+      "grad_norm": 0.5882472375305573,
+      "learning_rate": 1.9218198983834155e-05,
+      "loss": 0.4088,
+      "step": 2588
+    },
+    {
+      "epoch": 0.5104495268138801,
+      "grad_norm": 0.6522458410768266,
+      "learning_rate": 1.9217598054407598e-05,
+      "loss": 0.4515,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5106466876971609,
+      "grad_norm": 0.6181026041148553,
+      "learning_rate": 1.9216996903520827e-05,
+      "loss": 0.4571,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5108438485804416,
+      "grad_norm": 0.5899550767444578,
+      "learning_rate": 1.9216395531188277e-05,
+      "loss": 0.4303,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5110410094637224,
+      "grad_norm": 0.6129864766903566,
+      "learning_rate": 1.9215793937424404e-05,
+      "loss": 0.4711,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5112381703470031,
+      "grad_norm": 0.687929702723702,
+      "learning_rate": 1.9215192122243663e-05,
+      "loss": 0.4661,
+      "step": 2593
+    },
+    {
+      "epoch": 0.511435331230284,
+      "grad_norm": 0.5550118700233378,
+      "learning_rate": 1.921459008566051e-05,
+      "loss": 0.4122,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5116324921135647,
+      "grad_norm": 0.6517123976025492,
+      "learning_rate": 1.921398782768941e-05,
+      "loss": 0.4537,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5118296529968455,
+      "grad_norm": 0.5998912924516978,
+      "learning_rate": 1.9213385348344827e-05,
+      "loss": 0.4539,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5120268138801262,
+      "grad_norm": 0.6170261164147766,
+      "learning_rate": 1.9212782647641247e-05,
+      "loss": 0.4556,
+      "step": 2597
+    },
+    {
+      "epoch": 0.512223974763407,
+      "grad_norm": 0.6459407251957247,
+      "learning_rate": 1.9212179725593144e-05,
+      "loss": 0.4729,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5124211356466877,
+      "grad_norm": 0.6260843018867657,
+      "learning_rate": 1.9211576582215e-05,
+      "loss": 0.4851,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5126182965299685,
+      "grad_norm": 0.5489077348828413,
+      "learning_rate": 1.921097321752132e-05,
+      "loss": 0.4099,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5128154574132492,
+      "grad_norm": 0.903385996056016,
+      "learning_rate": 1.9210369631526583e-05,
+      "loss": 0.429,
+      "step": 2601
+    },
+    {
+      "epoch": 0.51301261829653,
+      "grad_norm": 0.608980064766385,
+      "learning_rate": 1.9209765824245302e-05,
+      "loss": 0.4643,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5132097791798107,
+      "grad_norm": 0.5824584984075388,
+      "learning_rate": 1.9209161795691975e-05,
+      "loss": 0.4359,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5134069400630915,
+      "grad_norm": 62.5920871913074,
+      "learning_rate": 1.9208557545881127e-05,
+      "loss": 0.7084,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5136041009463722,
+      "grad_norm": 0.7237096419484113,
+      "learning_rate": 1.9207953074827264e-05,
+      "loss": 0.4803,
+      "step": 2605
+    },
+    {
+      "epoch": 0.513801261829653,
+      "grad_norm": 0.6244628305928418,
+      "learning_rate": 1.9207348382544914e-05,
+      "loss": 0.4455,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5139984227129337,
+      "grad_norm": 0.6740117956920548,
+      "learning_rate": 1.9206743469048606e-05,
+      "loss": 0.4243,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5141955835962145,
+      "grad_norm": 0.64353274531661,
+      "learning_rate": 1.920613833435287e-05,
+      "loss": 0.4422,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5143927444794952,
+      "grad_norm": 0.6873646678800827,
+      "learning_rate": 1.920553297847225e-05,
+      "loss": 0.4708,
+      "step": 2609
+    },
+    {
+      "epoch": 0.514589905362776,
+      "grad_norm": 0.8576563607865519,
+      "learning_rate": 1.9204927401421284e-05,
+      "loss": 0.4518,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5147870662460567,
+      "grad_norm": 0.6288653111671121,
+      "learning_rate": 1.9204321603214523e-05,
+      "loss": 0.4579,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5149842271293376,
+      "grad_norm": 0.631342041680086,
+      "learning_rate": 1.9203715583866527e-05,
+      "loss": 0.4919,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5151813880126183,
+      "grad_norm": 0.7478480009750041,
+      "learning_rate": 1.920310934339185e-05,
+      "loss": 0.4642,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5153785488958991,
+      "grad_norm": 0.6757289333762163,
+      "learning_rate": 1.920250288180506e-05,
+      "loss": 0.4699,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5155757097791798,
+      "grad_norm": 0.569219529443845,
+      "learning_rate": 1.9201896199120728e-05,
+      "loss": 0.4038,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5157728706624606,
+      "grad_norm": 0.6266639043108642,
+      "learning_rate": 1.920128929535343e-05,
+      "loss": 0.4584,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5159700315457413,
+      "grad_norm": 0.6019770065246319,
+      "learning_rate": 1.9200682170517746e-05,
+      "loss": 0.4545,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5161671924290221,
+      "grad_norm": 0.7396548559691547,
+      "learning_rate": 1.9200074824628267e-05,
+      "loss": 0.4819,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5163643533123028,
+      "grad_norm": 0.6032049311937472,
+      "learning_rate": 1.9199467257699577e-05,
+      "loss": 0.4672,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5165615141955836,
+      "grad_norm": 0.5534977232280806,
+      "learning_rate": 1.919885946974628e-05,
+      "loss": 0.433,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5167586750788643,
+      "grad_norm": 0.6139976333469656,
+      "learning_rate": 1.9198251460782974e-05,
+      "loss": 0.4354,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5169558359621451,
+      "grad_norm": 0.6302918986448092,
+      "learning_rate": 1.9197643230824272e-05,
+      "loss": 0.4114,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5171529968454258,
+      "grad_norm": 0.6836953420990394,
+      "learning_rate": 1.9197034779884785e-05,
+      "loss": 0.4611,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5173501577287066,
+      "grad_norm": 0.5979385220267467,
+      "learning_rate": 1.919642610797913e-05,
+      "loss": 0.4061,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5175473186119873,
+      "grad_norm": 0.657979498979339,
+      "learning_rate": 1.9195817215121933e-05,
+      "loss": 0.4931,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5177444794952681,
+      "grad_norm": 0.9799814092912918,
+      "learning_rate": 1.9195208101327818e-05,
+      "loss": 0.4581,
+      "step": 2626
+    },
+    {
+      "epoch": 0.517941640378549,
+      "grad_norm": 0.6066811057663211,
+      "learning_rate": 1.9194598766611426e-05,
+      "loss": 0.4601,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5181388012618297,
+      "grad_norm": 0.6509679416434392,
+      "learning_rate": 1.9193989210987396e-05,
+      "loss": 0.503,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5183359621451105,
+      "grad_norm": 0.5884583173326658,
+      "learning_rate": 1.919337943447037e-05,
+      "loss": 0.4744,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5185331230283912,
+      "grad_norm": 0.6369404434163963,
+      "learning_rate": 1.9192769437075e-05,
+      "loss": 0.4354,
+      "step": 2630
+    },
+    {
+      "epoch": 0.518730283911672,
+      "grad_norm": 0.6280177351268637,
+      "learning_rate": 1.919215921881594e-05,
+      "loss": 0.4608,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5189274447949527,
+      "grad_norm": 0.6573633987043811,
+      "learning_rate": 1.9191548779707854e-05,
+      "loss": 0.4649,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5191246056782335,
+      "grad_norm": 0.5972245759129801,
+      "learning_rate": 1.9190938119765404e-05,
+      "loss": 0.4536,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5193217665615142,
+      "grad_norm": 0.625177544249115,
+      "learning_rate": 1.9190327239003267e-05,
+      "loss": 0.4509,
+      "step": 2634
+    },
+    {
+      "epoch": 0.519518927444795,
+      "grad_norm": 0.5920183204667294,
+      "learning_rate": 1.9189716137436118e-05,
+      "loss": 0.466,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5197160883280757,
+      "grad_norm": 0.5840691953513331,
+      "learning_rate": 1.9189104815078633e-05,
+      "loss": 0.4253,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5199132492113565,
+      "grad_norm": 0.5749628137522924,
+      "learning_rate": 1.918849327194551e-05,
+      "loss": 0.4263,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5201104100946372,
+      "grad_norm": 0.645892075328813,
+      "learning_rate": 1.9187881508051433e-05,
+      "loss": 0.4376,
+      "step": 2638
+    },
+    {
+      "epoch": 0.520307570977918,
+      "grad_norm": 0.5985813732201994,
+      "learning_rate": 1.9187269523411108e-05,
+      "loss": 0.469,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5205047318611987,
+      "grad_norm": 0.5662446538186162,
+      "learning_rate": 1.918665731803923e-05,
+      "loss": 0.4278,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5207018927444795,
+      "grad_norm": 0.5774151154266736,
+      "learning_rate": 1.9186044891950514e-05,
+      "loss": 0.4541,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5208990536277602,
+      "grad_norm": 0.5628515665275783,
+      "learning_rate": 1.9185432245159675e-05,
+      "loss": 0.4533,
+      "step": 2642
+    },
+    {
+      "epoch": 0.521096214511041,
+      "grad_norm": 0.602812437259254,
+      "learning_rate": 1.9184819377681425e-05,
+      "loss": 0.4411,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5212933753943217,
+      "grad_norm": 0.6047746007443482,
+      "learning_rate": 1.9184206289530496e-05,
+      "loss": 0.4664,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5214905362776026,
+      "grad_norm": 0.8256532576919925,
+      "learning_rate": 1.918359298072161e-05,
+      "loss": 0.4513,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5216876971608833,
+      "grad_norm": 0.5378632559684782,
+      "learning_rate": 1.9182979451269513e-05,
+      "loss": 0.4264,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5218848580441641,
+      "grad_norm": 0.6701591614996074,
+      "learning_rate": 1.9182365701188933e-05,
+      "loss": 0.4815,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5220820189274448,
+      "grad_norm": 0.6319925276235775,
+      "learning_rate": 1.918175173049463e-05,
+      "loss": 0.4799,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5222791798107256,
+      "grad_norm": 0.6247215611056411,
+      "learning_rate": 1.9181137539201343e-05,
+      "loss": 0.4698,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5224763406940063,
+      "grad_norm": 0.5842030623500385,
+      "learning_rate": 1.9180523127323834e-05,
+      "loss": 0.4394,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5226735015772871,
+      "grad_norm": 0.5932577854627189,
+      "learning_rate": 1.9179908494876863e-05,
+      "loss": 0.451,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5228706624605678,
+      "grad_norm": 0.6071689058545614,
+      "learning_rate": 1.91792936418752e-05,
+      "loss": 0.4458,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5230678233438486,
+      "grad_norm": 0.6417327540656751,
+      "learning_rate": 1.917867856833361e-05,
+      "loss": 0.4421,
+      "step": 2653
+    },
+    {
+      "epoch": 0.5232649842271293,
+      "grad_norm": 8.906897332645489,
+      "learning_rate": 1.9178063274266884e-05,
+      "loss": 0.4716,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5234621451104101,
+      "grad_norm": 0.76731746274502,
+      "learning_rate": 1.9177447759689792e-05,
+      "loss": 0.4537,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5236593059936908,
+      "grad_norm": 0.6082849024874425,
+      "learning_rate": 1.9176832024617125e-05,
+      "loss": 0.4746,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5238564668769716,
+      "grad_norm": 0.6698492507500728,
+      "learning_rate": 1.9176216069063683e-05,
+      "loss": 0.4377,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5240536277602523,
+      "grad_norm": 0.6683170097007389,
+      "learning_rate": 1.917559989304426e-05,
+      "loss": 0.4536,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5242507886435331,
+      "grad_norm": 0.6479408068752674,
+      "learning_rate": 1.9174983496573657e-05,
+      "loss": 0.4526,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5244479495268138,
+      "grad_norm": 0.6865722121413244,
+      "learning_rate": 1.917436687966669e-05,
+      "loss": 0.4687,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5246451104100947,
+      "grad_norm": 0.6689601140224801,
+      "learning_rate": 1.917375004233817e-05,
+      "loss": 0.4841,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5248422712933754,
+      "grad_norm": 0.6723818049998349,
+      "learning_rate": 1.9173132984602914e-05,
+      "loss": 0.4405,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5250394321766562,
+      "grad_norm": 0.6763870303856786,
+      "learning_rate": 1.9172515706475755e-05,
+      "loss": 0.4439,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5252365930599369,
+      "grad_norm": 0.7423211243997336,
+      "learning_rate": 1.9171898207971518e-05,
+      "loss": 0.466,
+      "step": 2664
+    },
+    {
+      "epoch": 0.5254337539432177,
+      "grad_norm": 0.5578326316635425,
+      "learning_rate": 1.9171280489105043e-05,
+      "loss": 0.4387,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5256309148264984,
+      "grad_norm": 0.6327176422438838,
+      "learning_rate": 1.9170662549891162e-05,
+      "loss": 0.4419,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5258280757097792,
+      "grad_norm": 0.6097938287122503,
+      "learning_rate": 1.9170044390344737e-05,
+      "loss": 0.4684,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5260252365930599,
+      "grad_norm": 0.6652570785788708,
+      "learning_rate": 1.9169426010480604e-05,
+      "loss": 0.4439,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5262223974763407,
+      "grad_norm": 0.5705839336000622,
+      "learning_rate": 1.916880741031363e-05,
+      "loss": 0.4416,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5264195583596214,
+      "grad_norm": 0.5966638281252562,
+      "learning_rate": 1.9168188589858675e-05,
+      "loss": 0.4328,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5266167192429022,
+      "grad_norm": 0.5661798360166143,
+      "learning_rate": 1.9167569549130604e-05,
+      "loss": 0.4222,
+      "step": 2671
+    },
+    {
+      "epoch": 0.526813880126183,
+      "grad_norm": 0.6409908246303587,
+      "learning_rate": 1.9166950288144296e-05,
+      "loss": 0.4262,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5270110410094637,
+      "grad_norm": 0.5846011872700821,
+      "learning_rate": 1.916633080691462e-05,
+      "loss": 0.4606,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5272082018927445,
+      "grad_norm": 0.6281868356100663,
+      "learning_rate": 1.9165711105456468e-05,
+      "loss": 0.4683,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5274053627760252,
+      "grad_norm": 0.6643878477972568,
+      "learning_rate": 1.9165091183784722e-05,
+      "loss": 0.4666,
+      "step": 2675
+    },
+    {
+      "epoch": 0.527602523659306,
+      "grad_norm": 0.877520567898525,
+      "learning_rate": 1.9164471041914283e-05,
+      "loss": 0.4992,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5277996845425867,
+      "grad_norm": 0.8982575542215294,
+      "learning_rate": 1.9163850679860046e-05,
+      "loss": 0.4169,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5279968454258676,
+      "grad_norm": 0.9377399621895196,
+      "learning_rate": 1.9163230097636917e-05,
+      "loss": 0.4429,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5281940063091483,
+      "grad_norm": 0.5944294971560273,
+      "learning_rate": 1.9162609295259805e-05,
+      "loss": 0.4963,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5283911671924291,
+      "grad_norm": 0.6623678780030434,
+      "learning_rate": 1.9161988272743627e-05,
+      "loss": 0.4514,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5285883280757098,
+      "grad_norm": 0.7305199400460632,
+      "learning_rate": 1.9161367030103303e-05,
+      "loss": 0.4526,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5287854889589906,
+      "grad_norm": 0.6299026603110343,
+      "learning_rate": 1.9160745567353758e-05,
+      "loss": 0.4495,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5289826498422713,
+      "grad_norm": 0.606070194816596,
+      "learning_rate": 1.9160123884509923e-05,
+      "loss": 0.4555,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5291798107255521,
+      "grad_norm": 0.5864541808065882,
+      "learning_rate": 1.9159501981586738e-05,
+      "loss": 0.4447,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5293769716088328,
+      "grad_norm": 0.5378633706513408,
+      "learning_rate": 1.9158879858599138e-05,
+      "loss": 0.4205,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5295741324921136,
+      "grad_norm": 0.5692796059391859,
+      "learning_rate": 1.9158257515562075e-05,
+      "loss": 0.446,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5297712933753943,
+      "grad_norm": 0.5718248315346496,
+      "learning_rate": 1.91576349524905e-05,
+      "loss": 0.4417,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5299684542586751,
+      "grad_norm": 0.5934922303137548,
+      "learning_rate": 1.9157012169399372e-05,
+      "loss": 0.4271,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5301656151419558,
+      "grad_norm": 0.7026701042054814,
+      "learning_rate": 1.9156389166303652e-05,
+      "loss": 0.4772,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5303627760252366,
+      "grad_norm": 0.5525996284668334,
+      "learning_rate": 1.9155765943218304e-05,
+      "loss": 0.3835,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5305599369085173,
+      "grad_norm": 0.6691635308990771,
+      "learning_rate": 1.9155142500158312e-05,
+      "loss": 0.4962,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5307570977917981,
+      "grad_norm": 0.6586289639914643,
+      "learning_rate": 1.9154518837138644e-05,
+      "loss": 0.4623,
+      "step": 2692
+    },
+    {
+      "epoch": 0.5309542586750788,
+      "grad_norm": 0.5627097183725107,
+      "learning_rate": 1.9153894954174294e-05,
+      "loss": 0.4543,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5311514195583596,
+      "grad_norm": 0.5790687836358622,
+      "learning_rate": 1.9153270851280245e-05,
+      "loss": 0.4632,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5313485804416404,
+      "grad_norm": 0.6605341566789122,
+      "learning_rate": 1.915264652847149e-05,
+      "loss": 0.4662,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5315457413249212,
+      "grad_norm": 0.5991069128744081,
+      "learning_rate": 1.9152021985763035e-05,
+      "loss": 0.4425,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5317429022082019,
+      "grad_norm": 0.6215310692072236,
+      "learning_rate": 1.9151397223169877e-05,
+      "loss": 0.4386,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5319400630914827,
+      "grad_norm": 0.5638306576019859,
+      "learning_rate": 1.9150772240707038e-05,
+      "loss": 0.4245,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5321372239747634,
+      "grad_norm": 0.6748512477251645,
+      "learning_rate": 1.915014703838952e-05,
+      "loss": 0.4187,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5323343848580442,
+      "grad_norm": 0.6108680696319301,
+      "learning_rate": 1.9149521616232354e-05,
+      "loss": 0.4595,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5325315457413249,
+      "grad_norm": 0.596636107232528,
+      "learning_rate": 1.9148895974250562e-05,
+      "loss": 0.4858,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5327287066246057,
+      "grad_norm": 0.5630137990858612,
+      "learning_rate": 1.9148270112459178e-05,
+      "loss": 0.4191,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5329258675078864,
+      "grad_norm": 0.6635599997569216,
+      "learning_rate": 1.9147644030873236e-05,
+      "loss": 0.4665,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5331230283911672,
+      "grad_norm": 0.792415273908661,
+      "learning_rate": 1.914701772950778e-05,
+      "loss": 0.4355,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5333201892744479,
+      "grad_norm": 0.6016005658016478,
+      "learning_rate": 1.9146391208377856e-05,
+      "loss": 0.4361,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5335173501577287,
+      "grad_norm": 0.58541615627604,
+      "learning_rate": 1.914576446749852e-05,
+      "loss": 0.4288,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5337145110410094,
+      "grad_norm": 0.6092999107646474,
+      "learning_rate": 1.9145137506884826e-05,
+      "loss": 0.4174,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5339116719242902,
+      "grad_norm": 0.7227982055069082,
+      "learning_rate": 1.914451032655184e-05,
+      "loss": 0.4822,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5341088328075709,
+      "grad_norm": 0.6246387176004011,
+      "learning_rate": 1.914388292651463e-05,
+      "loss": 0.436,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5343059936908517,
+      "grad_norm": 0.573293012232519,
+      "learning_rate": 1.9143255306788266e-05,
+      "loss": 0.4834,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5345031545741324,
+      "grad_norm": 0.5767714766984778,
+      "learning_rate": 1.9142627467387833e-05,
+      "loss": 0.4374,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5347003154574133,
+      "grad_norm": 0.5672517687751784,
+      "learning_rate": 1.9141999408328412e-05,
+      "loss": 0.4461,
+      "step": 2712
+    },
+    {
+      "epoch": 0.534897476340694,
+      "grad_norm": 0.5888251273732579,
+      "learning_rate": 1.914137112962509e-05,
+      "loss": 0.4775,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5350946372239748,
+      "grad_norm": 0.633920406740737,
+      "learning_rate": 1.914074263129297e-05,
+      "loss": 0.4626,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5352917981072555,
+      "grad_norm": 0.6205409845878437,
+      "learning_rate": 1.9140113913347145e-05,
+      "loss": 0.4504,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5354889589905363,
+      "grad_norm": 1.6126724762687141,
+      "learning_rate": 1.9139484975802723e-05,
+      "loss": 0.4977,
+      "step": 2716
+    },
+    {
+      "epoch": 0.535686119873817,
+      "grad_norm": 0.6349889203833219,
+      "learning_rate": 1.9138855818674814e-05,
+      "loss": 0.4596,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5358832807570978,
+      "grad_norm": 0.5920992239234482,
+      "learning_rate": 1.9138226441978533e-05,
+      "loss": 0.4302,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5360804416403786,
+      "grad_norm": 0.5585845890938809,
+      "learning_rate": 1.9137596845729005e-05,
+      "loss": 0.4447,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5362776025236593,
+      "grad_norm": 0.6075692038210041,
+      "learning_rate": 1.9136967029941354e-05,
+      "loss": 0.4849,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5364747634069401,
+      "grad_norm": 0.6678891128497162,
+      "learning_rate": 1.9136336994630712e-05,
+      "loss": 0.4498,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5366719242902208,
+      "grad_norm": 0.6172599424463148,
+      "learning_rate": 1.9135706739812217e-05,
+      "loss": 0.4605,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5368690851735016,
+      "grad_norm": 0.5878777248662985,
+      "learning_rate": 1.913507626550101e-05,
+      "loss": 0.4588,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5370662460567823,
+      "grad_norm": 0.6514199512488131,
+      "learning_rate": 1.9134445571712237e-05,
+      "loss": 0.3952,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5372634069400631,
+      "grad_norm": 0.5691982494960713,
+      "learning_rate": 1.9133814658461056e-05,
+      "loss": 0.419,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5374605678233438,
+      "grad_norm": 0.6165356310225942,
+      "learning_rate": 1.9133183525762622e-05,
+      "loss": 0.4368,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5376577287066246,
+      "grad_norm": 0.6483155000970097,
+      "learning_rate": 1.9132552173632097e-05,
+      "loss": 0.4526,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5378548895899053,
+      "grad_norm": 0.6668794939528021,
+      "learning_rate": 1.9131920602084656e-05,
+      "loss": 0.4663,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5380520504731862,
+      "grad_norm": 0.7674580396358098,
+      "learning_rate": 1.9131288811135465e-05,
+      "loss": 0.4869,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5382492113564669,
+      "grad_norm": 0.7119703322980993,
+      "learning_rate": 1.9130656800799706e-05,
+      "loss": 0.4495,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5384463722397477,
+      "grad_norm": 0.5810364329746894,
+      "learning_rate": 1.9130024571092565e-05,
+      "loss": 0.4188,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5386435331230284,
+      "grad_norm": 0.74091178804136,
+      "learning_rate": 1.9129392122029233e-05,
+      "loss": 0.4444,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5388406940063092,
+      "grad_norm": 0.5752138416355487,
+      "learning_rate": 1.9128759453624904e-05,
+      "loss": 0.4434,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5390378548895899,
+      "grad_norm": 0.7080034594600536,
+      "learning_rate": 1.9128126565894776e-05,
+      "loss": 0.4436,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5392350157728707,
+      "grad_norm": 0.5927536597698786,
+      "learning_rate": 1.9127493458854055e-05,
+      "loss": 0.4587,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5394321766561514,
+      "grad_norm": 0.7051671166398329,
+      "learning_rate": 1.9126860132517958e-05,
+      "loss": 0.4431,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5396293375394322,
+      "grad_norm": 0.6126458882890922,
+      "learning_rate": 1.9126226586901693e-05,
+      "loss": 0.4926,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5398264984227129,
+      "grad_norm": 0.6919952144838898,
+      "learning_rate": 1.9125592822020485e-05,
+      "loss": 0.4734,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5400236593059937,
+      "grad_norm": 0.6350862584452707,
+      "learning_rate": 1.912495883788956e-05,
+      "loss": 0.4682,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5402208201892744,
+      "grad_norm": 0.653383083517542,
+      "learning_rate": 1.9124324634524153e-05,
+      "loss": 0.4427,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5404179810725552,
+      "grad_norm": 0.6323326038671259,
+      "learning_rate": 1.91236902119395e-05,
+      "loss": 0.4433,
+      "step": 2741
+    },
+    {
+      "epoch": 0.5406151419558359,
+      "grad_norm": 0.7015096284263576,
+      "learning_rate": 1.912305557015084e-05,
+      "loss": 0.4368,
+      "step": 2742
+    },
+    {
+      "epoch": 0.5408123028391167,
+      "grad_norm": 0.6210161210833985,
+      "learning_rate": 1.9122420709173422e-05,
+      "loss": 0.409,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5410094637223974,
+      "grad_norm": 0.7037911524098324,
+      "learning_rate": 1.9121785629022502e-05,
+      "loss": 0.4446,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5412066246056783,
+      "grad_norm": 0.6143364835057887,
+      "learning_rate": 1.9121150329713334e-05,
+      "loss": 0.4485,
+      "step": 2745
+    },
+    {
+      "epoch": 0.541403785488959,
+      "grad_norm": 0.6298390892240697,
+      "learning_rate": 1.9120514811261187e-05,
+      "loss": 0.4014,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5416009463722398,
+      "grad_norm": 0.6010553806758819,
+      "learning_rate": 1.9119879073681328e-05,
+      "loss": 0.4275,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5417981072555205,
+      "grad_norm": 0.6014151169068335,
+      "learning_rate": 1.9119243116989022e-05,
+      "loss": 0.438,
+      "step": 2748
+    },
+    {
+      "epoch": 0.5419952681388013,
+      "grad_norm": 0.743894988167943,
+      "learning_rate": 1.9118606941199565e-05,
+      "loss": 0.4777,
+      "step": 2749
+    },
+    {
+      "epoch": 0.542192429022082,
+      "grad_norm": 0.612605392695115,
+      "learning_rate": 1.911797054632823e-05,
+      "loss": 0.4606,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5423895899053628,
+      "grad_norm": 0.6393149368458824,
+      "learning_rate": 1.911733393239031e-05,
+      "loss": 0.4622,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5425867507886435,
+      "grad_norm": 0.6406348967467802,
+      "learning_rate": 1.9116697099401103e-05,
+      "loss": 0.4575,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5427839116719243,
+      "grad_norm": 0.6210504950614151,
+      "learning_rate": 1.9116060047375903e-05,
+      "loss": 0.4361,
+      "step": 2753
+    },
+    {
+      "epoch": 0.542981072555205,
+      "grad_norm": 0.5936925930435107,
+      "learning_rate": 1.911542277633002e-05,
+      "loss": 0.4278,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5431782334384858,
+      "grad_norm": 0.6521721561800861,
+      "learning_rate": 1.9114785286278767e-05,
+      "loss": 0.4693,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5433753943217665,
+      "grad_norm": 0.5816573139740299,
+      "learning_rate": 1.9114147577237452e-05,
+      "loss": 0.4478,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5435725552050473,
+      "grad_norm": 0.5927002843696301,
+      "learning_rate": 1.9113509649221403e-05,
+      "loss": 0.4289,
+      "step": 2757
+    },
+    {
+      "epoch": 0.543769716088328,
+      "grad_norm": 0.6312074883108042,
+      "learning_rate": 1.911287150224595e-05,
+      "loss": 0.4581,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5439668769716088,
+      "grad_norm": 0.63059238176542,
+      "learning_rate": 1.9112233136326416e-05,
+      "loss": 0.4336,
+      "step": 2759
+    },
+    {
+      "epoch": 0.5441640378548895,
+      "grad_norm": 0.722821516571449,
+      "learning_rate": 1.9111594551478146e-05,
+      "loss": 0.4523,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5443611987381703,
+      "grad_norm": 0.5865645060807954,
+      "learning_rate": 1.9110955747716478e-05,
+      "loss": 0.4407,
+      "step": 2761
+    },
+    {
+      "epoch": 0.544558359621451,
+      "grad_norm": 0.6247985614675826,
+      "learning_rate": 1.911031672505676e-05,
+      "loss": 0.4229,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5447555205047319,
+      "grad_norm": 0.6542180266531745,
+      "learning_rate": 1.9109677483514346e-05,
+      "loss": 0.5063,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5449526813880127,
+      "grad_norm": 0.5911009525801598,
+      "learning_rate": 1.91090380231046e-05,
+      "loss": 0.4443,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5451498422712934,
+      "grad_norm": 0.8124320978143662,
+      "learning_rate": 1.9108398343842873e-05,
+      "loss": 0.4413,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5453470031545742,
+      "grad_norm": 0.5550449237988123,
+      "learning_rate": 1.910775844574454e-05,
+      "loss": 0.4284,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5455441640378549,
+      "grad_norm": 0.597109232711348,
+      "learning_rate": 1.910711832882498e-05,
+      "loss": 0.4775,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5457413249211357,
+      "grad_norm": 0.7408895073247752,
+      "learning_rate": 1.910647799309957e-05,
+      "loss": 0.4716,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5459384858044164,
+      "grad_norm": 0.5592810141688882,
+      "learning_rate": 1.9105837438583693e-05,
+      "loss": 0.4494,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5461356466876972,
+      "grad_norm": 0.6407004174091825,
+      "learning_rate": 1.9105196665292735e-05,
+      "loss": 0.4471,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5463328075709779,
+      "grad_norm": 0.5920197717658957,
+      "learning_rate": 1.9104555673242092e-05,
+      "loss": 0.4632,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5465299684542587,
+      "grad_norm": 0.5534555677305497,
+      "learning_rate": 1.9103914462447172e-05,
+      "loss": 0.4683,
+      "step": 2772
+    },
+    {
+      "epoch": 0.5467271293375394,
+      "grad_norm": 0.6333249803496639,
+      "learning_rate": 1.9103273032923378e-05,
+      "loss": 0.4167,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5469242902208202,
+      "grad_norm": 0.5735766563287512,
+      "learning_rate": 1.9102631384686116e-05,
+      "loss": 0.4556,
+      "step": 2774
+    },
+    {
+      "epoch": 0.5471214511041009,
+      "grad_norm": 0.6225909847389804,
+      "learning_rate": 1.91019895177508e-05,
+      "loss": 0.4532,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5473186119873817,
+      "grad_norm": 0.6265493049371124,
+      "learning_rate": 1.910134743213286e-05,
+      "loss": 0.4352,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5475157728706624,
+      "grad_norm": 0.7433505225427158,
+      "learning_rate": 1.910070512784772e-05,
+      "loss": 0.4435,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5477129337539433,
+      "grad_norm": 0.6861455293800391,
+      "learning_rate": 1.910006260491081e-05,
+      "loss": 0.4573,
+      "step": 2778
+    },
+    {
+      "epoch": 0.547910094637224,
+      "grad_norm": 0.6005840206295373,
+      "learning_rate": 1.9099419863337567e-05,
+      "loss": 0.4565,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5481072555205048,
+      "grad_norm": 0.5980496434263233,
+      "learning_rate": 1.909877690314343e-05,
+      "loss": 0.4452,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5483044164037855,
+      "grad_norm": 0.6336383497213709,
+      "learning_rate": 1.9098133724343853e-05,
+      "loss": 0.4745,
+      "step": 2781
+    },
+    {
+      "epoch": 0.5485015772870663,
+      "grad_norm": 0.600357744618508,
+      "learning_rate": 1.9097490326954288e-05,
+      "loss": 0.4485,
+      "step": 2782
+    },
+    {
+      "epoch": 0.548698738170347,
+      "grad_norm": 0.6216658860334479,
+      "learning_rate": 1.9096846710990192e-05,
+      "loss": 0.4393,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5488958990536278,
+      "grad_norm": 0.5921780263629273,
+      "learning_rate": 1.9096202876467028e-05,
+      "loss": 0.4365,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5490930599369085,
+      "grad_norm": 0.5793518128764239,
+      "learning_rate": 1.9095558823400266e-05,
+      "loss": 0.458,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5492902208201893,
+      "grad_norm": 0.5941381807403091,
+      "learning_rate": 1.9094914551805377e-05,
+      "loss": 0.4269,
+      "step": 2786
+    },
+    {
+      "epoch": 0.54948738170347,
+      "grad_norm": 0.5243117956195389,
+      "learning_rate": 1.909427006169784e-05,
+      "loss": 0.4263,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5496845425867508,
+      "grad_norm": 0.5990209951331408,
+      "learning_rate": 1.9093625353093146e-05,
+      "loss": 0.4689,
+      "step": 2788
+    },
+    {
+      "epoch": 0.5498817034700315,
+      "grad_norm": 0.6974111022304779,
+      "learning_rate": 1.9092980426006774e-05,
+      "loss": 0.4537,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5500788643533123,
+      "grad_norm": 0.5355199493063655,
+      "learning_rate": 1.909233528045423e-05,
+      "loss": 0.4156,
+      "step": 2790
+    },
+    {
+      "epoch": 0.550276025236593,
+      "grad_norm": 0.6415663325485629,
+      "learning_rate": 1.9091689916451006e-05,
+      "loss": 0.4639,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5504731861198738,
+      "grad_norm": 0.5466690904751329,
+      "learning_rate": 1.909104433401261e-05,
+      "loss": 0.4197,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5506703470031545,
+      "grad_norm": 0.5535455971550053,
+      "learning_rate": 1.9090398533154552e-05,
+      "loss": 0.4221,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5508675078864353,
+      "grad_norm": 0.6772223151490671,
+      "learning_rate": 1.908975251389235e-05,
+      "loss": 0.4584,
+      "step": 2794
+    },
+    {
+      "epoch": 0.551064668769716,
+      "grad_norm": 0.639977189301138,
+      "learning_rate": 1.9089106276241523e-05,
+      "loss": 0.479,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5512618296529969,
+      "grad_norm": 0.5718559924246178,
+      "learning_rate": 1.9088459820217602e-05,
+      "loss": 0.428,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5514589905362776,
+      "grad_norm": 0.5777292639541047,
+      "learning_rate": 1.908781314583611e-05,
+      "loss": 0.4451,
+      "step": 2797
+    },
+    {
+      "epoch": 0.5516561514195584,
+      "grad_norm": 0.5900189358667072,
+      "learning_rate": 1.9087166253112594e-05,
+      "loss": 0.4692,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5518533123028391,
+      "grad_norm": 0.6961573118708897,
+      "learning_rate": 1.9086519142062587e-05,
+      "loss": 0.4743,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5520504731861199,
+      "grad_norm": 0.5509038753437092,
+      "learning_rate": 1.9085871812701642e-05,
+      "loss": 0.4182,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5522476340694006,
+      "grad_norm": 0.5723685332377061,
+      "learning_rate": 1.908522426504531e-05,
+      "loss": 0.4357,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5524447949526814,
+      "grad_norm": 0.6424941228887284,
+      "learning_rate": 1.9084576499109148e-05,
+      "loss": 0.4767,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5526419558359621,
+      "grad_norm": 0.6223021214852441,
+      "learning_rate": 1.908392851490872e-05,
+      "loss": 0.459,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5528391167192429,
+      "grad_norm": 0.5876602293739934,
+      "learning_rate": 1.9083280312459595e-05,
+      "loss": 0.4468,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5530362776025236,
+      "grad_norm": 0.6188653567310646,
+      "learning_rate": 1.9082631891777345e-05,
+      "loss": 0.42,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5532334384858044,
+      "grad_norm": 0.6259526543976492,
+      "learning_rate": 1.9081983252877548e-05,
+      "loss": 0.4388,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5534305993690851,
+      "grad_norm": 0.5477244518993755,
+      "learning_rate": 1.9081334395775788e-05,
+      "loss": 0.4735,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5536277602523659,
+      "grad_norm": 0.6135819658127445,
+      "learning_rate": 1.908068532048766e-05,
+      "loss": 0.4716,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5538249211356467,
+      "grad_norm": 0.5562204082368883,
+      "learning_rate": 1.9080036027028752e-05,
+      "loss": 0.455,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5540220820189274,
+      "grad_norm": 0.5792847495400975,
+      "learning_rate": 1.9079386515414667e-05,
+      "loss": 0.421,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5542192429022083,
+      "grad_norm": 0.6003218566784827,
+      "learning_rate": 1.9078736785661012e-05,
+      "loss": 0.427,
+      "step": 2811
+    },
+    {
+      "epoch": 0.554416403785489,
+      "grad_norm": 0.5977583508502151,
+      "learning_rate": 1.9078086837783393e-05,
+      "loss": 0.4535,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5546135646687698,
+      "grad_norm": 0.5699405638630519,
+      "learning_rate": 1.9077436671797426e-05,
+      "loss": 0.4305,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5548107255520505,
+      "grad_norm": 0.5800436763185806,
+      "learning_rate": 1.9076786287718734e-05,
+      "loss": 0.4306,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5550078864353313,
+      "grad_norm": 0.5825228330765003,
+      "learning_rate": 1.9076135685562942e-05,
+      "loss": 0.4059,
+      "step": 2815
+    },
+    {
+      "epoch": 0.555205047318612,
+      "grad_norm": 0.5807662460549189,
+      "learning_rate": 1.9075484865345678e-05,
+      "loss": 0.4399,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5554022082018928,
+      "grad_norm": 0.6218425966437778,
+      "learning_rate": 1.9074833827082586e-05,
+      "loss": 0.4857,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5555993690851735,
+      "grad_norm": 0.5785450171816167,
+      "learning_rate": 1.90741825707893e-05,
+      "loss": 0.4473,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5557965299684543,
+      "grad_norm": 0.5632479711426183,
+      "learning_rate": 1.9073531096481475e-05,
+      "loss": 0.4494,
+      "step": 2819
+    },
+    {
+      "epoch": 0.555993690851735,
+      "grad_norm": 0.6791829687064139,
+      "learning_rate": 1.9072879404174755e-05,
+      "loss": 0.4438,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5561908517350158,
+      "grad_norm": 0.6216151824829282,
+      "learning_rate": 1.90722274938848e-05,
+      "loss": 0.4486,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5563880126182965,
+      "grad_norm": 0.6115679004172528,
+      "learning_rate": 1.9071575365627274e-05,
+      "loss": 0.4846,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5565851735015773,
+      "grad_norm": 0.6026873566499071,
+      "learning_rate": 1.9070923019417848e-05,
+      "loss": 0.4395,
+      "step": 2823
+    },
+    {
+      "epoch": 0.556782334384858,
+      "grad_norm": 0.6123086454851155,
+      "learning_rate": 1.907027045527219e-05,
+      "loss": 0.406,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5569794952681388,
+      "grad_norm": 0.5882773742606572,
+      "learning_rate": 1.906961767320598e-05,
+      "loss": 0.4679,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5571766561514195,
+      "grad_norm": 0.5795368478426368,
+      "learning_rate": 1.90689646732349e-05,
+      "loss": 0.4508,
+      "step": 2826
+    },
+    {
+      "epoch": 0.5573738170347003,
+      "grad_norm": 0.6306886418135007,
+      "learning_rate": 1.9068311455374638e-05,
+      "loss": 0.4591,
+      "step": 2827
+    },
+    {
+      "epoch": 0.557570977917981,
+      "grad_norm": 0.6081030199830474,
+      "learning_rate": 1.9067658019640897e-05,
+      "loss": 0.4718,
+      "step": 2828
+    },
+    {
+      "epoch": 0.5577681388012619,
+      "grad_norm": 0.6178637399147585,
+      "learning_rate": 1.9067004366049367e-05,
+      "loss": 0.4833,
+      "step": 2829
+    },
+    {
+      "epoch": 0.5579652996845426,
+      "grad_norm": 0.6385458957420991,
+      "learning_rate": 1.9066350494615756e-05,
+      "loss": 0.4273,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5581624605678234,
+      "grad_norm": 0.645226093197602,
+      "learning_rate": 1.9065696405355774e-05,
+      "loss": 0.4959,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5583596214511041,
+      "grad_norm": 0.6114367401980242,
+      "learning_rate": 1.9065042098285132e-05,
+      "loss": 0.48,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5585567823343849,
+      "grad_norm": 0.60055106272059,
+      "learning_rate": 1.9064387573419555e-05,
+      "loss": 0.4782,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5587539432176656,
+      "grad_norm": 0.6195745515122923,
+      "learning_rate": 1.906373283077477e-05,
+      "loss": 0.4494,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5589511041009464,
+      "grad_norm": 0.7119201520791544,
+      "learning_rate": 1.9063077870366504e-05,
+      "loss": 0.4394,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5591482649842271,
+      "grad_norm": 0.5747295277509834,
+      "learning_rate": 1.906242269221049e-05,
+      "loss": 0.4137,
+      "step": 2836
+    },
+    {
+      "epoch": 0.5593454258675079,
+      "grad_norm": 0.5453408326530682,
+      "learning_rate": 1.9061767296322477e-05,
+      "loss": 0.415,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5595425867507886,
+      "grad_norm": 1.321620620951851,
+      "learning_rate": 1.9061111682718204e-05,
+      "loss": 0.4296,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5597397476340694,
+      "grad_norm": 0.562269189961482,
+      "learning_rate": 1.9060455851413424e-05,
+      "loss": 0.4354,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5599369085173501,
+      "grad_norm": 0.550600831319121,
+      "learning_rate": 1.90597998024239e-05,
+      "loss": 0.4179,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5601340694006309,
+      "grad_norm": 0.5594944562329867,
+      "learning_rate": 1.905914353576539e-05,
+      "loss": 0.4536,
+      "step": 2841
+    },
+    {
+      "epoch": 0.5603312302839116,
+      "grad_norm": 0.5764617009479925,
+      "learning_rate": 1.9058487051453662e-05,
+      "loss": 0.4629,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5605283911671924,
+      "grad_norm": 11.946607473336949,
+      "learning_rate": 1.9057830349504484e-05,
+      "loss": 0.4617,
+      "step": 2843
+    },
+    {
+      "epoch": 0.5607255520504731,
+      "grad_norm": 0.6562336888788557,
+      "learning_rate": 1.9057173429933636e-05,
+      "loss": 0.4496,
+      "step": 2844
+    },
+    {
+      "epoch": 0.560922712933754,
+      "grad_norm": 0.5696593285659094,
+      "learning_rate": 1.905651629275691e-05,
+      "loss": 0.4374,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5611198738170347,
+      "grad_norm": 0.6107715466947922,
+      "learning_rate": 1.9055858937990083e-05,
+      "loss": 0.4478,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5613170347003155,
+      "grad_norm": 0.6442549288365613,
+      "learning_rate": 1.905520136564895e-05,
+      "loss": 0.4537,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5615141955835962,
+      "grad_norm": 0.5746797728643914,
+      "learning_rate": 1.9054543575749317e-05,
+      "loss": 0.4498,
+      "step": 2848
+    },
+    {
+      "epoch": 0.561711356466877,
+      "grad_norm": 0.9157971817240291,
+      "learning_rate": 1.905388556830698e-05,
+      "loss": 0.4024,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5619085173501577,
+      "grad_norm": 0.5954555864900205,
+      "learning_rate": 1.905322734333775e-05,
+      "loss": 0.4712,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5621056782334385,
+      "grad_norm": 0.68954636579152,
+      "learning_rate": 1.9052568900857443e-05,
+      "loss": 0.4205,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5623028391167192,
+      "grad_norm": 0.7460974392901148,
+      "learning_rate": 1.9051910240881883e-05,
+      "loss": 0.4483,
+      "step": 2852
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 0.5355564846309835,
+      "learning_rate": 1.9051251363426883e-05,
+      "loss": 0.4232,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5626971608832808,
+      "grad_norm": 0.7155811623930997,
+      "learning_rate": 1.9050592268508284e-05,
+      "loss": 0.4553,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5628943217665615,
+      "grad_norm": 0.6320781522351271,
+      "learning_rate": 1.9049932956141917e-05,
+      "loss": 0.4436,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5630914826498423,
+      "grad_norm": 0.6801664356997779,
+      "learning_rate": 1.9049273426343622e-05,
+      "loss": 0.4637,
+      "step": 2856
+    },
+    {
+      "epoch": 0.563288643533123,
+      "grad_norm": 0.5803175479602963,
+      "learning_rate": 1.9048613679129246e-05,
+      "loss": 0.466,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5634858044164038,
+      "grad_norm": 0.6126039116402423,
+      "learning_rate": 1.904795371451464e-05,
+      "loss": 0.4587,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5636829652996845,
+      "grad_norm": 0.5827030340943455,
+      "learning_rate": 1.904729353251566e-05,
+      "loss": 0.4518,
+      "step": 2859
+    },
+    {
+      "epoch": 0.5638801261829653,
+      "grad_norm": 0.601328312382053,
+      "learning_rate": 1.9046633133148164e-05,
+      "loss": 0.4675,
+      "step": 2860
+    },
+    {
+      "epoch": 0.564077287066246,
+      "grad_norm": 0.5717955796305233,
+      "learning_rate": 1.9045972516428026e-05,
+      "loss": 0.4549,
+      "step": 2861
+    },
+    {
+      "epoch": 0.5642744479495269,
+      "grad_norm": 0.5613732974457137,
+      "learning_rate": 1.904531168237111e-05,
+      "loss": 0.443,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5644716088328076,
+      "grad_norm": 0.5744128359613806,
+      "learning_rate": 1.90446506309933e-05,
+      "loss": 0.4439,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5646687697160884,
+      "grad_norm": 0.6119814238497652,
+      "learning_rate": 1.9043989362310472e-05,
+      "loss": 0.4707,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5648659305993691,
+      "grad_norm": 0.5567314391501524,
+      "learning_rate": 1.9043327876338517e-05,
+      "loss": 0.4271,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5650630914826499,
+      "grad_norm": 0.6354333087115706,
+      "learning_rate": 1.904266617309333e-05,
+      "loss": 0.4968,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5652602523659306,
+      "grad_norm": 0.5773958814672833,
+      "learning_rate": 1.9042004252590804e-05,
+      "loss": 0.407,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5654574132492114,
+      "grad_norm": 0.5615283708079536,
+      "learning_rate": 1.9041342114846844e-05,
+      "loss": 0.4448,
+      "step": 2868
+    },
+    {
+      "epoch": 0.5656545741324921,
+      "grad_norm": 0.634498935968448,
+      "learning_rate": 1.9040679759877358e-05,
+      "loss": 0.4633,
+      "step": 2869
+    },
+    {
+      "epoch": 0.5658517350157729,
+      "grad_norm": 0.6007126242862416,
+      "learning_rate": 1.904001718769826e-05,
+      "loss": 0.477,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5660488958990536,
+      "grad_norm": 0.5903936190376452,
+      "learning_rate": 1.903935439832547e-05,
+      "loss": 0.4441,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5662460567823344,
+      "grad_norm": 0.6404835660073024,
+      "learning_rate": 1.9038691391774913e-05,
+      "loss": 0.4424,
+      "step": 2872
+    },
+    {
+      "epoch": 0.5664432176656151,
+      "grad_norm": 0.524427319947031,
+      "learning_rate": 1.9038028168062517e-05,
+      "loss": 0.4137,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5666403785488959,
+      "grad_norm": 0.7149877450692915,
+      "learning_rate": 1.9037364727204216e-05,
+      "loss": 0.4914,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5668375394321766,
+      "grad_norm": 0.5728331748504152,
+      "learning_rate": 1.9036701069215947e-05,
+      "loss": 0.4609,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5670347003154574,
+      "grad_norm": 0.6344827370627714,
+      "learning_rate": 1.9036037194113656e-05,
+      "loss": 0.4456,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5672318611987381,
+      "grad_norm": 0.606955120653267,
+      "learning_rate": 1.90353731019133e-05,
+      "loss": 0.4341,
+      "step": 2877
+    },
+    {
+      "epoch": 0.567429022082019,
+      "grad_norm": 0.7183503039613219,
+      "learning_rate": 1.9034708792630824e-05,
+      "loss": 0.4661,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5676261829652997,
+      "grad_norm": 0.6264186805484021,
+      "learning_rate": 1.9034044266282196e-05,
+      "loss": 0.4745,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5678233438485805,
+      "grad_norm": 0.686093957072564,
+      "learning_rate": 1.903337952288338e-05,
+      "loss": 0.4532,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5680205047318612,
+      "grad_norm": 0.5783475117989667,
+      "learning_rate": 1.9032714562450345e-05,
+      "loss": 0.449,
+      "step": 2881
+    },
+    {
+      "epoch": 0.568217665615142,
+      "grad_norm": 0.6403276197455571,
+      "learning_rate": 1.903204938499907e-05,
+      "loss": 0.4453,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5684148264984227,
+      "grad_norm": 0.561414469598965,
+      "learning_rate": 1.9031383990545532e-05,
+      "loss": 0.4499,
+      "step": 2883
+    },
+    {
+      "epoch": 0.5686119873817035,
+      "grad_norm": 0.6500041432778779,
+      "learning_rate": 1.9030718379105726e-05,
+      "loss": 0.4347,
+      "step": 2884
+    },
+    {
+      "epoch": 0.5688091482649842,
+      "grad_norm": 0.504940933975182,
+      "learning_rate": 1.9030052550695636e-05,
+      "loss": 0.3846,
+      "step": 2885
+    },
+    {
+      "epoch": 0.569006309148265,
+      "grad_norm": 0.6281846779360019,
+      "learning_rate": 1.902938650533126e-05,
+      "loss": 0.4713,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5692034700315457,
+      "grad_norm": 0.5974696177502193,
+      "learning_rate": 1.9028720243028604e-05,
+      "loss": 0.4417,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5694006309148265,
+      "grad_norm": 0.5826680180672237,
+      "learning_rate": 1.9028053763803673e-05,
+      "loss": 0.4566,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5695977917981072,
+      "grad_norm": 0.5875996261062907,
+      "learning_rate": 1.902738706767248e-05,
+      "loss": 0.4656,
+      "step": 2889
+    },
+    {
+      "epoch": 0.569794952681388,
+      "grad_norm": 0.5602438327843606,
+      "learning_rate": 1.902672015465104e-05,
+      "loss": 0.4268,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5699921135646687,
+      "grad_norm": 0.6034100519824507,
+      "learning_rate": 1.9026053024755384e-05,
+      "loss": 0.499,
+      "step": 2891
+    },
+    {
+      "epoch": 0.5701892744479495,
+      "grad_norm": 0.5961260542439667,
+      "learning_rate": 1.902538567800153e-05,
+      "loss": 0.4556,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5703864353312302,
+      "grad_norm": 0.5847619792244266,
+      "learning_rate": 1.902471811440552e-05,
+      "loss": 0.444,
+      "step": 2893
+    },
+    {
+      "epoch": 0.570583596214511,
+      "grad_norm": 0.5717255423270757,
+      "learning_rate": 1.902405033398339e-05,
+      "loss": 0.4604,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5707807570977917,
+      "grad_norm": 0.5775461408981852,
+      "learning_rate": 1.9023382336751185e-05,
+      "loss": 0.4372,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5709779179810726,
+      "grad_norm": 0.5830553771363007,
+      "learning_rate": 1.902271412272495e-05,
+      "loss": 0.4547,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5711750788643533,
+      "grad_norm": 0.6305168162558938,
+      "learning_rate": 1.9022045691920742e-05,
+      "loss": 0.4703,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5713722397476341,
+      "grad_norm": 0.5848064263309912,
+      "learning_rate": 1.9021377044354624e-05,
+      "loss": 0.4542,
+      "step": 2898
+    },
+    {
+      "epoch": 0.5715694006309149,
+      "grad_norm": 0.6638321170730301,
+      "learning_rate": 1.9020708180042654e-05,
+      "loss": 0.4936,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5717665615141956,
+      "grad_norm": 0.637624069032922,
+      "learning_rate": 1.902003909900091e-05,
+      "loss": 0.4551,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5719637223974764,
+      "grad_norm": 0.623095105439941,
+      "learning_rate": 1.9019369801245458e-05,
+      "loss": 0.4491,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5721608832807571,
+      "grad_norm": 0.548146806710325,
+      "learning_rate": 1.9018700286792388e-05,
+      "loss": 0.4262,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5723580441640379,
+      "grad_norm": 0.6045037208451278,
+      "learning_rate": 1.9018030555657776e-05,
+      "loss": 0.4825,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5725552050473186,
+      "grad_norm": 0.5859988705690671,
+      "learning_rate": 1.9017360607857724e-05,
+      "loss": 0.4446,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5727523659305994,
+      "grad_norm": 0.565267471448854,
+      "learning_rate": 1.9016690443408314e-05,
+      "loss": 0.4103,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5729495268138801,
+      "grad_norm": 0.6135424852694339,
+      "learning_rate": 1.901602006232566e-05,
+      "loss": 0.4822,
+      "step": 2906
+    },
+    {
+      "epoch": 0.5731466876971609,
+      "grad_norm": 0.5918245843382935,
+      "learning_rate": 1.901534946462586e-05,
+      "loss": 0.4231,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5733438485804416,
+      "grad_norm": 0.5728688344893811,
+      "learning_rate": 1.9014678650325035e-05,
+      "loss": 0.4464,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5735410094637224,
+      "grad_norm": 0.6157367306965482,
+      "learning_rate": 1.901400761943929e-05,
+      "loss": 0.4399,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5737381703470031,
+      "grad_norm": 0.8263123254378486,
+      "learning_rate": 1.9013336371984756e-05,
+      "loss": 0.4274,
+      "step": 2910
+    },
+    {
+      "epoch": 0.573935331230284,
+      "grad_norm": 0.685238179627279,
+      "learning_rate": 1.9012664907977557e-05,
+      "loss": 0.4683,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5741324921135647,
+      "grad_norm": 0.5358383697418044,
+      "learning_rate": 1.9011993227433826e-05,
+      "loss": 0.3689,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5743296529968455,
+      "grad_norm": 0.6077623131242404,
+      "learning_rate": 1.9011321330369696e-05,
+      "loss": 0.4245,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5745268138801262,
+      "grad_norm": 0.5994256634418688,
+      "learning_rate": 1.9010649216801316e-05,
+      "loss": 0.4768,
+      "step": 2914
+    },
+    {
+      "epoch": 0.574723974763407,
+      "grad_norm": 0.6159531989573963,
+      "learning_rate": 1.9009976886744837e-05,
+      "loss": 0.4501,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5749211356466877,
+      "grad_norm": 0.6096049034892379,
+      "learning_rate": 1.9009304340216403e-05,
+      "loss": 0.4464,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5751182965299685,
+      "grad_norm": 0.5603268106696802,
+      "learning_rate": 1.900863157723218e-05,
+      "loss": 0.458,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5753154574132492,
+      "grad_norm": 0.5558645790716764,
+      "learning_rate": 1.9007958597808326e-05,
+      "loss": 0.3959,
+      "step": 2918
+    },
+    {
+      "epoch": 0.57551261829653,
+      "grad_norm": 0.6195875338599605,
+      "learning_rate": 1.9007285401961016e-05,
+      "loss": 0.5038,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5757097791798107,
+      "grad_norm": 0.5783958420849634,
+      "learning_rate": 1.9006611989706417e-05,
+      "loss": 0.4647,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5759069400630915,
+      "grad_norm": 0.5986709701715736,
+      "learning_rate": 1.9005938361060714e-05,
+      "loss": 0.457,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5761041009463722,
+      "grad_norm": 0.5992935813507028,
+      "learning_rate": 1.900526451604009e-05,
+      "loss": 0.4146,
+      "step": 2922
+    },
+    {
+      "epoch": 0.576301261829653,
+      "grad_norm": 0.6114620175570423,
+      "learning_rate": 1.900459045466073e-05,
+      "loss": 0.4699,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5764984227129337,
+      "grad_norm": 0.5778870682581325,
+      "learning_rate": 1.9003916176938837e-05,
+      "loss": 0.4132,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5766955835962145,
+      "grad_norm": 0.556437544810432,
+      "learning_rate": 1.9003241682890607e-05,
+      "loss": 0.4196,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5768927444794952,
+      "grad_norm": 0.5654030288060881,
+      "learning_rate": 1.9002566972532242e-05,
+      "loss": 0.4016,
+      "step": 2926
+    },
+    {
+      "epoch": 0.577089905362776,
+      "grad_norm": 0.5874171442829121,
+      "learning_rate": 1.9001892045879963e-05,
+      "loss": 0.4617,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5772870662460567,
+      "grad_norm": 0.5875515040090509,
+      "learning_rate": 1.9001216902949974e-05,
+      "loss": 0.4511,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5774842271293376,
+      "grad_norm": 0.5568685656549751,
+      "learning_rate": 1.9000541543758497e-05,
+      "loss": 0.4347,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5776813880126183,
+      "grad_norm": 0.5746594931865293,
+      "learning_rate": 1.8999865968321765e-05,
+      "loss": 0.4523,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5778785488958991,
+      "grad_norm": 0.5500934493880131,
+      "learning_rate": 1.8999190176656004e-05,
+      "loss": 0.401,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5780757097791798,
+      "grad_norm": 0.5988645511870025,
+      "learning_rate": 1.8998514168777453e-05,
+      "loss": 0.4418,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5782728706624606,
+      "grad_norm": 0.5787533040554713,
+      "learning_rate": 1.8997837944702352e-05,
+      "loss": 0.4552,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5784700315457413,
+      "grad_norm": 0.5924302662548973,
+      "learning_rate": 1.899716150444695e-05,
+      "loss": 0.423,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5786671924290221,
+      "grad_norm": 2.4787691128805935,
+      "learning_rate": 1.8996484848027496e-05,
+      "loss": 0.4051,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5788643533123028,
+      "grad_norm": 0.722111103146418,
+      "learning_rate": 1.8995807975460246e-05,
+      "loss": 0.4383,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5790615141955836,
+      "grad_norm": 0.7452661199567964,
+      "learning_rate": 1.8995130886761468e-05,
+      "loss": 0.4328,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5792586750788643,
+      "grad_norm": 0.6272537937748907,
+      "learning_rate": 1.8994453581947428e-05,
+      "loss": 0.4501,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5794558359621451,
+      "grad_norm": 0.5871428411077138,
+      "learning_rate": 1.8993776061034394e-05,
+      "loss": 0.4197,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5796529968454258,
+      "grad_norm": 0.6082870109603066,
+      "learning_rate": 1.899309832403865e-05,
+      "loss": 0.4397,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5798501577287066,
+      "grad_norm": 0.5698173415968192,
+      "learning_rate": 1.8992420370976476e-05,
+      "loss": 0.4297,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5800473186119873,
+      "grad_norm": 0.659454422363326,
+      "learning_rate": 1.899174220186416e-05,
+      "loss": 0.4565,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5802444794952681,
+      "grad_norm": 0.5921121311398183,
+      "learning_rate": 1.8991063816717998e-05,
+      "loss": 0.4357,
+      "step": 2943
+    },
+    {
+      "epoch": 0.580441640378549,
+      "grad_norm": 0.6950468872106844,
+      "learning_rate": 1.899038521555429e-05,
+      "loss": 0.4087,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5806388012618297,
+      "grad_norm": 0.5936144532941147,
+      "learning_rate": 1.8989706398389335e-05,
+      "loss": 0.4237,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5808359621451105,
+      "grad_norm": 0.5868915401669754,
+      "learning_rate": 1.8989027365239443e-05,
+      "loss": 0.4006,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5810331230283912,
+      "grad_norm": 0.6069291322319756,
+      "learning_rate": 1.8988348116120926e-05,
+      "loss": 0.4487,
+      "step": 2947
+    },
+    {
+      "epoch": 0.581230283911672,
+      "grad_norm": 0.5598237023370913,
+      "learning_rate": 1.8987668651050117e-05,
+      "loss": 0.4363,
+      "step": 2948
+    },
+    {
+      "epoch": 0.5814274447949527,
+      "grad_norm": 0.6162522565496759,
+      "learning_rate": 1.8986988970043324e-05,
+      "loss": 0.4778,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5816246056782335,
+      "grad_norm": 0.5922390245389011,
+      "learning_rate": 1.8986309073116883e-05,
+      "loss": 0.4492,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5818217665615142,
+      "grad_norm": 0.5993508377014622,
+      "learning_rate": 1.8985628960287134e-05,
+      "loss": 0.4637,
+      "step": 2951
+    },
+    {
+      "epoch": 0.582018927444795,
+      "grad_norm": 0.549571504491819,
+      "learning_rate": 1.898494863157041e-05,
+      "loss": 0.425,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5822160883280757,
+      "grad_norm": 0.5759972011389546,
+      "learning_rate": 1.898426808698306e-05,
+      "loss": 0.4408,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5824132492113565,
+      "grad_norm": 0.5482466547534278,
+      "learning_rate": 1.8983587326541437e-05,
+      "loss": 0.4289,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5826104100946372,
+      "grad_norm": 0.5913250211398492,
+      "learning_rate": 1.898290635026189e-05,
+      "loss": 0.4629,
+      "step": 2955
+    },
+    {
+      "epoch": 0.582807570977918,
+      "grad_norm": 0.570705758804096,
+      "learning_rate": 1.8982225158160788e-05,
+      "loss": 0.4442,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5830047318611987,
+      "grad_norm": 0.5878782769052985,
+      "learning_rate": 1.898154375025449e-05,
+      "loss": 0.3961,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5832018927444795,
+      "grad_norm": 0.5735902991837972,
+      "learning_rate": 1.8980862126559373e-05,
+      "loss": 0.4702,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5833990536277602,
+      "grad_norm": 0.5657885791348485,
+      "learning_rate": 1.898018028709181e-05,
+      "loss": 0.455,
+      "step": 2959
+    },
+    {
+      "epoch": 0.583596214511041,
+      "grad_norm": 1.3679534419258494,
+      "learning_rate": 1.8979498231868183e-05,
+      "loss": 0.4948,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5837933753943217,
+      "grad_norm": 0.6017442927840219,
+      "learning_rate": 1.897881596090488e-05,
+      "loss": 0.4728,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5839905362776026,
+      "grad_norm": 0.5555623988000693,
+      "learning_rate": 1.8978133474218294e-05,
+      "loss": 0.4455,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5841876971608833,
+      "grad_norm": 0.6041207462728285,
+      "learning_rate": 1.897745077182482e-05,
+      "loss": 0.4544,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5843848580441641,
+      "grad_norm": 0.6250067192179154,
+      "learning_rate": 1.8976767853740866e-05,
+      "loss": 0.4752,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5845820189274448,
+      "grad_norm": 0.571207168389721,
+      "learning_rate": 1.897608471998283e-05,
+      "loss": 0.4342,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5847791798107256,
+      "grad_norm": 2.8095095602359463,
+      "learning_rate": 1.897540137056713e-05,
+      "loss": 0.4852,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5849763406940063,
+      "grad_norm": 0.6427772218920856,
+      "learning_rate": 1.897471780551019e-05,
+      "loss": 0.4679,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5851735015772871,
+      "grad_norm": 0.5594426060918991,
+      "learning_rate": 1.897403402482842e-05,
+      "loss": 0.4083,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5853706624605678,
+      "grad_norm": 0.5801095527319007,
+      "learning_rate": 1.897335002853826e-05,
+      "loss": 0.423,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5855678233438486,
+      "grad_norm": 0.7313301838049326,
+      "learning_rate": 1.897266581665614e-05,
+      "loss": 0.4418,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5857649842271293,
+      "grad_norm": 0.615578548987388,
+      "learning_rate": 1.8971981389198495e-05,
+      "loss": 0.4555,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5859621451104101,
+      "grad_norm": 0.5588797678763053,
+      "learning_rate": 1.8971296746181774e-05,
+      "loss": 0.423,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5861593059936908,
+      "grad_norm": 0.6472108314860092,
+      "learning_rate": 1.8970611887622425e-05,
+      "loss": 0.4733,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5863564668769716,
+      "grad_norm": 0.6277578303497903,
+      "learning_rate": 1.89699268135369e-05,
+      "loss": 0.4319,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5865536277602523,
+      "grad_norm": 8.45953869638077,
+      "learning_rate": 1.8969241523941662e-05,
+      "loss": 0.4668,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5867507886435331,
+      "grad_norm": 0.751366864143633,
+      "learning_rate": 1.896855601885317e-05,
+      "loss": 0.4836,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5869479495268138,
+      "grad_norm": 0.5909034853303601,
+      "learning_rate": 1.89678702982879e-05,
+      "loss": 0.5005,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5871451104100947,
+      "grad_norm": 0.7792417477818725,
+      "learning_rate": 1.8967184362262324e-05,
+      "loss": 0.4447,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5873422712933754,
+      "grad_norm": 1.0145547450801644,
+      "learning_rate": 1.896649821079292e-05,
+      "loss": 0.4624,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5875394321766562,
+      "grad_norm": 0.5792099704275391,
+      "learning_rate": 1.8965811843896178e-05,
+      "loss": 0.4701,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5877365930599369,
+      "grad_norm": 0.6324701038532613,
+      "learning_rate": 1.8965125261588586e-05,
+      "loss": 0.4543,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5879337539432177,
+      "grad_norm": 0.6298268546935512,
+      "learning_rate": 1.8964438463886638e-05,
+      "loss": 0.4484,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5881309148264984,
+      "grad_norm": 0.6842407738734424,
+      "learning_rate": 1.896375145080684e-05,
+      "loss": 0.4355,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5883280757097792,
+      "grad_norm": 0.6522884562722095,
+      "learning_rate": 1.8963064222365694e-05,
+      "loss": 0.469,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5885252365930599,
+      "grad_norm": 0.6206326292297457,
+      "learning_rate": 1.896237677857971e-05,
+      "loss": 0.471,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5887223974763407,
+      "grad_norm": 0.5650099483761517,
+      "learning_rate": 1.896168911946541e-05,
+      "loss": 0.4313,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5889195583596214,
+      "grad_norm": 0.653001742089255,
+      "learning_rate": 1.896100124503931e-05,
+      "loss": 0.4548,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5891167192429022,
+      "grad_norm": 0.5751129122223437,
+      "learning_rate": 1.896031315531794e-05,
+      "loss": 0.4307,
+      "step": 2988
+    },
+    {
+      "epoch": 0.589313880126183,
+      "grad_norm": 0.5731286729374794,
+      "learning_rate": 1.895962485031783e-05,
+      "loss": 0.429,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5895110410094637,
+      "grad_norm": 0.5812391295799345,
+      "learning_rate": 1.8958936330055516e-05,
+      "loss": 0.4326,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5897082018927445,
+      "grad_norm": 0.6341740958943612,
+      "learning_rate": 1.8958247594547543e-05,
+      "loss": 0.4421,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5899053627760252,
+      "grad_norm": 0.5855292136590585,
+      "learning_rate": 1.895755864381046e-05,
+      "loss": 0.4476,
+      "step": 2992
+    },
+    {
+      "epoch": 0.590102523659306,
+      "grad_norm": 0.5897754099618571,
+      "learning_rate": 1.8956869477860813e-05,
+      "loss": 0.4347,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5902996845425867,
+      "grad_norm": 0.5651589224007416,
+      "learning_rate": 1.8956180096715166e-05,
+      "loss": 0.47,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5904968454258676,
+      "grad_norm": 0.5901739340372232,
+      "learning_rate": 1.895549050039008e-05,
+      "loss": 0.4463,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5906940063091483,
+      "grad_norm": 0.5918462673605263,
+      "learning_rate": 1.8954800688902125e-05,
+      "loss": 0.4616,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5908911671924291,
+      "grad_norm": 0.5857619648713869,
+      "learning_rate": 1.8954110662267868e-05,
+      "loss": 0.4655,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5910883280757098,
+      "grad_norm": 0.5749020201058925,
+      "learning_rate": 1.8953420420503894e-05,
+      "loss": 0.4538,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5912854889589906,
+      "grad_norm": 0.6364951608672293,
+      "learning_rate": 1.8952729963626783e-05,
+      "loss": 0.4516,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5914826498422713,
+      "grad_norm": 0.5455105233117238,
+      "learning_rate": 1.8952039291653126e-05,
+      "loss": 0.4129,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5916798107255521,
+      "grad_norm": 0.5688101335852396,
+      "learning_rate": 1.8951348404599518e-05,
+      "loss": 0.436,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5918769716088328,
+      "grad_norm": 0.5869599963349247,
+      "learning_rate": 1.895065730248255e-05,
+      "loss": 0.4625,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5920741324921136,
+      "grad_norm": 0.583433314102157,
+      "learning_rate": 1.8949965985318835e-05,
+      "loss": 0.4619,
+      "step": 3003
+    },
+    {
+      "epoch": 0.5922712933753943,
+      "grad_norm": 1.8179141375028536,
+      "learning_rate": 1.8949274453124985e-05,
+      "loss": 0.4457,
+      "step": 3004
+    },
+    {
+      "epoch": 0.5924684542586751,
+      "grad_norm": 1.0052816097153328,
+      "learning_rate": 1.8948582705917605e-05,
+      "loss": 0.3993,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5926656151419558,
+      "grad_norm": 0.6655550567863493,
+      "learning_rate": 1.8947890743713316e-05,
+      "loss": 0.4633,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5928627760252366,
+      "grad_norm": 0.6343745778626908,
+      "learning_rate": 1.8947198566528752e-05,
+      "loss": 0.4534,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5930599369085173,
+      "grad_norm": 0.8146895268571629,
+      "learning_rate": 1.894650617438054e-05,
+      "loss": 0.4622,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5932570977917981,
+      "grad_norm": 0.5919980337355466,
+      "learning_rate": 1.8945813567285303e-05,
+      "loss": 0.4444,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5934542586750788,
+      "grad_norm": 0.6233642830617198,
+      "learning_rate": 1.8945120745259696e-05,
+      "loss": 0.4585,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5936514195583596,
+      "grad_norm": 0.6560319675501647,
+      "learning_rate": 1.894442770832036e-05,
+      "loss": 0.4733,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5938485804416404,
+      "grad_norm": 0.6361924491098656,
+      "learning_rate": 1.8943734456483944e-05,
+      "loss": 0.463,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5940457413249212,
+      "grad_norm": 0.5872105410259241,
+      "learning_rate": 1.8943040989767104e-05,
+      "loss": 0.4613,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5942429022082019,
+      "grad_norm": 0.9249850168680832,
+      "learning_rate": 1.8942347308186506e-05,
+      "loss": 0.4223,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5944400630914827,
+      "grad_norm": 0.5669241268386221,
+      "learning_rate": 1.8941653411758813e-05,
+      "loss": 0.4231,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5946372239747634,
+      "grad_norm": 0.6196619172043841,
+      "learning_rate": 1.894095930050069e-05,
+      "loss": 0.4403,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5948343848580442,
+      "grad_norm": 0.563726313430164,
+      "learning_rate": 1.8940264974428827e-05,
+      "loss": 0.4407,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5950315457413249,
+      "grad_norm": 0.6200267950921632,
+      "learning_rate": 1.8939570433559894e-05,
+      "loss": 0.4129,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5952287066246057,
+      "grad_norm": 0.6093402086204679,
+      "learning_rate": 1.8938875677910586e-05,
+      "loss": 0.464,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5954258675078864,
+      "grad_norm": 0.5932927503972287,
+      "learning_rate": 1.8938180707497588e-05,
+      "loss": 0.4361,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5956230283911672,
+      "grad_norm": 0.57662299638576,
+      "learning_rate": 1.89374855223376e-05,
+      "loss": 0.4501,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5958201892744479,
+      "grad_norm": 0.5619926542960811,
+      "learning_rate": 1.8936790122447327e-05,
+      "loss": 0.4682,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5960173501577287,
+      "grad_norm": 0.7525793185982179,
+      "learning_rate": 1.8936094507843476e-05,
+      "loss": 0.4779,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5962145110410094,
+      "grad_norm": 0.5548877208195396,
+      "learning_rate": 1.8935398678542752e-05,
+      "loss": 0.4321,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5964116719242902,
+      "grad_norm": 0.5988617075988154,
+      "learning_rate": 1.8934702634561887e-05,
+      "loss": 0.4293,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5966088328075709,
+      "grad_norm": 0.6782737257196965,
+      "learning_rate": 1.893400637591759e-05,
+      "loss": 0.4032,
+      "step": 3026
+    },
+    {
+      "epoch": 0.5968059936908517,
+      "grad_norm": 0.5513926310853049,
+      "learning_rate": 1.8933309902626598e-05,
+      "loss": 0.4134,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5970031545741324,
+      "grad_norm": 0.5598624884027122,
+      "learning_rate": 1.893261321470564e-05,
+      "loss": 0.4409,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5972003154574133,
+      "grad_norm": 0.5745829727781053,
+      "learning_rate": 1.893191631217146e-05,
+      "loss": 0.4699,
+      "step": 3029
+    },
+    {
+      "epoch": 0.597397476340694,
+      "grad_norm": 0.6008234547988593,
+      "learning_rate": 1.8931219195040796e-05,
+      "loss": 0.4526,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5975946372239748,
+      "grad_norm": 0.9177487857817422,
+      "learning_rate": 1.8930521863330395e-05,
+      "loss": 0.4306,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5977917981072555,
+      "grad_norm": 0.5675597820246167,
+      "learning_rate": 1.892982431705702e-05,
+      "loss": 0.4458,
+      "step": 3032
+    },
+    {
+      "epoch": 0.5979889589905363,
+      "grad_norm": 18.36759892361865,
+      "learning_rate": 1.892912655623742e-05,
+      "loss": 0.4854,
+      "step": 3033
+    },
+    {
+      "epoch": 0.598186119873817,
+      "grad_norm": 0.9555573861343462,
+      "learning_rate": 1.8928428580888365e-05,
+      "loss": 0.4249,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5983832807570978,
+      "grad_norm": 0.6045541668448458,
+      "learning_rate": 1.8927730391026625e-05,
+      "loss": 0.4369,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5985804416403786,
+      "grad_norm": 0.5905804898359281,
+      "learning_rate": 1.8927031986668973e-05,
+      "loss": 0.4367,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5987776025236593,
+      "grad_norm": 0.6171855342108601,
+      "learning_rate": 1.8926333367832188e-05,
+      "loss": 0.4311,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5989747634069401,
+      "grad_norm": 0.561383493809823,
+      "learning_rate": 1.8925634534533054e-05,
+      "loss": 0.4225,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5991719242902208,
+      "grad_norm": 0.6623749634079279,
+      "learning_rate": 1.8924935486788362e-05,
+      "loss": 0.4703,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5993690851735016,
+      "grad_norm": 0.6116882687261561,
+      "learning_rate": 1.892423622461491e-05,
+      "loss": 0.4566,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5995662460567823,
+      "grad_norm": 0.5713571661306532,
+      "learning_rate": 1.8923536748029495e-05,
+      "loss": 0.4275,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5997634069400631,
+      "grad_norm": 0.6069279944169308,
+      "learning_rate": 1.8922837057048925e-05,
+      "loss": 0.4318,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5999605678233438,
+      "grad_norm": 0.5706152548682657,
+      "learning_rate": 1.8922137151690005e-05,
+      "loss": 0.4448,
+      "step": 3043
+    },
+    {
+      "epoch": 0.6001577287066246,
+      "grad_norm": 0.6191149624010278,
+      "learning_rate": 1.8921437031969557e-05,
+      "loss": 0.4668,
+      "step": 3044
+    },
+    {
+      "epoch": 0.6003548895899053,
+      "grad_norm": 0.537186705194169,
+      "learning_rate": 1.8920736697904406e-05,
+      "loss": 0.3969,
+      "step": 3045
+    },
+    {
+      "epoch": 0.6005520504731862,
+      "grad_norm": 0.5823679438434386,
+      "learning_rate": 1.8920036149511365e-05,
+      "loss": 0.4349,
+      "step": 3046
+    },
+    {
+      "epoch": 0.6007492113564669,
+      "grad_norm": 0.5764684381732735,
+      "learning_rate": 1.8919335386807275e-05,
+      "loss": 0.449,
+      "step": 3047
+    },
+    {
+      "epoch": 0.6009463722397477,
+      "grad_norm": 0.5972458224831989,
+      "learning_rate": 1.891863440980897e-05,
+      "loss": 0.4386,
+      "step": 3048
+    },
+    {
+      "epoch": 0.6011435331230284,
+      "grad_norm": 0.6843819029628029,
+      "learning_rate": 1.8917933218533285e-05,
+      "loss": 0.4148,
+      "step": 3049
+    },
+    {
+      "epoch": 0.6013406940063092,
+      "grad_norm": 23.053542751701265,
+      "learning_rate": 1.891723181299708e-05,
+      "loss": 0.4442,
+      "step": 3050
+    },
+    {
+      "epoch": 0.6015378548895899,
+      "grad_norm": 0.6757852367698026,
+      "learning_rate": 1.8916530193217197e-05,
+      "loss": 0.411,
+      "step": 3051
+    },
+    {
+      "epoch": 0.6017350157728707,
+      "grad_norm": 0.6378151249084146,
+      "learning_rate": 1.8915828359210494e-05,
+      "loss": 0.4503,
+      "step": 3052
+    },
+    {
+      "epoch": 0.6019321766561514,
+      "grad_norm": 0.8860348502775296,
+      "learning_rate": 1.8915126310993838e-05,
+      "loss": 0.4358,
+      "step": 3053
+    },
+    {
+      "epoch": 0.6021293375394322,
+      "grad_norm": 0.6733989234081417,
+      "learning_rate": 1.891442404858409e-05,
+      "loss": 0.4989,
+      "step": 3054
+    },
+    {
+      "epoch": 0.6023264984227129,
+      "grad_norm": 0.6616823955387168,
+      "learning_rate": 1.891372157199813e-05,
+      "loss": 0.4833,
+      "step": 3055
+    },
+    {
+      "epoch": 0.6025236593059937,
+      "grad_norm": 0.613822177088368,
+      "learning_rate": 1.8913018881252827e-05,
+      "loss": 0.4309,
+      "step": 3056
+    },
+    {
+      "epoch": 0.6027208201892744,
+      "grad_norm": 0.7450747392214685,
+      "learning_rate": 1.891231597636507e-05,
+      "loss": 0.4752,
+      "step": 3057
+    },
+    {
+      "epoch": 0.6029179810725552,
+      "grad_norm": 0.5822085751848699,
+      "learning_rate": 1.8911612857351743e-05,
+      "loss": 0.4617,
+      "step": 3058
+    },
+    {
+      "epoch": 0.6031151419558359,
+      "grad_norm": 0.5976169928739026,
+      "learning_rate": 1.891090952422974e-05,
+      "loss": 0.4175,
+      "step": 3059
+    },
+    {
+      "epoch": 0.6033123028391167,
+      "grad_norm": 0.6141756747523544,
+      "learning_rate": 1.891020597701596e-05,
+      "loss": 0.4467,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6035094637223974,
+      "grad_norm": 0.7375579702582761,
+      "learning_rate": 1.890950221572731e-05,
+      "loss": 0.4993,
+      "step": 3061
+    },
+    {
+      "epoch": 0.6037066246056783,
+      "grad_norm": 0.6447777519058864,
+      "learning_rate": 1.8908798240380692e-05,
+      "loss": 0.4402,
+      "step": 3062
+    },
+    {
+      "epoch": 0.603903785488959,
+      "grad_norm": 0.572469473088046,
+      "learning_rate": 1.890809405099302e-05,
+      "loss": 0.4245,
+      "step": 3063
+    },
+    {
+      "epoch": 0.6041009463722398,
+      "grad_norm": 6.873905360016103,
+      "learning_rate": 1.8907389647581216e-05,
+      "loss": 0.5418,
+      "step": 3064
+    },
+    {
+      "epoch": 0.6042981072555205,
+      "grad_norm": 0.7088428326506333,
+      "learning_rate": 1.89066850301622e-05,
+      "loss": 0.4588,
+      "step": 3065
+    },
+    {
+      "epoch": 0.6044952681388013,
+      "grad_norm": 0.6487251096155348,
+      "learning_rate": 1.890598019875291e-05,
+      "loss": 0.47,
+      "step": 3066
+    },
+    {
+      "epoch": 0.604692429022082,
+      "grad_norm": 0.7195687916217166,
+      "learning_rate": 1.8905275153370272e-05,
+      "loss": 0.4444,
+      "step": 3067
+    },
+    {
+      "epoch": 0.6048895899053628,
+      "grad_norm": 0.6405425829888592,
+      "learning_rate": 1.890456989403122e-05,
+      "loss": 0.4242,
+      "step": 3068
+    },
+    {
+      "epoch": 0.6050867507886435,
+      "grad_norm": 1.1334558627993632,
+      "learning_rate": 1.8903864420752712e-05,
+      "loss": 0.497,
+      "step": 3069
+    },
+    {
+      "epoch": 0.6052839116719243,
+      "grad_norm": 0.612356344424891,
+      "learning_rate": 1.890315873355169e-05,
+      "loss": 0.3887,
+      "step": 3070
+    },
+    {
+      "epoch": 0.605481072555205,
+      "grad_norm": 0.666337306996091,
+      "learning_rate": 1.890245283244511e-05,
+      "loss": 0.4491,
+      "step": 3071
+    },
+    {
+      "epoch": 0.6056782334384858,
+      "grad_norm": 0.6409164321823476,
+      "learning_rate": 1.8901746717449932e-05,
+      "loss": 0.4323,
+      "step": 3072
+    },
+    {
+      "epoch": 0.6058753943217665,
+      "grad_norm": 0.6288119419802719,
+      "learning_rate": 1.8901040388583117e-05,
+      "loss": 0.4444,
+      "step": 3073
+    },
+    {
+      "epoch": 0.6060725552050473,
+      "grad_norm": 0.6764142603017349,
+      "learning_rate": 1.8900333845861643e-05,
+      "loss": 0.4598,
+      "step": 3074
+    },
+    {
+      "epoch": 0.606269716088328,
+      "grad_norm": 0.6209145176055464,
+      "learning_rate": 1.889962708930248e-05,
+      "loss": 0.4582,
+      "step": 3075
+    },
+    {
+      "epoch": 0.6064668769716088,
+      "grad_norm": 0.6837083145298952,
+      "learning_rate": 1.8898920118922607e-05,
+      "loss": 0.4073,
+      "step": 3076
+    },
+    {
+      "epoch": 0.6066640378548895,
+      "grad_norm": 0.6516951362240437,
+      "learning_rate": 1.8898212934739012e-05,
+      "loss": 0.4418,
+      "step": 3077
+    },
+    {
+      "epoch": 0.6068611987381703,
+      "grad_norm": 0.6228422659793167,
+      "learning_rate": 1.889750553676869e-05,
+      "loss": 0.4025,
+      "step": 3078
+    },
+    {
+      "epoch": 0.607058359621451,
+      "grad_norm": 0.9492283175824956,
+      "learning_rate": 1.8896797925028626e-05,
+      "loss": 0.4269,
+      "step": 3079
+    },
+    {
+      "epoch": 0.6072555205047319,
+      "grad_norm": 0.6988466626741697,
+      "learning_rate": 1.8896090099535834e-05,
+      "loss": 0.4966,
+      "step": 3080
+    },
+    {
+      "epoch": 0.6074526813880127,
+      "grad_norm": 1.6653405313238836,
+      "learning_rate": 1.889538206030731e-05,
+      "loss": 0.4327,
+      "step": 3081
+    },
+    {
+      "epoch": 0.6076498422712934,
+      "grad_norm": 0.6950411466750152,
+      "learning_rate": 1.8894673807360065e-05,
+      "loss": 0.4513,
+      "step": 3082
+    },
+    {
+      "epoch": 0.6078470031545742,
+      "grad_norm": 0.5702620043948126,
+      "learning_rate": 1.8893965340711126e-05,
+      "loss": 0.4481,
+      "step": 3083
+    },
+    {
+      "epoch": 0.6080441640378549,
+      "grad_norm": 0.6716600497276086,
+      "learning_rate": 1.8893256660377505e-05,
+      "loss": 0.4765,
+      "step": 3084
+    },
+    {
+      "epoch": 0.6082413249211357,
+      "grad_norm": 0.602345996688224,
+      "learning_rate": 1.8892547766376228e-05,
+      "loss": 0.4651,
+      "step": 3085
+    },
+    {
+      "epoch": 0.6084384858044164,
+      "grad_norm": 0.5853869541989889,
+      "learning_rate": 1.889183865872433e-05,
+      "loss": 0.4199,
+      "step": 3086
+    },
+    {
+      "epoch": 0.6086356466876972,
+      "grad_norm": 0.6334963552684939,
+      "learning_rate": 1.889112933743885e-05,
+      "loss": 0.4836,
+      "step": 3087
+    },
+    {
+      "epoch": 0.6088328075709779,
+      "grad_norm": 0.6800088424135952,
+      "learning_rate": 1.8890419802536826e-05,
+      "loss": 0.4875,
+      "step": 3088
+    },
+    {
+      "epoch": 0.6090299684542587,
+      "grad_norm": 0.6475039848768125,
+      "learning_rate": 1.888971005403531e-05,
+      "loss": 0.4679,
+      "step": 3089
+    },
+    {
+      "epoch": 0.6092271293375394,
+      "grad_norm": 0.6133780661367402,
+      "learning_rate": 1.8889000091951347e-05,
+      "loss": 0.4805,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6094242902208202,
+      "grad_norm": 0.6087154224375381,
+      "learning_rate": 1.8888289916302e-05,
+      "loss": 0.4193,
+      "step": 3091
+    },
+    {
+      "epoch": 0.6096214511041009,
+      "grad_norm": 0.6392859491176853,
+      "learning_rate": 1.8887579527104332e-05,
+      "loss": 0.4499,
+      "step": 3092
+    },
+    {
+      "epoch": 0.6098186119873817,
+      "grad_norm": 0.561076270997986,
+      "learning_rate": 1.8886868924375407e-05,
+      "loss": 0.4607,
+      "step": 3093
+    },
+    {
+      "epoch": 0.6100157728706624,
+      "grad_norm": 0.6104019667478391,
+      "learning_rate": 1.8886158108132298e-05,
+      "loss": 0.448,
+      "step": 3094
+    },
+    {
+      "epoch": 0.6102129337539433,
+      "grad_norm": 0.5694020567275695,
+      "learning_rate": 1.888544707839209e-05,
+      "loss": 0.4318,
+      "step": 3095
+    },
+    {
+      "epoch": 0.610410094637224,
+      "grad_norm": 0.6155013174910009,
+      "learning_rate": 1.888473583517185e-05,
+      "loss": 0.4129,
+      "step": 3096
+    },
+    {
+      "epoch": 0.6106072555205048,
+      "grad_norm": 0.8830383227880512,
+      "learning_rate": 1.8884024378488686e-05,
+      "loss": 0.4668,
+      "step": 3097
+    },
+    {
+      "epoch": 0.6108044164037855,
+      "grad_norm": 0.5411137975815842,
+      "learning_rate": 1.888331270835968e-05,
+      "loss": 0.4269,
+      "step": 3098
+    },
+    {
+      "epoch": 0.6110015772870663,
+      "grad_norm": 0.8549193019773949,
+      "learning_rate": 1.8882600824801932e-05,
+      "loss": 0.4349,
+      "step": 3099
+    },
+    {
+      "epoch": 0.611198738170347,
+      "grad_norm": 5.778893003349931,
+      "learning_rate": 1.888188872783255e-05,
+      "loss": 0.4817,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6113958990536278,
+      "grad_norm": 1.1949763025723064,
+      "learning_rate": 1.888117641746863e-05,
+      "loss": 0.4478,
+      "step": 3101
+    },
+    {
+      "epoch": 0.6115930599369085,
+      "grad_norm": 0.6510700265765879,
+      "learning_rate": 1.8880463893727297e-05,
+      "loss": 0.4496,
+      "step": 3102
+    },
+    {
+      "epoch": 0.6117902208201893,
+      "grad_norm": 0.5839708585126449,
+      "learning_rate": 1.8879751156625673e-05,
+      "loss": 0.4457,
+      "step": 3103
+    },
+    {
+      "epoch": 0.61198738170347,
+      "grad_norm": 17.48414582449077,
+      "learning_rate": 1.887903820618087e-05,
+      "loss": 0.4768,
+      "step": 3104
+    },
+    {
+      "epoch": 0.6121845425867508,
+      "grad_norm": 1.437092612882128,
+      "learning_rate": 1.887832504241003e-05,
+      "loss": 0.4374,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6123817034700315,
+      "grad_norm": 4.569861399699513,
+      "learning_rate": 1.887761166533028e-05,
+      "loss": 0.4938,
+      "step": 3106
+    },
+    {
+      "epoch": 0.6125788643533123,
+      "grad_norm": 0.8415642887494247,
+      "learning_rate": 1.8876898074958757e-05,
+      "loss": 0.4235,
+      "step": 3107
+    },
+    {
+      "epoch": 0.612776025236593,
+      "grad_norm": 0.7258259413681535,
+      "learning_rate": 1.887618427131261e-05,
+      "loss": 0.4572,
+      "step": 3108
+    },
+    {
+      "epoch": 0.6129731861198738,
+      "grad_norm": 0.6290488996802963,
+      "learning_rate": 1.887547025440899e-05,
+      "loss": 0.425,
+      "step": 3109
+    },
+    {
+      "epoch": 0.6131703470031545,
+      "grad_norm": 0.7318494963979986,
+      "learning_rate": 1.8874756024265045e-05,
+      "loss": 0.4631,
+      "step": 3110
+    },
+    {
+      "epoch": 0.6133675078864353,
+      "grad_norm": 0.6443190758237272,
+      "learning_rate": 1.8874041580897944e-05,
+      "loss": 0.4751,
+      "step": 3111
+    },
+    {
+      "epoch": 0.613564668769716,
+      "grad_norm": 0.6538478656310595,
+      "learning_rate": 1.887332692432485e-05,
+      "loss": 0.4124,
+      "step": 3112
+    },
+    {
+      "epoch": 0.6137618296529969,
+      "grad_norm": 0.6822067929384322,
+      "learning_rate": 1.8872612054562927e-05,
+      "loss": 0.4744,
+      "step": 3113
+    },
+    {
+      "epoch": 0.6139589905362776,
+      "grad_norm": 0.5861492452369412,
+      "learning_rate": 1.8871896971629356e-05,
+      "loss": 0.4494,
+      "step": 3114
+    },
+    {
+      "epoch": 0.6141561514195584,
+      "grad_norm": 0.6402603357742715,
+      "learning_rate": 1.8871181675541316e-05,
+      "loss": 0.4203,
+      "step": 3115
+    },
+    {
+      "epoch": 0.6143533123028391,
+      "grad_norm": 0.5974596546772335,
+      "learning_rate": 1.8870466166315992e-05,
+      "loss": 0.4682,
+      "step": 3116
+    },
+    {
+      "epoch": 0.6145504731861199,
+      "grad_norm": 0.6645668240438647,
+      "learning_rate": 1.8869750443970574e-05,
+      "loss": 0.461,
+      "step": 3117
+    },
+    {
+      "epoch": 0.6147476340694006,
+      "grad_norm": 0.5555905216873552,
+      "learning_rate": 1.8869034508522255e-05,
+      "loss": 0.4412,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6149447949526814,
+      "grad_norm": 0.6339546381509664,
+      "learning_rate": 1.8868318359988247e-05,
+      "loss": 0.4321,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6151419558359621,
+      "grad_norm": 0.5848309865797923,
+      "learning_rate": 1.8867601998385746e-05,
+      "loss": 0.4476,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6153391167192429,
+      "grad_norm": 0.6339063416959785,
+      "learning_rate": 1.8866885423731965e-05,
+      "loss": 0.4726,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6155362776025236,
+      "grad_norm": 0.6250487989240934,
+      "learning_rate": 1.8866168636044123e-05,
+      "loss": 0.4438,
+      "step": 3122
+    },
+    {
+      "epoch": 0.6157334384858044,
+      "grad_norm": 0.5912498590576298,
+      "learning_rate": 1.8865451635339437e-05,
+      "loss": 0.4249,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6159305993690851,
+      "grad_norm": 0.9964033422451408,
+      "learning_rate": 1.8864734421635138e-05,
+      "loss": 0.4533,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6161277602523659,
+      "grad_norm": 0.6783125045850023,
+      "learning_rate": 1.8864016994948456e-05,
+      "loss": 0.4399,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6163249211356467,
+      "grad_norm": 0.5640318304653738,
+      "learning_rate": 1.8863299355296626e-05,
+      "loss": 0.417,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6165220820189274,
+      "grad_norm": 0.7270371763719108,
+      "learning_rate": 1.8862581502696893e-05,
+      "loss": 0.4581,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6167192429022083,
+      "grad_norm": 0.5953459293919341,
+      "learning_rate": 1.8861863437166503e-05,
+      "loss": 0.4629,
+      "step": 3128
+    },
+    {
+      "epoch": 0.616916403785489,
+      "grad_norm": 0.7333601695695047,
+      "learning_rate": 1.8861145158722703e-05,
+      "loss": 0.4449,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6171135646687698,
+      "grad_norm": 0.6093823968435338,
+      "learning_rate": 1.886042666738276e-05,
+      "loss": 0.4538,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6173107255520505,
+      "grad_norm": 0.7379382803953564,
+      "learning_rate": 1.885970796316393e-05,
+      "loss": 0.422,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6175078864353313,
+      "grad_norm": 0.5910536949790164,
+      "learning_rate": 1.885898904608348e-05,
+      "loss": 0.4416,
+      "step": 3132
+    },
+    {
+      "epoch": 0.617705047318612,
+      "grad_norm": 0.7181285290226457,
+      "learning_rate": 1.8858269916158683e-05,
+      "loss": 0.4505,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6179022082018928,
+      "grad_norm": 0.6153838686175938,
+      "learning_rate": 1.885755057340682e-05,
+      "loss": 0.462,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6180993690851735,
+      "grad_norm": 0.6682026146168444,
+      "learning_rate": 1.8856831017845172e-05,
+      "loss": 0.4394,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6182965299684543,
+      "grad_norm": 0.6870178773294343,
+      "learning_rate": 1.885611124949102e-05,
+      "loss": 0.4573,
+      "step": 3136
+    },
+    {
+      "epoch": 0.618493690851735,
+      "grad_norm": 0.734375693308938,
+      "learning_rate": 1.8855391268361672e-05,
+      "loss": 0.4893,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6186908517350158,
+      "grad_norm": 0.6503117986853993,
+      "learning_rate": 1.8854671074474415e-05,
+      "loss": 0.4491,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6188880126182965,
+      "grad_norm": 0.6902398712205959,
+      "learning_rate": 1.8853950667846552e-05,
+      "loss": 0.4619,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6190851735015773,
+      "grad_norm": 0.5584656484231819,
+      "learning_rate": 1.8853230048495397e-05,
+      "loss": 0.4073,
+      "step": 3140
+    },
+    {
+      "epoch": 0.619282334384858,
+      "grad_norm": 0.6705513309518663,
+      "learning_rate": 1.885250921643826e-05,
+      "loss": 0.4545,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6194794952681388,
+      "grad_norm": 0.5384726667184218,
+      "learning_rate": 1.885178817169246e-05,
+      "loss": 0.3684,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6196766561514195,
+      "grad_norm": 0.706927829405689,
+      "learning_rate": 1.885106691427532e-05,
+      "loss": 0.4447,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6198738170347003,
+      "grad_norm": 0.553276508635115,
+      "learning_rate": 1.885034544420417e-05,
+      "loss": 0.4144,
+      "step": 3144
+    },
+    {
+      "epoch": 0.620070977917981,
+      "grad_norm": 0.6314672976429081,
+      "learning_rate": 1.8849623761496344e-05,
+      "loss": 0.4518,
+      "step": 3145
+    },
+    {
+      "epoch": 0.6202681388012619,
+      "grad_norm": 0.5953908729469908,
+      "learning_rate": 1.884890186616918e-05,
+      "loss": 0.4872,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6204652996845426,
+      "grad_norm": 0.6073305974604772,
+      "learning_rate": 1.884817975824002e-05,
+      "loss": 0.441,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6206624605678234,
+      "grad_norm": 0.6526389975490265,
+      "learning_rate": 1.884745743772622e-05,
+      "loss": 0.435,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6208596214511041,
+      "grad_norm": 0.6677687240737739,
+      "learning_rate": 1.884673490464513e-05,
+      "loss": 0.4322,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6210567823343849,
+      "grad_norm": 0.5838277331167949,
+      "learning_rate": 1.884601215901411e-05,
+      "loss": 0.4541,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6212539432176656,
+      "grad_norm": 0.5916911486373814,
+      "learning_rate": 1.8845289200850523e-05,
+      "loss": 0.4284,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6214511041009464,
+      "grad_norm": 0.5790112173160624,
+      "learning_rate": 1.8844566030171737e-05,
+      "loss": 0.4481,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6216482649842271,
+      "grad_norm": 1.1931766299083246,
+      "learning_rate": 1.8843842646995135e-05,
+      "loss": 0.4262,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6218454258675079,
+      "grad_norm": 0.6719849832897626,
+      "learning_rate": 1.884311905133809e-05,
+      "loss": 0.4364,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6220425867507886,
+      "grad_norm": 0.6366996603140915,
+      "learning_rate": 1.8842395243217986e-05,
+      "loss": 0.452,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6222397476340694,
+      "grad_norm": 0.6564134246631125,
+      "learning_rate": 1.884167122265222e-05,
+      "loss": 0.4725,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6224369085173501,
+      "grad_norm": 0.6718054415711968,
+      "learning_rate": 1.8840946989658175e-05,
+      "loss": 0.4179,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6226340694006309,
+      "grad_norm": 0.6323002874688755,
+      "learning_rate": 1.8840222544253265e-05,
+      "loss": 0.4829,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6228312302839116,
+      "grad_norm": 0.5594006708403959,
+      "learning_rate": 1.883949788645489e-05,
+      "loss": 0.4349,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6230283911671924,
+      "grad_norm": 0.6793405116890823,
+      "learning_rate": 1.8838773016280457e-05,
+      "loss": 0.4814,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6232255520504731,
+      "grad_norm": 0.5889819240809804,
+      "learning_rate": 1.8838047933747386e-05,
+      "loss": 0.477,
+      "step": 3161
+    },
+    {
+      "epoch": 0.623422712933754,
+      "grad_norm": 0.6015700310464448,
+      "learning_rate": 1.8837322638873093e-05,
+      "loss": 0.4701,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6236198738170347,
+      "grad_norm": 0.537156319132712,
+      "learning_rate": 1.883659713167501e-05,
+      "loss": 0.3994,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6238170347003155,
+      "grad_norm": 0.6309490765989089,
+      "learning_rate": 1.8835871412170563e-05,
+      "loss": 0.4657,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6240141955835962,
+      "grad_norm": 0.5932465191692712,
+      "learning_rate": 1.8835145480377194e-05,
+      "loss": 0.4406,
+      "step": 3165
+    },
+    {
+      "epoch": 0.624211356466877,
+      "grad_norm": 0.5843771388751763,
+      "learning_rate": 1.8834419336312334e-05,
+      "loss": 0.4644,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6244085173501577,
+      "grad_norm": 0.5350744617276989,
+      "learning_rate": 1.8833692979993437e-05,
+      "loss": 0.4334,
+      "step": 3167
+    },
+    {
+      "epoch": 0.6246056782334385,
+      "grad_norm": 0.6032969293393299,
+      "learning_rate": 1.8832966411437958e-05,
+      "loss": 0.4516,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6248028391167192,
+      "grad_norm": 0.5402329623043484,
+      "learning_rate": 1.883223963066334e-05,
+      "loss": 0.4669,
+      "step": 3169
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.6420727536456771,
+      "learning_rate": 1.8831512637687054e-05,
+      "loss": 0.4933,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6251971608832808,
+      "grad_norm": 0.6837145058399116,
+      "learning_rate": 1.8830785432526568e-05,
+      "loss": 0.4316,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6253943217665615,
+      "grad_norm": 0.5669487608780952,
+      "learning_rate": 1.883005801519935e-05,
+      "loss": 0.4435,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6255914826498423,
+      "grad_norm": 0.5500128986494645,
+      "learning_rate": 1.8829330385722875e-05,
+      "loss": 0.4352,
+      "step": 3173
+    },
+    {
+      "epoch": 0.625788643533123,
+      "grad_norm": 0.659708848221679,
+      "learning_rate": 1.882860254411463e-05,
+      "loss": 0.4302,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6259858044164038,
+      "grad_norm": 0.5634015211579477,
+      "learning_rate": 1.8827874490392095e-05,
+      "loss": 0.4344,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6261829652996845,
+      "grad_norm": 0.5511225422280275,
+      "learning_rate": 1.882714622457277e-05,
+      "loss": 0.4493,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6263801261829653,
+      "grad_norm": 0.5535762556867952,
+      "learning_rate": 1.882641774667415e-05,
+      "loss": 0.4786,
+      "step": 3177
+    },
+    {
+      "epoch": 0.626577287066246,
+      "grad_norm": 0.7296756084321033,
+      "learning_rate": 1.8825689056713733e-05,
+      "loss": 0.4107,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6267744479495269,
+      "grad_norm": 0.5691759534943049,
+      "learning_rate": 1.8824960154709027e-05,
+      "loss": 0.4937,
+      "step": 3179
+    },
+    {
+      "epoch": 0.6269716088328076,
+      "grad_norm": 0.5643398502645974,
+      "learning_rate": 1.882423104067755e-05,
+      "loss": 0.4371,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6271687697160884,
+      "grad_norm": 0.5414110934753018,
+      "learning_rate": 1.8823501714636815e-05,
+      "loss": 0.4372,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6273659305993691,
+      "grad_norm": 0.5716798244654584,
+      "learning_rate": 1.8822772176604346e-05,
+      "loss": 0.4386,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6275630914826499,
+      "grad_norm": 0.5594732072778538,
+      "learning_rate": 1.882204242659767e-05,
+      "loss": 0.4415,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6277602523659306,
+      "grad_norm": 0.5968240035493283,
+      "learning_rate": 1.8821312464634318e-05,
+      "loss": 0.4651,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6279574132492114,
+      "grad_norm": 0.5050689025421782,
+      "learning_rate": 1.8820582290731836e-05,
+      "loss": 0.3813,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6281545741324921,
+      "grad_norm": 0.571720125704504,
+      "learning_rate": 1.8819851904907756e-05,
+      "loss": 0.4117,
+      "step": 3186
+    },
+    {
+      "epoch": 0.6283517350157729,
+      "grad_norm": 0.6099795491237106,
+      "learning_rate": 1.8819121307179634e-05,
+      "loss": 0.4739,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6285488958990536,
+      "grad_norm": 0.5516740130300875,
+      "learning_rate": 1.881839049756502e-05,
+      "loss": 0.4132,
+      "step": 3188
+    },
+    {
+      "epoch": 0.6287460567823344,
+      "grad_norm": 0.6078983072011943,
+      "learning_rate": 1.8817659476081474e-05,
+      "loss": 0.4612,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6289432176656151,
+      "grad_norm": 0.5561987130798062,
+      "learning_rate": 1.8816928242746554e-05,
+      "loss": 0.4411,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6291403785488959,
+      "grad_norm": 0.5499939289696452,
+      "learning_rate": 1.8816196797577838e-05,
+      "loss": 0.4462,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6293375394321766,
+      "grad_norm": 0.565448729109912,
+      "learning_rate": 1.881546514059289e-05,
+      "loss": 0.4426,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6295347003154574,
+      "grad_norm": 0.5466556054488702,
+      "learning_rate": 1.8814733271809296e-05,
+      "loss": 0.4633,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6297318611987381,
+      "grad_norm": 0.6054836321256308,
+      "learning_rate": 1.8814001191244636e-05,
+      "loss": 0.4592,
+      "step": 3194
+    },
+    {
+      "epoch": 0.629929022082019,
+      "grad_norm": 0.6252405670685508,
+      "learning_rate": 1.8813268898916498e-05,
+      "loss": 0.4588,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6301261829652997,
+      "grad_norm": 0.5210693946541807,
+      "learning_rate": 1.881253639484248e-05,
+      "loss": 0.4461,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6303233438485805,
+      "grad_norm": 0.5953514887490388,
+      "learning_rate": 1.8811803679040178e-05,
+      "loss": 0.4862,
+      "step": 3197
+    },
+    {
+      "epoch": 0.6305205047318612,
+      "grad_norm": 0.6102071586071421,
+      "learning_rate": 1.8811070751527196e-05,
+      "loss": 0.4214,
+      "step": 3198
+    },
+    {
+      "epoch": 0.630717665615142,
+      "grad_norm": 0.6100495554396542,
+      "learning_rate": 1.8810337612321144e-05,
+      "loss": 0.4648,
+      "step": 3199
+    },
+    {
+      "epoch": 0.6309148264984227,
+      "grad_norm": 0.5197551849910016,
+      "learning_rate": 1.8809604261439634e-05,
+      "loss": 0.4382,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6311119873817035,
+      "grad_norm": 0.5490144964409054,
+      "learning_rate": 1.8808870698900288e-05,
+      "loss": 0.4477,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6313091482649842,
+      "grad_norm": 0.5628554485158465,
+      "learning_rate": 1.880813692472073e-05,
+      "loss": 0.4605,
+      "step": 3202
+    },
+    {
+      "epoch": 0.631506309148265,
+      "grad_norm": 0.8650405899982267,
+      "learning_rate": 1.8807402938918588e-05,
+      "loss": 0.4288,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6317034700315457,
+      "grad_norm": 0.5546723666140645,
+      "learning_rate": 1.88066687415115e-05,
+      "loss": 0.4412,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6319006309148265,
+      "grad_norm": 0.5808634529336613,
+      "learning_rate": 1.8805934332517104e-05,
+      "loss": 0.4733,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6320977917981072,
+      "grad_norm": 0.6305165897108991,
+      "learning_rate": 1.880519971195304e-05,
+      "loss": 0.4469,
+      "step": 3206
+    },
+    {
+      "epoch": 0.632294952681388,
+      "grad_norm": 0.5672855153253277,
+      "learning_rate": 1.880446487983697e-05,
+      "loss": 0.4699,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6324921135646687,
+      "grad_norm": 0.5959371940317938,
+      "learning_rate": 1.880372983618653e-05,
+      "loss": 0.481,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6326892744479495,
+      "grad_norm": 0.5264066379704362,
+      "learning_rate": 1.88029945810194e-05,
+      "loss": 0.4093,
+      "step": 3209
+    },
+    {
+      "epoch": 0.6328864353312302,
+      "grad_norm": 0.5705739208816702,
+      "learning_rate": 1.880225911435323e-05,
+      "loss": 0.4517,
+      "step": 3210
+    },
+    {
+      "epoch": 0.633083596214511,
+      "grad_norm": 0.6114799226361008,
+      "learning_rate": 1.88015234362057e-05,
+      "loss": 0.4527,
+      "step": 3211
+    },
+    {
+      "epoch": 0.6332807570977917,
+      "grad_norm": 0.5837271501297863,
+      "learning_rate": 1.8800787546594487e-05,
+      "loss": 0.4596,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6334779179810726,
+      "grad_norm": 0.5577201101577096,
+      "learning_rate": 1.8800051445537256e-05,
+      "loss": 0.4577,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6336750788643533,
+      "grad_norm": 0.5736984526638703,
+      "learning_rate": 1.8799315133051707e-05,
+      "loss": 0.4451,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6338722397476341,
+      "grad_norm": 0.7258407481830798,
+      "learning_rate": 1.8798578609155528e-05,
+      "loss": 0.4459,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6340694006309149,
+      "grad_norm": 0.5742874979476474,
+      "learning_rate": 1.8797841873866406e-05,
+      "loss": 0.4688,
+      "step": 3216
+    },
+    {
+      "epoch": 0.6342665615141956,
+      "grad_norm": 0.5334223008485237,
+      "learning_rate": 1.8797104927202055e-05,
+      "loss": 0.4496,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6344637223974764,
+      "grad_norm": 0.5706148443360702,
+      "learning_rate": 1.879636776918017e-05,
+      "loss": 0.4617,
+      "step": 3218
+    },
+    {
+      "epoch": 0.6346608832807571,
+      "grad_norm": 0.8529042316450798,
+      "learning_rate": 1.8795630399818466e-05,
+      "loss": 0.4704,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6348580441640379,
+      "grad_norm": 0.5624250580860771,
+      "learning_rate": 1.8794892819134657e-05,
+      "loss": 0.4725,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6350552050473186,
+      "grad_norm": 0.5838011246869812,
+      "learning_rate": 1.8794155027146468e-05,
+      "loss": 0.4641,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6352523659305994,
+      "grad_norm": 0.564379408051252,
+      "learning_rate": 1.879341702387162e-05,
+      "loss": 0.4536,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6354495268138801,
+      "grad_norm": 0.5730447220643163,
+      "learning_rate": 1.8792678809327852e-05,
+      "loss": 0.4228,
+      "step": 3223
+    },
+    {
+      "epoch": 0.6356466876971609,
+      "grad_norm": 0.9816181582279039,
+      "learning_rate": 1.879194038353289e-05,
+      "loss": 0.4869,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6358438485804416,
+      "grad_norm": 0.5470063459656487,
+      "learning_rate": 1.8791201746504485e-05,
+      "loss": 0.4252,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6360410094637224,
+      "grad_norm": 0.5839122977250125,
+      "learning_rate": 1.8790462898260373e-05,
+      "loss": 0.3985,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6362381703470031,
+      "grad_norm": 0.601019249469374,
+      "learning_rate": 1.8789723838818314e-05,
+      "loss": 0.4639,
+      "step": 3227
+    },
+    {
+      "epoch": 0.636435331230284,
+      "grad_norm": 0.6426956793821633,
+      "learning_rate": 1.878898456819606e-05,
+      "loss": 0.4722,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6366324921135647,
+      "grad_norm": 0.5652057471720214,
+      "learning_rate": 1.878824508641137e-05,
+      "loss": 0.4385,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6368296529968455,
+      "grad_norm": 0.634391516253938,
+      "learning_rate": 1.8787505393482023e-05,
+      "loss": 0.4861,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6370268138801262,
+      "grad_norm": 0.5276142848529989,
+      "learning_rate": 1.8786765489425776e-05,
+      "loss": 0.4167,
+      "step": 3231
+    },
+    {
+      "epoch": 0.637223974763407,
+      "grad_norm": 0.642092919755005,
+      "learning_rate": 1.8786025374260418e-05,
+      "loss": 0.4447,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6374211356466877,
+      "grad_norm": 0.6622984389500397,
+      "learning_rate": 1.8785285048003722e-05,
+      "loss": 0.4509,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6376182965299685,
+      "grad_norm": 0.6059786233435729,
+      "learning_rate": 1.8784544510673477e-05,
+      "loss": 0.4328,
+      "step": 3234
+    },
+    {
+      "epoch": 0.6378154574132492,
+      "grad_norm": 0.5608277860083667,
+      "learning_rate": 1.8783803762287477e-05,
+      "loss": 0.4403,
+      "step": 3235
+    },
+    {
+      "epoch": 0.63801261829653,
+      "grad_norm": 0.5705812031919998,
+      "learning_rate": 1.8783062802863516e-05,
+      "loss": 0.447,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6382097791798107,
+      "grad_norm": 0.6099064166899052,
+      "learning_rate": 1.8782321632419402e-05,
+      "loss": 0.4635,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6384069400630915,
+      "grad_norm": 0.5789456791219298,
+      "learning_rate": 1.8781580250972933e-05,
+      "loss": 0.453,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6386041009463722,
+      "grad_norm": 0.5435966426313446,
+      "learning_rate": 1.8780838658541932e-05,
+      "loss": 0.4088,
+      "step": 3239
+    },
+    {
+      "epoch": 0.638801261829653,
+      "grad_norm": 0.5637574474073005,
+      "learning_rate": 1.878009685514421e-05,
+      "loss": 0.4516,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6389984227129337,
+      "grad_norm": 0.6583384784739884,
+      "learning_rate": 1.8779354840797588e-05,
+      "loss": 0.454,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6391955835962145,
+      "grad_norm": 0.5554247237363399,
+      "learning_rate": 1.87786126155199e-05,
+      "loss": 0.4567,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6393927444794952,
+      "grad_norm": 0.5657140987582041,
+      "learning_rate": 1.877787017932897e-05,
+      "loss": 0.4716,
+      "step": 3243
+    },
+    {
+      "epoch": 0.639589905362776,
+      "grad_norm": 0.5498999624708573,
+      "learning_rate": 1.8777127532242643e-05,
+      "loss": 0.4527,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6397870662460567,
+      "grad_norm": 0.597593319418814,
+      "learning_rate": 1.8776384674278756e-05,
+      "loss": 0.4409,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6399842271293376,
+      "grad_norm": 0.5970410292124881,
+      "learning_rate": 1.8775641605455162e-05,
+      "loss": 0.4716,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6401813880126183,
+      "grad_norm": 0.5712017944124336,
+      "learning_rate": 1.877489832578971e-05,
+      "loss": 0.4408,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6403785488958991,
+      "grad_norm": 0.5264381336103525,
+      "learning_rate": 1.877415483530026e-05,
+      "loss": 0.4419,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6405757097791798,
+      "grad_norm": 0.6131589523064833,
+      "learning_rate": 1.8773411134004677e-05,
+      "loss": 0.4544,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6407728706624606,
+      "grad_norm": 0.5718419218962923,
+      "learning_rate": 1.8772667221920823e-05,
+      "loss": 0.4296,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6409700315457413,
+      "grad_norm": 0.578624341262867,
+      "learning_rate": 1.8771923099066573e-05,
+      "loss": 0.4605,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6411671924290221,
+      "grad_norm": 0.5787190935133748,
+      "learning_rate": 1.877117876545981e-05,
+      "loss": 0.4434,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6413643533123028,
+      "grad_norm": 0.5634817233410009,
+      "learning_rate": 1.877043422111841e-05,
+      "loss": 0.4614,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6415615141955836,
+      "grad_norm": 0.5510386384849804,
+      "learning_rate": 1.876968946606027e-05,
+      "loss": 0.4477,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6417586750788643,
+      "grad_norm": 0.6031650127924183,
+      "learning_rate": 1.8768944500303276e-05,
+      "loss": 0.4528,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6419558359621451,
+      "grad_norm": 0.563947528442472,
+      "learning_rate": 1.876819932386533e-05,
+      "loss": 0.4464,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6421529968454258,
+      "grad_norm": 0.562104581809266,
+      "learning_rate": 1.8767453936764332e-05,
+      "loss": 0.4448,
+      "step": 3257
+    },
+    {
+      "epoch": 0.6423501577287066,
+      "grad_norm": 0.5711382767124764,
+      "learning_rate": 1.876670833901819e-05,
+      "loss": 0.432,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6425473186119873,
+      "grad_norm": 0.5697395824924175,
+      "learning_rate": 1.8765962530644826e-05,
+      "loss": 0.446,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6427444794952681,
+      "grad_norm": 0.5940770780237805,
+      "learning_rate": 1.8765216511662153e-05,
+      "loss": 0.4551,
+      "step": 3260
+    },
+    {
+      "epoch": 0.642941640378549,
+      "grad_norm": 0.9903145042557886,
+      "learning_rate": 1.876447028208809e-05,
+      "loss": 0.4633,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6431388012618297,
+      "grad_norm": 0.5566812407384747,
+      "learning_rate": 1.8763723841940576e-05,
+      "loss": 0.4058,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6433359621451105,
+      "grad_norm": 0.6250449584024177,
+      "learning_rate": 1.8762977191237536e-05,
+      "loss": 0.4706,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6435331230283912,
+      "grad_norm": 2.210243090857098,
+      "learning_rate": 1.876223032999691e-05,
+      "loss": 0.4406,
+      "step": 3264
+    },
+    {
+      "epoch": 0.643730283911672,
+      "grad_norm": 0.6481804636618524,
+      "learning_rate": 1.876148325823665e-05,
+      "loss": 0.432,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6439274447949527,
+      "grad_norm": 0.6303860129263393,
+      "learning_rate": 1.8760735975974693e-05,
+      "loss": 0.4137,
+      "step": 3266
+    },
+    {
+      "epoch": 0.6441246056782335,
+      "grad_norm": 0.6383795817729836,
+      "learning_rate": 1.8759988483229e-05,
+      "loss": 0.4554,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6443217665615142,
+      "grad_norm": 0.6875260817255804,
+      "learning_rate": 1.8759240780017534e-05,
+      "loss": 0.443,
+      "step": 3268
+    },
+    {
+      "epoch": 0.644518927444795,
+      "grad_norm": 0.6662104585984299,
+      "learning_rate": 1.875849286635825e-05,
+      "loss": 0.4445,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6447160883280757,
+      "grad_norm": 0.6171930410732172,
+      "learning_rate": 1.8757744742269123e-05,
+      "loss": 0.4374,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6449132492113565,
+      "grad_norm": 0.7933975678175971,
+      "learning_rate": 1.8756996407768128e-05,
+      "loss": 0.4766,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6451104100946372,
+      "grad_norm": 1.0409394957016507,
+      "learning_rate": 1.875624786287324e-05,
+      "loss": 0.4418,
+      "step": 3272
+    },
+    {
+      "epoch": 0.645307570977918,
+      "grad_norm": 0.5947622881166746,
+      "learning_rate": 1.875549910760245e-05,
+      "loss": 0.4575,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6455047318611987,
+      "grad_norm": 0.5670316521357006,
+      "learning_rate": 1.875475014197374e-05,
+      "loss": 0.4488,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6457018927444795,
+      "grad_norm": 0.6158749237920658,
+      "learning_rate": 1.8754000966005105e-05,
+      "loss": 0.4173,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6458990536277602,
+      "grad_norm": 0.6765411423930585,
+      "learning_rate": 1.8753251579714548e-05,
+      "loss": 0.4131,
+      "step": 3276
+    },
+    {
+      "epoch": 0.646096214511041,
+      "grad_norm": 0.5338915194924335,
+      "learning_rate": 1.8752501983120076e-05,
+      "loss": 0.4262,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6462933753943217,
+      "grad_norm": 0.566003677186254,
+      "learning_rate": 1.8751752176239693e-05,
+      "loss": 0.4208,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6464905362776026,
+      "grad_norm": 0.6018668670309205,
+      "learning_rate": 1.8751002159091415e-05,
+      "loss": 0.4591,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6466876971608833,
+      "grad_norm": 0.6534805827884597,
+      "learning_rate": 1.8750251931693265e-05,
+      "loss": 0.4538,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6468848580441641,
+      "grad_norm": 0.5692551760672482,
+      "learning_rate": 1.8749501494063266e-05,
+      "loss": 0.4171,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6470820189274448,
+      "grad_norm": 0.6447690381178922,
+      "learning_rate": 1.874875084621945e-05,
+      "loss": 0.4436,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6472791798107256,
+      "grad_norm": 0.531474927660677,
+      "learning_rate": 1.8747999988179846e-05,
+      "loss": 0.4287,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6474763406940063,
+      "grad_norm": 0.5470438621167268,
+      "learning_rate": 1.8747248919962498e-05,
+      "loss": 0.4417,
+      "step": 3284
+    },
+    {
+      "epoch": 0.6476735015772871,
+      "grad_norm": 0.5741916924212119,
+      "learning_rate": 1.874649764158545e-05,
+      "loss": 0.425,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6478706624605678,
+      "grad_norm": 0.5348263622053581,
+      "learning_rate": 1.8745746153066756e-05,
+      "loss": 0.4169,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6480678233438486,
+      "grad_norm": 0.5607951226714928,
+      "learning_rate": 1.8744994454424463e-05,
+      "loss": 0.4375,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6482649842271293,
+      "grad_norm": 0.6004093976459436,
+      "learning_rate": 1.874424254567664e-05,
+      "loss": 0.4482,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6484621451104101,
+      "grad_norm": 0.5668829924890922,
+      "learning_rate": 1.8743490426841346e-05,
+      "loss": 0.4462,
+      "step": 3289
+    },
+    {
+      "epoch": 0.6486593059936908,
+      "grad_norm": 0.7152077392581754,
+      "learning_rate": 1.8742738097936653e-05,
+      "loss": 0.4899,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6488564668769716,
+      "grad_norm": 0.5722777962307597,
+      "learning_rate": 1.874198555898064e-05,
+      "loss": 0.4618,
+      "step": 3291
+    },
+    {
+      "epoch": 0.6490536277602523,
+      "grad_norm": 0.6284802948397332,
+      "learning_rate": 1.874123280999138e-05,
+      "loss": 0.415,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6492507886435331,
+      "grad_norm": 0.6162363187667032,
+      "learning_rate": 1.8740479850986962e-05,
+      "loss": 0.4557,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6494479495268138,
+      "grad_norm": 0.616624529665098,
+      "learning_rate": 1.8739726681985478e-05,
+      "loss": 0.4604,
+      "step": 3294
+    },
+    {
+      "epoch": 0.6496451104100947,
+      "grad_norm": 0.5855562263616807,
+      "learning_rate": 1.8738973303005024e-05,
+      "loss": 0.4637,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6498422712933754,
+      "grad_norm": 0.6041841740545615,
+      "learning_rate": 1.87382197140637e-05,
+      "loss": 0.4623,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6500394321766562,
+      "grad_norm": 4.533496170121018,
+      "learning_rate": 1.873746591517961e-05,
+      "loss": 0.4471,
+      "step": 3297
+    },
+    {
+      "epoch": 0.6502365930599369,
+      "grad_norm": 0.6624890927859284,
+      "learning_rate": 1.873671190637086e-05,
+      "loss": 0.4267,
+      "step": 3298
+    },
+    {
+      "epoch": 0.6504337539432177,
+      "grad_norm": 0.5905974277110145,
+      "learning_rate": 1.8735957687655577e-05,
+      "loss": 0.4125,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6506309148264984,
+      "grad_norm": 0.6041600898310406,
+      "learning_rate": 1.8735203259051872e-05,
+      "loss": 0.454,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6508280757097792,
+      "grad_norm": 0.6390231671362254,
+      "learning_rate": 1.8734448620577875e-05,
+      "loss": 0.4887,
+      "step": 3301
+    },
+    {
+      "epoch": 0.6510252365930599,
+      "grad_norm": 0.6117066782843804,
+      "learning_rate": 1.8733693772251716e-05,
+      "loss": 0.4556,
+      "step": 3302
+    },
+    {
+      "epoch": 0.6512223974763407,
+      "grad_norm": 0.5711747386901173,
+      "learning_rate": 1.873293871409153e-05,
+      "loss": 0.4204,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6514195583596214,
+      "grad_norm": 0.6143464512727876,
+      "learning_rate": 1.8732183446115462e-05,
+      "loss": 0.4434,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6516167192429022,
+      "grad_norm": 0.5417403420673808,
+      "learning_rate": 1.8731427968341654e-05,
+      "loss": 0.4246,
+      "step": 3305
+    },
+    {
+      "epoch": 0.651813880126183,
+      "grad_norm": 0.6890669641904964,
+      "learning_rate": 1.8730672280788254e-05,
+      "loss": 0.4989,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6520110410094637,
+      "grad_norm": 0.5499879128399455,
+      "learning_rate": 1.8729916383473427e-05,
+      "loss": 0.4529,
+      "step": 3307
+    },
+    {
+      "epoch": 0.6522082018927445,
+      "grad_norm": 0.5830578121646517,
+      "learning_rate": 1.8729160276415325e-05,
+      "loss": 0.4023,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6524053627760252,
+      "grad_norm": 0.5667181804512541,
+      "learning_rate": 1.872840395963212e-05,
+      "loss": 0.4491,
+      "step": 3309
+    },
+    {
+      "epoch": 0.652602523659306,
+      "grad_norm": 0.5766356519567429,
+      "learning_rate": 1.872764743314198e-05,
+      "loss": 0.4666,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6527996845425867,
+      "grad_norm": 0.5432561051881504,
+      "learning_rate": 1.872689069696308e-05,
+      "loss": 0.4311,
+      "step": 3311
+    },
+    {
+      "epoch": 0.6529968454258676,
+      "grad_norm": 0.5811877061331849,
+      "learning_rate": 1.8726133751113605e-05,
+      "loss": 0.4351,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6531940063091483,
+      "grad_norm": 0.5756974097066245,
+      "learning_rate": 1.872537659561174e-05,
+      "loss": 0.4561,
+      "step": 3313
+    },
+    {
+      "epoch": 0.6533911671924291,
+      "grad_norm": 0.5313182062056155,
+      "learning_rate": 1.8724619230475675e-05,
+      "loss": 0.3902,
+      "step": 3314
+    },
+    {
+      "epoch": 0.6535883280757098,
+      "grad_norm": 0.5497309033214044,
+      "learning_rate": 1.872386165572361e-05,
+      "loss": 0.4263,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6537854889589906,
+      "grad_norm": 0.5466469389230997,
+      "learning_rate": 1.872310387137374e-05,
+      "loss": 0.4224,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6539826498422713,
+      "grad_norm": 2.202702375918148,
+      "learning_rate": 1.872234587744427e-05,
+      "loss": 0.4472,
+      "step": 3317
+    },
+    {
+      "epoch": 0.6541798107255521,
+      "grad_norm": 0.5869148492580374,
+      "learning_rate": 1.8721587673953425e-05,
+      "loss": 0.4671,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6543769716088328,
+      "grad_norm": 0.5427996690221996,
+      "learning_rate": 1.8720829260919407e-05,
+      "loss": 0.4289,
+      "step": 3319
+    },
+    {
+      "epoch": 0.6545741324921136,
+      "grad_norm": 0.5656017555414672,
+      "learning_rate": 1.8720070638360447e-05,
+      "loss": 0.4294,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6547712933753943,
+      "grad_norm": 4.358651338020947,
+      "learning_rate": 1.8719311806294768e-05,
+      "loss": 0.5382,
+      "step": 3321
+    },
+    {
+      "epoch": 0.6549684542586751,
+      "grad_norm": 0.6865845679707644,
+      "learning_rate": 1.87185527647406e-05,
+      "loss": 0.4885,
+      "step": 3322
+    },
+    {
+      "epoch": 0.6551656151419558,
+      "grad_norm": 1.9321821824405176,
+      "learning_rate": 1.871779351371618e-05,
+      "loss": 0.4279,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6553627760252366,
+      "grad_norm": 0.6207436028024012,
+      "learning_rate": 1.8717034053239748e-05,
+      "loss": 0.4415,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6555599369085173,
+      "grad_norm": 0.7726441299810708,
+      "learning_rate": 1.8716274383329556e-05,
+      "loss": 0.4166,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6557570977917981,
+      "grad_norm": 1.3763107640662533,
+      "learning_rate": 1.8715514504003854e-05,
+      "loss": 0.4449,
+      "step": 3326
+    },
+    {
+      "epoch": 0.6559542586750788,
+      "grad_norm": 0.6496821405637063,
+      "learning_rate": 1.8714754415280894e-05,
+      "loss": 0.451,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6561514195583596,
+      "grad_norm": 0.6130036203951224,
+      "learning_rate": 1.8713994117178945e-05,
+      "loss": 0.4593,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6563485804416404,
+      "grad_norm": 0.5725117299905743,
+      "learning_rate": 1.8713233609716266e-05,
+      "loss": 0.4278,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6565457413249212,
+      "grad_norm": 0.5445827811729219,
+      "learning_rate": 1.8712472892911132e-05,
+      "loss": 0.4101,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6567429022082019,
+      "grad_norm": 0.7500084982788475,
+      "learning_rate": 1.8711711966781826e-05,
+      "loss": 0.4248,
+      "step": 3331
+    },
+    {
+      "epoch": 0.6569400630914827,
+      "grad_norm": 0.6621959429007946,
+      "learning_rate": 1.8710950831346623e-05,
+      "loss": 0.4908,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6571372239747634,
+      "grad_norm": 2.2676905068146054,
+      "learning_rate": 1.871018948662381e-05,
+      "loss": 0.498,
+      "step": 3333
+    },
+    {
+      "epoch": 0.6573343848580442,
+      "grad_norm": 0.7294312287786445,
+      "learning_rate": 1.870942793263168e-05,
+      "loss": 0.4693,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6575315457413249,
+      "grad_norm": 0.5923749861108738,
+      "learning_rate": 1.870866616938853e-05,
+      "loss": 0.4634,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6577287066246057,
+      "grad_norm": 0.6623239066356625,
+      "learning_rate": 1.870790419691266e-05,
+      "loss": 0.4434,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6579258675078864,
+      "grad_norm": 0.6507535977225328,
+      "learning_rate": 1.8707142015222386e-05,
+      "loss": 0.4619,
+      "step": 3337
+    },
+    {
+      "epoch": 0.6581230283911672,
+      "grad_norm": 3.387004003722433,
+      "learning_rate": 1.870637962433601e-05,
+      "loss": 0.5185,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6583201892744479,
+      "grad_norm": 0.7986315924274782,
+      "learning_rate": 1.870561702427185e-05,
+      "loss": 0.441,
+      "step": 3339
+    },
+    {
+      "epoch": 0.6585173501577287,
+      "grad_norm": 0.5803415335213525,
+      "learning_rate": 1.870485421504823e-05,
+      "loss": 0.4293,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6587145110410094,
+      "grad_norm": 0.7703993364474427,
+      "learning_rate": 1.870409119668348e-05,
+      "loss": 0.4929,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6589116719242902,
+      "grad_norm": 0.6513075944835972,
+      "learning_rate": 1.870332796919593e-05,
+      "loss": 0.4607,
+      "step": 3342
+    },
+    {
+      "epoch": 0.6591088328075709,
+      "grad_norm": 0.6750849955698458,
+      "learning_rate": 1.8702564532603917e-05,
+      "loss": 0.439,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6593059936908517,
+      "grad_norm": 0.6375890019239625,
+      "learning_rate": 1.8701800886925784e-05,
+      "loss": 0.4726,
+      "step": 3344
+    },
+    {
+      "epoch": 0.6595031545741324,
+      "grad_norm": 0.65991248253205,
+      "learning_rate": 1.8701037032179873e-05,
+      "loss": 0.4631,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6597003154574133,
+      "grad_norm": 0.8742592115103047,
+      "learning_rate": 1.870027296838454e-05,
+      "loss": 0.4502,
+      "step": 3346
+    },
+    {
+      "epoch": 0.659897476340694,
+      "grad_norm": 0.6581694810065215,
+      "learning_rate": 1.8699508695558145e-05,
+      "loss": 0.4371,
+      "step": 3347
+    },
+    {
+      "epoch": 0.6600946372239748,
+      "grad_norm": 0.7348305014231767,
+      "learning_rate": 1.869874421371905e-05,
+      "loss": 0.4287,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6602917981072555,
+      "grad_norm": 0.6216191228474292,
+      "learning_rate": 1.8697979522885617e-05,
+      "loss": 0.441,
+      "step": 3349
+    },
+    {
+      "epoch": 0.6604889589905363,
+      "grad_norm": 0.7451526119076238,
+      "learning_rate": 1.8697214623076222e-05,
+      "loss": 0.4419,
+      "step": 3350
+    },
+    {
+      "epoch": 0.660686119873817,
+      "grad_norm": 0.5518756188697328,
+      "learning_rate": 1.8696449514309244e-05,
+      "loss": 0.4176,
+      "step": 3351
+    },
+    {
+      "epoch": 0.6608832807570978,
+      "grad_norm": 0.6734594739824642,
+      "learning_rate": 1.869568419660306e-05,
+      "loss": 0.4669,
+      "step": 3352
+    },
+    {
+      "epoch": 0.6610804416403786,
+      "grad_norm": 0.679456190018415,
+      "learning_rate": 1.8694918669976063e-05,
+      "loss": 0.456,
+      "step": 3353
+    },
+    {
+      "epoch": 0.6612776025236593,
+      "grad_norm": 0.5841633102935858,
+      "learning_rate": 1.8694152934446642e-05,
+      "loss": 0.4252,
+      "step": 3354
+    },
+    {
+      "epoch": 0.6614747634069401,
+      "grad_norm": 0.5737135644711173,
+      "learning_rate": 1.8693386990033194e-05,
+      "loss": 0.4152,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6616719242902208,
+      "grad_norm": 0.6231408225081773,
+      "learning_rate": 1.8692620836754124e-05,
+      "loss": 0.4393,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6618690851735016,
+      "grad_norm": 0.7428753795745229,
+      "learning_rate": 1.8691854474627838e-05,
+      "loss": 0.4596,
+      "step": 3357
+    },
+    {
+      "epoch": 0.6620662460567823,
+      "grad_norm": 0.6536356718789794,
+      "learning_rate": 1.8691087903672752e-05,
+      "loss": 0.4548,
+      "step": 3358
+    },
+    {
+      "epoch": 0.6622634069400631,
+      "grad_norm": 0.6121171634606317,
+      "learning_rate": 1.8690321123907277e-05,
+      "loss": 0.4633,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6624605678233438,
+      "grad_norm": 0.5626255419013269,
+      "learning_rate": 1.868955413534984e-05,
+      "loss": 0.443,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6626577287066246,
+      "grad_norm": 0.5728350774998016,
+      "learning_rate": 1.8688786938018866e-05,
+      "loss": 0.4273,
+      "step": 3361
+    },
+    {
+      "epoch": 0.6628548895899053,
+      "grad_norm": 0.5765941930596808,
+      "learning_rate": 1.8688019531932788e-05,
+      "loss": 0.4511,
+      "step": 3362
+    },
+    {
+      "epoch": 0.6630520504731862,
+      "grad_norm": 0.5622738400789903,
+      "learning_rate": 1.8687251917110045e-05,
+      "loss": 0.4643,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6632492113564669,
+      "grad_norm": 0.5415771750284268,
+      "learning_rate": 1.8686484093569078e-05,
+      "loss": 0.4228,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6634463722397477,
+      "grad_norm": 0.5508756543268856,
+      "learning_rate": 1.868571606132834e-05,
+      "loss": 0.452,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6636435331230284,
+      "grad_norm": 0.6135272218599022,
+      "learning_rate": 1.8684947820406273e-05,
+      "loss": 0.4667,
+      "step": 3366
+    },
+    {
+      "epoch": 0.6638406940063092,
+      "grad_norm": 0.5960455949438687,
+      "learning_rate": 1.8684179370821343e-05,
+      "loss": 0.4686,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6640378548895899,
+      "grad_norm": 0.5480090387207861,
+      "learning_rate": 1.8683410712592015e-05,
+      "loss": 0.448,
+      "step": 3368
+    },
+    {
+      "epoch": 0.6642350157728707,
+      "grad_norm": 0.6217390232273009,
+      "learning_rate": 1.8682641845736748e-05,
+      "loss": 0.4106,
+      "step": 3369
+    },
+    {
+      "epoch": 0.6644321766561514,
+      "grad_norm": 0.5922247859816271,
+      "learning_rate": 1.8681872770274013e-05,
+      "loss": 0.4577,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6646293375394322,
+      "grad_norm": 0.556590095986522,
+      "learning_rate": 1.86811034862223e-05,
+      "loss": 0.4541,
+      "step": 3371
+    },
+    {
+      "epoch": 0.6648264984227129,
+      "grad_norm": 0.5422277567079224,
+      "learning_rate": 1.8680333993600084e-05,
+      "loss": 0.4413,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6650236593059937,
+      "grad_norm": 0.609846737333361,
+      "learning_rate": 1.867956429242585e-05,
+      "loss": 0.4439,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6652208201892744,
+      "grad_norm": 0.5366562298747772,
+      "learning_rate": 1.86787943827181e-05,
+      "loss": 0.4227,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6654179810725552,
+      "grad_norm": 0.59012017930504,
+      "learning_rate": 1.8678024264495323e-05,
+      "loss": 0.4351,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6656151419558359,
+      "grad_norm": 0.5771194135297009,
+      "learning_rate": 1.8677253937776024e-05,
+      "loss": 0.4206,
+      "step": 3376
+    },
+    {
+      "epoch": 0.6658123028391167,
+      "grad_norm": 0.5678453290335862,
+      "learning_rate": 1.8676483402578714e-05,
+      "loss": 0.458,
+      "step": 3377
+    },
+    {
+      "epoch": 0.6660094637223974,
+      "grad_norm": 0.5848984447281752,
+      "learning_rate": 1.86757126589219e-05,
+      "loss": 0.4294,
+      "step": 3378
+    },
+    {
+      "epoch": 0.6662066246056783,
+      "grad_norm": 0.5612739049068476,
+      "learning_rate": 1.8674941706824104e-05,
+      "loss": 0.4335,
+      "step": 3379
+    },
+    {
+      "epoch": 0.666403785488959,
+      "grad_norm": 0.5554127683707469,
+      "learning_rate": 1.8674170546303846e-05,
+      "loss": 0.4216,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6666009463722398,
+      "grad_norm": 0.5356096683041639,
+      "learning_rate": 1.8673399177379657e-05,
+      "loss": 0.417,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6667981072555205,
+      "grad_norm": 0.5413485188215648,
+      "learning_rate": 1.8672627600070068e-05,
+      "loss": 0.4244,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6669952681388013,
+      "grad_norm": 0.5305208256677184,
+      "learning_rate": 1.8671855814393617e-05,
+      "loss": 0.4153,
+      "step": 3383
+    },
+    {
+      "epoch": 0.667192429022082,
+      "grad_norm": 0.5903750497070852,
+      "learning_rate": 1.8671083820368846e-05,
+      "loss": 0.4633,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6673895899053628,
+      "grad_norm": 1.389773255933898,
+      "learning_rate": 1.8670311618014307e-05,
+      "loss": 0.4518,
+      "step": 3385
+    },
+    {
+      "epoch": 0.6675867507886435,
+      "grad_norm": 0.6734664022072558,
+      "learning_rate": 1.8669539207348544e-05,
+      "loss": 0.4303,
+      "step": 3386
+    },
+    {
+      "epoch": 0.6677839116719243,
+      "grad_norm": 0.5925325374467875,
+      "learning_rate": 1.8668766588390122e-05,
+      "loss": 0.4222,
+      "step": 3387
+    },
+    {
+      "epoch": 0.667981072555205,
+      "grad_norm": 0.5505241807065062,
+      "learning_rate": 1.8667993761157602e-05,
+      "loss": 0.4407,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6681782334384858,
+      "grad_norm": 0.7188824261126605,
+      "learning_rate": 1.866722072566955e-05,
+      "loss": 0.4378,
+      "step": 3389
+    },
+    {
+      "epoch": 0.6683753943217665,
+      "grad_norm": 1.6862636115429694,
+      "learning_rate": 1.8666447481944542e-05,
+      "loss": 0.4472,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6685725552050473,
+      "grad_norm": 0.535071580188235,
+      "learning_rate": 1.8665674030001154e-05,
+      "loss": 0.4324,
+      "step": 3391
+    },
+    {
+      "epoch": 0.668769716088328,
+      "grad_norm": 0.5720087515109528,
+      "learning_rate": 1.866490036985797e-05,
+      "loss": 0.4285,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6689668769716088,
+      "grad_norm": 0.5753023247784349,
+      "learning_rate": 1.8664126501533576e-05,
+      "loss": 0.471,
+      "step": 3393
+    },
+    {
+      "epoch": 0.6691640378548895,
+      "grad_norm": 0.5438723970201407,
+      "learning_rate": 1.8663352425046564e-05,
+      "loss": 0.4349,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6693611987381703,
+      "grad_norm": 0.565118624393037,
+      "learning_rate": 1.8662578140415535e-05,
+      "loss": 0.4534,
+      "step": 3395
+    },
+    {
+      "epoch": 0.669558359621451,
+      "grad_norm": 0.6635843691593936,
+      "learning_rate": 1.866180364765909e-05,
+      "loss": 0.4541,
+      "step": 3396
+    },
+    {
+      "epoch": 0.6697555205047319,
+      "grad_norm": 0.5553454488274417,
+      "learning_rate": 1.8661028946795837e-05,
+      "loss": 0.4164,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6699526813880127,
+      "grad_norm": 2.036398178901347,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.4708,
+      "step": 3398
+    },
+    {
+      "epoch": 0.6701498422712934,
+      "grad_norm": 0.5825487984269258,
+      "learning_rate": 1.8659478920823364e-05,
+      "loss": 0.4196,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6703470031545742,
+      "grad_norm": 0.5882347623158389,
+      "learning_rate": 1.865870359575138e-05,
+      "loss": 0.4376,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6705441640378549,
+      "grad_norm": 0.5537130860684163,
+      "learning_rate": 1.8657928062647075e-05,
+      "loss": 0.4292,
+      "step": 3401
+    },
+    {
+      "epoch": 0.6707413249211357,
+      "grad_norm": 0.6278956340355503,
+      "learning_rate": 1.8657152321529075e-05,
+      "loss": 0.4216,
+      "step": 3402
+    },
+    {
+      "epoch": 0.6709384858044164,
+      "grad_norm": 0.5849954721222042,
+      "learning_rate": 1.8656376372416017e-05,
+      "loss": 0.472,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6711356466876972,
+      "grad_norm": 0.6501311313277848,
+      "learning_rate": 1.8655600215326547e-05,
+      "loss": 0.4157,
+      "step": 3404
+    },
+    {
+      "epoch": 0.6713328075709779,
+      "grad_norm": 0.60533862716353,
+      "learning_rate": 1.8654823850279312e-05,
+      "loss": 0.4249,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6715299684542587,
+      "grad_norm": 0.7045177259010618,
+      "learning_rate": 1.8654047277292962e-05,
+      "loss": 0.5035,
+      "step": 3406
+    },
+    {
+      "epoch": 0.6717271293375394,
+      "grad_norm": 0.5709556862264039,
+      "learning_rate": 1.8653270496386163e-05,
+      "loss": 0.4189,
+      "step": 3407
+    },
+    {
+      "epoch": 0.6719242902208202,
+      "grad_norm": 0.657277026760384,
+      "learning_rate": 1.8652493507577564e-05,
+      "loss": 0.4936,
+      "step": 3408
+    },
+    {
+      "epoch": 0.6721214511041009,
+      "grad_norm": 0.8211869369723545,
+      "learning_rate": 1.8651716310885845e-05,
+      "loss": 0.4697,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6723186119873817,
+      "grad_norm": 0.5634640318066427,
+      "learning_rate": 1.8650938906329674e-05,
+      "loss": 0.4266,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6725157728706624,
+      "grad_norm": 0.6140117782031614,
+      "learning_rate": 1.865016129392773e-05,
+      "loss": 0.4469,
+      "step": 3411
+    },
+    {
+      "epoch": 0.6727129337539433,
+      "grad_norm": 0.6385762621009046,
+      "learning_rate": 1.864938347369869e-05,
+      "loss": 0.471,
+      "step": 3412
+    },
+    {
+      "epoch": 0.672910094637224,
+      "grad_norm": 0.5758795206307361,
+      "learning_rate": 1.8648605445661256e-05,
+      "loss": 0.4321,
+      "step": 3413
+    },
+    {
+      "epoch": 0.6731072555205048,
+      "grad_norm": 0.633725193056609,
+      "learning_rate": 1.8647827209834105e-05,
+      "loss": 0.469,
+      "step": 3414
+    },
+    {
+      "epoch": 0.6733044164037855,
+      "grad_norm": 0.5649636256512869,
+      "learning_rate": 1.864704876623594e-05,
+      "loss": 0.3962,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6735015772870663,
+      "grad_norm": 0.5541617307127727,
+      "learning_rate": 1.8646270114885467e-05,
+      "loss": 0.4245,
+      "step": 3416
+    },
+    {
+      "epoch": 0.673698738170347,
+      "grad_norm": 0.5576784865143702,
+      "learning_rate": 1.864549125580139e-05,
+      "loss": 0.4479,
+      "step": 3417
+    },
+    {
+      "epoch": 0.6738958990536278,
+      "grad_norm": 0.6606800054432356,
+      "learning_rate": 1.8644712189002426e-05,
+      "loss": 0.4209,
+      "step": 3418
+    },
+    {
+      "epoch": 0.6740930599369085,
+      "grad_norm": 0.5996093302287164,
+      "learning_rate": 1.864393291450729e-05,
+      "loss": 0.4806,
+      "step": 3419
+    },
+    {
+      "epoch": 0.6742902208201893,
+      "grad_norm": 0.5796582195346354,
+      "learning_rate": 1.8643153432334703e-05,
+      "loss": 0.4231,
+      "step": 3420
+    },
+    {
+      "epoch": 0.67448738170347,
+      "grad_norm": 0.6804211689949103,
+      "learning_rate": 1.8642373742503395e-05,
+      "loss": 0.469,
+      "step": 3421
+    },
+    {
+      "epoch": 0.6746845425867508,
+      "grad_norm": 0.8018195345405356,
+      "learning_rate": 1.8641593845032098e-05,
+      "loss": 0.4612,
+      "step": 3422
+    },
+    {
+      "epoch": 0.6748817034700315,
+      "grad_norm": 0.5430532048956593,
+      "learning_rate": 1.864081373993955e-05,
+      "loss": 0.447,
+      "step": 3423
+    },
+    {
+      "epoch": 0.6750788643533123,
+      "grad_norm": 0.5081290249970105,
+      "learning_rate": 1.8640033427244497e-05,
+      "loss": 0.3789,
+      "step": 3424
+    },
+    {
+      "epoch": 0.675276025236593,
+      "grad_norm": 0.6128416516260621,
+      "learning_rate": 1.863925290696568e-05,
+      "loss": 0.4795,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6754731861198738,
+      "grad_norm": 0.5433092528759821,
+      "learning_rate": 1.8638472179121855e-05,
+      "loss": 0.4399,
+      "step": 3426
+    },
+    {
+      "epoch": 0.6756703470031545,
+      "grad_norm": 6.721875316952158,
+      "learning_rate": 1.863769124373178e-05,
+      "loss": 0.4689,
+      "step": 3427
+    },
+    {
+      "epoch": 0.6758675078864353,
+      "grad_norm": 0.7065414307044805,
+      "learning_rate": 1.8636910100814216e-05,
+      "loss": 0.457,
+      "step": 3428
+    },
+    {
+      "epoch": 0.676064668769716,
+      "grad_norm": 0.5525113473136398,
+      "learning_rate": 1.863612875038793e-05,
+      "loss": 0.4101,
+      "step": 3429
+    },
+    {
+      "epoch": 0.6762618296529969,
+      "grad_norm": 0.6783813336542089,
+      "learning_rate": 1.86353471924717e-05,
+      "loss": 0.4869,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6764589905362776,
+      "grad_norm": 0.7824903261317586,
+      "learning_rate": 1.8634565427084295e-05,
+      "loss": 0.4584,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6766561514195584,
+      "grad_norm": 0.6900514518883072,
+      "learning_rate": 1.8633783454244506e-05,
+      "loss": 0.4408,
+      "step": 3432
+    },
+    {
+      "epoch": 0.6768533123028391,
+      "grad_norm": 0.6615900688611679,
+      "learning_rate": 1.8633001273971115e-05,
+      "loss": 0.4558,
+      "step": 3433
+    },
+    {
+      "epoch": 0.6770504731861199,
+      "grad_norm": 0.5567854903968781,
+      "learning_rate": 1.863221888628292e-05,
+      "loss": 0.4031,
+      "step": 3434
+    },
+    {
+      "epoch": 0.6772476340694006,
+      "grad_norm": 0.6417523757757003,
+      "learning_rate": 1.8631436291198707e-05,
+      "loss": 0.4672,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6774447949526814,
+      "grad_norm": 0.6024057489717937,
+      "learning_rate": 1.863065348873729e-05,
+      "loss": 0.4381,
+      "step": 3436
+    },
+    {
+      "epoch": 0.6776419558359621,
+      "grad_norm": 0.6645203432793743,
+      "learning_rate": 1.8629870478917477e-05,
+      "loss": 0.4424,
+      "step": 3437
+    },
+    {
+      "epoch": 0.6778391167192429,
+      "grad_norm": 0.5526525607402512,
+      "learning_rate": 1.8629087261758072e-05,
+      "loss": 0.4154,
+      "step": 3438
+    },
+    {
+      "epoch": 0.6780362776025236,
+      "grad_norm": 0.5995871293018885,
+      "learning_rate": 1.8628303837277893e-05,
+      "loss": 0.4304,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6782334384858044,
+      "grad_norm": 0.5743129816438846,
+      "learning_rate": 1.8627520205495772e-05,
+      "loss": 0.448,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6784305993690851,
+      "grad_norm": 0.5829067346816689,
+      "learning_rate": 1.862673636643053e-05,
+      "loss": 0.4211,
+      "step": 3441
+    },
+    {
+      "epoch": 0.6786277602523659,
+      "grad_norm": 0.6220895944597309,
+      "learning_rate": 1.8625952320100998e-05,
+      "loss": 0.4241,
+      "step": 3442
+    },
+    {
+      "epoch": 0.6788249211356467,
+      "grad_norm": 0.8700470116975232,
+      "learning_rate": 1.8625168066526017e-05,
+      "loss": 0.4635,
+      "step": 3443
+    },
+    {
+      "epoch": 0.6790220820189274,
+      "grad_norm": 0.572249467365147,
+      "learning_rate": 1.8624383605724422e-05,
+      "loss": 0.4684,
+      "step": 3444
+    },
+    {
+      "epoch": 0.6792192429022083,
+      "grad_norm": 0.5329766253780658,
+      "learning_rate": 1.8623598937715072e-05,
+      "loss": 0.4132,
+      "step": 3445
+    },
+    {
+      "epoch": 0.679416403785489,
+      "grad_norm": 0.5378353899707224,
+      "learning_rate": 1.8622814062516807e-05,
+      "loss": 0.4514,
+      "step": 3446
+    },
+    {
+      "epoch": 0.6796135646687698,
+      "grad_norm": 0.6043076943443687,
+      "learning_rate": 1.8622028980148494e-05,
+      "loss": 0.4195,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6798107255520505,
+      "grad_norm": 0.5415001575820032,
+      "learning_rate": 1.8621243690628993e-05,
+      "loss": 0.4388,
+      "step": 3448
+    },
+    {
+      "epoch": 0.6800078864353313,
+      "grad_norm": 0.5715084310965,
+      "learning_rate": 1.8620458193977166e-05,
+      "loss": 0.4377,
+      "step": 3449
+    },
+    {
+      "epoch": 0.680205047318612,
+      "grad_norm": 0.5477132137010473,
+      "learning_rate": 1.861967249021189e-05,
+      "loss": 0.4166,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6804022082018928,
+      "grad_norm": 0.513438305978723,
+      "learning_rate": 1.861888657935204e-05,
+      "loss": 0.4098,
+      "step": 3451
+    },
+    {
+      "epoch": 0.6805993690851735,
+      "grad_norm": 0.7909891794930992,
+      "learning_rate": 1.8618100461416503e-05,
+      "loss": 0.4311,
+      "step": 3452
+    },
+    {
+      "epoch": 0.6807965299684543,
+      "grad_norm": 0.5263405808562206,
+      "learning_rate": 1.8617314136424157e-05,
+      "loss": 0.4381,
+      "step": 3453
+    },
+    {
+      "epoch": 0.680993690851735,
+      "grad_norm": 0.5858050604116924,
+      "learning_rate": 1.8616527604393903e-05,
+      "loss": 0.4383,
+      "step": 3454
+    },
+    {
+      "epoch": 0.6811908517350158,
+      "grad_norm": 0.56256951114871,
+      "learning_rate": 1.8615740865344632e-05,
+      "loss": 0.4517,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6813880126182965,
+      "grad_norm": 1.956983000612891,
+      "learning_rate": 1.861495391929525e-05,
+      "loss": 0.4863,
+      "step": 3456
+    },
+    {
+      "epoch": 0.6815851735015773,
+      "grad_norm": 224.166458708411,
+      "learning_rate": 1.8614166766264662e-05,
+      "loss": 0.8088,
+      "step": 3457
+    },
+    {
+      "epoch": 0.681782334384858,
+      "grad_norm": 0.7066319163990039,
+      "learning_rate": 1.8613379406271784e-05,
+      "loss": 0.4568,
+      "step": 3458
+    },
+    {
+      "epoch": 0.6819794952681388,
+      "grad_norm": 0.530318401808118,
+      "learning_rate": 1.8612591839335526e-05,
+      "loss": 0.4104,
+      "step": 3459
+    },
+    {
+      "epoch": 0.6821766561514195,
+      "grad_norm": 0.6416231249123863,
+      "learning_rate": 1.861180406547481e-05,
+      "loss": 0.4504,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6823738170347003,
+      "grad_norm": 0.6016468271692968,
+      "learning_rate": 1.8611016084708572e-05,
+      "loss": 0.4866,
+      "step": 3461
+    },
+    {
+      "epoch": 0.682570977917981,
+      "grad_norm": 0.5665263512082256,
+      "learning_rate": 1.8610227897055736e-05,
+      "loss": 0.4272,
+      "step": 3462
+    },
+    {
+      "epoch": 0.6827681388012619,
+      "grad_norm": 0.5951978340725875,
+      "learning_rate": 1.8609439502535244e-05,
+      "loss": 0.47,
+      "step": 3463
+    },
+    {
+      "epoch": 0.6829652996845426,
+      "grad_norm": 0.5145073667769414,
+      "learning_rate": 1.8608650901166034e-05,
+      "loss": 0.3968,
+      "step": 3464
+    },
+    {
+      "epoch": 0.6831624605678234,
+      "grad_norm": 0.6596358236905893,
+      "learning_rate": 1.8607862092967048e-05,
+      "loss": 0.4825,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6833596214511041,
+      "grad_norm": 0.5633284833079664,
+      "learning_rate": 1.8607073077957246e-05,
+      "loss": 0.4295,
+      "step": 3466
+    },
+    {
+      "epoch": 0.6835567823343849,
+      "grad_norm": 0.5986090013281019,
+      "learning_rate": 1.8606283856155585e-05,
+      "loss": 0.4758,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6837539432176656,
+      "grad_norm": 0.6051379255945182,
+      "learning_rate": 1.8605494427581022e-05,
+      "loss": 0.4411,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6839511041009464,
+      "grad_norm": 0.5119522262994246,
+      "learning_rate": 1.8604704792252524e-05,
+      "loss": 0.4063,
+      "step": 3469
+    },
+    {
+      "epoch": 0.6841482649842271,
+      "grad_norm": 0.5838237384793697,
+      "learning_rate": 1.8603914950189063e-05,
+      "loss": 0.4083,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6843454258675079,
+      "grad_norm": 0.5495244130738517,
+      "learning_rate": 1.860312490140962e-05,
+      "loss": 0.4398,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6845425867507886,
+      "grad_norm": 0.8373973638550604,
+      "learning_rate": 1.860233464593317e-05,
+      "loss": 0.4274,
+      "step": 3472
+    },
+    {
+      "epoch": 0.6847397476340694,
+      "grad_norm": 0.6095069405419896,
+      "learning_rate": 1.8601544183778707e-05,
+      "loss": 0.4282,
+      "step": 3473
+    },
+    {
+      "epoch": 0.6849369085173501,
+      "grad_norm": 0.6313696871900962,
+      "learning_rate": 1.8600753514965215e-05,
+      "loss": 0.4644,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6851340694006309,
+      "grad_norm": 0.6331745807987341,
+      "learning_rate": 1.8599962639511692e-05,
+      "loss": 0.4227,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6853312302839116,
+      "grad_norm": 0.6271712528145608,
+      "learning_rate": 1.8599171557437147e-05,
+      "loss": 0.4182,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6855283911671924,
+      "grad_norm": 0.6822672026358032,
+      "learning_rate": 1.8598380268760573e-05,
+      "loss": 0.4609,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6857255520504731,
+      "grad_norm": 0.6394977318604866,
+      "learning_rate": 1.8597588773500997e-05,
+      "loss": 0.4657,
+      "step": 3478
+    },
+    {
+      "epoch": 0.685922712933754,
+      "grad_norm": 0.5473583517133394,
+      "learning_rate": 1.8596797071677422e-05,
+      "loss": 0.4207,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6861198738170347,
+      "grad_norm": 0.5332831738588625,
+      "learning_rate": 1.8596005163308874e-05,
+      "loss": 0.393,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6863170347003155,
+      "grad_norm": 0.5850491469644991,
+      "learning_rate": 1.859521304841438e-05,
+      "loss": 0.4544,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6865141955835962,
+      "grad_norm": 0.5298743941736833,
+      "learning_rate": 1.859442072701297e-05,
+      "loss": 0.4558,
+      "step": 3482
+    },
+    {
+      "epoch": 0.686711356466877,
+      "grad_norm": 0.5914438587794221,
+      "learning_rate": 1.8593628199123684e-05,
+      "loss": 0.4696,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6869085173501577,
+      "grad_norm": 0.5849436915983998,
+      "learning_rate": 1.8592835464765557e-05,
+      "loss": 0.468,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6871056782334385,
+      "grad_norm": 0.5517879662711394,
+      "learning_rate": 1.859204252395764e-05,
+      "loss": 0.4404,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6873028391167192,
+      "grad_norm": 0.5432311116041397,
+      "learning_rate": 1.8591249376718984e-05,
+      "loss": 0.4182,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 0.5778253507921007,
+      "learning_rate": 1.859045602306864e-05,
+      "loss": 0.4714,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6876971608832808,
+      "grad_norm": 0.5626709550212079,
+      "learning_rate": 1.8589662463025674e-05,
+      "loss": 0.4214,
+      "step": 3488
+    },
+    {
+      "epoch": 0.6878943217665615,
+      "grad_norm": 0.5828536367056488,
+      "learning_rate": 1.858886869660915e-05,
+      "loss": 0.4687,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6880914826498423,
+      "grad_norm": 0.5546642607606428,
+      "learning_rate": 1.8588074723838136e-05,
+      "loss": 0.4452,
+      "step": 3490
+    },
+    {
+      "epoch": 0.688288643533123,
+      "grad_norm": 0.5748243298511659,
+      "learning_rate": 1.8587280544731712e-05,
+      "loss": 0.4597,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6884858044164038,
+      "grad_norm": 0.5614432197404846,
+      "learning_rate": 1.858648615930896e-05,
+      "loss": 0.4094,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6886829652996845,
+      "grad_norm": 0.5310867099166175,
+      "learning_rate": 1.8585691567588964e-05,
+      "loss": 0.4,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6888801261829653,
+      "grad_norm": 0.5487952682416996,
+      "learning_rate": 1.858489676959081e-05,
+      "loss": 0.4263,
+      "step": 3494
+    },
+    {
+      "epoch": 0.689077287066246,
+      "grad_norm": 0.6769572158367133,
+      "learning_rate": 1.85841017653336e-05,
+      "loss": 0.4654,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6892744479495269,
+      "grad_norm": 0.6393020125226531,
+      "learning_rate": 1.8583306554836432e-05,
+      "loss": 0.4583,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6894716088328076,
+      "grad_norm": 0.559712814410196,
+      "learning_rate": 1.8582511138118413e-05,
+      "loss": 0.4298,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6896687697160884,
+      "grad_norm": 0.602123934295947,
+      "learning_rate": 1.8581715515198652e-05,
+      "loss": 0.4675,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6898659305993691,
+      "grad_norm": 0.5733778203272168,
+      "learning_rate": 1.8580919686096263e-05,
+      "loss": 0.4297,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6900630914826499,
+      "grad_norm": 0.6829776698050027,
+      "learning_rate": 1.858012365083037e-05,
+      "loss": 0.4478,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6902602523659306,
+      "grad_norm": 0.9792198683444031,
+      "learning_rate": 1.8579327409420094e-05,
+      "loss": 0.4772,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6904574132492114,
+      "grad_norm": 0.594469172261072,
+      "learning_rate": 1.8578530961884574e-05,
+      "loss": 0.4306,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6906545741324921,
+      "grad_norm": 0.6517540399351949,
+      "learning_rate": 1.8577734308242936e-05,
+      "loss": 0.4324,
+      "step": 3503
+    },
+    {
+      "epoch": 0.6908517350157729,
+      "grad_norm": 0.5804034407531462,
+      "learning_rate": 1.8576937448514323e-05,
+      "loss": 0.4582,
+      "step": 3504
+    },
+    {
+      "epoch": 0.6910488958990536,
+      "grad_norm": 0.5754381054150248,
+      "learning_rate": 1.857614038271788e-05,
+      "loss": 0.4442,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6912460567823344,
+      "grad_norm": 0.5394282365890705,
+      "learning_rate": 1.857534311087276e-05,
+      "loss": 0.3904,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6914432176656151,
+      "grad_norm": 0.5714584536292786,
+      "learning_rate": 1.8574545632998116e-05,
+      "loss": 0.438,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6916403785488959,
+      "grad_norm": 0.6005244403686105,
+      "learning_rate": 1.857374794911311e-05,
+      "loss": 0.4698,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6918375394321766,
+      "grad_norm": 0.5475318354156559,
+      "learning_rate": 1.85729500592369e-05,
+      "loss": 0.4425,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6920347003154574,
+      "grad_norm": 0.5867289944866695,
+      "learning_rate": 1.857215196338866e-05,
+      "loss": 0.4601,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6922318611987381,
+      "grad_norm": 0.5465094788446564,
+      "learning_rate": 1.8571353661587573e-05,
+      "loss": 0.4496,
+      "step": 3511
+    },
+    {
+      "epoch": 0.692429022082019,
+      "grad_norm": 0.574286072227518,
+      "learning_rate": 1.8570555153852806e-05,
+      "loss": 0.4449,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6926261829652997,
+      "grad_norm": 0.6073127421651645,
+      "learning_rate": 1.8569756440203554e-05,
+      "loss": 0.449,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6928233438485805,
+      "grad_norm": 0.5771548485295239,
+      "learning_rate": 1.8568957520659e-05,
+      "loss": 0.4555,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6930205047318612,
+      "grad_norm": 0.5445589778122638,
+      "learning_rate": 1.856815839523834e-05,
+      "loss": 0.4319,
+      "step": 3515
+    },
+    {
+      "epoch": 0.693217665615142,
+      "grad_norm": 0.5629986928595075,
+      "learning_rate": 1.8567359063960778e-05,
+      "loss": 0.439,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6934148264984227,
+      "grad_norm": 0.5596461637975891,
+      "learning_rate": 1.8566559526845512e-05,
+      "loss": 0.4341,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6936119873817035,
+      "grad_norm": 0.5375000041425667,
+      "learning_rate": 1.8565759783911756e-05,
+      "loss": 0.3897,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6938091482649842,
+      "grad_norm": 0.503604882436781,
+      "learning_rate": 1.8564959835178725e-05,
+      "loss": 0.3891,
+      "step": 3519
+    },
+    {
+      "epoch": 0.694006309148265,
+      "grad_norm": 0.6060651930247586,
+      "learning_rate": 1.8564159680665633e-05,
+      "loss": 0.4572,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6942034700315457,
+      "grad_norm": 0.5897040744548951,
+      "learning_rate": 1.856335932039171e-05,
+      "loss": 0.4341,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6944006309148265,
+      "grad_norm": 0.5726098247448242,
+      "learning_rate": 1.8562558754376182e-05,
+      "loss": 0.4337,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6945977917981072,
+      "grad_norm": 0.5241962440051866,
+      "learning_rate": 1.8561757982638285e-05,
+      "loss": 0.4222,
+      "step": 3523
+    },
+    {
+      "epoch": 0.694794952681388,
+      "grad_norm": 0.5767874135195625,
+      "learning_rate": 1.856095700519726e-05,
+      "loss": 0.4669,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6949921135646687,
+      "grad_norm": 0.605586730484475,
+      "learning_rate": 1.856015582207235e-05,
+      "loss": 0.4965,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6951892744479495,
+      "grad_norm": 0.5293811874847645,
+      "learning_rate": 1.8559354433282795e-05,
+      "loss": 0.4187,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6953864353312302,
+      "grad_norm": 0.557624349231964,
+      "learning_rate": 1.8558552838847862e-05,
+      "loss": 0.4413,
+      "step": 3527
+    },
+    {
+      "epoch": 0.695583596214511,
+      "grad_norm": 0.5724753571082727,
+      "learning_rate": 1.8557751038786807e-05,
+      "loss": 0.4325,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6957807570977917,
+      "grad_norm": 0.5642589878108021,
+      "learning_rate": 1.8556949033118886e-05,
+      "loss": 0.4346,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6959779179810726,
+      "grad_norm": 0.587492971862397,
+      "learning_rate": 1.855614682186338e-05,
+      "loss": 0.4709,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6961750788643533,
+      "grad_norm": 0.5465280182977807,
+      "learning_rate": 1.8555344405039553e-05,
+      "loss": 0.4184,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6963722397476341,
+      "grad_norm": 0.6480278193209652,
+      "learning_rate": 1.8554541782666685e-05,
+      "loss": 0.4649,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6965694006309149,
+      "grad_norm": 0.5135839915061162,
+      "learning_rate": 1.8553738954764068e-05,
+      "loss": 0.4147,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6967665615141956,
+      "grad_norm": 0.5495932553673605,
+      "learning_rate": 1.855293592135098e-05,
+      "loss": 0.3947,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6969637223974764,
+      "grad_norm": 0.5759718073274976,
+      "learning_rate": 1.8552132682446716e-05,
+      "loss": 0.4613,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6971608832807571,
+      "grad_norm": 0.5693917333392535,
+      "learning_rate": 1.8551329238070583e-05,
+      "loss": 0.4661,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6973580441640379,
+      "grad_norm": 0.633091177062187,
+      "learning_rate": 1.8550525588241878e-05,
+      "loss": 0.4114,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6975552050473186,
+      "grad_norm": 0.5719956129908553,
+      "learning_rate": 1.8549721732979904e-05,
+      "loss": 0.4638,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6977523659305994,
+      "grad_norm": 0.5734766418241823,
+      "learning_rate": 1.8548917672303987e-05,
+      "loss": 0.4279,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6979495268138801,
+      "grad_norm": 0.5817240921924031,
+      "learning_rate": 1.8548113406233436e-05,
+      "loss": 0.4595,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6981466876971609,
+      "grad_norm": 0.5498165766780686,
+      "learning_rate": 1.8547308934787576e-05,
+      "loss": 0.4183,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6983438485804416,
+      "grad_norm": 0.5771901083506239,
+      "learning_rate": 1.8546504257985738e-05,
+      "loss": 0.4399,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6985410094637224,
+      "grad_norm": 0.8880467198374812,
+      "learning_rate": 1.8545699375847247e-05,
+      "loss": 0.4673,
+      "step": 3543
+    },
+    {
+      "epoch": 0.6987381703470031,
+      "grad_norm": 0.8215695607693159,
+      "learning_rate": 1.8544894288391452e-05,
+      "loss": 0.3944,
+      "step": 3544
+    },
+    {
+      "epoch": 0.698935331230284,
+      "grad_norm": 2.190816475198652,
+      "learning_rate": 1.8544088995637693e-05,
+      "loss": 0.43,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6991324921135647,
+      "grad_norm": 0.6619407960935211,
+      "learning_rate": 1.854328349760531e-05,
+      "loss": 0.474,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6993296529968455,
+      "grad_norm": 0.7707630061873859,
+      "learning_rate": 1.8542477794313662e-05,
+      "loss": 0.4369,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6995268138801262,
+      "grad_norm": 0.6668874413715213,
+      "learning_rate": 1.8541671885782106e-05,
+      "loss": 0.4226,
+      "step": 3548
+    },
+    {
+      "epoch": 0.699723974763407,
+      "grad_norm": 0.5933363551194153,
+      "learning_rate": 1.8540865772030004e-05,
+      "loss": 0.44,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6999211356466877,
+      "grad_norm": 0.5411012362808466,
+      "learning_rate": 1.8540059453076728e-05,
+      "loss": 0.4505,
+      "step": 3550
+    },
+    {
+      "epoch": 0.7001182965299685,
+      "grad_norm": 2.092070387748023,
+      "learning_rate": 1.853925292894164e-05,
+      "loss": 0.4816,
+      "step": 3551
+    },
+    {
+      "epoch": 0.7003154574132492,
+      "grad_norm": 0.6015879767455612,
+      "learning_rate": 1.853844619964413e-05,
+      "loss": 0.4578,
+      "step": 3552
+    },
+    {
+      "epoch": 0.70051261829653,
+      "grad_norm": 0.6227707549366479,
+      "learning_rate": 1.853763926520357e-05,
+      "loss": 0.4621,
+      "step": 3553
+    },
+    {
+      "epoch": 0.7007097791798107,
+      "grad_norm": 0.6158286256739309,
+      "learning_rate": 1.8536832125639353e-05,
+      "loss": 0.4615,
+      "step": 3554
+    },
+    {
+      "epoch": 0.7009069400630915,
+      "grad_norm": 4.099666789727112,
+      "learning_rate": 1.8536024780970868e-05,
+      "loss": 0.4471,
+      "step": 3555
+    },
+    {
+      "epoch": 0.7011041009463722,
+      "grad_norm": 1.3865107372029537,
+      "learning_rate": 1.8535217231217512e-05,
+      "loss": 0.4835,
+      "step": 3556
+    },
+    {
+      "epoch": 0.701301261829653,
+      "grad_norm": 0.6437956304730564,
+      "learning_rate": 1.8534409476398693e-05,
+      "loss": 0.4976,
+      "step": 3557
+    },
+    {
+      "epoch": 0.7014984227129337,
+      "grad_norm": 0.809900272846465,
+      "learning_rate": 1.853360151653381e-05,
+      "loss": 0.4337,
+      "step": 3558
+    },
+    {
+      "epoch": 0.7016955835962145,
+      "grad_norm": 0.5732397543697593,
+      "learning_rate": 1.8532793351642283e-05,
+      "loss": 0.447,
+      "step": 3559
+    },
+    {
+      "epoch": 0.7018927444794952,
+      "grad_norm": 0.6180992287814596,
+      "learning_rate": 1.853198498174352e-05,
+      "loss": 0.4483,
+      "step": 3560
+    },
+    {
+      "epoch": 0.702089905362776,
+      "grad_norm": 0.5836355816115922,
+      "learning_rate": 1.853117640685695e-05,
+      "loss": 0.4698,
+      "step": 3561
+    },
+    {
+      "epoch": 0.7022870662460567,
+      "grad_norm": 0.6266470149737501,
+      "learning_rate": 1.853036762700199e-05,
+      "loss": 0.4526,
+      "step": 3562
+    },
+    {
+      "epoch": 0.7024842271293376,
+      "grad_norm": 0.5545308522500857,
+      "learning_rate": 1.8529558642198085e-05,
+      "loss": 0.4275,
+      "step": 3563
+    },
+    {
+      "epoch": 0.7026813880126183,
+      "grad_norm": 0.5997846025265139,
+      "learning_rate": 1.8528749452464667e-05,
+      "loss": 0.4495,
+      "step": 3564
+    },
+    {
+      "epoch": 0.7028785488958991,
+      "grad_norm": 0.5932966298805183,
+      "learning_rate": 1.8527940057821168e-05,
+      "loss": 0.4404,
+      "step": 3565
+    },
+    {
+      "epoch": 0.7030757097791798,
+      "grad_norm": 0.6264233832414875,
+      "learning_rate": 1.8527130458287047e-05,
+      "loss": 0.4668,
+      "step": 3566
+    },
+    {
+      "epoch": 0.7032728706624606,
+      "grad_norm": 0.5422421411631603,
+      "learning_rate": 1.8526320653881745e-05,
+      "loss": 0.4352,
+      "step": 3567
+    },
+    {
+      "epoch": 0.7034700315457413,
+      "grad_norm": 0.5335967174546794,
+      "learning_rate": 1.8525510644624726e-05,
+      "loss": 0.4091,
+      "step": 3568
+    },
+    {
+      "epoch": 0.7036671924290221,
+      "grad_norm": 0.8352069980523148,
+      "learning_rate": 1.852470043053545e-05,
+      "loss": 0.4723,
+      "step": 3569
+    },
+    {
+      "epoch": 0.7038643533123028,
+      "grad_norm": 2.2486627770555585,
+      "learning_rate": 1.8523890011633377e-05,
+      "loss": 0.4209,
+      "step": 3570
+    },
+    {
+      "epoch": 0.7040615141955836,
+      "grad_norm": 0.5577369897883039,
+      "learning_rate": 1.8523079387937984e-05,
+      "loss": 0.4209,
+      "step": 3571
+    },
+    {
+      "epoch": 0.7042586750788643,
+      "grad_norm": 0.7637425020315376,
+      "learning_rate": 1.8522268559468744e-05,
+      "loss": 0.44,
+      "step": 3572
+    },
+    {
+      "epoch": 0.7044558359621451,
+      "grad_norm": 0.6449349203733895,
+      "learning_rate": 1.8521457526245142e-05,
+      "loss": 0.4603,
+      "step": 3573
+    },
+    {
+      "epoch": 0.7046529968454258,
+      "grad_norm": 0.5791189282772926,
+      "learning_rate": 1.852064628828666e-05,
+      "loss": 0.4327,
+      "step": 3574
+    },
+    {
+      "epoch": 0.7048501577287066,
+      "grad_norm": 0.7439087018417354,
+      "learning_rate": 1.851983484561279e-05,
+      "loss": 0.4688,
+      "step": 3575
+    },
+    {
+      "epoch": 0.7050473186119873,
+      "grad_norm": 0.5678987372658572,
+      "learning_rate": 1.8519023198243023e-05,
+      "loss": 0.4532,
+      "step": 3576
+    },
+    {
+      "epoch": 0.7052444794952681,
+      "grad_norm": 0.6893907579878671,
+      "learning_rate": 1.8518211346196865e-05,
+      "loss": 0.4482,
+      "step": 3577
+    },
+    {
+      "epoch": 0.705441640378549,
+      "grad_norm": 0.6063321929451956,
+      "learning_rate": 1.851739928949382e-05,
+      "loss": 0.452,
+      "step": 3578
+    },
+    {
+      "epoch": 0.7056388012618297,
+      "grad_norm": 0.5937895590659021,
+      "learning_rate": 1.85165870281534e-05,
+      "loss": 0.4067,
+      "step": 3579
+    },
+    {
+      "epoch": 0.7058359621451105,
+      "grad_norm": 0.571875582830355,
+      "learning_rate": 1.8515774562195115e-05,
+      "loss": 0.4296,
+      "step": 3580
+    },
+    {
+      "epoch": 0.7060331230283912,
+      "grad_norm": 0.6350546066367247,
+      "learning_rate": 1.851496189163849e-05,
+      "loss": 0.4802,
+      "step": 3581
+    },
+    {
+      "epoch": 0.706230283911672,
+      "grad_norm": 0.5728556146008228,
+      "learning_rate": 1.8514149016503048e-05,
+      "loss": 0.396,
+      "step": 3582
+    },
+    {
+      "epoch": 0.7064274447949527,
+      "grad_norm": 0.5607106332513473,
+      "learning_rate": 1.851333593680832e-05,
+      "loss": 0.395,
+      "step": 3583
+    },
+    {
+      "epoch": 0.7066246056782335,
+      "grad_norm": 0.6303565279440619,
+      "learning_rate": 1.851252265257384e-05,
+      "loss": 0.4606,
+      "step": 3584
+    },
+    {
+      "epoch": 0.7068217665615142,
+      "grad_norm": 0.5656202726110657,
+      "learning_rate": 1.8511709163819146e-05,
+      "loss": 0.4347,
+      "step": 3585
+    },
+    {
+      "epoch": 0.707018927444795,
+      "grad_norm": 0.5691199982028473,
+      "learning_rate": 1.851089547056379e-05,
+      "loss": 0.4799,
+      "step": 3586
+    },
+    {
+      "epoch": 0.7072160883280757,
+      "grad_norm": 0.5132663071610697,
+      "learning_rate": 1.851008157282731e-05,
+      "loss": 0.4254,
+      "step": 3587
+    },
+    {
+      "epoch": 0.7074132492113565,
+      "grad_norm": 0.599385661554487,
+      "learning_rate": 1.8509267470629275e-05,
+      "loss": 0.4807,
+      "step": 3588
+    },
+    {
+      "epoch": 0.7076104100946372,
+      "grad_norm": 0.5953814395962257,
+      "learning_rate": 1.850845316398923e-05,
+      "loss": 0.4638,
+      "step": 3589
+    },
+    {
+      "epoch": 0.707807570977918,
+      "grad_norm": 0.5296035745748813,
+      "learning_rate": 1.8507638652926748e-05,
+      "loss": 0.4341,
+      "step": 3590
+    },
+    {
+      "epoch": 0.7080047318611987,
+      "grad_norm": 0.5836957939032046,
+      "learning_rate": 1.85068239374614e-05,
+      "loss": 0.4664,
+      "step": 3591
+    },
+    {
+      "epoch": 0.7082018927444795,
+      "grad_norm": 0.5383659896543327,
+      "learning_rate": 1.8506009017612752e-05,
+      "loss": 0.4342,
+      "step": 3592
+    },
+    {
+      "epoch": 0.7083990536277602,
+      "grad_norm": 0.5802394458133452,
+      "learning_rate": 1.850519389340039e-05,
+      "loss": 0.4459,
+      "step": 3593
+    },
+    {
+      "epoch": 0.708596214511041,
+      "grad_norm": 0.6351661320355233,
+      "learning_rate": 1.850437856484389e-05,
+      "loss": 0.4401,
+      "step": 3594
+    },
+    {
+      "epoch": 0.7087933753943217,
+      "grad_norm": 0.5835902309373062,
+      "learning_rate": 1.850356303196285e-05,
+      "loss": 0.4412,
+      "step": 3595
+    },
+    {
+      "epoch": 0.7089905362776026,
+      "grad_norm": 0.5574546489985274,
+      "learning_rate": 1.850274729477686e-05,
+      "loss": 0.4558,
+      "step": 3596
+    },
+    {
+      "epoch": 0.7091876971608833,
+      "grad_norm": 0.6989042029450458,
+      "learning_rate": 1.850193135330552e-05,
+      "loss": 0.4467,
+      "step": 3597
+    },
+    {
+      "epoch": 0.7093848580441641,
+      "grad_norm": 0.5679323561115319,
+      "learning_rate": 1.850111520756843e-05,
+      "loss": 0.4235,
+      "step": 3598
+    },
+    {
+      "epoch": 0.7095820189274448,
+      "grad_norm": 2.385763474041582,
+      "learning_rate": 1.8500298857585207e-05,
+      "loss": 0.442,
+      "step": 3599
+    },
+    {
+      "epoch": 0.7097791798107256,
+      "grad_norm": 0.6566291998053769,
+      "learning_rate": 1.8499482303375454e-05,
+      "loss": 0.4537,
+      "step": 3600
+    },
+    {
+      "epoch": 0.7099763406940063,
+      "grad_norm": 0.5737937851393634,
+      "learning_rate": 1.8498665544958793e-05,
+      "loss": 0.4083,
+      "step": 3601
+    },
+    {
+      "epoch": 0.7101735015772871,
+      "grad_norm": 0.5528769295761875,
+      "learning_rate": 1.8497848582354852e-05,
+      "loss": 0.4165,
+      "step": 3602
+    },
+    {
+      "epoch": 0.7103706624605678,
+      "grad_norm": 0.5798728107229193,
+      "learning_rate": 1.8497031415583252e-05,
+      "loss": 0.4289,
+      "step": 3603
+    },
+    {
+      "epoch": 0.7105678233438486,
+      "grad_norm": 0.5389476339142518,
+      "learning_rate": 1.8496214044663633e-05,
+      "loss": 0.4062,
+      "step": 3604
+    },
+    {
+      "epoch": 0.7107649842271293,
+      "grad_norm": 0.6195250954116948,
+      "learning_rate": 1.8495396469615627e-05,
+      "loss": 0.4685,
+      "step": 3605
+    },
+    {
+      "epoch": 0.7109621451104101,
+      "grad_norm": 0.6193545738894738,
+      "learning_rate": 1.849457869045888e-05,
+      "loss": 0.4704,
+      "step": 3606
+    },
+    {
+      "epoch": 0.7111593059936908,
+      "grad_norm": 0.5963823991477826,
+      "learning_rate": 1.849376070721304e-05,
+      "loss": 0.4376,
+      "step": 3607
+    },
+    {
+      "epoch": 0.7113564668769716,
+      "grad_norm": 0.5649286189482374,
+      "learning_rate": 1.849294251989776e-05,
+      "loss": 0.442,
+      "step": 3608
+    },
+    {
+      "epoch": 0.7115536277602523,
+      "grad_norm": 0.5655868715091625,
+      "learning_rate": 1.8492124128532697e-05,
+      "loss": 0.4472,
+      "step": 3609
+    },
+    {
+      "epoch": 0.7117507886435331,
+      "grad_norm": 0.5624574778497905,
+      "learning_rate": 1.849130553313751e-05,
+      "loss": 0.4198,
+      "step": 3610
+    },
+    {
+      "epoch": 0.7119479495268138,
+      "grad_norm": 0.5918835988067689,
+      "learning_rate": 1.849048673373187e-05,
+      "loss": 0.4553,
+      "step": 3611
+    },
+    {
+      "epoch": 0.7121451104100947,
+      "grad_norm": 0.5450609283163641,
+      "learning_rate": 1.848966773033545e-05,
+      "loss": 0.408,
+      "step": 3612
+    },
+    {
+      "epoch": 0.7123422712933754,
+      "grad_norm": 0.6211662601865566,
+      "learning_rate": 1.8488848522967926e-05,
+      "loss": 0.4925,
+      "step": 3613
+    },
+    {
+      "epoch": 0.7125394321766562,
+      "grad_norm": 0.9407135144432643,
+      "learning_rate": 1.848802911164898e-05,
+      "loss": 0.4563,
+      "step": 3614
+    },
+    {
+      "epoch": 0.7127365930599369,
+      "grad_norm": 0.6364553089373497,
+      "learning_rate": 1.8487209496398298e-05,
+      "loss": 0.4556,
+      "step": 3615
+    },
+    {
+      "epoch": 0.7129337539432177,
+      "grad_norm": 0.6193764318415926,
+      "learning_rate": 1.8486389677235577e-05,
+      "loss": 0.5091,
+      "step": 3616
+    },
+    {
+      "epoch": 0.7131309148264984,
+      "grad_norm": 0.5810125836188512,
+      "learning_rate": 1.8485569654180506e-05,
+      "loss": 0.4331,
+      "step": 3617
+    },
+    {
+      "epoch": 0.7133280757097792,
+      "grad_norm": 0.5647496664846747,
+      "learning_rate": 1.8484749427252794e-05,
+      "loss": 0.4669,
+      "step": 3618
+    },
+    {
+      "epoch": 0.7135252365930599,
+      "grad_norm": 0.5966913216409047,
+      "learning_rate": 1.848392899647214e-05,
+      "loss": 0.4058,
+      "step": 3619
+    },
+    {
+      "epoch": 0.7137223974763407,
+      "grad_norm": 0.5710136188088236,
+      "learning_rate": 1.8483108361858263e-05,
+      "loss": 0.4354,
+      "step": 3620
+    },
+    {
+      "epoch": 0.7139195583596214,
+      "grad_norm": 0.5532446856241241,
+      "learning_rate": 1.8482287523430876e-05,
+      "loss": 0.4198,
+      "step": 3621
+    },
+    {
+      "epoch": 0.7141167192429022,
+      "grad_norm": 0.6722276257279677,
+      "learning_rate": 1.8481466481209696e-05,
+      "loss": 0.46,
+      "step": 3622
+    },
+    {
+      "epoch": 0.714313880126183,
+      "grad_norm": 0.5639143923716924,
+      "learning_rate": 1.848064523521446e-05,
+      "loss": 0.47,
+      "step": 3623
+    },
+    {
+      "epoch": 0.7145110410094637,
+      "grad_norm": 0.651640797548847,
+      "learning_rate": 1.847982378546489e-05,
+      "loss": 0.4575,
+      "step": 3624
+    },
+    {
+      "epoch": 0.7147082018927445,
+      "grad_norm": 0.6678536066940324,
+      "learning_rate": 1.8479002131980726e-05,
+      "loss": 0.4547,
+      "step": 3625
+    },
+    {
+      "epoch": 0.7149053627760252,
+      "grad_norm": 0.5640299919371896,
+      "learning_rate": 1.8478180274781707e-05,
+      "loss": 0.4302,
+      "step": 3626
+    },
+    {
+      "epoch": 0.715102523659306,
+      "grad_norm": 0.5625943114415008,
+      "learning_rate": 1.8477358213887578e-05,
+      "loss": 0.4295,
+      "step": 3627
+    },
+    {
+      "epoch": 0.7152996845425867,
+      "grad_norm": 0.8717418579189087,
+      "learning_rate": 1.8476535949318092e-05,
+      "loss": 0.4288,
+      "step": 3628
+    },
+    {
+      "epoch": 0.7154968454258676,
+      "grad_norm": 0.7111204636633626,
+      "learning_rate": 1.8475713481093005e-05,
+      "loss": 0.4779,
+      "step": 3629
+    },
+    {
+      "epoch": 0.7156940063091483,
+      "grad_norm": 0.5934289467421391,
+      "learning_rate": 1.8474890809232073e-05,
+      "loss": 0.4132,
+      "step": 3630
+    },
+    {
+      "epoch": 0.7158911671924291,
+      "grad_norm": 0.5316183763624316,
+      "learning_rate": 1.8474067933755067e-05,
+      "loss": 0.4343,
+      "step": 3631
+    },
+    {
+      "epoch": 0.7160883280757098,
+      "grad_norm": 0.6347964274352421,
+      "learning_rate": 1.8473244854681755e-05,
+      "loss": 0.4702,
+      "step": 3632
+    },
+    {
+      "epoch": 0.7162854889589906,
+      "grad_norm": 0.5307412228255478,
+      "learning_rate": 1.847242157203191e-05,
+      "loss": 0.4272,
+      "step": 3633
+    },
+    {
+      "epoch": 0.7164826498422713,
+      "grad_norm": 0.6030911118878556,
+      "learning_rate": 1.8471598085825318e-05,
+      "loss": 0.4644,
+      "step": 3634
+    },
+    {
+      "epoch": 0.7166798107255521,
+      "grad_norm": 0.5631771711931902,
+      "learning_rate": 1.8470774396081756e-05,
+      "loss": 0.445,
+      "step": 3635
+    },
+    {
+      "epoch": 0.7168769716088328,
+      "grad_norm": 0.528315324776306,
+      "learning_rate": 1.846995050282102e-05,
+      "loss": 0.413,
+      "step": 3636
+    },
+    {
+      "epoch": 0.7170741324921136,
+      "grad_norm": 0.542371055858913,
+      "learning_rate": 1.84691264060629e-05,
+      "loss": 0.4628,
+      "step": 3637
+    },
+    {
+      "epoch": 0.7172712933753943,
+      "grad_norm": 0.5752445019945468,
+      "learning_rate": 1.8468302105827195e-05,
+      "loss": 0.4715,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7174684542586751,
+      "grad_norm": 0.5556213004935148,
+      "learning_rate": 1.8467477602133716e-05,
+      "loss": 0.451,
+      "step": 3639
+    },
+    {
+      "epoch": 0.7176656151419558,
+      "grad_norm": 0.5529878854385556,
+      "learning_rate": 1.8466652895002272e-05,
+      "loss": 0.4484,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7178627760252366,
+      "grad_norm": 0.5491695139904219,
+      "learning_rate": 1.846582798445267e-05,
+      "loss": 0.4218,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7180599369085173,
+      "grad_norm": 0.5467401669297904,
+      "learning_rate": 1.8465002870504734e-05,
+      "loss": 0.4229,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7182570977917981,
+      "grad_norm": 0.557372492282451,
+      "learning_rate": 1.8464177553178287e-05,
+      "loss": 0.426,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7184542586750788,
+      "grad_norm": 0.5362074017531132,
+      "learning_rate": 1.8463352032493162e-05,
+      "loss": 0.41,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7186514195583596,
+      "grad_norm": 0.5685836955448099,
+      "learning_rate": 1.8462526308469182e-05,
+      "loss": 0.4272,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7188485804416404,
+      "grad_norm": 0.5370351970866275,
+      "learning_rate": 1.8461700381126198e-05,
+      "loss": 0.4278,
+      "step": 3646
+    },
+    {
+      "epoch": 0.7190457413249212,
+      "grad_norm": 0.5227153310031122,
+      "learning_rate": 1.8460874250484045e-05,
+      "loss": 0.4478,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7192429022082019,
+      "grad_norm": 0.5582355226303339,
+      "learning_rate": 1.8460047916562573e-05,
+      "loss": 0.412,
+      "step": 3648
+    },
+    {
+      "epoch": 0.7194400630914827,
+      "grad_norm": 0.9636937931489509,
+      "learning_rate": 1.845922137938164e-05,
+      "loss": 0.4217,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7196372239747634,
+      "grad_norm": 0.5507780948989435,
+      "learning_rate": 1.8458394638961102e-05,
+      "loss": 0.46,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7198343848580442,
+      "grad_norm": 0.5726098792524265,
+      "learning_rate": 1.845756769532082e-05,
+      "loss": 0.4523,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7200315457413249,
+      "grad_norm": 0.5717699113943913,
+      "learning_rate": 1.8456740548480666e-05,
+      "loss": 0.4571,
+      "step": 3652
+    },
+    {
+      "epoch": 0.7202287066246057,
+      "grad_norm": 0.6191561110356281,
+      "learning_rate": 1.8455913198460503e-05,
+      "loss": 0.4511,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7204258675078864,
+      "grad_norm": 0.707869839857266,
+      "learning_rate": 1.845508564528022e-05,
+      "loss": 0.4854,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7206230283911672,
+      "grad_norm": 0.5271413420379327,
+      "learning_rate": 1.8454257888959695e-05,
+      "loss": 0.4362,
+      "step": 3655
+    },
+    {
+      "epoch": 0.7208201892744479,
+      "grad_norm": 0.6354322809953766,
+      "learning_rate": 1.845342992951882e-05,
+      "loss": 0.4481,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7210173501577287,
+      "grad_norm": 0.5688074081822942,
+      "learning_rate": 1.845260176697748e-05,
+      "loss": 0.4423,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7212145110410094,
+      "grad_norm": 0.5351090443779083,
+      "learning_rate": 1.8451773401355576e-05,
+      "loss": 0.4057,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7214116719242902,
+      "grad_norm": 0.550968760823806,
+      "learning_rate": 1.845094483267301e-05,
+      "loss": 0.4556,
+      "step": 3659
+    },
+    {
+      "epoch": 0.7216088328075709,
+      "grad_norm": 0.5487359171002028,
+      "learning_rate": 1.845011606094969e-05,
+      "loss": 0.4499,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7218059936908517,
+      "grad_norm": 0.544581400232813,
+      "learning_rate": 1.8449287086205525e-05,
+      "loss": 0.4242,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7220031545741324,
+      "grad_norm": 0.5414669158249051,
+      "learning_rate": 1.8448457908460434e-05,
+      "loss": 0.4179,
+      "step": 3662
+    },
+    {
+      "epoch": 0.7222003154574133,
+      "grad_norm": 0.5796223289254395,
+      "learning_rate": 1.844762852773434e-05,
+      "loss": 0.4388,
+      "step": 3663
+    },
+    {
+      "epoch": 0.722397476340694,
+      "grad_norm": 0.5501910206556003,
+      "learning_rate": 1.8446798944047163e-05,
+      "loss": 0.4614,
+      "step": 3664
+    },
+    {
+      "epoch": 0.7225946372239748,
+      "grad_norm": 0.7117890177693054,
+      "learning_rate": 1.8445969157418845e-05,
+      "loss": 0.4557,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7227917981072555,
+      "grad_norm": 0.5801289426969807,
+      "learning_rate": 1.844513916786931e-05,
+      "loss": 0.4469,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7229889589905363,
+      "grad_norm": 0.565154672928485,
+      "learning_rate": 1.844430897541851e-05,
+      "loss": 0.411,
+      "step": 3667
+    },
+    {
+      "epoch": 0.723186119873817,
+      "grad_norm": 0.5512031287037732,
+      "learning_rate": 1.8443478580086388e-05,
+      "loss": 0.4382,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7233832807570978,
+      "grad_norm": 0.5610735735424272,
+      "learning_rate": 1.844264798189289e-05,
+      "loss": 0.4713,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7235804416403786,
+      "grad_norm": 0.5706453195319626,
+      "learning_rate": 1.8441817180857977e-05,
+      "loss": 0.4797,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7237776025236593,
+      "grad_norm": 0.5321887393174666,
+      "learning_rate": 1.844098617700161e-05,
+      "loss": 0.4554,
+      "step": 3671
+    },
+    {
+      "epoch": 0.7239747634069401,
+      "grad_norm": 0.54415284289807,
+      "learning_rate": 1.8440154970343747e-05,
+      "loss": 0.4344,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7241719242902208,
+      "grad_norm": 0.5250957449330941,
+      "learning_rate": 1.843932356090437e-05,
+      "loss": 0.4252,
+      "step": 3673
+    },
+    {
+      "epoch": 0.7243690851735016,
+      "grad_norm": 0.5266852693718858,
+      "learning_rate": 1.8438491948703445e-05,
+      "loss": 0.4257,
+      "step": 3674
+    },
+    {
+      "epoch": 0.7245662460567823,
+      "grad_norm": 0.5325304829293485,
+      "learning_rate": 1.8437660133760955e-05,
+      "loss": 0.4373,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7247634069400631,
+      "grad_norm": 0.5793560541922426,
+      "learning_rate": 1.8436828116096886e-05,
+      "loss": 0.4349,
+      "step": 3676
+    },
+    {
+      "epoch": 0.7249605678233438,
+      "grad_norm": 0.560121961862576,
+      "learning_rate": 1.843599589573123e-05,
+      "loss": 0.4494,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7251577287066246,
+      "grad_norm": 0.5249351610445621,
+      "learning_rate": 1.843516347268398e-05,
+      "loss": 0.4246,
+      "step": 3678
+    },
+    {
+      "epoch": 0.7253548895899053,
+      "grad_norm": 0.5412985163854938,
+      "learning_rate": 1.8434330846975128e-05,
+      "loss": 0.4273,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7255520504731862,
+      "grad_norm": 0.5315287597533882,
+      "learning_rate": 1.843349801862469e-05,
+      "loss": 0.4564,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7257492113564669,
+      "grad_norm": 0.5682525884568123,
+      "learning_rate": 1.843266498765267e-05,
+      "loss": 0.4446,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7259463722397477,
+      "grad_norm": 0.5374471401172799,
+      "learning_rate": 1.8431831754079084e-05,
+      "loss": 0.4586,
+      "step": 3682
+    },
+    {
+      "epoch": 0.7261435331230284,
+      "grad_norm": 0.5687422399150629,
+      "learning_rate": 1.843099831792395e-05,
+      "loss": 0.4518,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7263406940063092,
+      "grad_norm": 0.5448483856895682,
+      "learning_rate": 1.843016467920729e-05,
+      "loss": 0.4522,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7265378548895899,
+      "grad_norm": 0.5303752136858124,
+      "learning_rate": 1.8429330837949134e-05,
+      "loss": 0.4303,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7267350157728707,
+      "grad_norm": 0.5441603901459253,
+      "learning_rate": 1.842849679416952e-05,
+      "loss": 0.4265,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7269321766561514,
+      "grad_norm": 0.5565608021825121,
+      "learning_rate": 1.842766254788848e-05,
+      "loss": 0.4704,
+      "step": 3687
+    },
+    {
+      "epoch": 0.7271293375394322,
+      "grad_norm": 0.5392512013397398,
+      "learning_rate": 1.8426828099126058e-05,
+      "loss": 0.4282,
+      "step": 3688
+    },
+    {
+      "epoch": 0.7273264984227129,
+      "grad_norm": 0.5228171340025033,
+      "learning_rate": 1.8425993447902312e-05,
+      "loss": 0.4262,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7275236593059937,
+      "grad_norm": 0.5057395283789563,
+      "learning_rate": 1.8425158594237285e-05,
+      "loss": 0.3781,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7277208201892744,
+      "grad_norm": 0.5637585116791193,
+      "learning_rate": 1.8424323538151038e-05,
+      "loss": 0.4578,
+      "step": 3691
+    },
+    {
+      "epoch": 0.7279179810725552,
+      "grad_norm": 0.5770703957710751,
+      "learning_rate": 1.842348827966363e-05,
+      "loss": 0.4474,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7281151419558359,
+      "grad_norm": 0.5757553217171197,
+      "learning_rate": 1.8422652818795136e-05,
+      "loss": 0.4266,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7283123028391167,
+      "grad_norm": 0.5212848130468406,
+      "learning_rate": 1.8421817155565627e-05,
+      "loss": 0.4297,
+      "step": 3694
+    },
+    {
+      "epoch": 0.7285094637223974,
+      "grad_norm": 0.6230547006563527,
+      "learning_rate": 1.8420981289995174e-05,
+      "loss": 0.5156,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7287066246056783,
+      "grad_norm": 0.5232291475507522,
+      "learning_rate": 1.842014522210387e-05,
+      "loss": 0.4114,
+      "step": 3696
+    },
+    {
+      "epoch": 0.728903785488959,
+      "grad_norm": 0.5347786514736478,
+      "learning_rate": 1.841930895191179e-05,
+      "loss": 0.4235,
+      "step": 3697
+    },
+    {
+      "epoch": 0.7291009463722398,
+      "grad_norm": 0.5462129983268524,
+      "learning_rate": 1.841847247943904e-05,
+      "loss": 0.4421,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7292981072555205,
+      "grad_norm": 0.49073103119405975,
+      "learning_rate": 1.84176358047057e-05,
+      "loss": 0.3842,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7294952681388013,
+      "grad_norm": 0.6150333675068365,
+      "learning_rate": 1.8416798927731888e-05,
+      "loss": 0.4886,
+      "step": 3700
+    },
+    {
+      "epoch": 0.729692429022082,
+      "grad_norm": 0.5361700966964398,
+      "learning_rate": 1.8415961848537702e-05,
+      "loss": 0.4386,
+      "step": 3701
+    },
+    {
+      "epoch": 0.7298895899053628,
+      "grad_norm": 0.5347576067338814,
+      "learning_rate": 1.8415124567143258e-05,
+      "loss": 0.4184,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7300867507886435,
+      "grad_norm": 0.5621215334987752,
+      "learning_rate": 1.8414287083568666e-05,
+      "loss": 0.4261,
+      "step": 3703
+    },
+    {
+      "epoch": 0.7302839116719243,
+      "grad_norm": 0.5700646061060279,
+      "learning_rate": 1.841344939783405e-05,
+      "loss": 0.4179,
+      "step": 3704
+    },
+    {
+      "epoch": 0.730481072555205,
+      "grad_norm": 0.518131991780773,
+      "learning_rate": 1.841261150995954e-05,
+      "loss": 0.4339,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7306782334384858,
+      "grad_norm": 0.5459761294030725,
+      "learning_rate": 1.8411773419965263e-05,
+      "loss": 0.397,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7308753943217665,
+      "grad_norm": 0.9314288519022254,
+      "learning_rate": 1.8410935127871356e-05,
+      "loss": 0.442,
+      "step": 3707
+    },
+    {
+      "epoch": 0.7310725552050473,
+      "grad_norm": 0.5636871116155396,
+      "learning_rate": 1.8410096633697956e-05,
+      "loss": 0.4173,
+      "step": 3708
+    },
+    {
+      "epoch": 0.731269716088328,
+      "grad_norm": 0.5193160515915649,
+      "learning_rate": 1.8409257937465216e-05,
+      "loss": 0.3913,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7314668769716088,
+      "grad_norm": 0.712727581510582,
+      "learning_rate": 1.840841903919328e-05,
+      "loss": 0.4502,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7316640378548895,
+      "grad_norm": 0.5769818802867137,
+      "learning_rate": 1.8407579938902302e-05,
+      "loss": 0.4527,
+      "step": 3711
+    },
+    {
+      "epoch": 0.7318611987381703,
+      "grad_norm": 0.5392705119979897,
+      "learning_rate": 1.8406740636612447e-05,
+      "loss": 0.4538,
+      "step": 3712
+    },
+    {
+      "epoch": 0.732058359621451,
+      "grad_norm": 0.6098765553285055,
+      "learning_rate": 1.8405901132343882e-05,
+      "loss": 0.4413,
+      "step": 3713
+    },
+    {
+      "epoch": 0.7322555205047319,
+      "grad_norm": 0.5771216137328318,
+      "learning_rate": 1.840506142611677e-05,
+      "loss": 0.4454,
+      "step": 3714
+    },
+    {
+      "epoch": 0.7324526813880127,
+      "grad_norm": 0.62152619615878,
+      "learning_rate": 1.840422151795129e-05,
+      "loss": 0.4713,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7326498422712934,
+      "grad_norm": 0.5895477541474831,
+      "learning_rate": 1.840338140786762e-05,
+      "loss": 0.4415,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7328470031545742,
+      "grad_norm": 0.7408210800785281,
+      "learning_rate": 1.8402541095885943e-05,
+      "loss": 0.4733,
+      "step": 3717
+    },
+    {
+      "epoch": 0.7330441640378549,
+      "grad_norm": 0.56864244318536,
+      "learning_rate": 1.8401700582026452e-05,
+      "loss": 0.4047,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7332413249211357,
+      "grad_norm": 0.6695180859601412,
+      "learning_rate": 1.8400859866309337e-05,
+      "loss": 0.4215,
+      "step": 3719
+    },
+    {
+      "epoch": 0.7334384858044164,
+      "grad_norm": 0.6556762484776179,
+      "learning_rate": 1.84000189487548e-05,
+      "loss": 0.5069,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7336356466876972,
+      "grad_norm": 0.5610750056727031,
+      "learning_rate": 1.8399177829383043e-05,
+      "loss": 0.4505,
+      "step": 3721
+    },
+    {
+      "epoch": 0.7338328075709779,
+      "grad_norm": 0.5779111361169296,
+      "learning_rate": 1.839833650821427e-05,
+      "loss": 0.4109,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7340299684542587,
+      "grad_norm": 0.5512782925585601,
+      "learning_rate": 1.8397494985268705e-05,
+      "loss": 0.4559,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7342271293375394,
+      "grad_norm": 0.5691768441468529,
+      "learning_rate": 1.839665326056656e-05,
+      "loss": 0.4112,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7344242902208202,
+      "grad_norm": 0.5624881189319345,
+      "learning_rate": 1.8395811334128058e-05,
+      "loss": 0.4293,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7346214511041009,
+      "grad_norm": 0.7923653826550998,
+      "learning_rate": 1.8394969205973426e-05,
+      "loss": 0.418,
+      "step": 3726
+    },
+    {
+      "epoch": 0.7348186119873817,
+      "grad_norm": 0.5815459567626416,
+      "learning_rate": 1.8394126876122896e-05,
+      "loss": 0.4526,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7350157728706624,
+      "grad_norm": 0.6090000924351008,
+      "learning_rate": 1.8393284344596715e-05,
+      "loss": 0.4544,
+      "step": 3728
+    },
+    {
+      "epoch": 0.7352129337539433,
+      "grad_norm": 0.5232417470669563,
+      "learning_rate": 1.8392441611415113e-05,
+      "loss": 0.4496,
+      "step": 3729
+    },
+    {
+      "epoch": 0.735410094637224,
+      "grad_norm": 0.7399663560330513,
+      "learning_rate": 1.8391598676598344e-05,
+      "loss": 0.4242,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7356072555205048,
+      "grad_norm": 0.5615281033101944,
+      "learning_rate": 1.8390755540166663e-05,
+      "loss": 0.4376,
+      "step": 3731
+    },
+    {
+      "epoch": 0.7358044164037855,
+      "grad_norm": 0.599463559470186,
+      "learning_rate": 1.8389912202140318e-05,
+      "loss": 0.446,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7360015772870663,
+      "grad_norm": 0.5377177978774068,
+      "learning_rate": 1.838906866253958e-05,
+      "loss": 0.4516,
+      "step": 3733
+    },
+    {
+      "epoch": 0.736198738170347,
+      "grad_norm": 0.5681460967594567,
+      "learning_rate": 1.8388224921384707e-05,
+      "loss": 0.4066,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7363958990536278,
+      "grad_norm": 0.5141694762603598,
+      "learning_rate": 1.8387380978695977e-05,
+      "loss": 0.4322,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7365930599369085,
+      "grad_norm": 0.8238886062154376,
+      "learning_rate": 1.8386536834493667e-05,
+      "loss": 0.4507,
+      "step": 3736
+    },
+    {
+      "epoch": 0.7367902208201893,
+      "grad_norm": 0.535199442346653,
+      "learning_rate": 1.8385692488798056e-05,
+      "loss": 0.4553,
+      "step": 3737
+    },
+    {
+      "epoch": 0.73698738170347,
+      "grad_norm": 0.6483426681293745,
+      "learning_rate": 1.8384847941629423e-05,
+      "loss": 0.4634,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7371845425867508,
+      "grad_norm": 0.5764953582261776,
+      "learning_rate": 1.8384003193008072e-05,
+      "loss": 0.442,
+      "step": 3739
+    },
+    {
+      "epoch": 0.7373817034700315,
+      "grad_norm": 0.5572119301548107,
+      "learning_rate": 1.8383158242954296e-05,
+      "loss": 0.4508,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7375788643533123,
+      "grad_norm": 0.5681528660063264,
+      "learning_rate": 1.8382313091488385e-05,
+      "loss": 0.4532,
+      "step": 3741
+    },
+    {
+      "epoch": 0.737776025236593,
+      "grad_norm": 0.5891890950746687,
+      "learning_rate": 1.8381467738630656e-05,
+      "loss": 0.4669,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7379731861198738,
+      "grad_norm": 0.5049888382066734,
+      "learning_rate": 1.8380622184401416e-05,
+      "loss": 0.3829,
+      "step": 3743
+    },
+    {
+      "epoch": 0.7381703470031545,
+      "grad_norm": 0.563208762050558,
+      "learning_rate": 1.8379776428820974e-05,
+      "loss": 0.4047,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7383675078864353,
+      "grad_norm": 0.5779457902661825,
+      "learning_rate": 1.8378930471909658e-05,
+      "loss": 0.4206,
+      "step": 3745
+    },
+    {
+      "epoch": 0.738564668769716,
+      "grad_norm": 0.7774905699447151,
+      "learning_rate": 1.837808431368779e-05,
+      "loss": 0.4779,
+      "step": 3746
+    },
+    {
+      "epoch": 0.7387618296529969,
+      "grad_norm": 0.5629107990200406,
+      "learning_rate": 1.83772379541757e-05,
+      "loss": 0.3925,
+      "step": 3747
+    },
+    {
+      "epoch": 0.7389589905362776,
+      "grad_norm": 0.5970620392228986,
+      "learning_rate": 1.837639139339372e-05,
+      "loss": 0.4488,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7391561514195584,
+      "grad_norm": 0.5557125511475749,
+      "learning_rate": 1.8375544631362195e-05,
+      "loss": 0.452,
+      "step": 3749
+    },
+    {
+      "epoch": 0.7393533123028391,
+      "grad_norm": 0.5872831543728793,
+      "learning_rate": 1.8374697668101463e-05,
+      "loss": 0.4594,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7395504731861199,
+      "grad_norm": 0.5286161720159008,
+      "learning_rate": 1.8373850503631872e-05,
+      "loss": 0.4322,
+      "step": 3751
+    },
+    {
+      "epoch": 0.7397476340694006,
+      "grad_norm": 0.560014306046355,
+      "learning_rate": 1.8373003137973783e-05,
+      "loss": 0.4316,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7399447949526814,
+      "grad_norm": 0.5456631340877649,
+      "learning_rate": 1.8372155571147554e-05,
+      "loss": 0.4331,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7401419558359621,
+      "grad_norm": 0.5985522047951393,
+      "learning_rate": 1.837130780317354e-05,
+      "loss": 0.4628,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7403391167192429,
+      "grad_norm": 0.5672181573442395,
+      "learning_rate": 1.8370459834072118e-05,
+      "loss": 0.4454,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7405362776025236,
+      "grad_norm": 0.5635085742932182,
+      "learning_rate": 1.8369611663863656e-05,
+      "loss": 0.4223,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7407334384858044,
+      "grad_norm": 0.5873022605290665,
+      "learning_rate": 1.8368763292568532e-05,
+      "loss": 0.4478,
+      "step": 3757
+    },
+    {
+      "epoch": 0.7409305993690851,
+      "grad_norm": 0.5346924744494808,
+      "learning_rate": 1.8367914720207137e-05,
+      "loss": 0.3994,
+      "step": 3758
+    },
+    {
+      "epoch": 0.7411277602523659,
+      "grad_norm": 0.5791612680965824,
+      "learning_rate": 1.8367065946799845e-05,
+      "loss": 0.417,
+      "step": 3759
+    },
+    {
+      "epoch": 0.7413249211356467,
+      "grad_norm": 0.5887526324679428,
+      "learning_rate": 1.8366216972367058e-05,
+      "loss": 0.4324,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7415220820189274,
+      "grad_norm": 0.631982631731162,
+      "learning_rate": 1.836536779692917e-05,
+      "loss": 0.4398,
+      "step": 3761
+    },
+    {
+      "epoch": 0.7417192429022083,
+      "grad_norm": 0.5498266504156066,
+      "learning_rate": 1.836451842050659e-05,
+      "loss": 0.444,
+      "step": 3762
+    },
+    {
+      "epoch": 0.741916403785489,
+      "grad_norm": 0.5248298760030489,
+      "learning_rate": 1.8363668843119713e-05,
+      "loss": 0.4397,
+      "step": 3763
+    },
+    {
+      "epoch": 0.7421135646687698,
+      "grad_norm": 0.5574383621046671,
+      "learning_rate": 1.8362819064788956e-05,
+      "loss": 0.4321,
+      "step": 3764
+    },
+    {
+      "epoch": 0.7423107255520505,
+      "grad_norm": 0.5509414305533834,
+      "learning_rate": 1.8361969085534742e-05,
+      "loss": 0.4449,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7425078864353313,
+      "grad_norm": 0.6218658695377531,
+      "learning_rate": 1.8361118905377483e-05,
+      "loss": 0.4698,
+      "step": 3766
+    },
+    {
+      "epoch": 0.742705047318612,
+      "grad_norm": 0.6025705112487485,
+      "learning_rate": 1.8360268524337606e-05,
+      "loss": 0.4425,
+      "step": 3767
+    },
+    {
+      "epoch": 0.7429022082018928,
+      "grad_norm": 0.5354499032310642,
+      "learning_rate": 1.835941794243555e-05,
+      "loss": 0.4212,
+      "step": 3768
+    },
+    {
+      "epoch": 0.7430993690851735,
+      "grad_norm": 0.6198349365571671,
+      "learning_rate": 1.8358567159691745e-05,
+      "loss": 0.4173,
+      "step": 3769
+    },
+    {
+      "epoch": 0.7432965299684543,
+      "grad_norm": 0.5633007693206912,
+      "learning_rate": 1.8357716176126633e-05,
+      "loss": 0.4407,
+      "step": 3770
+    },
+    {
+      "epoch": 0.743493690851735,
+      "grad_norm": 0.6033616935263962,
+      "learning_rate": 1.8356864991760658e-05,
+      "loss": 0.4832,
+      "step": 3771
+    },
+    {
+      "epoch": 0.7436908517350158,
+      "grad_norm": 0.5318673355760555,
+      "learning_rate": 1.8356013606614277e-05,
+      "loss": 0.4512,
+      "step": 3772
+    },
+    {
+      "epoch": 0.7438880126182965,
+      "grad_norm": 0.9499333131372167,
+      "learning_rate": 1.8355162020707932e-05,
+      "loss": 0.4394,
+      "step": 3773
+    },
+    {
+      "epoch": 0.7440851735015773,
+      "grad_norm": 0.5559579298686984,
+      "learning_rate": 1.8354310234062097e-05,
+      "loss": 0.4233,
+      "step": 3774
+    },
+    {
+      "epoch": 0.744282334384858,
+      "grad_norm": 0.6229971755454574,
+      "learning_rate": 1.835345824669723e-05,
+      "loss": 0.44,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7444794952681388,
+      "grad_norm": 0.5923143947918056,
+      "learning_rate": 1.83526060586338e-05,
+      "loss": 0.4606,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7446766561514195,
+      "grad_norm": 0.5501204352607113,
+      "learning_rate": 1.8351753669892284e-05,
+      "loss": 0.4291,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7448738170347003,
+      "grad_norm": 0.6932256286950385,
+      "learning_rate": 1.8350901080493158e-05,
+      "loss": 0.461,
+      "step": 3778
+    },
+    {
+      "epoch": 0.745070977917981,
+      "grad_norm": 0.6073398760332815,
+      "learning_rate": 1.8350048290456912e-05,
+      "loss": 0.4141,
+      "step": 3779
+    },
+    {
+      "epoch": 0.7452681388012619,
+      "grad_norm": 0.5251923915616548,
+      "learning_rate": 1.834919529980403e-05,
+      "loss": 0.4481,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7454652996845426,
+      "grad_norm": 0.5442185285627984,
+      "learning_rate": 1.8348342108555007e-05,
+      "loss": 0.4026,
+      "step": 3781
+    },
+    {
+      "epoch": 0.7456624605678234,
+      "grad_norm": 0.5693867434023069,
+      "learning_rate": 1.8347488716730343e-05,
+      "loss": 0.4311,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7458596214511041,
+      "grad_norm": 0.5938661250513164,
+      "learning_rate": 1.834663512435054e-05,
+      "loss": 0.4613,
+      "step": 3783
+    },
+    {
+      "epoch": 0.7460567823343849,
+      "grad_norm": 0.5455676983214097,
+      "learning_rate": 1.8345781331436106e-05,
+      "loss": 0.4195,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7462539432176656,
+      "grad_norm": 0.5860282571566524,
+      "learning_rate": 1.8344927338007554e-05,
+      "loss": 0.4211,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7464511041009464,
+      "grad_norm": 0.5333294020645926,
+      "learning_rate": 1.8344073144085406e-05,
+      "loss": 0.4371,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7466482649842271,
+      "grad_norm": 0.5706728165649482,
+      "learning_rate": 1.834321874969018e-05,
+      "loss": 0.4389,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7468454258675079,
+      "grad_norm": 0.7074581360169672,
+      "learning_rate": 1.8342364154842404e-05,
+      "loss": 0.4937,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7470425867507886,
+      "grad_norm": 0.5778783270484186,
+      "learning_rate": 1.8341509359562608e-05,
+      "loss": 0.4377,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7472397476340694,
+      "grad_norm": 0.5270159539912247,
+      "learning_rate": 1.8340654363871334e-05,
+      "loss": 0.4117,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7474369085173501,
+      "grad_norm": 0.5223441256989405,
+      "learning_rate": 1.8339799167789127e-05,
+      "loss": 0.4471,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7476340694006309,
+      "grad_norm": 0.5781995373444279,
+      "learning_rate": 1.8338943771336522e-05,
+      "loss": 0.4378,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7478312302839116,
+      "grad_norm": 0.5634947765385621,
+      "learning_rate": 1.8338088174534083e-05,
+      "loss": 0.4454,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7480283911671924,
+      "grad_norm": 0.5059725150862037,
+      "learning_rate": 1.833723237740236e-05,
+      "loss": 0.4245,
+      "step": 3794
+    },
+    {
+      "epoch": 0.7482255520504731,
+      "grad_norm": 0.5326325739466916,
+      "learning_rate": 1.833637637996191e-05,
+      "loss": 0.4425,
+      "step": 3795
+    },
+    {
+      "epoch": 0.748422712933754,
+      "grad_norm": 0.5574154238044264,
+      "learning_rate": 1.833552018223331e-05,
+      "loss": 0.4492,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7486198738170347,
+      "grad_norm": 0.5402149077761322,
+      "learning_rate": 1.8334663784237124e-05,
+      "loss": 0.4006,
+      "step": 3797
+    },
+    {
+      "epoch": 0.7488170347003155,
+      "grad_norm": 0.5791969213862481,
+      "learning_rate": 1.8333807185993927e-05,
+      "loss": 0.4434,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7490141955835962,
+      "grad_norm": 0.6018305196834619,
+      "learning_rate": 1.8332950387524304e-05,
+      "loss": 0.4588,
+      "step": 3799
+    },
+    {
+      "epoch": 0.749211356466877,
+      "grad_norm": 0.6572227006369701,
+      "learning_rate": 1.8332093388848836e-05,
+      "loss": 0.446,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7494085173501577,
+      "grad_norm": 0.5684378740198406,
+      "learning_rate": 1.8331236189988115e-05,
+      "loss": 0.4071,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7496056782334385,
+      "grad_norm": 0.5338624607390893,
+      "learning_rate": 1.8330378790962734e-05,
+      "loss": 0.4635,
+      "step": 3802
+    },
+    {
+      "epoch": 0.7498028391167192,
+      "grad_norm": 0.5649660572252639,
+      "learning_rate": 1.8329521191793293e-05,
+      "loss": 0.4277,
+      "step": 3803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5526706058734381,
+      "learning_rate": 1.83286633925004e-05,
+      "loss": 0.4404,
+      "step": 3804
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.44013890624046326,
+      "eval_runtime": 343.9735,
+      "eval_samples_per_second": 23.636,
+      "eval_steps_per_second": 1.48,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7501971608832808,
+      "grad_norm": 0.6372546717506914,
+      "learning_rate": 1.8327805393104658e-05,
+      "loss": 0.4548,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7503943217665615,
+      "grad_norm": 0.5913842512169649,
+      "learning_rate": 1.832694719362669e-05,
+      "loss": 0.4518,
+      "step": 3806
+    },
+    {
+      "epoch": 0.7505914826498423,
+      "grad_norm": 0.5826300542247199,
+      "learning_rate": 1.8326088794087108e-05,
+      "loss": 0.4818,
+      "step": 3807
+    },
+    {
+      "epoch": 0.750788643533123,
+      "grad_norm": 0.5907815365065618,
+      "learning_rate": 1.8325230194506538e-05,
+      "loss": 0.4365,
+      "step": 3808
+    },
+    {
+      "epoch": 0.7509858044164038,
+      "grad_norm": 0.5496356683694023,
+      "learning_rate": 1.8324371394905606e-05,
+      "loss": 0.4574,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7511829652996845,
+      "grad_norm": 0.5711613351725243,
+      "learning_rate": 1.832351239530495e-05,
+      "loss": 0.4584,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7513801261829653,
+      "grad_norm": 0.9200633659180221,
+      "learning_rate": 1.8322653195725206e-05,
+      "loss": 0.4781,
+      "step": 3811
+    },
+    {
+      "epoch": 0.751577287066246,
+      "grad_norm": 0.6033375447437546,
+      "learning_rate": 1.832179379618702e-05,
+      "loss": 0.4412,
+      "step": 3812
+    },
+    {
+      "epoch": 0.7517744479495269,
+      "grad_norm": 0.5171494431027918,
+      "learning_rate": 1.832093419671103e-05,
+      "loss": 0.4153,
+      "step": 3813
+    },
+    {
+      "epoch": 0.7519716088328076,
+      "grad_norm": 0.7257919932365838,
+      "learning_rate": 1.83200743973179e-05,
+      "loss": 0.4529,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7521687697160884,
+      "grad_norm": 0.6164155494736133,
+      "learning_rate": 1.831921439802828e-05,
+      "loss": 0.4239,
+      "step": 3815
+    },
+    {
+      "epoch": 0.7523659305993691,
+      "grad_norm": 0.5742726759356669,
+      "learning_rate": 1.8318354198862836e-05,
+      "loss": 0.4595,
+      "step": 3816
+    },
+    {
+      "epoch": 0.7525630914826499,
+      "grad_norm": 0.5364358313603149,
+      "learning_rate": 1.831749379984223e-05,
+      "loss": 0.3946,
+      "step": 3817
+    },
+    {
+      "epoch": 0.7527602523659306,
+      "grad_norm": 0.5931144322460239,
+      "learning_rate": 1.8316633200987143e-05,
+      "loss": 0.4723,
+      "step": 3818
+    },
+    {
+      "epoch": 0.7529574132492114,
+      "grad_norm": 0.6631211961273565,
+      "learning_rate": 1.8315772402318243e-05,
+      "loss": 0.5013,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7531545741324921,
+      "grad_norm": 0.555817939093465,
+      "learning_rate": 1.8314911403856212e-05,
+      "loss": 0.4234,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7533517350157729,
+      "grad_norm": 0.5699005102970771,
+      "learning_rate": 1.8314050205621742e-05,
+      "loss": 0.4629,
+      "step": 3821
+    },
+    {
+      "epoch": 0.7535488958990536,
+      "grad_norm": 0.5456220237338646,
+      "learning_rate": 1.831318880763552e-05,
+      "loss": 0.4227,
+      "step": 3822
+    },
+    {
+      "epoch": 0.7537460567823344,
+      "grad_norm": 0.564506446004603,
+      "learning_rate": 1.8312327209918242e-05,
+      "loss": 0.4478,
+      "step": 3823
+    },
+    {
+      "epoch": 0.7539432176656151,
+      "grad_norm": 0.6007442236000919,
+      "learning_rate": 1.831146541249061e-05,
+      "loss": 0.4404,
+      "step": 3824
+    },
+    {
+      "epoch": 0.7541403785488959,
+      "grad_norm": 0.5756483770605955,
+      "learning_rate": 1.8310603415373328e-05,
+      "loss": 0.436,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7543375394321766,
+      "grad_norm": 0.5730763258268717,
+      "learning_rate": 1.8309741218587102e-05,
+      "loss": 0.4509,
+      "step": 3826
+    },
+    {
+      "epoch": 0.7545347003154574,
+      "grad_norm": 0.573198374198784,
+      "learning_rate": 1.8308878822152655e-05,
+      "loss": 0.437,
+      "step": 3827
+    },
+    {
+      "epoch": 0.7547318611987381,
+      "grad_norm": 0.5616109549442615,
+      "learning_rate": 1.8308016226090704e-05,
+      "loss": 0.4439,
+      "step": 3828
+    },
+    {
+      "epoch": 0.754929022082019,
+      "grad_norm": 0.5895791098794968,
+      "learning_rate": 1.8307153430421972e-05,
+      "loss": 0.4744,
+      "step": 3829
+    },
+    {
+      "epoch": 0.7551261829652997,
+      "grad_norm": 0.5565201237486335,
+      "learning_rate": 1.830629043516719e-05,
+      "loss": 0.4657,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7553233438485805,
+      "grad_norm": 0.5061359477774192,
+      "learning_rate": 1.830542724034709e-05,
+      "loss": 0.4204,
+      "step": 3831
+    },
+    {
+      "epoch": 0.7555205047318612,
+      "grad_norm": 0.5660890953859956,
+      "learning_rate": 1.8304563845982413e-05,
+      "loss": 0.4396,
+      "step": 3832
+    },
+    {
+      "epoch": 0.755717665615142,
+      "grad_norm": 0.5503977078210044,
+      "learning_rate": 1.83037002520939e-05,
+      "loss": 0.4393,
+      "step": 3833
+    },
+    {
+      "epoch": 0.7559148264984227,
+      "grad_norm": 0.6550111174917882,
+      "learning_rate": 1.8302836458702302e-05,
+      "loss": 0.4523,
+      "step": 3834
+    },
+    {
+      "epoch": 0.7561119873817035,
+      "grad_norm": 2.619858501191366,
+      "learning_rate": 1.8301972465828373e-05,
+      "loss": 0.469,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7563091482649842,
+      "grad_norm": 0.6235970138061254,
+      "learning_rate": 1.830110827349287e-05,
+      "loss": 0.4333,
+      "step": 3836
+    },
+    {
+      "epoch": 0.756506309148265,
+      "grad_norm": 0.532577589765572,
+      "learning_rate": 1.8300243881716553e-05,
+      "loss": 0.4646,
+      "step": 3837
+    },
+    {
+      "epoch": 0.7567034700315457,
+      "grad_norm": 0.6198914137370217,
+      "learning_rate": 1.8299379290520197e-05,
+      "loss": 0.4148,
+      "step": 3838
+    },
+    {
+      "epoch": 0.7569006309148265,
+      "grad_norm": 0.5916239893137991,
+      "learning_rate": 1.8298514499924567e-05,
+      "loss": 0.4566,
+      "step": 3839
+    },
+    {
+      "epoch": 0.7570977917981072,
+      "grad_norm": 0.5602675170057476,
+      "learning_rate": 1.8297649509950446e-05,
+      "loss": 0.4242,
+      "step": 3840
+    },
+    {
+      "epoch": 0.757294952681388,
+      "grad_norm": 0.493735580049777,
+      "learning_rate": 1.829678432061861e-05,
+      "loss": 0.3774,
+      "step": 3841
+    },
+    {
+      "epoch": 0.7574921135646687,
+      "grad_norm": 0.5400394381153409,
+      "learning_rate": 1.829591893194985e-05,
+      "loss": 0.4206,
+      "step": 3842
+    },
+    {
+      "epoch": 0.7576892744479495,
+      "grad_norm": 0.5643283428425839,
+      "learning_rate": 1.829505334396496e-05,
+      "loss": 0.4571,
+      "step": 3843
+    },
+    {
+      "epoch": 0.7578864353312302,
+      "grad_norm": 0.5377612017847214,
+      "learning_rate": 1.8294187556684733e-05,
+      "loss": 0.4133,
+      "step": 3844
+    },
+    {
+      "epoch": 0.758083596214511,
+      "grad_norm": 0.5497122296347786,
+      "learning_rate": 1.829332157012997e-05,
+      "loss": 0.4332,
+      "step": 3845
+    },
+    {
+      "epoch": 0.7582807570977917,
+      "grad_norm": 0.5849200039580923,
+      "learning_rate": 1.8292455384321476e-05,
+      "loss": 0.4284,
+      "step": 3846
+    },
+    {
+      "epoch": 0.7584779179810726,
+      "grad_norm": 0.5543841134301215,
+      "learning_rate": 1.8291588999280065e-05,
+      "loss": 0.4532,
+      "step": 3847
+    },
+    {
+      "epoch": 0.7586750788643533,
+      "grad_norm": 0.5615983003588527,
+      "learning_rate": 1.8290722415026548e-05,
+      "loss": 0.4302,
+      "step": 3848
+    },
+    {
+      "epoch": 0.7588722397476341,
+      "grad_norm": 0.551746761958415,
+      "learning_rate": 1.828985563158175e-05,
+      "loss": 0.3985,
+      "step": 3849
+    },
+    {
+      "epoch": 0.7590694006309149,
+      "grad_norm": 0.6412908877953788,
+      "learning_rate": 1.8288988648966498e-05,
+      "loss": 0.415,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7592665615141956,
+      "grad_norm": 0.5308071077261949,
+      "learning_rate": 1.8288121467201615e-05,
+      "loss": 0.4216,
+      "step": 3851
+    },
+    {
+      "epoch": 0.7594637223974764,
+      "grad_norm": 0.5623371567129707,
+      "learning_rate": 1.8287254086307942e-05,
+      "loss": 0.4205,
+      "step": 3852
+    },
+    {
+      "epoch": 0.7596608832807571,
+      "grad_norm": 0.6609929780062805,
+      "learning_rate": 1.8286386506306314e-05,
+      "loss": 0.4483,
+      "step": 3853
+    },
+    {
+      "epoch": 0.7598580441640379,
+      "grad_norm": 0.5571621496040055,
+      "learning_rate": 1.8285518727217578e-05,
+      "loss": 0.3971,
+      "step": 3854
+    },
+    {
+      "epoch": 0.7600552050473186,
+      "grad_norm": 0.6151096972545813,
+      "learning_rate": 1.8284650749062583e-05,
+      "loss": 0.4393,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7602523659305994,
+      "grad_norm": 0.565063463574547,
+      "learning_rate": 1.8283782571862182e-05,
+      "loss": 0.4521,
+      "step": 3856
+    },
+    {
+      "epoch": 0.7604495268138801,
+      "grad_norm": 0.5410714930908408,
+      "learning_rate": 1.828291419563723e-05,
+      "loss": 0.4147,
+      "step": 3857
+    },
+    {
+      "epoch": 0.7606466876971609,
+      "grad_norm": 0.5396095207165513,
+      "learning_rate": 1.8282045620408596e-05,
+      "loss": 0.4591,
+      "step": 3858
+    },
+    {
+      "epoch": 0.7608438485804416,
+      "grad_norm": 0.6397377487608418,
+      "learning_rate": 1.828117684619715e-05,
+      "loss": 0.4952,
+      "step": 3859
+    },
+    {
+      "epoch": 0.7610410094637224,
+      "grad_norm": 0.5360037148692628,
+      "learning_rate": 1.8280307873023758e-05,
+      "loss": 0.4169,
+      "step": 3860
+    },
+    {
+      "epoch": 0.7612381703470031,
+      "grad_norm": 0.5634653164334658,
+      "learning_rate": 1.8279438700909305e-05,
+      "loss": 0.4209,
+      "step": 3861
+    },
+    {
+      "epoch": 0.761435331230284,
+      "grad_norm": 0.5642735061715564,
+      "learning_rate": 1.8278569329874667e-05,
+      "loss": 0.4156,
+      "step": 3862
+    },
+    {
+      "epoch": 0.7616324921135647,
+      "grad_norm": 0.5346697867413164,
+      "learning_rate": 1.8277699759940732e-05,
+      "loss": 0.4157,
+      "step": 3863
+    },
+    {
+      "epoch": 0.7618296529968455,
+      "grad_norm": 0.5458496115616259,
+      "learning_rate": 1.8276829991128397e-05,
+      "loss": 0.4412,
+      "step": 3864
+    },
+    {
+      "epoch": 0.7620268138801262,
+      "grad_norm": 0.542780014402022,
+      "learning_rate": 1.8275960023458554e-05,
+      "loss": 0.4051,
+      "step": 3865
+    },
+    {
+      "epoch": 0.762223974763407,
+      "grad_norm": 0.5179693655492266,
+      "learning_rate": 1.827508985695211e-05,
+      "loss": 0.4127,
+      "step": 3866
+    },
+    {
+      "epoch": 0.7624211356466877,
+      "grad_norm": 0.5553660779995085,
+      "learning_rate": 1.8274219491629965e-05,
+      "loss": 0.4401,
+      "step": 3867
+    },
+    {
+      "epoch": 0.7626182965299685,
+      "grad_norm": 0.5712424475051656,
+      "learning_rate": 1.827334892751304e-05,
+      "loss": 0.4267,
+      "step": 3868
+    },
+    {
+      "epoch": 0.7628154574132492,
+      "grad_norm": 0.5521939908028256,
+      "learning_rate": 1.8272478164622237e-05,
+      "loss": 0.4525,
+      "step": 3869
+    },
+    {
+      "epoch": 0.76301261829653,
+      "grad_norm": 0.6059666034435798,
+      "learning_rate": 1.827160720297849e-05,
+      "loss": 0.4352,
+      "step": 3870
+    },
+    {
+      "epoch": 0.7632097791798107,
+      "grad_norm": 0.5568390943861937,
+      "learning_rate": 1.827073604260271e-05,
+      "loss": 0.4435,
+      "step": 3871
+    },
+    {
+      "epoch": 0.7634069400630915,
+      "grad_norm": 0.6154734211328192,
+      "learning_rate": 1.8269864683515847e-05,
+      "loss": 0.478,
+      "step": 3872
+    },
+    {
+      "epoch": 0.7636041009463722,
+      "grad_norm": 0.6336972990527457,
+      "learning_rate": 1.8268993125738817e-05,
+      "loss": 0.488,
+      "step": 3873
+    },
+    {
+      "epoch": 0.763801261829653,
+      "grad_norm": 0.5158784852798877,
+      "learning_rate": 1.826812136929257e-05,
+      "loss": 0.3997,
+      "step": 3874
+    },
+    {
+      "epoch": 0.7639984227129337,
+      "grad_norm": 0.5545942044525611,
+      "learning_rate": 1.8267249414198055e-05,
+      "loss": 0.4042,
+      "step": 3875
+    },
+    {
+      "epoch": 0.7641955835962145,
+      "grad_norm": 0.7691672609377771,
+      "learning_rate": 1.8266377260476206e-05,
+      "loss": 0.4804,
+      "step": 3876
+    },
+    {
+      "epoch": 0.7643927444794952,
+      "grad_norm": 0.607087584955189,
+      "learning_rate": 1.826550490814799e-05,
+      "loss": 0.4558,
+      "step": 3877
+    },
+    {
+      "epoch": 0.764589905362776,
+      "grad_norm": 0.62388863749188,
+      "learning_rate": 1.8264632357234366e-05,
+      "loss": 0.4604,
+      "step": 3878
+    },
+    {
+      "epoch": 0.7647870662460567,
+      "grad_norm": 0.5639061589375046,
+      "learning_rate": 1.826375960775629e-05,
+      "loss": 0.4197,
+      "step": 3879
+    },
+    {
+      "epoch": 0.7649842271293376,
+      "grad_norm": 0.575114173679805,
+      "learning_rate": 1.8262886659734738e-05,
+      "loss": 0.466,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7651813880126183,
+      "grad_norm": 1.0686366669808436,
+      "learning_rate": 1.8262013513190677e-05,
+      "loss": 0.4063,
+      "step": 3881
+    },
+    {
+      "epoch": 0.7653785488958991,
+      "grad_norm": 0.6064570679491103,
+      "learning_rate": 1.826114016814509e-05,
+      "loss": 0.3918,
+      "step": 3882
+    },
+    {
+      "epoch": 0.7655757097791798,
+      "grad_norm": 0.5647896972824821,
+      "learning_rate": 1.8260266624618957e-05,
+      "loss": 0.4329,
+      "step": 3883
+    },
+    {
+      "epoch": 0.7657728706624606,
+      "grad_norm": 0.6946858259139369,
+      "learning_rate": 1.8259392882633266e-05,
+      "loss": 0.4704,
+      "step": 3884
+    },
+    {
+      "epoch": 0.7659700315457413,
+      "grad_norm": 0.5720833541883139,
+      "learning_rate": 1.825851894220901e-05,
+      "loss": 0.4309,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7661671924290221,
+      "grad_norm": 0.5967274277243448,
+      "learning_rate": 1.8257644803367186e-05,
+      "loss": 0.4341,
+      "step": 3886
+    },
+    {
+      "epoch": 0.7663643533123028,
+      "grad_norm": 0.5985333908745908,
+      "learning_rate": 1.8256770466128793e-05,
+      "loss": 0.4523,
+      "step": 3887
+    },
+    {
+      "epoch": 0.7665615141955836,
+      "grad_norm": 0.5560392931956263,
+      "learning_rate": 1.8255895930514843e-05,
+      "loss": 0.4228,
+      "step": 3888
+    },
+    {
+      "epoch": 0.7667586750788643,
+      "grad_norm": 0.5832994995378911,
+      "learning_rate": 1.8255021196546346e-05,
+      "loss": 0.4299,
+      "step": 3889
+    },
+    {
+      "epoch": 0.7669558359621451,
+      "grad_norm": 1.4002961334330222,
+      "learning_rate": 1.8254146264244316e-05,
+      "loss": 0.4941,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7671529968454258,
+      "grad_norm": 0.5534367003995347,
+      "learning_rate": 1.8253271133629775e-05,
+      "loss": 0.4185,
+      "step": 3891
+    },
+    {
+      "epoch": 0.7673501577287066,
+      "grad_norm": 0.5935042775545306,
+      "learning_rate": 1.8252395804723744e-05,
+      "loss": 0.4807,
+      "step": 3892
+    },
+    {
+      "epoch": 0.7675473186119873,
+      "grad_norm": 0.8877026134448658,
+      "learning_rate": 1.8251520277547267e-05,
+      "loss": 0.4268,
+      "step": 3893
+    },
+    {
+      "epoch": 0.7677444794952681,
+      "grad_norm": 1.2517353246924268,
+      "learning_rate": 1.8250644552121362e-05,
+      "loss": 0.4519,
+      "step": 3894
+    },
+    {
+      "epoch": 0.767941640378549,
+      "grad_norm": 0.5929121870930241,
+      "learning_rate": 1.8249768628467085e-05,
+      "loss": 0.4413,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7681388012618297,
+      "grad_norm": 0.5782475207334447,
+      "learning_rate": 1.8248892506605468e-05,
+      "loss": 0.4157,
+      "step": 3896
+    },
+    {
+      "epoch": 0.7683359621451105,
+      "grad_norm": 0.6117468432248259,
+      "learning_rate": 1.8248016186557566e-05,
+      "loss": 0.4333,
+      "step": 3897
+    },
+    {
+      "epoch": 0.7685331230283912,
+      "grad_norm": 0.5927409035058885,
+      "learning_rate": 1.8247139668344432e-05,
+      "loss": 0.4869,
+      "step": 3898
+    },
+    {
+      "epoch": 0.768730283911672,
+      "grad_norm": 1.158455927574009,
+      "learning_rate": 1.824626295198713e-05,
+      "loss": 0.4575,
+      "step": 3899
+    },
+    {
+      "epoch": 0.7689274447949527,
+      "grad_norm": 0.8392945391726799,
+      "learning_rate": 1.824538603750672e-05,
+      "loss": 0.4607,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7691246056782335,
+      "grad_norm": 0.5998059892020197,
+      "learning_rate": 1.824450892492427e-05,
+      "loss": 0.4594,
+      "step": 3901
+    },
+    {
+      "epoch": 0.7693217665615142,
+      "grad_norm": 0.5695316389310182,
+      "learning_rate": 1.824363161426085e-05,
+      "loss": 0.4175,
+      "step": 3902
+    },
+    {
+      "epoch": 0.769518927444795,
+      "grad_norm": 0.6240626330754476,
+      "learning_rate": 1.8242754105537542e-05,
+      "loss": 0.454,
+      "step": 3903
+    },
+    {
+      "epoch": 0.7697160883280757,
+      "grad_norm": 0.5912514018969862,
+      "learning_rate": 1.8241876398775434e-05,
+      "loss": 0.4311,
+      "step": 3904
+    },
+    {
+      "epoch": 0.7699132492113565,
+      "grad_norm": 0.6624090571377835,
+      "learning_rate": 1.8240998493995607e-05,
+      "loss": 0.4719,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7701104100946372,
+      "grad_norm": 0.9203541875889368,
+      "learning_rate": 1.8240120391219148e-05,
+      "loss": 0.4803,
+      "step": 3906
+    },
+    {
+      "epoch": 0.770307570977918,
+      "grad_norm": 0.5914744368657348,
+      "learning_rate": 1.823924209046717e-05,
+      "loss": 0.418,
+      "step": 3907
+    },
+    {
+      "epoch": 0.7705047318611987,
+      "grad_norm": 0.6070617757379179,
+      "learning_rate": 1.8238363591760758e-05,
+      "loss": 0.4429,
+      "step": 3908
+    },
+    {
+      "epoch": 0.7707018927444795,
+      "grad_norm": 0.6325247392458869,
+      "learning_rate": 1.8237484895121033e-05,
+      "loss": 0.4688,
+      "step": 3909
+    },
+    {
+      "epoch": 0.7708990536277602,
+      "grad_norm": 0.573852165719456,
+      "learning_rate": 1.8236606000569095e-05,
+      "loss": 0.4126,
+      "step": 3910
+    },
+    {
+      "epoch": 0.771096214511041,
+      "grad_norm": 0.5567521119492211,
+      "learning_rate": 1.823572690812607e-05,
+      "loss": 0.4415,
+      "step": 3911
+    },
+    {
+      "epoch": 0.7712933753943217,
+      "grad_norm": 0.6224892204399546,
+      "learning_rate": 1.8234847617813067e-05,
+      "loss": 0.4473,
+      "step": 3912
+    },
+    {
+      "epoch": 0.7714905362776026,
+      "grad_norm": 0.6241807022753024,
+      "learning_rate": 1.823396812965122e-05,
+      "loss": 0.4632,
+      "step": 3913
+    },
+    {
+      "epoch": 0.7716876971608833,
+      "grad_norm": 0.564677599383395,
+      "learning_rate": 1.8233088443661665e-05,
+      "loss": 0.4242,
+      "step": 3914
+    },
+    {
+      "epoch": 0.7718848580441641,
+      "grad_norm": 0.5882925220136831,
+      "learning_rate": 1.8232208559865522e-05,
+      "loss": 0.431,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7720820189274448,
+      "grad_norm": 0.574227496801443,
+      "learning_rate": 1.823132847828394e-05,
+      "loss": 0.451,
+      "step": 3916
+    },
+    {
+      "epoch": 0.7722791798107256,
+      "grad_norm": 0.5222564423269557,
+      "learning_rate": 1.8230448198938067e-05,
+      "loss": 0.4296,
+      "step": 3917
+    },
+    {
+      "epoch": 0.7724763406940063,
+      "grad_norm": 0.57083859662196,
+      "learning_rate": 1.8229567721849046e-05,
+      "loss": 0.4257,
+      "step": 3918
+    },
+    {
+      "epoch": 0.7726735015772871,
+      "grad_norm": 1.6592668807628246,
+      "learning_rate": 1.822868704703803e-05,
+      "loss": 0.4921,
+      "step": 3919
+    },
+    {
+      "epoch": 0.7728706624605678,
+      "grad_norm": 0.6157770095064646,
+      "learning_rate": 1.8227806174526187e-05,
+      "loss": 0.4314,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7730678233438486,
+      "grad_norm": 0.5852063956961934,
+      "learning_rate": 1.822692510433467e-05,
+      "loss": 0.4654,
+      "step": 3921
+    },
+    {
+      "epoch": 0.7732649842271293,
+      "grad_norm": 0.5866034997283565,
+      "learning_rate": 1.8226043836484655e-05,
+      "loss": 0.4346,
+      "step": 3922
+    },
+    {
+      "epoch": 0.7734621451104101,
+      "grad_norm": 0.5510812356437924,
+      "learning_rate": 1.8225162370997313e-05,
+      "loss": 0.4801,
+      "step": 3923
+    },
+    {
+      "epoch": 0.7736593059936908,
+      "grad_norm": 0.5472377435499002,
+      "learning_rate": 1.822428070789382e-05,
+      "loss": 0.4502,
+      "step": 3924
+    },
+    {
+      "epoch": 0.7738564668769716,
+      "grad_norm": 0.5620895652186938,
+      "learning_rate": 1.8223398847195358e-05,
+      "loss": 0.4428,
+      "step": 3925
+    },
+    {
+      "epoch": 0.7740536277602523,
+      "grad_norm": 0.5573421453852319,
+      "learning_rate": 1.822251678892312e-05,
+      "loss": 0.4397,
+      "step": 3926
+    },
+    {
+      "epoch": 0.7742507886435331,
+      "grad_norm": 0.5756487306350889,
+      "learning_rate": 1.822163453309829e-05,
+      "loss": 0.4255,
+      "step": 3927
+    },
+    {
+      "epoch": 0.7744479495268138,
+      "grad_norm": 0.5374818660181303,
+      "learning_rate": 1.8220752079742072e-05,
+      "loss": 0.4208,
+      "step": 3928
+    },
+    {
+      "epoch": 0.7746451104100947,
+      "grad_norm": 0.5294345547402046,
+      "learning_rate": 1.8219869428875668e-05,
+      "loss": 0.4217,
+      "step": 3929
+    },
+    {
+      "epoch": 0.7748422712933754,
+      "grad_norm": 0.56730405884325,
+      "learning_rate": 1.8218986580520276e-05,
+      "loss": 0.4768,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7750394321766562,
+      "grad_norm": 0.5604501496661701,
+      "learning_rate": 1.8218103534697116e-05,
+      "loss": 0.4596,
+      "step": 3931
+    },
+    {
+      "epoch": 0.7752365930599369,
+      "grad_norm": 0.5579910747422517,
+      "learning_rate": 1.8217220291427398e-05,
+      "loss": 0.4527,
+      "step": 3932
+    },
+    {
+      "epoch": 0.7754337539432177,
+      "grad_norm": 0.5270550522697447,
+      "learning_rate": 1.821633685073235e-05,
+      "loss": 0.4559,
+      "step": 3933
+    },
+    {
+      "epoch": 0.7756309148264984,
+      "grad_norm": 0.5923687114112185,
+      "learning_rate": 1.8215453212633188e-05,
+      "loss": 0.4548,
+      "step": 3934
+    },
+    {
+      "epoch": 0.7758280757097792,
+      "grad_norm": 0.6303703716242776,
+      "learning_rate": 1.821456937715115e-05,
+      "loss": 0.5182,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7760252365930599,
+      "grad_norm": 0.5753784721230842,
+      "learning_rate": 1.8213685344307465e-05,
+      "loss": 0.4475,
+      "step": 3936
+    },
+    {
+      "epoch": 0.7762223974763407,
+      "grad_norm": 0.7373579959797398,
+      "learning_rate": 1.8212801114123377e-05,
+      "loss": 0.4316,
+      "step": 3937
+    },
+    {
+      "epoch": 0.7764195583596214,
+      "grad_norm": 0.5762319703064297,
+      "learning_rate": 1.8211916686620128e-05,
+      "loss": 0.4546,
+      "step": 3938
+    },
+    {
+      "epoch": 0.7766167192429022,
+      "grad_norm": 0.5920585040131835,
+      "learning_rate": 1.8211032061818968e-05,
+      "loss": 0.4471,
+      "step": 3939
+    },
+    {
+      "epoch": 0.776813880126183,
+      "grad_norm": 0.6803659907474547,
+      "learning_rate": 1.8210147239741148e-05,
+      "loss": 0.4755,
+      "step": 3940
+    },
+    {
+      "epoch": 0.7770110410094637,
+      "grad_norm": 0.5509653813630704,
+      "learning_rate": 1.8209262220407932e-05,
+      "loss": 0.4586,
+      "step": 3941
+    },
+    {
+      "epoch": 0.7772082018927445,
+      "grad_norm": 0.5581868920790423,
+      "learning_rate": 1.820837700384058e-05,
+      "loss": 0.4219,
+      "step": 3942
+    },
+    {
+      "epoch": 0.7774053627760252,
+      "grad_norm": 0.5592867285417933,
+      "learning_rate": 1.8207491590060356e-05,
+      "loss": 0.4193,
+      "step": 3943
+    },
+    {
+      "epoch": 0.777602523659306,
+      "grad_norm": 0.6501883656371863,
+      "learning_rate": 1.8206605979088545e-05,
+      "loss": 0.4511,
+      "step": 3944
+    },
+    {
+      "epoch": 0.7777996845425867,
+      "grad_norm": 0.5572979380305522,
+      "learning_rate": 1.820572017094641e-05,
+      "loss": 0.4403,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7779968454258676,
+      "grad_norm": 0.6146644203648315,
+      "learning_rate": 1.8204834165655242e-05,
+      "loss": 0.4852,
+      "step": 3946
+    },
+    {
+      "epoch": 0.7781940063091483,
+      "grad_norm": 0.5413773415764146,
+      "learning_rate": 1.8203947963236322e-05,
+      "loss": 0.4437,
+      "step": 3947
+    },
+    {
+      "epoch": 0.7783911671924291,
+      "grad_norm": 0.5605314804048697,
+      "learning_rate": 1.8203061563710952e-05,
+      "loss": 0.4294,
+      "step": 3948
+    },
+    {
+      "epoch": 0.7785883280757098,
+      "grad_norm": 0.5316072340317646,
+      "learning_rate": 1.820217496710042e-05,
+      "loss": 0.4243,
+      "step": 3949
+    },
+    {
+      "epoch": 0.7787854889589906,
+      "grad_norm": 0.6492198435447817,
+      "learning_rate": 1.8201288173426027e-05,
+      "loss": 0.4659,
+      "step": 3950
+    },
+    {
+      "epoch": 0.7789826498422713,
+      "grad_norm": 0.5713530844391045,
+      "learning_rate": 1.820040118270908e-05,
+      "loss": 0.4579,
+      "step": 3951
+    },
+    {
+      "epoch": 0.7791798107255521,
+      "grad_norm": 0.5363420111437506,
+      "learning_rate": 1.8199513994970893e-05,
+      "loss": 0.425,
+      "step": 3952
+    },
+    {
+      "epoch": 0.7793769716088328,
+      "grad_norm": 0.5542539770275866,
+      "learning_rate": 1.819862661023278e-05,
+      "loss": 0.4234,
+      "step": 3953
+    },
+    {
+      "epoch": 0.7795741324921136,
+      "grad_norm": 0.5285872082909737,
+      "learning_rate": 1.8197739028516062e-05,
+      "loss": 0.4238,
+      "step": 3954
+    },
+    {
+      "epoch": 0.7797712933753943,
+      "grad_norm": 0.5456326663863709,
+      "learning_rate": 1.819685124984206e-05,
+      "loss": 0.4301,
+      "step": 3955
+    },
+    {
+      "epoch": 0.7799684542586751,
+      "grad_norm": 0.5664354597499723,
+      "learning_rate": 1.8195963274232106e-05,
+      "loss": 0.4555,
+      "step": 3956
+    },
+    {
+      "epoch": 0.7801656151419558,
+      "grad_norm": 0.5315491155568633,
+      "learning_rate": 1.8195075101707535e-05,
+      "loss": 0.4272,
+      "step": 3957
+    },
+    {
+      "epoch": 0.7803627760252366,
+      "grad_norm": 0.8663503651895043,
+      "learning_rate": 1.819418673228968e-05,
+      "loss": 0.4816,
+      "step": 3958
+    },
+    {
+      "epoch": 0.7805599369085173,
+      "grad_norm": 0.5526623269158064,
+      "learning_rate": 1.8193298165999896e-05,
+      "loss": 0.462,
+      "step": 3959
+    },
+    {
+      "epoch": 0.7807570977917981,
+      "grad_norm": 0.5973908735760364,
+      "learning_rate": 1.8192409402859526e-05,
+      "loss": 0.4452,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7809542586750788,
+      "grad_norm": 0.5521161874463199,
+      "learning_rate": 1.819152044288992e-05,
+      "loss": 0.4203,
+      "step": 3961
+    },
+    {
+      "epoch": 0.7811514195583596,
+      "grad_norm": 0.6040478645836748,
+      "learning_rate": 1.819063128611244e-05,
+      "loss": 0.4598,
+      "step": 3962
+    },
+    {
+      "epoch": 0.7813485804416404,
+      "grad_norm": 0.5454361035093315,
+      "learning_rate": 1.8189741932548447e-05,
+      "loss": 0.426,
+      "step": 3963
+    },
+    {
+      "epoch": 0.7815457413249212,
+      "grad_norm": 0.6321003848230068,
+      "learning_rate": 1.8188852382219308e-05,
+      "loss": 0.4685,
+      "step": 3964
+    },
+    {
+      "epoch": 0.7817429022082019,
+      "grad_norm": 0.6137933316061781,
+      "learning_rate": 1.8187962635146397e-05,
+      "loss": 0.4376,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7819400630914827,
+      "grad_norm": 0.5396185891653529,
+      "learning_rate": 1.8187072691351088e-05,
+      "loss": 0.421,
+      "step": 3966
+    },
+    {
+      "epoch": 0.7821372239747634,
+      "grad_norm": 0.6005283575463105,
+      "learning_rate": 1.8186182550854768e-05,
+      "loss": 0.4782,
+      "step": 3967
+    },
+    {
+      "epoch": 0.7823343848580442,
+      "grad_norm": 0.516827483855773,
+      "learning_rate": 1.8185292213678818e-05,
+      "loss": 0.4034,
+      "step": 3968
+    },
+    {
+      "epoch": 0.7825315457413249,
+      "grad_norm": 0.5753697155306382,
+      "learning_rate": 1.818440167984463e-05,
+      "loss": 0.4585,
+      "step": 3969
+    },
+    {
+      "epoch": 0.7827287066246057,
+      "grad_norm": 0.5542615113432319,
+      "learning_rate": 1.8183510949373603e-05,
+      "loss": 0.4504,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7829258675078864,
+      "grad_norm": 0.5333879318693766,
+      "learning_rate": 1.8182620022287133e-05,
+      "loss": 0.4213,
+      "step": 3971
+    },
+    {
+      "epoch": 0.7831230283911672,
+      "grad_norm": 0.6072385415952638,
+      "learning_rate": 1.8181728898606628e-05,
+      "loss": 0.4395,
+      "step": 3972
+    },
+    {
+      "epoch": 0.7833201892744479,
+      "grad_norm": 0.543662857256321,
+      "learning_rate": 1.81808375783535e-05,
+      "loss": 0.4385,
+      "step": 3973
+    },
+    {
+      "epoch": 0.7835173501577287,
+      "grad_norm": 5.542609848424303,
+      "learning_rate": 1.817994606154916e-05,
+      "loss": 0.5795,
+      "step": 3974
+    },
+    {
+      "epoch": 0.7837145110410094,
+      "grad_norm": 0.5948140810991429,
+      "learning_rate": 1.8179054348215025e-05,
+      "loss": 0.4474,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7839116719242902,
+      "grad_norm": 0.519423366632737,
+      "learning_rate": 1.8178162438372528e-05,
+      "loss": 0.4145,
+      "step": 3976
+    },
+    {
+      "epoch": 0.7841088328075709,
+      "grad_norm": 0.5902188567847949,
+      "learning_rate": 1.817727033204309e-05,
+      "loss": 0.4315,
+      "step": 3977
+    },
+    {
+      "epoch": 0.7843059936908517,
+      "grad_norm": 0.5948977039074232,
+      "learning_rate": 1.8176378029248147e-05,
+      "loss": 0.458,
+      "step": 3978
+    },
+    {
+      "epoch": 0.7845031545741324,
+      "grad_norm": 0.5611639183104089,
+      "learning_rate": 1.8175485530009137e-05,
+      "loss": 0.4531,
+      "step": 3979
+    },
+    {
+      "epoch": 0.7847003154574133,
+      "grad_norm": 1.0876807024291277,
+      "learning_rate": 1.8174592834347503e-05,
+      "loss": 0.4904,
+      "step": 3980
+    },
+    {
+      "epoch": 0.784897476340694,
+      "grad_norm": 0.5358410389890251,
+      "learning_rate": 1.8173699942284695e-05,
+      "loss": 0.4228,
+      "step": 3981
+    },
+    {
+      "epoch": 0.7850946372239748,
+      "grad_norm": 0.5943472824942584,
+      "learning_rate": 1.8172806853842163e-05,
+      "loss": 0.4777,
+      "step": 3982
+    },
+    {
+      "epoch": 0.7852917981072555,
+      "grad_norm": 0.5829558673091376,
+      "learning_rate": 1.8171913569041362e-05,
+      "loss": 0.4223,
+      "step": 3983
+    },
+    {
+      "epoch": 0.7854889589905363,
+      "grad_norm": 0.5306668965862121,
+      "learning_rate": 1.8171020087903762e-05,
+      "loss": 0.4089,
+      "step": 3984
+    },
+    {
+      "epoch": 0.785686119873817,
+      "grad_norm": 0.5451241305750418,
+      "learning_rate": 1.8170126410450823e-05,
+      "loss": 0.4361,
+      "step": 3985
+    },
+    {
+      "epoch": 0.7858832807570978,
+      "grad_norm": 0.5469353788262228,
+      "learning_rate": 1.8169232536704012e-05,
+      "loss": 0.4297,
+      "step": 3986
+    },
+    {
+      "epoch": 0.7860804416403786,
+      "grad_norm": 0.7105261028282002,
+      "learning_rate": 1.8168338466684817e-05,
+      "loss": 0.4311,
+      "step": 3987
+    },
+    {
+      "epoch": 0.7862776025236593,
+      "grad_norm": 0.5989580178979008,
+      "learning_rate": 1.816744420041471e-05,
+      "loss": 0.468,
+      "step": 3988
+    },
+    {
+      "epoch": 0.7864747634069401,
+      "grad_norm": 0.545707298003204,
+      "learning_rate": 1.816654973791518e-05,
+      "loss": 0.4527,
+      "step": 3989
+    },
+    {
+      "epoch": 0.7866719242902208,
+      "grad_norm": 0.6273199078420786,
+      "learning_rate": 1.8165655079207716e-05,
+      "loss": 0.449,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7868690851735016,
+      "grad_norm": 0.5611837255166214,
+      "learning_rate": 1.816476022431381e-05,
+      "loss": 0.4656,
+      "step": 3991
+    },
+    {
+      "epoch": 0.7870662460567823,
+      "grad_norm": 0.5518444727001759,
+      "learning_rate": 1.816386517325497e-05,
+      "loss": 0.4284,
+      "step": 3992
+    },
+    {
+      "epoch": 0.7872634069400631,
+      "grad_norm": 0.5892275816171331,
+      "learning_rate": 1.816296992605269e-05,
+      "loss": 0.4539,
+      "step": 3993
+    },
+    {
+      "epoch": 0.7874605678233438,
+      "grad_norm": 0.5254195560267715,
+      "learning_rate": 1.8162074482728487e-05,
+      "loss": 0.4565,
+      "step": 3994
+    },
+    {
+      "epoch": 0.7876577287066246,
+      "grad_norm": 0.5448949680828074,
+      "learning_rate": 1.816117884330387e-05,
+      "loss": 0.4178,
+      "step": 3995
+    },
+    {
+      "epoch": 0.7878548895899053,
+      "grad_norm": 0.5746576844950827,
+      "learning_rate": 1.816028300780036e-05,
+      "loss": 0.4735,
+      "step": 3996
+    },
+    {
+      "epoch": 0.7880520504731862,
+      "grad_norm": 0.537017218341343,
+      "learning_rate": 1.8159386976239478e-05,
+      "loss": 0.4047,
+      "step": 3997
+    },
+    {
+      "epoch": 0.7882492113564669,
+      "grad_norm": 0.6354220762186988,
+      "learning_rate": 1.815849074864275e-05,
+      "loss": 0.4155,
+      "step": 3998
+    },
+    {
+      "epoch": 0.7884463722397477,
+      "grad_norm": 0.578467197061079,
+      "learning_rate": 1.8157594325031716e-05,
+      "loss": 0.4183,
+      "step": 3999
+    },
+    {
+      "epoch": 0.7886435331230284,
+      "grad_norm": 0.6475745820310453,
+      "learning_rate": 1.8156697705427907e-05,
+      "loss": 0.4209,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7888406940063092,
+      "grad_norm": 0.6157428470516835,
+      "learning_rate": 1.815580088985287e-05,
+      "loss": 0.4531,
+      "step": 4001
+    },
+    {
+      "epoch": 0.7890378548895899,
+      "grad_norm": 0.5998799236699729,
+      "learning_rate": 1.815490387832815e-05,
+      "loss": 0.4751,
+      "step": 4002
+    },
+    {
+      "epoch": 0.7892350157728707,
+      "grad_norm": 0.5551797172113444,
+      "learning_rate": 1.8154006670875294e-05,
+      "loss": 0.444,
+      "step": 4003
+    },
+    {
+      "epoch": 0.7894321766561514,
+      "grad_norm": 0.7557929675891345,
+      "learning_rate": 1.815310926751586e-05,
+      "loss": 0.433,
+      "step": 4004
+    },
+    {
+      "epoch": 0.7896293375394322,
+      "grad_norm": 3.8174090731821937,
+      "learning_rate": 1.8152211668271413e-05,
+      "loss": 0.428,
+      "step": 4005
+    },
+    {
+      "epoch": 0.7898264984227129,
+      "grad_norm": 0.7034780817377414,
+      "learning_rate": 1.8151313873163513e-05,
+      "loss": 0.417,
+      "step": 4006
+    },
+    {
+      "epoch": 0.7900236593059937,
+      "grad_norm": 0.5926550848858589,
+      "learning_rate": 1.8150415882213735e-05,
+      "loss": 0.4756,
+      "step": 4007
+    },
+    {
+      "epoch": 0.7902208201892744,
+      "grad_norm": 0.6888172885215963,
+      "learning_rate": 1.814951769544365e-05,
+      "loss": 0.4438,
+      "step": 4008
+    },
+    {
+      "epoch": 0.7904179810725552,
+      "grad_norm": 0.5290234268356881,
+      "learning_rate": 1.8148619312874844e-05,
+      "loss": 0.4186,
+      "step": 4009
+    },
+    {
+      "epoch": 0.7906151419558359,
+      "grad_norm": 0.7662700497693185,
+      "learning_rate": 1.8147720734528893e-05,
+      "loss": 0.4871,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7908123028391167,
+      "grad_norm": 0.5957613132464488,
+      "learning_rate": 1.814682196042739e-05,
+      "loss": 0.4572,
+      "step": 4011
+    },
+    {
+      "epoch": 0.7910094637223974,
+      "grad_norm": 0.6866807817706564,
+      "learning_rate": 1.8145922990591932e-05,
+      "loss": 0.4296,
+      "step": 4012
+    },
+    {
+      "epoch": 0.7912066246056783,
+      "grad_norm": 0.5397278268058946,
+      "learning_rate": 1.8145023825044114e-05,
+      "loss": 0.4247,
+      "step": 4013
+    },
+    {
+      "epoch": 0.791403785488959,
+      "grad_norm": 0.7343980490854036,
+      "learning_rate": 1.8144124463805535e-05,
+      "loss": 0.4587,
+      "step": 4014
+    },
+    {
+      "epoch": 0.7916009463722398,
+      "grad_norm": 0.5376919265100979,
+      "learning_rate": 1.8143224906897812e-05,
+      "loss": 0.4111,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7917981072555205,
+      "grad_norm": 0.633488416178425,
+      "learning_rate": 1.814232515434255e-05,
+      "loss": 0.4592,
+      "step": 4016
+    },
+    {
+      "epoch": 0.7919952681388013,
+      "grad_norm": 0.5283179674486278,
+      "learning_rate": 1.814142520616137e-05,
+      "loss": 0.4288,
+      "step": 4017
+    },
+    {
+      "epoch": 0.792192429022082,
+      "grad_norm": 0.5780943767126454,
+      "learning_rate": 1.8140525062375894e-05,
+      "loss": 0.4239,
+      "step": 4018
+    },
+    {
+      "epoch": 0.7923895899053628,
+      "grad_norm": 0.5489901267012213,
+      "learning_rate": 1.8139624723007748e-05,
+      "loss": 0.423,
+      "step": 4019
+    },
+    {
+      "epoch": 0.7925867507886435,
+      "grad_norm": 0.5878569214560349,
+      "learning_rate": 1.813872418807856e-05,
+      "loss": 0.4265,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7927839116719243,
+      "grad_norm": 0.5737545069149199,
+      "learning_rate": 1.813782345760997e-05,
+      "loss": 0.4491,
+      "step": 4021
+    },
+    {
+      "epoch": 0.792981072555205,
+      "grad_norm": 0.823990029369412,
+      "learning_rate": 1.813692253162362e-05,
+      "loss": 0.4082,
+      "step": 4022
+    },
+    {
+      "epoch": 0.7931782334384858,
+      "grad_norm": 0.5299434682153719,
+      "learning_rate": 1.8136021410141154e-05,
+      "loss": 0.3884,
+      "step": 4023
+    },
+    {
+      "epoch": 0.7933753943217665,
+      "grad_norm": 0.551899978074941,
+      "learning_rate": 1.813512009318422e-05,
+      "loss": 0.4369,
+      "step": 4024
+    },
+    {
+      "epoch": 0.7935725552050473,
+      "grad_norm": 8.176889365950668,
+      "learning_rate": 1.8134218580774475e-05,
+      "loss": 0.4485,
+      "step": 4025
+    },
+    {
+      "epoch": 0.793769716088328,
+      "grad_norm": 0.6125163347743271,
+      "learning_rate": 1.813331687293358e-05,
+      "loss": 0.4474,
+      "step": 4026
+    },
+    {
+      "epoch": 0.7939668769716088,
+      "grad_norm": 0.5484917124775043,
+      "learning_rate": 1.8132414969683197e-05,
+      "loss": 0.4205,
+      "step": 4027
+    },
+    {
+      "epoch": 0.7941640378548895,
+      "grad_norm": 0.557875142995521,
+      "learning_rate": 1.8131512871044993e-05,
+      "loss": 0.4225,
+      "step": 4028
+    },
+    {
+      "epoch": 0.7943611987381703,
+      "grad_norm": 0.637672714031013,
+      "learning_rate": 1.8130610577040646e-05,
+      "loss": 0.483,
+      "step": 4029
+    },
+    {
+      "epoch": 0.794558359621451,
+      "grad_norm": 0.5142583048499065,
+      "learning_rate": 1.812970808769183e-05,
+      "loss": 0.42,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7947555205047319,
+      "grad_norm": 0.5565242219925195,
+      "learning_rate": 1.812880540302023e-05,
+      "loss": 0.4521,
+      "step": 4031
+    },
+    {
+      "epoch": 0.7949526813880127,
+      "grad_norm": 0.5468093785090244,
+      "learning_rate": 1.812790252304754e-05,
+      "loss": 0.4285,
+      "step": 4032
+    },
+    {
+      "epoch": 0.7951498422712934,
+      "grad_norm": 0.5792091705278589,
+      "learning_rate": 1.8126999447795438e-05,
+      "loss": 0.4666,
+      "step": 4033
+    },
+    {
+      "epoch": 0.7953470031545742,
+      "grad_norm": 4.072216890062226,
+      "learning_rate": 1.8126096177285637e-05,
+      "loss": 0.5029,
+      "step": 4034
+    },
+    {
+      "epoch": 0.7955441640378549,
+      "grad_norm": 0.6296907138779285,
+      "learning_rate": 1.8125192711539828e-05,
+      "loss": 0.4361,
+      "step": 4035
+    },
+    {
+      "epoch": 0.7957413249211357,
+      "grad_norm": 0.5788781690916837,
+      "learning_rate": 1.812428905057972e-05,
+      "loss": 0.4399,
+      "step": 4036
+    },
+    {
+      "epoch": 0.7959384858044164,
+      "grad_norm": 0.5884757591246823,
+      "learning_rate": 1.8123385194427027e-05,
+      "loss": 0.4888,
+      "step": 4037
+    },
+    {
+      "epoch": 0.7961356466876972,
+      "grad_norm": 0.6261353284400933,
+      "learning_rate": 1.8122481143103465e-05,
+      "loss": 0.4474,
+      "step": 4038
+    },
+    {
+      "epoch": 0.7963328075709779,
+      "grad_norm": 0.5682139648963489,
+      "learning_rate": 1.812157689663075e-05,
+      "loss": 0.4466,
+      "step": 4039
+    },
+    {
+      "epoch": 0.7965299684542587,
+      "grad_norm": 0.6006306057576911,
+      "learning_rate": 1.8120672455030606e-05,
+      "loss": 0.4287,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7967271293375394,
+      "grad_norm": 0.6096153492041305,
+      "learning_rate": 1.8119767818324773e-05,
+      "loss": 0.4854,
+      "step": 4041
+    },
+    {
+      "epoch": 0.7969242902208202,
+      "grad_norm": 0.5816536621919691,
+      "learning_rate": 1.8118862986534974e-05,
+      "loss": 0.4314,
+      "step": 4042
+    },
+    {
+      "epoch": 0.7971214511041009,
+      "grad_norm": 0.5915477679203885,
+      "learning_rate": 1.811795795968296e-05,
+      "loss": 0.4392,
+      "step": 4043
+    },
+    {
+      "epoch": 0.7973186119873817,
+      "grad_norm": 0.5550175414784126,
+      "learning_rate": 1.8117052737790463e-05,
+      "loss": 0.3972,
+      "step": 4044
+    },
+    {
+      "epoch": 0.7975157728706624,
+      "grad_norm": 0.5778936646533668,
+      "learning_rate": 1.8116147320879238e-05,
+      "loss": 0.4382,
+      "step": 4045
+    },
+    {
+      "epoch": 0.7977129337539433,
+      "grad_norm": 0.6157010473714177,
+      "learning_rate": 1.811524170897104e-05,
+      "loss": 0.4333,
+      "step": 4046
+    },
+    {
+      "epoch": 0.797910094637224,
+      "grad_norm": 0.5637162571951514,
+      "learning_rate": 1.8114335902087625e-05,
+      "loss": 0.4542,
+      "step": 4047
+    },
+    {
+      "epoch": 0.7981072555205048,
+      "grad_norm": 0.584581727444574,
+      "learning_rate": 1.811342990025075e-05,
+      "loss": 0.454,
+      "step": 4048
+    },
+    {
+      "epoch": 0.7983044164037855,
+      "grad_norm": 0.5543640727335405,
+      "learning_rate": 1.8112523703482194e-05,
+      "loss": 0.4145,
+      "step": 4049
+    },
+    {
+      "epoch": 0.7985015772870663,
+      "grad_norm": 0.6099912064912423,
+      "learning_rate": 1.8111617311803722e-05,
+      "loss": 0.4478,
+      "step": 4050
+    },
+    {
+      "epoch": 0.798698738170347,
+      "grad_norm": 0.5700094595349666,
+      "learning_rate": 1.8110710725237114e-05,
+      "loss": 0.4459,
+      "step": 4051
+    },
+    {
+      "epoch": 0.7988958990536278,
+      "grad_norm": 0.5495028696341402,
+      "learning_rate": 1.8109803943804146e-05,
+      "loss": 0.424,
+      "step": 4052
+    },
+    {
+      "epoch": 0.7990930599369085,
+      "grad_norm": 0.5306200168382031,
+      "learning_rate": 1.8108896967526607e-05,
+      "loss": 0.4018,
+      "step": 4053
+    },
+    {
+      "epoch": 0.7992902208201893,
+      "grad_norm": 1.038998012553428,
+      "learning_rate": 1.810798979642629e-05,
+      "loss": 0.4464,
+      "step": 4054
+    },
+    {
+      "epoch": 0.79948738170347,
+      "grad_norm": 1.0234857955785743,
+      "learning_rate": 1.8107082430524986e-05,
+      "loss": 0.4581,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7996845425867508,
+      "grad_norm": 0.5159790767288523,
+      "learning_rate": 1.81061748698445e-05,
+      "loss": 0.433,
+      "step": 4056
+    },
+    {
+      "epoch": 0.7998817034700315,
+      "grad_norm": 1.0841772174572797,
+      "learning_rate": 1.8105267114406633e-05,
+      "loss": 0.4212,
+      "step": 4057
+    },
+    {
+      "epoch": 0.8000788643533123,
+      "grad_norm": 0.5786065173723118,
+      "learning_rate": 1.81043591642332e-05,
+      "loss": 0.4607,
+      "step": 4058
+    },
+    {
+      "epoch": 0.800276025236593,
+      "grad_norm": 0.68262028393148,
+      "learning_rate": 1.810345101934601e-05,
+      "loss": 0.41,
+      "step": 4059
+    },
+    {
+      "epoch": 0.8004731861198738,
+      "grad_norm": 1.292583207003667,
+      "learning_rate": 1.8102542679766884e-05,
+      "loss": 0.4715,
+      "step": 4060
+    },
+    {
+      "epoch": 0.8006703470031545,
+      "grad_norm": 0.5803586064757899,
+      "learning_rate": 1.8101634145517644e-05,
+      "loss": 0.4231,
+      "step": 4061
+    },
+    {
+      "epoch": 0.8008675078864353,
+      "grad_norm": 0.6051334610135585,
+      "learning_rate": 1.810072541662012e-05,
+      "loss": 0.5071,
+      "step": 4062
+    },
+    {
+      "epoch": 0.801064668769716,
+      "grad_norm": 0.6153429571062644,
+      "learning_rate": 1.8099816493096144e-05,
+      "loss": 0.4541,
+      "step": 4063
+    },
+    {
+      "epoch": 0.8012618296529969,
+      "grad_norm": 0.6245215575502755,
+      "learning_rate": 1.8098907374967557e-05,
+      "loss": 0.4805,
+      "step": 4064
+    },
+    {
+      "epoch": 0.8014589905362776,
+      "grad_norm": 1.371389249743891,
+      "learning_rate": 1.8097998062256193e-05,
+      "loss": 0.4497,
+      "step": 4065
+    },
+    {
+      "epoch": 0.8016561514195584,
+      "grad_norm": 0.5974203617721795,
+      "learning_rate": 1.8097088554983906e-05,
+      "loss": 0.4546,
+      "step": 4066
+    },
+    {
+      "epoch": 0.8018533123028391,
+      "grad_norm": 0.8325017489031189,
+      "learning_rate": 1.8096178853172548e-05,
+      "loss": 0.4492,
+      "step": 4067
+    },
+    {
+      "epoch": 0.8020504731861199,
+      "grad_norm": 0.5910351760004731,
+      "learning_rate": 1.809526895684397e-05,
+      "loss": 0.4848,
+      "step": 4068
+    },
+    {
+      "epoch": 0.8022476340694006,
+      "grad_norm": 0.5900803892149564,
+      "learning_rate": 1.809435886602004e-05,
+      "loss": 0.4601,
+      "step": 4069
+    },
+    {
+      "epoch": 0.8024447949526814,
+      "grad_norm": 0.8379969783401379,
+      "learning_rate": 1.8093448580722617e-05,
+      "loss": 0.4094,
+      "step": 4070
+    },
+    {
+      "epoch": 0.8026419558359621,
+      "grad_norm": 0.6961897937500606,
+      "learning_rate": 1.809253810097358e-05,
+      "loss": 0.4621,
+      "step": 4071
+    },
+    {
+      "epoch": 0.8028391167192429,
+      "grad_norm": 0.5613024287580417,
+      "learning_rate": 1.809162742679479e-05,
+      "loss": 0.4727,
+      "step": 4072
+    },
+    {
+      "epoch": 0.8030362776025236,
+      "grad_norm": 0.5597933707348918,
+      "learning_rate": 1.8090716558208136e-05,
+      "loss": 0.4608,
+      "step": 4073
+    },
+    {
+      "epoch": 0.8032334384858044,
+      "grad_norm": 0.5506039857706837,
+      "learning_rate": 1.8089805495235507e-05,
+      "loss": 0.4177,
+      "step": 4074
+    },
+    {
+      "epoch": 0.8034305993690851,
+      "grad_norm": 0.6412143999415043,
+      "learning_rate": 1.808889423789878e-05,
+      "loss": 0.45,
+      "step": 4075
+    },
+    {
+      "epoch": 0.8036277602523659,
+      "grad_norm": 0.5355910563904579,
+      "learning_rate": 1.808798278621986e-05,
+      "loss": 0.4033,
+      "step": 4076
+    },
+    {
+      "epoch": 0.8038249211356467,
+      "grad_norm": 1.393193920661567,
+      "learning_rate": 1.808707114022064e-05,
+      "loss": 0.4729,
+      "step": 4077
+    },
+    {
+      "epoch": 0.8040220820189274,
+      "grad_norm": 0.5347566428548295,
+      "learning_rate": 1.808615929992302e-05,
+      "loss": 0.4392,
+      "step": 4078
+    },
+    {
+      "epoch": 0.8042192429022083,
+      "grad_norm": 0.5701756563883921,
+      "learning_rate": 1.8085247265348913e-05,
+      "loss": 0.4447,
+      "step": 4079
+    },
+    {
+      "epoch": 0.804416403785489,
+      "grad_norm": 0.5336017684518319,
+      "learning_rate": 1.808433503652023e-05,
+      "loss": 0.4357,
+      "step": 4080
+    },
+    {
+      "epoch": 0.8046135646687698,
+      "grad_norm": 0.5128271293806619,
+      "learning_rate": 1.8083422613458886e-05,
+      "loss": 0.4149,
+      "step": 4081
+    },
+    {
+      "epoch": 0.8048107255520505,
+      "grad_norm": 0.5813112643112214,
+      "learning_rate": 1.8082509996186802e-05,
+      "loss": 0.475,
+      "step": 4082
+    },
+    {
+      "epoch": 0.8050078864353313,
+      "grad_norm": 0.5645562044886906,
+      "learning_rate": 1.808159718472591e-05,
+      "loss": 0.449,
+      "step": 4083
+    },
+    {
+      "epoch": 0.805205047318612,
+      "grad_norm": 0.5209108203099843,
+      "learning_rate": 1.8080684179098135e-05,
+      "loss": 0.4282,
+      "step": 4084
+    },
+    {
+      "epoch": 0.8054022082018928,
+      "grad_norm": 0.5442950978703488,
+      "learning_rate": 1.807977097932542e-05,
+      "loss": 0.4519,
+      "step": 4085
+    },
+    {
+      "epoch": 0.8055993690851735,
+      "grad_norm": 0.5397055857931383,
+      "learning_rate": 1.8078857585429698e-05,
+      "loss": 0.4182,
+      "step": 4086
+    },
+    {
+      "epoch": 0.8057965299684543,
+      "grad_norm": 0.5457157266372396,
+      "learning_rate": 1.8077943997432913e-05,
+      "loss": 0.4589,
+      "step": 4087
+    },
+    {
+      "epoch": 0.805993690851735,
+      "grad_norm": 0.5415377375158944,
+      "learning_rate": 1.8077030215357024e-05,
+      "loss": 0.4621,
+      "step": 4088
+    },
+    {
+      "epoch": 0.8061908517350158,
+      "grad_norm": 0.5301693008219718,
+      "learning_rate": 1.8076116239223976e-05,
+      "loss": 0.4085,
+      "step": 4089
+    },
+    {
+      "epoch": 0.8063880126182965,
+      "grad_norm": 0.5719849084421536,
+      "learning_rate": 1.807520206905573e-05,
+      "loss": 0.4478,
+      "step": 4090
+    },
+    {
+      "epoch": 0.8065851735015773,
+      "grad_norm": 0.5068601560398842,
+      "learning_rate": 1.8074287704874258e-05,
+      "loss": 0.4119,
+      "step": 4091
+    },
+    {
+      "epoch": 0.806782334384858,
+      "grad_norm": 0.5667924640136568,
+      "learning_rate": 1.8073373146701517e-05,
+      "loss": 0.4064,
+      "step": 4092
+    },
+    {
+      "epoch": 0.8069794952681388,
+      "grad_norm": 0.5134527641968423,
+      "learning_rate": 1.8072458394559485e-05,
+      "loss": 0.3986,
+      "step": 4093
+    },
+    {
+      "epoch": 0.8071766561514195,
+      "grad_norm": 0.6144715703002955,
+      "learning_rate": 1.807154344847014e-05,
+      "loss": 0.4866,
+      "step": 4094
+    },
+    {
+      "epoch": 0.8073738170347003,
+      "grad_norm": 0.6012333197786712,
+      "learning_rate": 1.8070628308455463e-05,
+      "loss": 0.4539,
+      "step": 4095
+    },
+    {
+      "epoch": 0.807570977917981,
+      "grad_norm": 0.5267747372001423,
+      "learning_rate": 1.8069712974537444e-05,
+      "loss": 0.4468,
+      "step": 4096
+    },
+    {
+      "epoch": 0.8077681388012619,
+      "grad_norm": 0.5566674138070858,
+      "learning_rate": 1.8068797446738072e-05,
+      "loss": 0.4231,
+      "step": 4097
+    },
+    {
+      "epoch": 0.8079652996845426,
+      "grad_norm": 0.5254269426936413,
+      "learning_rate": 1.806788172507934e-05,
+      "loss": 0.4015,
+      "step": 4098
+    },
+    {
+      "epoch": 0.8081624605678234,
+      "grad_norm": 0.5478509815630053,
+      "learning_rate": 1.8066965809583255e-05,
+      "loss": 0.4472,
+      "step": 4099
+    },
+    {
+      "epoch": 0.8083596214511041,
+      "grad_norm": 0.5715957253192107,
+      "learning_rate": 1.8066049700271818e-05,
+      "loss": 0.4465,
+      "step": 4100
+    },
+    {
+      "epoch": 0.8085567823343849,
+      "grad_norm": 0.5584176748983026,
+      "learning_rate": 1.8065133397167045e-05,
+      "loss": 0.4425,
+      "step": 4101
+    },
+    {
+      "epoch": 0.8087539432176656,
+      "grad_norm": 0.5738027949569142,
+      "learning_rate": 1.8064216900290943e-05,
+      "loss": 0.4376,
+      "step": 4102
+    },
+    {
+      "epoch": 0.8089511041009464,
+      "grad_norm": 0.5582915978440195,
+      "learning_rate": 1.806330020966554e-05,
+      "loss": 0.4309,
+      "step": 4103
+    },
+    {
+      "epoch": 0.8091482649842271,
+      "grad_norm": 0.6063188844006441,
+      "learning_rate": 1.8062383325312855e-05,
+      "loss": 0.4636,
+      "step": 4104
+    },
+    {
+      "epoch": 0.8093454258675079,
+      "grad_norm": 0.5657855330946964,
+      "learning_rate": 1.8061466247254914e-05,
+      "loss": 0.4632,
+      "step": 4105
+    },
+    {
+      "epoch": 0.8095425867507886,
+      "grad_norm": 0.5670328332375311,
+      "learning_rate": 1.806054897551376e-05,
+      "loss": 0.439,
+      "step": 4106
+    },
+    {
+      "epoch": 0.8097397476340694,
+      "grad_norm": 0.6265658420538915,
+      "learning_rate": 1.8059631510111424e-05,
+      "loss": 0.4424,
+      "step": 4107
+    },
+    {
+      "epoch": 0.8099369085173501,
+      "grad_norm": 0.7579410291352082,
+      "learning_rate": 1.805871385106995e-05,
+      "loss": 0.4335,
+      "step": 4108
+    },
+    {
+      "epoch": 0.8101340694006309,
+      "grad_norm": 0.5063248042221858,
+      "learning_rate": 1.8057795998411384e-05,
+      "loss": 0.4214,
+      "step": 4109
+    },
+    {
+      "epoch": 0.8103312302839116,
+      "grad_norm": 0.5236497593759417,
+      "learning_rate": 1.8056877952157786e-05,
+      "loss": 0.43,
+      "step": 4110
+    },
+    {
+      "epoch": 0.8105283911671924,
+      "grad_norm": 0.6034930272092489,
+      "learning_rate": 1.80559597123312e-05,
+      "loss": 0.4211,
+      "step": 4111
+    },
+    {
+      "epoch": 0.8107255520504731,
+      "grad_norm": 0.5882438633902384,
+      "learning_rate": 1.80550412789537e-05,
+      "loss": 0.443,
+      "step": 4112
+    },
+    {
+      "epoch": 0.810922712933754,
+      "grad_norm": 0.5366584995675621,
+      "learning_rate": 1.8054122652047342e-05,
+      "loss": 0.4112,
+      "step": 4113
+    },
+    {
+      "epoch": 0.8111198738170347,
+      "grad_norm": 0.6250197267195037,
+      "learning_rate": 1.8053203831634207e-05,
+      "loss": 0.4361,
+      "step": 4114
+    },
+    {
+      "epoch": 0.8113170347003155,
+      "grad_norm": 0.5654169621581985,
+      "learning_rate": 1.805228481773636e-05,
+      "loss": 0.4699,
+      "step": 4115
+    },
+    {
+      "epoch": 0.8115141955835962,
+      "grad_norm": 0.5753974888094099,
+      "learning_rate": 1.8051365610375884e-05,
+      "loss": 0.4418,
+      "step": 4116
+    },
+    {
+      "epoch": 0.811711356466877,
+      "grad_norm": 0.5264404059073737,
+      "learning_rate": 1.8050446209574872e-05,
+      "loss": 0.4286,
+      "step": 4117
+    },
+    {
+      "epoch": 0.8119085173501577,
+      "grad_norm": 0.6190117803945342,
+      "learning_rate": 1.8049526615355404e-05,
+      "loss": 0.4704,
+      "step": 4118
+    },
+    {
+      "epoch": 0.8121056782334385,
+      "grad_norm": 0.5747549365566348,
+      "learning_rate": 1.8048606827739578e-05,
+      "loss": 0.4404,
+      "step": 4119
+    },
+    {
+      "epoch": 0.8123028391167192,
+      "grad_norm": 0.6355096601857466,
+      "learning_rate": 1.8047686846749488e-05,
+      "loss": 0.453,
+      "step": 4120
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 0.5199655678601448,
+      "learning_rate": 1.8046766672407244e-05,
+      "loss": 0.4341,
+      "step": 4121
+    },
+    {
+      "epoch": 0.8126971608832808,
+      "grad_norm": 0.6314812726234823,
+      "learning_rate": 1.8045846304734948e-05,
+      "loss": 0.4633,
+      "step": 4122
+    },
+    {
+      "epoch": 0.8128943217665615,
+      "grad_norm": 0.5432564307359536,
+      "learning_rate": 1.8044925743754717e-05,
+      "loss": 0.4788,
+      "step": 4123
+    },
+    {
+      "epoch": 0.8130914826498423,
+      "grad_norm": 0.7570361059774975,
+      "learning_rate": 1.8044004989488662e-05,
+      "loss": 0.4695,
+      "step": 4124
+    },
+    {
+      "epoch": 0.813288643533123,
+      "grad_norm": 0.5383797237758886,
+      "learning_rate": 1.8043084041958915e-05,
+      "loss": 0.4453,
+      "step": 4125
+    },
+    {
+      "epoch": 0.8134858044164038,
+      "grad_norm": 0.647994374550417,
+      "learning_rate": 1.8042162901187596e-05,
+      "loss": 0.4632,
+      "step": 4126
+    },
+    {
+      "epoch": 0.8136829652996845,
+      "grad_norm": 0.5849265654509385,
+      "learning_rate": 1.8041241567196834e-05,
+      "loss": 0.4533,
+      "step": 4127
+    },
+    {
+      "epoch": 0.8138801261829653,
+      "grad_norm": 0.5450500959334965,
+      "learning_rate": 1.804032004000877e-05,
+      "loss": 0.4477,
+      "step": 4128
+    },
+    {
+      "epoch": 0.814077287066246,
+      "grad_norm": 0.5608733688063781,
+      "learning_rate": 1.803939831964554e-05,
+      "loss": 0.4586,
+      "step": 4129
+    },
+    {
+      "epoch": 0.8142744479495269,
+      "grad_norm": 0.4822206632784062,
+      "learning_rate": 1.8038476406129294e-05,
+      "loss": 0.4067,
+      "step": 4130
+    },
+    {
+      "epoch": 0.8144716088328076,
+      "grad_norm": 0.5171796277326578,
+      "learning_rate": 1.803755429948218e-05,
+      "loss": 0.4101,
+      "step": 4131
+    },
+    {
+      "epoch": 0.8146687697160884,
+      "grad_norm": 0.5373199928293783,
+      "learning_rate": 1.8036631999726348e-05,
+      "loss": 0.4328,
+      "step": 4132
+    },
+    {
+      "epoch": 0.8148659305993691,
+      "grad_norm": 0.5649352218109933,
+      "learning_rate": 1.8035709506883962e-05,
+      "loss": 0.4896,
+      "step": 4133
+    },
+    {
+      "epoch": 0.8150630914826499,
+      "grad_norm": 0.5095327224371953,
+      "learning_rate": 1.8034786820977184e-05,
+      "loss": 0.4626,
+      "step": 4134
+    },
+    {
+      "epoch": 0.8152602523659306,
+      "grad_norm": 0.5296243774919772,
+      "learning_rate": 1.8033863942028183e-05,
+      "loss": 0.4478,
+      "step": 4135
+    },
+    {
+      "epoch": 0.8154574132492114,
+      "grad_norm": 0.5532856082609808,
+      "learning_rate": 1.803294087005913e-05,
+      "loss": 0.4353,
+      "step": 4136
+    },
+    {
+      "epoch": 0.8156545741324921,
+      "grad_norm": 0.5241033299839589,
+      "learning_rate": 1.8032017605092202e-05,
+      "loss": 0.4305,
+      "step": 4137
+    },
+    {
+      "epoch": 0.8158517350157729,
+      "grad_norm": 0.47170306660651057,
+      "learning_rate": 1.8031094147149587e-05,
+      "loss": 0.3626,
+      "step": 4138
+    },
+    {
+      "epoch": 0.8160488958990536,
+      "grad_norm": 0.5677681072099129,
+      "learning_rate": 1.8030170496253463e-05,
+      "loss": 0.4935,
+      "step": 4139
+    },
+    {
+      "epoch": 0.8162460567823344,
+      "grad_norm": 0.5340874190338516,
+      "learning_rate": 1.802924665242603e-05,
+      "loss": 0.42,
+      "step": 4140
+    },
+    {
+      "epoch": 0.8164432176656151,
+      "grad_norm": 0.5247377226187014,
+      "learning_rate": 1.8028322615689477e-05,
+      "loss": 0.4289,
+      "step": 4141
+    },
+    {
+      "epoch": 0.8166403785488959,
+      "grad_norm": 0.5542330994880117,
+      "learning_rate": 1.802739838606601e-05,
+      "loss": 0.4464,
+      "step": 4142
+    },
+    {
+      "epoch": 0.8168375394321766,
+      "grad_norm": 19.89904924544576,
+      "learning_rate": 1.8026473963577834e-05,
+      "loss": 0.5853,
+      "step": 4143
+    },
+    {
+      "epoch": 0.8170347003154574,
+      "grad_norm": 0.6171631795186983,
+      "learning_rate": 1.8025549348247154e-05,
+      "loss": 0.4306,
+      "step": 4144
+    },
+    {
+      "epoch": 0.8172318611987381,
+      "grad_norm": 0.5504405635665522,
+      "learning_rate": 1.802462454009619e-05,
+      "loss": 0.4298,
+      "step": 4145
+    },
+    {
+      "epoch": 0.817429022082019,
+      "grad_norm": 0.5886602077518848,
+      "learning_rate": 1.802369953914716e-05,
+      "loss": 0.4274,
+      "step": 4146
+    },
+    {
+      "epoch": 0.8176261829652997,
+      "grad_norm": 0.5150160069331389,
+      "learning_rate": 1.8022774345422284e-05,
+      "loss": 0.4456,
+      "step": 4147
+    },
+    {
+      "epoch": 0.8178233438485805,
+      "grad_norm": 0.5736174914873833,
+      "learning_rate": 1.8021848958943796e-05,
+      "loss": 0.4318,
+      "step": 4148
+    },
+    {
+      "epoch": 0.8180205047318612,
+      "grad_norm": 0.553943761555841,
+      "learning_rate": 1.8020923379733925e-05,
+      "loss": 0.4331,
+      "step": 4149
+    },
+    {
+      "epoch": 0.818217665615142,
+      "grad_norm": 0.5424039413522739,
+      "learning_rate": 1.801999760781491e-05,
+      "loss": 0.4062,
+      "step": 4150
+    },
+    {
+      "epoch": 0.8184148264984227,
+      "grad_norm": 0.6049951163196499,
+      "learning_rate": 1.8019071643208996e-05,
+      "loss": 0.4153,
+      "step": 4151
+    },
+    {
+      "epoch": 0.8186119873817035,
+      "grad_norm": 0.5680607234809361,
+      "learning_rate": 1.8018145485938427e-05,
+      "loss": 0.4338,
+      "step": 4152
+    },
+    {
+      "epoch": 0.8188091482649842,
+      "grad_norm": 0.5291584704930397,
+      "learning_rate": 1.8017219136025458e-05,
+      "loss": 0.4267,
+      "step": 4153
+    },
+    {
+      "epoch": 0.819006309148265,
+      "grad_norm": 0.5806259018554695,
+      "learning_rate": 1.801629259349234e-05,
+      "loss": 0.4584,
+      "step": 4154
+    },
+    {
+      "epoch": 0.8192034700315457,
+      "grad_norm": 0.5109979743220913,
+      "learning_rate": 1.801536585836134e-05,
+      "loss": 0.4448,
+      "step": 4155
+    },
+    {
+      "epoch": 0.8194006309148265,
+      "grad_norm": 1.0089832585060787,
+      "learning_rate": 1.801443893065472e-05,
+      "loss": 0.4481,
+      "step": 4156
+    },
+    {
+      "epoch": 0.8195977917981072,
+      "grad_norm": 0.5400204179891461,
+      "learning_rate": 1.8013511810394747e-05,
+      "loss": 0.4355,
+      "step": 4157
+    },
+    {
+      "epoch": 0.819794952681388,
+      "grad_norm": 0.5585252526726473,
+      "learning_rate": 1.80125844976037e-05,
+      "loss": 0.4572,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8199921135646687,
+      "grad_norm": 0.5965272393146293,
+      "learning_rate": 1.8011656992303863e-05,
+      "loss": 0.4494,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8201892744479495,
+      "grad_norm": 0.5918746014179956,
+      "learning_rate": 1.801072929451751e-05,
+      "loss": 0.4272,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8203864353312302,
+      "grad_norm": 0.5389833452459519,
+      "learning_rate": 1.8009801404266936e-05,
+      "loss": 0.4041,
+      "step": 4161
+    },
+    {
+      "epoch": 0.820583596214511,
+      "grad_norm": 2.6854862231259338,
+      "learning_rate": 1.8008873321574435e-05,
+      "loss": 0.4462,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8207807570977917,
+      "grad_norm": 12.845011498981133,
+      "learning_rate": 1.8007945046462302e-05,
+      "loss": 0.4828,
+      "step": 4163
+    },
+    {
+      "epoch": 0.8209779179810726,
+      "grad_norm": 0.6285393491640829,
+      "learning_rate": 1.800701657895284e-05,
+      "loss": 0.4869,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8211750788643533,
+      "grad_norm": 0.5463921951908919,
+      "learning_rate": 1.8006087919068354e-05,
+      "loss": 0.4517,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8213722397476341,
+      "grad_norm": 3.3042430055230434,
+      "learning_rate": 1.800515906683116e-05,
+      "loss": 0.4767,
+      "step": 4166
+    },
+    {
+      "epoch": 0.8215694006309149,
+      "grad_norm": 0.608965250290735,
+      "learning_rate": 1.8004230022263575e-05,
+      "loss": 0.4708,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8217665615141956,
+      "grad_norm": 0.5644431847609628,
+      "learning_rate": 1.800330078538792e-05,
+      "loss": 0.4087,
+      "step": 4168
+    },
+    {
+      "epoch": 0.8219637223974764,
+      "grad_norm": 0.6120359597106482,
+      "learning_rate": 1.8002371356226512e-05,
+      "loss": 0.4715,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8221608832807571,
+      "grad_norm": 0.5666338187192854,
+      "learning_rate": 1.800144173480169e-05,
+      "loss": 0.463,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8223580441640379,
+      "grad_norm": 0.5992673957167014,
+      "learning_rate": 1.800051192113579e-05,
+      "loss": 0.4724,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8225552050473186,
+      "grad_norm": 0.5713699259593359,
+      "learning_rate": 1.799958191525115e-05,
+      "loss": 0.4568,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8227523659305994,
+      "grad_norm": 0.5796266634572839,
+      "learning_rate": 1.7998651717170105e-05,
+      "loss": 0.4501,
+      "step": 4173
+    },
+    {
+      "epoch": 0.8229495268138801,
+      "grad_norm": 0.5794145486533863,
+      "learning_rate": 1.7997721326915015e-05,
+      "loss": 0.4894,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8231466876971609,
+      "grad_norm": 0.548917204300831,
+      "learning_rate": 1.799679074450823e-05,
+      "loss": 0.4151,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8233438485804416,
+      "grad_norm": 0.6880509796679153,
+      "learning_rate": 1.7995859969972108e-05,
+      "loss": 0.4362,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8235410094637224,
+      "grad_norm": 0.6333808028462822,
+      "learning_rate": 1.7994929003329008e-05,
+      "loss": 0.4802,
+      "step": 4177
+    },
+    {
+      "epoch": 0.8237381703470031,
+      "grad_norm": 0.5717804452138856,
+      "learning_rate": 1.7993997844601305e-05,
+      "loss": 0.4161,
+      "step": 4178
+    },
+    {
+      "epoch": 0.823935331230284,
+      "grad_norm": 0.5904214094197301,
+      "learning_rate": 1.799306649381136e-05,
+      "loss": 0.4527,
+      "step": 4179
+    },
+    {
+      "epoch": 0.8241324921135647,
+      "grad_norm": 0.5224269790433933,
+      "learning_rate": 1.7992134950981562e-05,
+      "loss": 0.3997,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8243296529968455,
+      "grad_norm": 0.5926917719981647,
+      "learning_rate": 1.7991203216134283e-05,
+      "loss": 0.436,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8245268138801262,
+      "grad_norm": 0.5735078186499827,
+      "learning_rate": 1.7990271289291913e-05,
+      "loss": 0.4367,
+      "step": 4182
+    },
+    {
+      "epoch": 0.824723974763407,
+      "grad_norm": 0.6677743637655965,
+      "learning_rate": 1.798933917047684e-05,
+      "loss": 0.4423,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8249211356466877,
+      "grad_norm": 0.582150975318319,
+      "learning_rate": 1.7988406859711457e-05,
+      "loss": 0.4656,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8251182965299685,
+      "grad_norm": 0.4907564880534393,
+      "learning_rate": 1.7987474357018172e-05,
+      "loss": 0.3808,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8253154574132492,
+      "grad_norm": 0.6380319217129097,
+      "learning_rate": 1.7986541662419376e-05,
+      "loss": 0.422,
+      "step": 4186
+    },
+    {
+      "epoch": 0.82551261829653,
+      "grad_norm": 0.5678615187977981,
+      "learning_rate": 1.7985608775937492e-05,
+      "loss": 0.4317,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8257097791798107,
+      "grad_norm": 0.5265688118305606,
+      "learning_rate": 1.798467569759492e-05,
+      "loss": 0.4309,
+      "step": 4188
+    },
+    {
+      "epoch": 0.8259069400630915,
+      "grad_norm": 0.5925439253755009,
+      "learning_rate": 1.798374242741409e-05,
+      "loss": 0.4434,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8261041009463722,
+      "grad_norm": 0.6569515898788928,
+      "learning_rate": 1.7982808965417415e-05,
+      "loss": 0.4008,
+      "step": 4190
+    },
+    {
+      "epoch": 0.826301261829653,
+      "grad_norm": 0.5541366851451021,
+      "learning_rate": 1.7981875311627327e-05,
+      "loss": 0.4168,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8264984227129337,
+      "grad_norm": 0.5419308496525422,
+      "learning_rate": 1.7980941466066254e-05,
+      "loss": 0.4263,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8266955835962145,
+      "grad_norm": 0.5627041984651608,
+      "learning_rate": 1.798000742875664e-05,
+      "loss": 0.4471,
+      "step": 4193
+    },
+    {
+      "epoch": 0.8268927444794952,
+      "grad_norm": 0.5277704275115969,
+      "learning_rate": 1.797907319972092e-05,
+      "loss": 0.4245,
+      "step": 4194
+    },
+    {
+      "epoch": 0.827089905362776,
+      "grad_norm": 0.5673323456563254,
+      "learning_rate": 1.797813877898154e-05,
+      "loss": 0.437,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8272870662460567,
+      "grad_norm": 0.580531434044333,
+      "learning_rate": 1.7977204166560954e-05,
+      "loss": 0.4844,
+      "step": 4196
+    },
+    {
+      "epoch": 0.8274842271293376,
+      "grad_norm": 0.5301677248107065,
+      "learning_rate": 1.797626936248161e-05,
+      "loss": 0.4146,
+      "step": 4197
+    },
+    {
+      "epoch": 0.8276813880126183,
+      "grad_norm": 0.6170836844395501,
+      "learning_rate": 1.7975334366765974e-05,
+      "loss": 0.4555,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8278785488958991,
+      "grad_norm": 0.5481999082788359,
+      "learning_rate": 1.7974399179436502e-05,
+      "loss": 0.467,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8280757097791798,
+      "grad_norm": 0.5408148308068799,
+      "learning_rate": 1.7973463800515675e-05,
+      "loss": 0.4672,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8282728706624606,
+      "grad_norm": 0.5652366078680668,
+      "learning_rate": 1.7972528230025954e-05,
+      "loss": 0.4412,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8284700315457413,
+      "grad_norm": 0.538008456834178,
+      "learning_rate": 1.7971592467989824e-05,
+      "loss": 0.4455,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8286671924290221,
+      "grad_norm": 0.5497144473137757,
+      "learning_rate": 1.7970656514429767e-05,
+      "loss": 0.4495,
+      "step": 4203
+    },
+    {
+      "epoch": 0.8288643533123028,
+      "grad_norm": 0.5400822082606279,
+      "learning_rate": 1.7969720369368266e-05,
+      "loss": 0.4171,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8290615141955836,
+      "grad_norm": 0.5032870883447115,
+      "learning_rate": 1.796878403282782e-05,
+      "loss": 0.4039,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8292586750788643,
+      "grad_norm": 0.5077806097030966,
+      "learning_rate": 1.7967847504830914e-05,
+      "loss": 0.4433,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8294558359621451,
+      "grad_norm": 0.5809809540553955,
+      "learning_rate": 1.7966910785400058e-05,
+      "loss": 0.4481,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8296529968454258,
+      "grad_norm": 0.5469499961190881,
+      "learning_rate": 1.7965973874557754e-05,
+      "loss": 0.4356,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8298501577287066,
+      "grad_norm": 0.5124543710594913,
+      "learning_rate": 1.7965036772326515e-05,
+      "loss": 0.4394,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8300473186119873,
+      "grad_norm": 0.5569661416418024,
+      "learning_rate": 1.796409947872885e-05,
+      "loss": 0.4131,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8302444794952681,
+      "grad_norm": 0.566607051094716,
+      "learning_rate": 1.7963161993787285e-05,
+      "loss": 0.4291,
+      "step": 4211
+    },
+    {
+      "epoch": 0.830441640378549,
+      "grad_norm": 0.6317395544141171,
+      "learning_rate": 1.796222431752434e-05,
+      "loss": 0.4528,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8306388012618297,
+      "grad_norm": 0.5260509022156253,
+      "learning_rate": 1.796128644996254e-05,
+      "loss": 0.4194,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8308359621451105,
+      "grad_norm": 0.8039173856781239,
+      "learning_rate": 1.7960348391124422e-05,
+      "loss": 0.4602,
+      "step": 4214
+    },
+    {
+      "epoch": 0.8310331230283912,
+      "grad_norm": 0.5673951125473756,
+      "learning_rate": 1.7959410141032524e-05,
+      "loss": 0.4874,
+      "step": 4215
+    },
+    {
+      "epoch": 0.831230283911672,
+      "grad_norm": 0.5472093603546104,
+      "learning_rate": 1.795847169970939e-05,
+      "loss": 0.4226,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8314274447949527,
+      "grad_norm": 0.6600760076559387,
+      "learning_rate": 1.7957533067177565e-05,
+      "loss": 0.4507,
+      "step": 4217
+    },
+    {
+      "epoch": 0.8316246056782335,
+      "grad_norm": 0.5420048851350387,
+      "learning_rate": 1.7956594243459597e-05,
+      "loss": 0.4597,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8318217665615142,
+      "grad_norm": 0.5846319849001685,
+      "learning_rate": 1.7955655228578046e-05,
+      "loss": 0.4637,
+      "step": 4219
+    },
+    {
+      "epoch": 0.832018927444795,
+      "grad_norm": 0.5925117015754054,
+      "learning_rate": 1.7954716022555474e-05,
+      "loss": 0.475,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8322160883280757,
+      "grad_norm": 0.5371836503383141,
+      "learning_rate": 1.795377662541444e-05,
+      "loss": 0.4217,
+      "step": 4221
+    },
+    {
+      "epoch": 0.8324132492113565,
+      "grad_norm": 0.7483433485104296,
+      "learning_rate": 1.795283703717752e-05,
+      "loss": 0.4527,
+      "step": 4222
+    },
+    {
+      "epoch": 0.8326104100946372,
+      "grad_norm": 0.5384653534370835,
+      "learning_rate": 1.7951897257867284e-05,
+      "loss": 0.4125,
+      "step": 4223
+    },
+    {
+      "epoch": 0.832807570977918,
+      "grad_norm": 0.5538901981181154,
+      "learning_rate": 1.7950957287506313e-05,
+      "loss": 0.437,
+      "step": 4224
+    },
+    {
+      "epoch": 0.8330047318611987,
+      "grad_norm": 0.5448553268692006,
+      "learning_rate": 1.795001712611719e-05,
+      "loss": 0.432,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8332018927444795,
+      "grad_norm": 0.5608394906395617,
+      "learning_rate": 1.7949076773722505e-05,
+      "loss": 0.4812,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8333990536277602,
+      "grad_norm": 0.48592649216287015,
+      "learning_rate": 1.7948136230344847e-05,
+      "loss": 0.4042,
+      "step": 4227
+    },
+    {
+      "epoch": 0.833596214511041,
+      "grad_norm": 0.5149456873877976,
+      "learning_rate": 1.7947195496006817e-05,
+      "loss": 0.4183,
+      "step": 4228
+    },
+    {
+      "epoch": 0.8337933753943217,
+      "grad_norm": 0.5308679728526855,
+      "learning_rate": 1.7946254570731015e-05,
+      "loss": 0.4564,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8339905362776026,
+      "grad_norm": 0.5705369452847299,
+      "learning_rate": 1.7945313454540046e-05,
+      "loss": 0.4626,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8341876971608833,
+      "grad_norm": 0.5379448499030216,
+      "learning_rate": 1.7944372147456527e-05,
+      "loss": 0.449,
+      "step": 4231
+    },
+    {
+      "epoch": 0.8343848580441641,
+      "grad_norm": 0.536663991436941,
+      "learning_rate": 1.7943430649503065e-05,
+      "loss": 0.4614,
+      "step": 4232
+    },
+    {
+      "epoch": 0.8345820189274448,
+      "grad_norm": 0.5322133371923756,
+      "learning_rate": 1.794248896070229e-05,
+      "loss": 0.4251,
+      "step": 4233
+    },
+    {
+      "epoch": 0.8347791798107256,
+      "grad_norm": 0.5399982863538748,
+      "learning_rate": 1.7941547081076818e-05,
+      "loss": 0.4575,
+      "step": 4234
+    },
+    {
+      "epoch": 0.8349763406940063,
+      "grad_norm": 0.5375534373449318,
+      "learning_rate": 1.7940605010649284e-05,
+      "loss": 0.4295,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8351735015772871,
+      "grad_norm": 0.5478250406848311,
+      "learning_rate": 1.7939662749442317e-05,
+      "loss": 0.4352,
+      "step": 4236
+    },
+    {
+      "epoch": 0.8353706624605678,
+      "grad_norm": 0.5493522873611169,
+      "learning_rate": 1.7938720297478564e-05,
+      "loss": 0.4487,
+      "step": 4237
+    },
+    {
+      "epoch": 0.8355678233438486,
+      "grad_norm": 0.5461493071646538,
+      "learning_rate": 1.7937777654780656e-05,
+      "loss": 0.4621,
+      "step": 4238
+    },
+    {
+      "epoch": 0.8357649842271293,
+      "grad_norm": 0.5405378451575877,
+      "learning_rate": 1.793683482137125e-05,
+      "loss": 0.4167,
+      "step": 4239
+    },
+    {
+      "epoch": 0.8359621451104101,
+      "grad_norm": 0.5514595660368122,
+      "learning_rate": 1.7935891797272998e-05,
+      "loss": 0.4397,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8361593059936908,
+      "grad_norm": 0.5700483701942513,
+      "learning_rate": 1.7934948582508554e-05,
+      "loss": 0.4287,
+      "step": 4241
+    },
+    {
+      "epoch": 0.8363564668769716,
+      "grad_norm": 34.56133146719008,
+      "learning_rate": 1.793400517710058e-05,
+      "loss": 0.4441,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8365536277602523,
+      "grad_norm": 24.176632613926778,
+      "learning_rate": 1.7933061581071743e-05,
+      "loss": 0.4912,
+      "step": 4243
+    },
+    {
+      "epoch": 0.8367507886435331,
+      "grad_norm": 0.6360329434335535,
+      "learning_rate": 1.793211779444471e-05,
+      "loss": 0.4639,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8369479495268138,
+      "grad_norm": 0.6094525873812249,
+      "learning_rate": 1.7931173817242163e-05,
+      "loss": 0.4648,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8371451104100947,
+      "grad_norm": 0.6835903177747272,
+      "learning_rate": 1.7930229649486777e-05,
+      "loss": 0.4328,
+      "step": 4246
+    },
+    {
+      "epoch": 0.8373422712933754,
+      "grad_norm": 0.5554084747146647,
+      "learning_rate": 1.7929285291201237e-05,
+      "loss": 0.4385,
+      "step": 4247
+    },
+    {
+      "epoch": 0.8375394321766562,
+      "grad_norm": 0.5699771546905784,
+      "learning_rate": 1.7928340742408236e-05,
+      "loss": 0.4457,
+      "step": 4248
+    },
+    {
+      "epoch": 0.8377365930599369,
+      "grad_norm": 0.6851558987403872,
+      "learning_rate": 1.792739600313046e-05,
+      "loss": 0.4285,
+      "step": 4249
+    },
+    {
+      "epoch": 0.8379337539432177,
+      "grad_norm": 0.5710755557207288,
+      "learning_rate": 1.7926451073390612e-05,
+      "loss": 0.4773,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8381309148264984,
+      "grad_norm": 0.6090925217214507,
+      "learning_rate": 1.7925505953211394e-05,
+      "loss": 0.4597,
+      "step": 4251
+    },
+    {
+      "epoch": 0.8383280757097792,
+      "grad_norm": 0.5459422698014952,
+      "learning_rate": 1.792456064261551e-05,
+      "loss": 0.4478,
+      "step": 4252
+    },
+    {
+      "epoch": 0.8385252365930599,
+      "grad_norm": 0.5617534088004623,
+      "learning_rate": 1.7923615141625677e-05,
+      "loss": 0.4396,
+      "step": 4253
+    },
+    {
+      "epoch": 0.8387223974763407,
+      "grad_norm": 0.6076401507399524,
+      "learning_rate": 1.792266945026461e-05,
+      "loss": 0.4723,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8389195583596214,
+      "grad_norm": 0.5960645754305288,
+      "learning_rate": 1.792172356855503e-05,
+      "loss": 0.4686,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8391167192429022,
+      "grad_norm": 0.5490789054602314,
+      "learning_rate": 1.7920777496519665e-05,
+      "loss": 0.4362,
+      "step": 4256
+    },
+    {
+      "epoch": 0.839313880126183,
+      "grad_norm": 0.5627710935612827,
+      "learning_rate": 1.7919831234181234e-05,
+      "loss": 0.4299,
+      "step": 4257
+    },
+    {
+      "epoch": 0.8395110410094637,
+      "grad_norm": 0.5861974351027728,
+      "learning_rate": 1.7918884781562486e-05,
+      "loss": 0.444,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8397082018927445,
+      "grad_norm": 0.5865858099699106,
+      "learning_rate": 1.7917938138686152e-05,
+      "loss": 0.4527,
+      "step": 4259
+    },
+    {
+      "epoch": 0.8399053627760252,
+      "grad_norm": 0.6002524896560638,
+      "learning_rate": 1.791699130557498e-05,
+      "loss": 0.4515,
+      "step": 4260
+    },
+    {
+      "epoch": 0.840102523659306,
+      "grad_norm": 1.3685110134791254,
+      "learning_rate": 1.7916044282251713e-05,
+      "loss": 0.3769,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8402996845425867,
+      "grad_norm": 0.5968798080121339,
+      "learning_rate": 1.7915097068739108e-05,
+      "loss": 0.443,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8404968454258676,
+      "grad_norm": 0.916862721842038,
+      "learning_rate": 1.7914149665059922e-05,
+      "loss": 0.4274,
+      "step": 4263
+    },
+    {
+      "epoch": 0.8406940063091483,
+      "grad_norm": 0.5597685865087868,
+      "learning_rate": 1.791320207123692e-05,
+      "loss": 0.4155,
+      "step": 4264
+    },
+    {
+      "epoch": 0.8408911671924291,
+      "grad_norm": 0.7515342823351008,
+      "learning_rate": 1.7912254287292863e-05,
+      "loss": 0.4606,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8410883280757098,
+      "grad_norm": 0.5225927715886619,
+      "learning_rate": 1.7911306313250523e-05,
+      "loss": 0.4025,
+      "step": 4266
+    },
+    {
+      "epoch": 0.8412854889589906,
+      "grad_norm": 0.5557086212596624,
+      "learning_rate": 1.7910358149132682e-05,
+      "loss": 0.4528,
+      "step": 4267
+    },
+    {
+      "epoch": 0.8414826498422713,
+      "grad_norm": 0.5735148560669096,
+      "learning_rate": 1.7909409794962115e-05,
+      "loss": 0.4348,
+      "step": 4268
+    },
+    {
+      "epoch": 0.8416798107255521,
+      "grad_norm": 0.5535257331063189,
+      "learning_rate": 1.790846125076161e-05,
+      "loss": 0.4406,
+      "step": 4269
+    },
+    {
+      "epoch": 0.8418769716088328,
+      "grad_norm": 0.6394946265678507,
+      "learning_rate": 1.790751251655395e-05,
+      "loss": 0.4316,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8420741324921136,
+      "grad_norm": 0.5519239974498845,
+      "learning_rate": 1.7906563592361935e-05,
+      "loss": 0.4457,
+      "step": 4271
+    },
+    {
+      "epoch": 0.8422712933753943,
+      "grad_norm": 0.5760757282643882,
+      "learning_rate": 1.7905614478208363e-05,
+      "loss": 0.4369,
+      "step": 4272
+    },
+    {
+      "epoch": 0.8424684542586751,
+      "grad_norm": 0.5792308751257407,
+      "learning_rate": 1.7904665174116038e-05,
+      "loss": 0.4549,
+      "step": 4273
+    },
+    {
+      "epoch": 0.8426656151419558,
+      "grad_norm": 0.5683056364517756,
+      "learning_rate": 1.790371568010777e-05,
+      "loss": 0.4554,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8428627760252366,
+      "grad_norm": 0.5615230001406172,
+      "learning_rate": 1.7902765996206364e-05,
+      "loss": 0.4459,
+      "step": 4275
+    },
+    {
+      "epoch": 0.8430599369085173,
+      "grad_norm": 0.5564542215994093,
+      "learning_rate": 1.790181612243464e-05,
+      "loss": 0.45,
+      "step": 4276
+    },
+    {
+      "epoch": 0.8432570977917981,
+      "grad_norm": 0.5401445915473098,
+      "learning_rate": 1.7900866058815424e-05,
+      "loss": 0.4439,
+      "step": 4277
+    },
+    {
+      "epoch": 0.8434542586750788,
+      "grad_norm": 0.5386599600784427,
+      "learning_rate": 1.7899915805371536e-05,
+      "loss": 0.4422,
+      "step": 4278
+    },
+    {
+      "epoch": 0.8436514195583596,
+      "grad_norm": 0.5120099294044634,
+      "learning_rate": 1.789896536212581e-05,
+      "loss": 0.4206,
+      "step": 4279
+    },
+    {
+      "epoch": 0.8438485804416404,
+      "grad_norm": 0.5832867234998125,
+      "learning_rate": 1.7898014729101077e-05,
+      "loss": 0.4727,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8440457413249212,
+      "grad_norm": 0.5192015921845785,
+      "learning_rate": 1.7897063906320182e-05,
+      "loss": 0.4239,
+      "step": 4281
+    },
+    {
+      "epoch": 0.8442429022082019,
+      "grad_norm": 0.5700985443501763,
+      "learning_rate": 1.7896112893805967e-05,
+      "loss": 0.4813,
+      "step": 4282
+    },
+    {
+      "epoch": 0.8444400630914827,
+      "grad_norm": 0.5274314100347487,
+      "learning_rate": 1.789516169158128e-05,
+      "loss": 0.4251,
+      "step": 4283
+    },
+    {
+      "epoch": 0.8446372239747634,
+      "grad_norm": 0.5003375182641053,
+      "learning_rate": 1.7894210299668977e-05,
+      "loss": 0.3926,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8448343848580442,
+      "grad_norm": 0.5327663026050636,
+      "learning_rate": 1.7893258718091916e-05,
+      "loss": 0.4379,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8450315457413249,
+      "grad_norm": 0.5526141806841286,
+      "learning_rate": 1.7892306946872952e-05,
+      "loss": 0.4765,
+      "step": 4286
+    },
+    {
+      "epoch": 0.8452287066246057,
+      "grad_norm": 0.5253955764429079,
+      "learning_rate": 1.7891354986034964e-05,
+      "loss": 0.4083,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8454258675078864,
+      "grad_norm": 0.9128365999475107,
+      "learning_rate": 1.7890402835600814e-05,
+      "loss": 0.4293,
+      "step": 4288
+    },
+    {
+      "epoch": 0.8456230283911672,
+      "grad_norm": 0.50508451474393,
+      "learning_rate": 1.7889450495593386e-05,
+      "loss": 0.4281,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8458201892744479,
+      "grad_norm": 0.5667695632508356,
+      "learning_rate": 1.7888497966035552e-05,
+      "loss": 0.4551,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8460173501577287,
+      "grad_norm": 0.5140164942376172,
+      "learning_rate": 1.7887545246950204e-05,
+      "loss": 0.4283,
+      "step": 4291
+    },
+    {
+      "epoch": 0.8462145110410094,
+      "grad_norm": 0.5547368089832118,
+      "learning_rate": 1.7886592338360227e-05,
+      "loss": 0.4204,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8464116719242902,
+      "grad_norm": 0.5107722071951123,
+      "learning_rate": 1.7885639240288523e-05,
+      "loss": 0.4269,
+      "step": 4293
+    },
+    {
+      "epoch": 0.8466088328075709,
+      "grad_norm": 0.5099322891614341,
+      "learning_rate": 1.788468595275798e-05,
+      "loss": 0.3775,
+      "step": 4294
+    },
+    {
+      "epoch": 0.8468059936908517,
+      "grad_norm": 0.5424856816385994,
+      "learning_rate": 1.7883732475791512e-05,
+      "loss": 0.4256,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8470031545741324,
+      "grad_norm": 0.5241853108328052,
+      "learning_rate": 1.7882778809412024e-05,
+      "loss": 0.4292,
+      "step": 4296
+    },
+    {
+      "epoch": 0.8472003154574133,
+      "grad_norm": 0.542588881145842,
+      "learning_rate": 1.7881824953642423e-05,
+      "loss": 0.4074,
+      "step": 4297
+    },
+    {
+      "epoch": 0.847397476340694,
+      "grad_norm": 0.7626469717720238,
+      "learning_rate": 1.788087090850563e-05,
+      "loss": 0.4365,
+      "step": 4298
+    },
+    {
+      "epoch": 0.8475946372239748,
+      "grad_norm": 0.6333662401521095,
+      "learning_rate": 1.787991667402457e-05,
+      "loss": 0.4631,
+      "step": 4299
+    },
+    {
+      "epoch": 0.8477917981072555,
+      "grad_norm": 0.5258355942220182,
+      "learning_rate": 1.787896225022216e-05,
+      "loss": 0.4423,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8479889589905363,
+      "grad_norm": 0.572162435469079,
+      "learning_rate": 1.7878007637121344e-05,
+      "loss": 0.4635,
+      "step": 4301
+    },
+    {
+      "epoch": 0.848186119873817,
+      "grad_norm": 0.6051771551039156,
+      "learning_rate": 1.7877052834745048e-05,
+      "loss": 0.5096,
+      "step": 4302
+    },
+    {
+      "epoch": 0.8483832807570978,
+      "grad_norm": 0.5953955107671394,
+      "learning_rate": 1.7876097843116214e-05,
+      "loss": 0.4335,
+      "step": 4303
+    },
+    {
+      "epoch": 0.8485804416403786,
+      "grad_norm": 0.5157443177704121,
+      "learning_rate": 1.7875142662257788e-05,
+      "loss": 0.4226,
+      "step": 4304
+    },
+    {
+      "epoch": 0.8487776025236593,
+      "grad_norm": 0.5783953306524755,
+      "learning_rate": 1.7874187292192716e-05,
+      "loss": 0.4105,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8489747634069401,
+      "grad_norm": 3.512906038519687,
+      "learning_rate": 1.7873231732943954e-05,
+      "loss": 0.4794,
+      "step": 4306
+    },
+    {
+      "epoch": 0.8491719242902208,
+      "grad_norm": 0.6607264573754198,
+      "learning_rate": 1.787227598453446e-05,
+      "loss": 0.4226,
+      "step": 4307
+    },
+    {
+      "epoch": 0.8493690851735016,
+      "grad_norm": 0.5184733373485347,
+      "learning_rate": 1.7871320046987195e-05,
+      "loss": 0.4468,
+      "step": 4308
+    },
+    {
+      "epoch": 0.8495662460567823,
+      "grad_norm": 0.6207995208356605,
+      "learning_rate": 1.7870363920325126e-05,
+      "loss": 0.4391,
+      "step": 4309
+    },
+    {
+      "epoch": 0.8497634069400631,
+      "grad_norm": 0.9804052510102346,
+      "learning_rate": 1.7869407604571228e-05,
+      "loss": 0.4196,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8499605678233438,
+      "grad_norm": 0.5996292455245761,
+      "learning_rate": 1.7868451099748473e-05,
+      "loss": 0.4297,
+      "step": 4311
+    },
+    {
+      "epoch": 0.8501577287066246,
+      "grad_norm": 0.5472017368916122,
+      "learning_rate": 1.7867494405879847e-05,
+      "loss": 0.4032,
+      "step": 4312
+    },
+    {
+      "epoch": 0.8503548895899053,
+      "grad_norm": 0.7587762660324456,
+      "learning_rate": 1.786653752298833e-05,
+      "loss": 0.4218,
+      "step": 4313
+    },
+    {
+      "epoch": 0.8505520504731862,
+      "grad_norm": 0.5433647751228663,
+      "learning_rate": 1.7865580451096912e-05,
+      "loss": 0.4439,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8507492113564669,
+      "grad_norm": 0.5467356597687231,
+      "learning_rate": 1.7864623190228592e-05,
+      "loss": 0.4679,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8509463722397477,
+      "grad_norm": 0.5811637139524398,
+      "learning_rate": 1.7863665740406367e-05,
+      "loss": 0.4204,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8511435331230284,
+      "grad_norm": 0.5301517872977239,
+      "learning_rate": 1.786270810165324e-05,
+      "loss": 0.4555,
+      "step": 4317
+    },
+    {
+      "epoch": 0.8513406940063092,
+      "grad_norm": 0.5854692352465581,
+      "learning_rate": 1.7861750273992216e-05,
+      "loss": 0.4597,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8515378548895899,
+      "grad_norm": 0.5663117363358948,
+      "learning_rate": 1.7860792257446315e-05,
+      "loss": 0.4437,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8517350157728707,
+      "grad_norm": 0.5681785500111551,
+      "learning_rate": 1.785983405203855e-05,
+      "loss": 0.4747,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8519321766561514,
+      "grad_norm": 0.6863444614061993,
+      "learning_rate": 1.7858875657791937e-05,
+      "loss": 0.4423,
+      "step": 4321
+    },
+    {
+      "epoch": 0.8521293375394322,
+      "grad_norm": 0.8893916870868374,
+      "learning_rate": 1.7857917074729513e-05,
+      "loss": 0.4413,
+      "step": 4322
+    },
+    {
+      "epoch": 0.8523264984227129,
+      "grad_norm": 0.5878530748957967,
+      "learning_rate": 1.78569583028743e-05,
+      "loss": 0.4737,
+      "step": 4323
+    },
+    {
+      "epoch": 0.8525236593059937,
+      "grad_norm": 0.5738260255184614,
+      "learning_rate": 1.7855999342249338e-05,
+      "loss": 0.4325,
+      "step": 4324
+    },
+    {
+      "epoch": 0.8527208201892744,
+      "grad_norm": 0.5660103845510339,
+      "learning_rate": 1.7855040192877666e-05,
+      "loss": 0.4242,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8529179810725552,
+      "grad_norm": 0.5644476271832222,
+      "learning_rate": 1.7854080854782324e-05,
+      "loss": 0.4465,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8531151419558359,
+      "grad_norm": 0.5507139670335401,
+      "learning_rate": 1.7853121327986368e-05,
+      "loss": 0.452,
+      "step": 4327
+    },
+    {
+      "epoch": 0.8533123028391167,
+      "grad_norm": 0.5645022256444481,
+      "learning_rate": 1.785216161251285e-05,
+      "loss": 0.4225,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8535094637223974,
+      "grad_norm": 2.5107638932499734,
+      "learning_rate": 1.7851201708384823e-05,
+      "loss": 0.466,
+      "step": 4329
+    },
+    {
+      "epoch": 0.8537066246056783,
+      "grad_norm": 0.6115304333104712,
+      "learning_rate": 1.785024161562535e-05,
+      "loss": 0.4656,
+      "step": 4330
+    },
+    {
+      "epoch": 0.853903785488959,
+      "grad_norm": 0.5583226024777117,
+      "learning_rate": 1.7849281334257504e-05,
+      "loss": 0.4285,
+      "step": 4331
+    },
+    {
+      "epoch": 0.8541009463722398,
+      "grad_norm": 0.5123230160651777,
+      "learning_rate": 1.784832086430435e-05,
+      "loss": 0.3872,
+      "step": 4332
+    },
+    {
+      "epoch": 0.8542981072555205,
+      "grad_norm": 0.6949006788224281,
+      "learning_rate": 1.784736020578897e-05,
+      "loss": 0.4535,
+      "step": 4333
+    },
+    {
+      "epoch": 0.8544952681388013,
+      "grad_norm": 1.427386384185727,
+      "learning_rate": 1.784639935873444e-05,
+      "loss": 0.4704,
+      "step": 4334
+    },
+    {
+      "epoch": 0.854692429022082,
+      "grad_norm": 1.106108233363933,
+      "learning_rate": 1.784543832316385e-05,
+      "loss": 0.4372,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8548895899053628,
+      "grad_norm": 0.5901770882082108,
+      "learning_rate": 1.7844477099100282e-05,
+      "loss": 0.4699,
+      "step": 4336
+    },
+    {
+      "epoch": 0.8550867507886435,
+      "grad_norm": 0.544525043882823,
+      "learning_rate": 1.784351568656684e-05,
+      "loss": 0.4027,
+      "step": 4337
+    },
+    {
+      "epoch": 0.8552839116719243,
+      "grad_norm": 0.5934533266780927,
+      "learning_rate": 1.7842554085586613e-05,
+      "loss": 0.4838,
+      "step": 4338
+    },
+    {
+      "epoch": 0.855481072555205,
+      "grad_norm": 0.5596803431341593,
+      "learning_rate": 1.7841592296182705e-05,
+      "loss": 0.4621,
+      "step": 4339
+    },
+    {
+      "epoch": 0.8556782334384858,
+      "grad_norm": 0.5394622987888912,
+      "learning_rate": 1.7840630318378233e-05,
+      "loss": 0.397,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8558753943217665,
+      "grad_norm": 0.5455394824500505,
+      "learning_rate": 1.78396681521963e-05,
+      "loss": 0.4519,
+      "step": 4341
+    },
+    {
+      "epoch": 0.8560725552050473,
+      "grad_norm": 0.5203051216706605,
+      "learning_rate": 1.7838705797660033e-05,
+      "loss": 0.4413,
+      "step": 4342
+    },
+    {
+      "epoch": 0.856269716088328,
+      "grad_norm": 0.5539835315564265,
+      "learning_rate": 1.783774325479254e-05,
+      "loss": 0.4285,
+      "step": 4343
+    },
+    {
+      "epoch": 0.8564668769716088,
+      "grad_norm": 0.546518036363135,
+      "learning_rate": 1.7836780523616957e-05,
+      "loss": 0.4444,
+      "step": 4344
+    },
+    {
+      "epoch": 0.8566640378548895,
+      "grad_norm": 0.5381032645999315,
+      "learning_rate": 1.7835817604156407e-05,
+      "loss": 0.4187,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8568611987381703,
+      "grad_norm": 0.6118133777534611,
+      "learning_rate": 1.7834854496434032e-05,
+      "loss": 0.4176,
+      "step": 4346
+    },
+    {
+      "epoch": 0.857058359621451,
+      "grad_norm": 0.6427676574290028,
+      "learning_rate": 1.7833891200472967e-05,
+      "loss": 0.5017,
+      "step": 4347
+    },
+    {
+      "epoch": 0.8572555205047319,
+      "grad_norm": 0.5031944020117598,
+      "learning_rate": 1.7832927716296357e-05,
+      "loss": 0.3831,
+      "step": 4348
+    },
+    {
+      "epoch": 0.8574526813880127,
+      "grad_norm": 0.5370545377340153,
+      "learning_rate": 1.7831964043927355e-05,
+      "loss": 0.4165,
+      "step": 4349
+    },
+    {
+      "epoch": 0.8576498422712934,
+      "grad_norm": 0.5878309901877112,
+      "learning_rate": 1.7831000183389107e-05,
+      "loss": 0.4848,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8578470031545742,
+      "grad_norm": 0.5608856336311973,
+      "learning_rate": 1.783003613470477e-05,
+      "loss": 0.4418,
+      "step": 4351
+    },
+    {
+      "epoch": 0.8580441640378549,
+      "grad_norm": 0.5690843999597316,
+      "learning_rate": 1.7829071897897515e-05,
+      "loss": 0.4331,
+      "step": 4352
+    },
+    {
+      "epoch": 0.8582413249211357,
+      "grad_norm": 0.5542069316816697,
+      "learning_rate": 1.7828107472990498e-05,
+      "loss": 0.4454,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8584384858044164,
+      "grad_norm": 0.5267592858399887,
+      "learning_rate": 1.78271428600069e-05,
+      "loss": 0.4271,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8586356466876972,
+      "grad_norm": 0.5686441971248088,
+      "learning_rate": 1.7826178058969884e-05,
+      "loss": 0.4757,
+      "step": 4355
+    },
+    {
+      "epoch": 0.8588328075709779,
+      "grad_norm": 0.5227257642875571,
+      "learning_rate": 1.7825213069902646e-05,
+      "loss": 0.4574,
+      "step": 4356
+    },
+    {
+      "epoch": 0.8590299684542587,
+      "grad_norm": 0.5300979103639493,
+      "learning_rate": 1.782424789282836e-05,
+      "loss": 0.4289,
+      "step": 4357
+    },
+    {
+      "epoch": 0.8592271293375394,
+      "grad_norm": 0.5849541511404455,
+      "learning_rate": 1.7823282527770214e-05,
+      "loss": 0.4209,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8594242902208202,
+      "grad_norm": 0.48637468314860755,
+      "learning_rate": 1.782231697475141e-05,
+      "loss": 0.4429,
+      "step": 4359
+    },
+    {
+      "epoch": 0.8596214511041009,
+      "grad_norm": 0.5419865357900749,
+      "learning_rate": 1.7821351233795135e-05,
+      "loss": 0.4511,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8598186119873817,
+      "grad_norm": 1.4858320044760873,
+      "learning_rate": 1.7820385304924602e-05,
+      "loss": 0.4831,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8600157728706624,
+      "grad_norm": 0.5226021841678236,
+      "learning_rate": 1.7819419188163015e-05,
+      "loss": 0.4268,
+      "step": 4362
+    },
+    {
+      "epoch": 0.8602129337539433,
+      "grad_norm": 0.5958597819761025,
+      "learning_rate": 1.7818452883533587e-05,
+      "loss": 0.4404,
+      "step": 4363
+    },
+    {
+      "epoch": 0.860410094637224,
+      "grad_norm": 0.5282803263222494,
+      "learning_rate": 1.781748639105953e-05,
+      "loss": 0.4351,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8606072555205048,
+      "grad_norm": 0.5691278859949257,
+      "learning_rate": 1.7816519710764065e-05,
+      "loss": 0.4288,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8608044164037855,
+      "grad_norm": 0.8711980455369287,
+      "learning_rate": 1.7815552842670424e-05,
+      "loss": 0.4175,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8610015772870663,
+      "grad_norm": 0.48580191160854,
+      "learning_rate": 1.7814585786801826e-05,
+      "loss": 0.4037,
+      "step": 4367
+    },
+    {
+      "epoch": 0.861198738170347,
+      "grad_norm": 0.5079270568376504,
+      "learning_rate": 1.7813618543181515e-05,
+      "loss": 0.4153,
+      "step": 4368
+    },
+    {
+      "epoch": 0.8613958990536278,
+      "grad_norm": 0.5060650682920903,
+      "learning_rate": 1.781265111183273e-05,
+      "loss": 0.4223,
+      "step": 4369
+    },
+    {
+      "epoch": 0.8615930599369085,
+      "grad_norm": 0.5229358894402564,
+      "learning_rate": 1.7811683492778704e-05,
+      "loss": 0.4267,
+      "step": 4370
+    },
+    {
+      "epoch": 0.8617902208201893,
+      "grad_norm": 0.5485854868559356,
+      "learning_rate": 1.7810715686042694e-05,
+      "loss": 0.4611,
+      "step": 4371
+    },
+    {
+      "epoch": 0.86198738170347,
+      "grad_norm": 0.7504801541100838,
+      "learning_rate": 1.7809747691647947e-05,
+      "loss": 0.4411,
+      "step": 4372
+    },
+    {
+      "epoch": 0.8621845425867508,
+      "grad_norm": 0.5652888768887478,
+      "learning_rate": 1.7808779509617726e-05,
+      "loss": 0.4544,
+      "step": 4373
+    },
+    {
+      "epoch": 0.8623817034700315,
+      "grad_norm": 0.5229601149145549,
+      "learning_rate": 1.7807811139975287e-05,
+      "loss": 0.432,
+      "step": 4374
+    },
+    {
+      "epoch": 0.8625788643533123,
+      "grad_norm": 0.5693809303118241,
+      "learning_rate": 1.78068425827439e-05,
+      "loss": 0.4531,
+      "step": 4375
+    },
+    {
+      "epoch": 0.862776025236593,
+      "grad_norm": 0.5324905030720758,
+      "learning_rate": 1.7805873837946833e-05,
+      "loss": 0.4503,
+      "step": 4376
+    },
+    {
+      "epoch": 0.8629731861198738,
+      "grad_norm": 0.5981637783741762,
+      "learning_rate": 1.780490490560736e-05,
+      "loss": 0.4517,
+      "step": 4377
+    },
+    {
+      "epoch": 0.8631703470031545,
+      "grad_norm": 0.5398817339769384,
+      "learning_rate": 1.7803935785748758e-05,
+      "loss": 0.4098,
+      "step": 4378
+    },
+    {
+      "epoch": 0.8633675078864353,
+      "grad_norm": 0.6324773282334193,
+      "learning_rate": 1.7802966478394318e-05,
+      "loss": 0.4814,
+      "step": 4379
+    },
+    {
+      "epoch": 0.863564668769716,
+      "grad_norm": 0.5574386589246891,
+      "learning_rate": 1.7801996983567325e-05,
+      "loss": 0.446,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8637618296529969,
+      "grad_norm": 0.5038265095672974,
+      "learning_rate": 1.780102730129107e-05,
+      "loss": 0.4219,
+      "step": 4381
+    },
+    {
+      "epoch": 0.8639589905362776,
+      "grad_norm": 0.5471006623991824,
+      "learning_rate": 1.7800057431588852e-05,
+      "loss": 0.4301,
+      "step": 4382
+    },
+    {
+      "epoch": 0.8641561514195584,
+      "grad_norm": 0.562249614894374,
+      "learning_rate": 1.7799087374483974e-05,
+      "loss": 0.4629,
+      "step": 4383
+    },
+    {
+      "epoch": 0.8643533123028391,
+      "grad_norm": 0.5815465050068923,
+      "learning_rate": 1.7798117129999738e-05,
+      "loss": 0.4361,
+      "step": 4384
+    },
+    {
+      "epoch": 0.8645504731861199,
+      "grad_norm": 0.5771439496581714,
+      "learning_rate": 1.779714669815946e-05,
+      "loss": 0.4527,
+      "step": 4385
+    },
+    {
+      "epoch": 0.8647476340694006,
+      "grad_norm": 0.5527264398916031,
+      "learning_rate": 1.7796176078986458e-05,
+      "loss": 0.4552,
+      "step": 4386
+    },
+    {
+      "epoch": 0.8649447949526814,
+      "grad_norm": 0.564880759780719,
+      "learning_rate": 1.7795205272504044e-05,
+      "loss": 0.4528,
+      "step": 4387
+    },
+    {
+      "epoch": 0.8651419558359621,
+      "grad_norm": 0.5384822591936045,
+      "learning_rate": 1.7794234278735544e-05,
+      "loss": 0.4476,
+      "step": 4388
+    },
+    {
+      "epoch": 0.8653391167192429,
+      "grad_norm": 1.03147273350115,
+      "learning_rate": 1.779326309770429e-05,
+      "loss": 0.427,
+      "step": 4389
+    },
+    {
+      "epoch": 0.8655362776025236,
+      "grad_norm": 0.5033337533381047,
+      "learning_rate": 1.7792291729433615e-05,
+      "loss": 0.422,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8657334384858044,
+      "grad_norm": 0.8142231011342511,
+      "learning_rate": 1.7791320173946857e-05,
+      "loss": 0.4298,
+      "step": 4391
+    },
+    {
+      "epoch": 0.8659305993690851,
+      "grad_norm": 0.5307387314489819,
+      "learning_rate": 1.7790348431267353e-05,
+      "loss": 0.4549,
+      "step": 4392
+    },
+    {
+      "epoch": 0.8661277602523659,
+      "grad_norm": 0.5806932313493961,
+      "learning_rate": 1.7789376501418457e-05,
+      "loss": 0.4667,
+      "step": 4393
+    },
+    {
+      "epoch": 0.8663249211356467,
+      "grad_norm": 0.516262259612544,
+      "learning_rate": 1.778840438442352e-05,
+      "loss": 0.4056,
+      "step": 4394
+    },
+    {
+      "epoch": 0.8665220820189274,
+      "grad_norm": 0.5715000920752781,
+      "learning_rate": 1.7787432080305895e-05,
+      "loss": 0.4327,
+      "step": 4395
+    },
+    {
+      "epoch": 0.8667192429022083,
+      "grad_norm": 0.549118836810879,
+      "learning_rate": 1.7786459589088942e-05,
+      "loss": 0.4196,
+      "step": 4396
+    },
+    {
+      "epoch": 0.866916403785489,
+      "grad_norm": 0.5185987495530604,
+      "learning_rate": 1.778548691079603e-05,
+      "loss": 0.3998,
+      "step": 4397
+    },
+    {
+      "epoch": 0.8671135646687698,
+      "grad_norm": 0.594091330595555,
+      "learning_rate": 1.7784514045450518e-05,
+      "loss": 0.4366,
+      "step": 4398
+    },
+    {
+      "epoch": 0.8673107255520505,
+      "grad_norm": 0.5512970340203216,
+      "learning_rate": 1.7783540993075793e-05,
+      "loss": 0.4249,
+      "step": 4399
+    },
+    {
+      "epoch": 0.8675078864353313,
+      "grad_norm": 0.6495070525009865,
+      "learning_rate": 1.7782567753695227e-05,
+      "loss": 0.4518,
+      "step": 4400
+    },
+    {
+      "epoch": 0.867705047318612,
+      "grad_norm": 0.5475014771103609,
+      "learning_rate": 1.7781594327332203e-05,
+      "loss": 0.4358,
+      "step": 4401
+    },
+    {
+      "epoch": 0.8679022082018928,
+      "grad_norm": 0.5446824793376154,
+      "learning_rate": 1.7780620714010108e-05,
+      "loss": 0.4317,
+      "step": 4402
+    },
+    {
+      "epoch": 0.8680993690851735,
+      "grad_norm": 0.5672459127571157,
+      "learning_rate": 1.7779646913752334e-05,
+      "loss": 0.4306,
+      "step": 4403
+    },
+    {
+      "epoch": 0.8682965299684543,
+      "grad_norm": 0.5159309255044556,
+      "learning_rate": 1.7778672926582277e-05,
+      "loss": 0.4356,
+      "step": 4404
+    },
+    {
+      "epoch": 0.868493690851735,
+      "grad_norm": 0.5706253535740993,
+      "learning_rate": 1.777769875252334e-05,
+      "loss": 0.4624,
+      "step": 4405
+    },
+    {
+      "epoch": 0.8686908517350158,
+      "grad_norm": 0.5468735157960866,
+      "learning_rate": 1.7776724391598928e-05,
+      "loss": 0.4204,
+      "step": 4406
+    },
+    {
+      "epoch": 0.8688880126182965,
+      "grad_norm": 0.6041214464176105,
+      "learning_rate": 1.7775749843832454e-05,
+      "loss": 0.4493,
+      "step": 4407
+    },
+    {
+      "epoch": 0.8690851735015773,
+      "grad_norm": 0.5209820736307841,
+      "learning_rate": 1.777477510924732e-05,
+      "loss": 0.4042,
+      "step": 4408
+    },
+    {
+      "epoch": 0.869282334384858,
+      "grad_norm": 0.5604674535426473,
+      "learning_rate": 1.777380018786696e-05,
+      "loss": 0.4276,
+      "step": 4409
+    },
+    {
+      "epoch": 0.8694794952681388,
+      "grad_norm": 0.5079309439621922,
+      "learning_rate": 1.7772825079714788e-05,
+      "loss": 0.4476,
+      "step": 4410
+    },
+    {
+      "epoch": 0.8696766561514195,
+      "grad_norm": 0.6476704699217136,
+      "learning_rate": 1.7771849784814232e-05,
+      "loss": 0.4578,
+      "step": 4411
+    },
+    {
+      "epoch": 0.8698738170347003,
+      "grad_norm": 0.5924379738519122,
+      "learning_rate": 1.7770874303188727e-05,
+      "loss": 0.4311,
+      "step": 4412
+    },
+    {
+      "epoch": 0.870070977917981,
+      "grad_norm": 0.524148406557284,
+      "learning_rate": 1.776989863486171e-05,
+      "loss": 0.389,
+      "step": 4413
+    },
+    {
+      "epoch": 0.8702681388012619,
+      "grad_norm": 0.5327802595222653,
+      "learning_rate": 1.776892277985662e-05,
+      "loss": 0.4413,
+      "step": 4414
+    },
+    {
+      "epoch": 0.8704652996845426,
+      "grad_norm": 0.62581900914729,
+      "learning_rate": 1.7767946738196903e-05,
+      "loss": 0.4319,
+      "step": 4415
+    },
+    {
+      "epoch": 0.8706624605678234,
+      "grad_norm": 0.55431472515896,
+      "learning_rate": 1.7766970509906014e-05,
+      "loss": 0.4486,
+      "step": 4416
+    },
+    {
+      "epoch": 0.8708596214511041,
+      "grad_norm": 0.5876652626007826,
+      "learning_rate": 1.77659940950074e-05,
+      "loss": 0.4344,
+      "step": 4417
+    },
+    {
+      "epoch": 0.8710567823343849,
+      "grad_norm": 0.5551588366167218,
+      "learning_rate": 1.7765017493524526e-05,
+      "loss": 0.4568,
+      "step": 4418
+    },
+    {
+      "epoch": 0.8712539432176656,
+      "grad_norm": 0.5251199442258818,
+      "learning_rate": 1.776404070548085e-05,
+      "loss": 0.444,
+      "step": 4419
+    },
+    {
+      "epoch": 0.8714511041009464,
+      "grad_norm": 0.5098007039746633,
+      "learning_rate": 1.7763063730899846e-05,
+      "loss": 0.4078,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8716482649842271,
+      "grad_norm": 0.5270323086561973,
+      "learning_rate": 1.776208656980499e-05,
+      "loss": 0.4585,
+      "step": 4421
+    },
+    {
+      "epoch": 0.8718454258675079,
+      "grad_norm": 0.575464987069899,
+      "learning_rate": 1.7761109222219747e-05,
+      "loss": 0.4532,
+      "step": 4422
+    },
+    {
+      "epoch": 0.8720425867507886,
+      "grad_norm": 0.5312918974599236,
+      "learning_rate": 1.7760131688167606e-05,
+      "loss": 0.4247,
+      "step": 4423
+    },
+    {
+      "epoch": 0.8722397476340694,
+      "grad_norm": 0.5487949802158028,
+      "learning_rate": 1.775915396767205e-05,
+      "loss": 0.4636,
+      "step": 4424
+    },
+    {
+      "epoch": 0.8724369085173501,
+      "grad_norm": 0.6044316638894782,
+      "learning_rate": 1.7758176060756572e-05,
+      "loss": 0.4606,
+      "step": 4425
+    },
+    {
+      "epoch": 0.8726340694006309,
+      "grad_norm": 0.5816796983719151,
+      "learning_rate": 1.775719796744467e-05,
+      "loss": 0.4585,
+      "step": 4426
+    },
+    {
+      "epoch": 0.8728312302839116,
+      "grad_norm": 0.5731233568405086,
+      "learning_rate": 1.775621968775984e-05,
+      "loss": 0.4263,
+      "step": 4427
+    },
+    {
+      "epoch": 0.8730283911671924,
+      "grad_norm": 0.596194849375308,
+      "learning_rate": 1.7755241221725583e-05,
+      "loss": 0.4631,
+      "step": 4428
+    },
+    {
+      "epoch": 0.8732255520504731,
+      "grad_norm": 0.5635206736546373,
+      "learning_rate": 1.7754262569365413e-05,
+      "loss": 0.4385,
+      "step": 4429
+    },
+    {
+      "epoch": 0.873422712933754,
+      "grad_norm": 0.5479175042231295,
+      "learning_rate": 1.7753283730702837e-05,
+      "loss": 0.431,
+      "step": 4430
+    },
+    {
+      "epoch": 0.8736198738170347,
+      "grad_norm": 0.5279639808442017,
+      "learning_rate": 1.7752304705761377e-05,
+      "loss": 0.4326,
+      "step": 4431
+    },
+    {
+      "epoch": 0.8738170347003155,
+      "grad_norm": 0.5194599602636989,
+      "learning_rate": 1.7751325494564556e-05,
+      "loss": 0.4215,
+      "step": 4432
+    },
+    {
+      "epoch": 0.8740141955835962,
+      "grad_norm": 0.5615746974328736,
+      "learning_rate": 1.7750346097135896e-05,
+      "loss": 0.4653,
+      "step": 4433
+    },
+    {
+      "epoch": 0.874211356466877,
+      "grad_norm": 0.5428620740818375,
+      "learning_rate": 1.774936651349893e-05,
+      "loss": 0.4014,
+      "step": 4434
+    },
+    {
+      "epoch": 0.8744085173501577,
+      "grad_norm": 0.4968476884769345,
+      "learning_rate": 1.774838674367719e-05,
+      "loss": 0.4188,
+      "step": 4435
+    },
+    {
+      "epoch": 0.8746056782334385,
+      "grad_norm": 0.5130271919066233,
+      "learning_rate": 1.7747406787694222e-05,
+      "loss": 0.4324,
+      "step": 4436
+    },
+    {
+      "epoch": 0.8748028391167192,
+      "grad_norm": 0.7695088300581262,
+      "learning_rate": 1.774642664557357e-05,
+      "loss": 0.4302,
+      "step": 4437
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.5482718548743002,
+      "learning_rate": 1.7745446317338773e-05,
+      "loss": 0.4285,
+      "step": 4438
+    },
+    {
+      "epoch": 0.8751971608832808,
+      "grad_norm": 0.5602436646595499,
+      "learning_rate": 1.7744465803013394e-05,
+      "loss": 0.4519,
+      "step": 4439
+    },
+    {
+      "epoch": 0.8753943217665615,
+      "grad_norm": 0.6144891088977394,
+      "learning_rate": 1.774348510262099e-05,
+      "loss": 0.4275,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8755914826498423,
+      "grad_norm": 0.56029801618344,
+      "learning_rate": 1.774250421618511e-05,
+      "loss": 0.3993,
+      "step": 4441
+    },
+    {
+      "epoch": 0.875788643533123,
+      "grad_norm": 0.5547908674306443,
+      "learning_rate": 1.7741523143729344e-05,
+      "loss": 0.4449,
+      "step": 4442
+    },
+    {
+      "epoch": 0.8759858044164038,
+      "grad_norm": 0.5349957311996212,
+      "learning_rate": 1.7740541885277243e-05,
+      "loss": 0.4173,
+      "step": 4443
+    },
+    {
+      "epoch": 0.8761829652996845,
+      "grad_norm": 0.5813115159456966,
+      "learning_rate": 1.773956044085239e-05,
+      "loss": 0.4598,
+      "step": 4444
+    },
+    {
+      "epoch": 0.8763801261829653,
+      "grad_norm": 0.5607295785609846,
+      "learning_rate": 1.773857881047837e-05,
+      "loss": 0.4563,
+      "step": 4445
+    },
+    {
+      "epoch": 0.876577287066246,
+      "grad_norm": 0.535106190739021,
+      "learning_rate": 1.773759699417876e-05,
+      "loss": 0.4743,
+      "step": 4446
+    },
+    {
+      "epoch": 0.8767744479495269,
+      "grad_norm": 0.5125347386594389,
+      "learning_rate": 1.773661499197715e-05,
+      "loss": 0.4162,
+      "step": 4447
+    },
+    {
+      "epoch": 0.8769716088328076,
+      "grad_norm": 0.49637878407034497,
+      "learning_rate": 1.7735632803897135e-05,
+      "loss": 0.4293,
+      "step": 4448
+    },
+    {
+      "epoch": 0.8771687697160884,
+      "grad_norm": 0.5116773726384485,
+      "learning_rate": 1.773465042996231e-05,
+      "loss": 0.449,
+      "step": 4449
+    },
+    {
+      "epoch": 0.8773659305993691,
+      "grad_norm": 0.5442212905587093,
+      "learning_rate": 1.7733667870196282e-05,
+      "loss": 0.433,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8775630914826499,
+      "grad_norm": 0.5755384800113009,
+      "learning_rate": 1.7732685124622656e-05,
+      "loss": 0.4539,
+      "step": 4451
+    },
+    {
+      "epoch": 0.8777602523659306,
+      "grad_norm": 0.5372496581928871,
+      "learning_rate": 1.773170219326504e-05,
+      "loss": 0.436,
+      "step": 4452
+    },
+    {
+      "epoch": 0.8779574132492114,
+      "grad_norm": 0.5189362257725963,
+      "learning_rate": 1.7730719076147057e-05,
+      "loss": 0.4433,
+      "step": 4453
+    },
+    {
+      "epoch": 0.8781545741324921,
+      "grad_norm": 0.548560113980181,
+      "learning_rate": 1.7729735773292322e-05,
+      "loss": 0.4596,
+      "step": 4454
+    },
+    {
+      "epoch": 0.8783517350157729,
+      "grad_norm": 0.5494525538977173,
+      "learning_rate": 1.7728752284724454e-05,
+      "loss": 0.4452,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8785488958990536,
+      "grad_norm": 0.5269179148603672,
+      "learning_rate": 1.7727768610467097e-05,
+      "loss": 0.4617,
+      "step": 4456
+    },
+    {
+      "epoch": 0.8787460567823344,
+      "grad_norm": 0.6517832670327971,
+      "learning_rate": 1.7726784750543867e-05,
+      "loss": 0.4377,
+      "step": 4457
+    },
+    {
+      "epoch": 0.8789432176656151,
+      "grad_norm": 0.5812600325676386,
+      "learning_rate": 1.7725800704978416e-05,
+      "loss": 0.4803,
+      "step": 4458
+    },
+    {
+      "epoch": 0.8791403785488959,
+      "grad_norm": 0.5376422098313386,
+      "learning_rate": 1.772481647379438e-05,
+      "loss": 0.4467,
+      "step": 4459
+    },
+    {
+      "epoch": 0.8793375394321766,
+      "grad_norm": 0.5404291511379831,
+      "learning_rate": 1.7723832057015413e-05,
+      "loss": 0.4662,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8795347003154574,
+      "grad_norm": 0.6308719185145436,
+      "learning_rate": 1.7722847454665156e-05,
+      "loss": 0.456,
+      "step": 4461
+    },
+    {
+      "epoch": 0.8797318611987381,
+      "grad_norm": 0.5257173961466224,
+      "learning_rate": 1.772186266676727e-05,
+      "loss": 0.4427,
+      "step": 4462
+    },
+    {
+      "epoch": 0.879929022082019,
+      "grad_norm": 0.6053907583720101,
+      "learning_rate": 1.7720877693345414e-05,
+      "loss": 0.4045,
+      "step": 4463
+    },
+    {
+      "epoch": 0.8801261829652997,
+      "grad_norm": 0.6644310590594147,
+      "learning_rate": 1.7719892534423255e-05,
+      "loss": 0.4652,
+      "step": 4464
+    },
+    {
+      "epoch": 0.8803233438485805,
+      "grad_norm": 0.5704624224624693,
+      "learning_rate": 1.7718907190024462e-05,
+      "loss": 0.4775,
+      "step": 4465
+    },
+    {
+      "epoch": 0.8805205047318612,
+      "grad_norm": 0.5410836564451232,
+      "learning_rate": 1.7717921660172708e-05,
+      "loss": 0.4499,
+      "step": 4466
+    },
+    {
+      "epoch": 0.880717665615142,
+      "grad_norm": 0.5228027698546309,
+      "learning_rate": 1.771693594489167e-05,
+      "loss": 0.4313,
+      "step": 4467
+    },
+    {
+      "epoch": 0.8809148264984227,
+      "grad_norm": 0.5265433161743185,
+      "learning_rate": 1.771595004420503e-05,
+      "loss": 0.404,
+      "step": 4468
+    },
+    {
+      "epoch": 0.8811119873817035,
+      "grad_norm": 0.5812483523416486,
+      "learning_rate": 1.771496395813648e-05,
+      "loss": 0.4359,
+      "step": 4469
+    },
+    {
+      "epoch": 0.8813091482649842,
+      "grad_norm": 0.5446967314373959,
+      "learning_rate": 1.7713977686709706e-05,
+      "loss": 0.4476,
+      "step": 4470
+    },
+    {
+      "epoch": 0.881506309148265,
+      "grad_norm": 0.5031631759781324,
+      "learning_rate": 1.7712991229948405e-05,
+      "loss": 0.3942,
+      "step": 4471
+    },
+    {
+      "epoch": 0.8817034700315457,
+      "grad_norm": 0.5115052245818232,
+      "learning_rate": 1.7712004587876278e-05,
+      "loss": 0.4236,
+      "step": 4472
+    },
+    {
+      "epoch": 0.8819006309148265,
+      "grad_norm": 0.5212693828848795,
+      "learning_rate": 1.7711017760517033e-05,
+      "loss": 0.4075,
+      "step": 4473
+    },
+    {
+      "epoch": 0.8820977917981072,
+      "grad_norm": 0.5593095083849272,
+      "learning_rate": 1.7710030747894375e-05,
+      "loss": 0.4372,
+      "step": 4474
+    },
+    {
+      "epoch": 0.882294952681388,
+      "grad_norm": 0.5103075891986539,
+      "learning_rate": 1.770904355003202e-05,
+      "loss": 0.424,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8824921135646687,
+      "grad_norm": 0.5391202190499896,
+      "learning_rate": 1.7708056166953684e-05,
+      "loss": 0.4861,
+      "step": 4476
+    },
+    {
+      "epoch": 0.8826892744479495,
+      "grad_norm": 0.5138735176424362,
+      "learning_rate": 1.7707068598683095e-05,
+      "loss": 0.4461,
+      "step": 4477
+    },
+    {
+      "epoch": 0.8828864353312302,
+      "grad_norm": 0.5261969184630167,
+      "learning_rate": 1.7706080845243975e-05,
+      "loss": 0.4413,
+      "step": 4478
+    },
+    {
+      "epoch": 0.883083596214511,
+      "grad_norm": 0.5319173341858092,
+      "learning_rate": 1.7705092906660054e-05,
+      "loss": 0.4574,
+      "step": 4479
+    },
+    {
+      "epoch": 0.8832807570977917,
+      "grad_norm": 0.5410201601688506,
+      "learning_rate": 1.7704104782955074e-05,
+      "loss": 0.4396,
+      "step": 4480
+    },
+    {
+      "epoch": 0.8834779179810726,
+      "grad_norm": 0.5346889247807762,
+      "learning_rate": 1.770311647415277e-05,
+      "loss": 0.4465,
+      "step": 4481
+    },
+    {
+      "epoch": 0.8836750788643533,
+      "grad_norm": 0.5216713453775305,
+      "learning_rate": 1.7702127980276893e-05,
+      "loss": 0.439,
+      "step": 4482
+    },
+    {
+      "epoch": 0.8838722397476341,
+      "grad_norm": 0.5794595199940287,
+      "learning_rate": 1.7701139301351187e-05,
+      "loss": 0.4277,
+      "step": 4483
+    },
+    {
+      "epoch": 0.8840694006309149,
+      "grad_norm": 0.5173700020686466,
+      "learning_rate": 1.7700150437399405e-05,
+      "loss": 0.4404,
+      "step": 4484
+    },
+    {
+      "epoch": 0.8842665615141956,
+      "grad_norm": 0.7338285688890706,
+      "learning_rate": 1.7699161388445313e-05,
+      "loss": 0.4936,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8844637223974764,
+      "grad_norm": 0.49830492751234945,
+      "learning_rate": 1.7698172154512666e-05,
+      "loss": 0.4316,
+      "step": 4486
+    },
+    {
+      "epoch": 0.8846608832807571,
+      "grad_norm": 0.5574600419347723,
+      "learning_rate": 1.7697182735625233e-05,
+      "loss": 0.448,
+      "step": 4487
+    },
+    {
+      "epoch": 0.8848580441640379,
+      "grad_norm": 0.5158020909340794,
+      "learning_rate": 1.7696193131806786e-05,
+      "loss": 0.4157,
+      "step": 4488
+    },
+    {
+      "epoch": 0.8850552050473186,
+      "grad_norm": 0.5156523644377932,
+      "learning_rate": 1.76952033430811e-05,
+      "loss": 0.4277,
+      "step": 4489
+    },
+    {
+      "epoch": 0.8852523659305994,
+      "grad_norm": 0.4951200681741904,
+      "learning_rate": 1.769421336947196e-05,
+      "loss": 0.41,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8854495268138801,
+      "grad_norm": 0.5210924234785088,
+      "learning_rate": 1.769322321100315e-05,
+      "loss": 0.4321,
+      "step": 4491
+    },
+    {
+      "epoch": 0.8856466876971609,
+      "grad_norm": 0.5546404645348809,
+      "learning_rate": 1.769223286769845e-05,
+      "loss": 0.46,
+      "step": 4492
+    },
+    {
+      "epoch": 0.8858438485804416,
+      "grad_norm": 0.5116058758291511,
+      "learning_rate": 1.7691242339581664e-05,
+      "loss": 0.4255,
+      "step": 4493
+    },
+    {
+      "epoch": 0.8860410094637224,
+      "grad_norm": 0.5259991490113677,
+      "learning_rate": 1.769025162667659e-05,
+      "loss": 0.4601,
+      "step": 4494
+    },
+    {
+      "epoch": 0.8862381703470031,
+      "grad_norm": 0.6096603798487029,
+      "learning_rate": 1.7689260729007025e-05,
+      "loss": 0.4446,
+      "step": 4495
+    },
+    {
+      "epoch": 0.886435331230284,
+      "grad_norm": 0.5502457929004166,
+      "learning_rate": 1.768826964659678e-05,
+      "loss": 0.4437,
+      "step": 4496
+    },
+    {
+      "epoch": 0.8866324921135647,
+      "grad_norm": 0.4955351951661073,
+      "learning_rate": 1.7687278379469665e-05,
+      "loss": 0.4367,
+      "step": 4497
+    },
+    {
+      "epoch": 0.8868296529968455,
+      "grad_norm": 0.48794238779718313,
+      "learning_rate": 1.7686286927649493e-05,
+      "loss": 0.3898,
+      "step": 4498
+    },
+    {
+      "epoch": 0.8870268138801262,
+      "grad_norm": 0.5403525717354225,
+      "learning_rate": 1.768529529116009e-05,
+      "loss": 0.4431,
+      "step": 4499
+    },
+    {
+      "epoch": 0.887223974763407,
+      "grad_norm": 0.5376973399491151,
+      "learning_rate": 1.768430347002528e-05,
+      "loss": 0.444,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8874211356466877,
+      "grad_norm": 0.5272976538103022,
+      "learning_rate": 1.768331146426889e-05,
+      "loss": 0.4236,
+      "step": 4501
+    },
+    {
+      "epoch": 0.8876182965299685,
+      "grad_norm": 0.5344980125641151,
+      "learning_rate": 1.7682319273914755e-05,
+      "loss": 0.4426,
+      "step": 4502
+    },
+    {
+      "epoch": 0.8878154574132492,
+      "grad_norm": 0.49953537905007334,
+      "learning_rate": 1.7681326898986713e-05,
+      "loss": 0.4209,
+      "step": 4503
+    },
+    {
+      "epoch": 0.88801261829653,
+      "grad_norm": 0.5371135502177788,
+      "learning_rate": 1.7680334339508604e-05,
+      "loss": 0.4334,
+      "step": 4504
+    },
+    {
+      "epoch": 0.8882097791798107,
+      "grad_norm": 0.5315742083699853,
+      "learning_rate": 1.767934159550428e-05,
+      "loss": 0.4259,
+      "step": 4505
+    },
+    {
+      "epoch": 0.8884069400630915,
+      "grad_norm": 0.5965034470720473,
+      "learning_rate": 1.767834866699759e-05,
+      "loss": 0.4426,
+      "step": 4506
+    },
+    {
+      "epoch": 0.8886041009463722,
+      "grad_norm": 0.5836671704413614,
+      "learning_rate": 1.767735555401239e-05,
+      "loss": 0.4287,
+      "step": 4507
+    },
+    {
+      "epoch": 0.888801261829653,
+      "grad_norm": 0.5625765431157921,
+      "learning_rate": 1.767636225657254e-05,
+      "loss": 0.4459,
+      "step": 4508
+    },
+    {
+      "epoch": 0.8889984227129337,
+      "grad_norm": 0.7370740412563955,
+      "learning_rate": 1.7675368774701906e-05,
+      "loss": 0.426,
+      "step": 4509
+    },
+    {
+      "epoch": 0.8891955835962145,
+      "grad_norm": 0.5477180104542236,
+      "learning_rate": 1.7674375108424354e-05,
+      "loss": 0.4451,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8893927444794952,
+      "grad_norm": 0.5330930261842161,
+      "learning_rate": 1.767338125776376e-05,
+      "loss": 0.3926,
+      "step": 4511
+    },
+    {
+      "epoch": 0.889589905362776,
+      "grad_norm": 0.5310418495745474,
+      "learning_rate": 1.7672387222744e-05,
+      "loss": 0.4434,
+      "step": 4512
+    },
+    {
+      "epoch": 0.8897870662460567,
+      "grad_norm": 0.5341421770335281,
+      "learning_rate": 1.7671393003388964e-05,
+      "loss": 0.4476,
+      "step": 4513
+    },
+    {
+      "epoch": 0.8899842271293376,
+      "grad_norm": 0.50982098147757,
+      "learning_rate": 1.7670398599722533e-05,
+      "loss": 0.4392,
+      "step": 4514
+    },
+    {
+      "epoch": 0.8901813880126183,
+      "grad_norm": 0.6050964967523266,
+      "learning_rate": 1.7669404011768596e-05,
+      "loss": 0.466,
+      "step": 4515
+    },
+    {
+      "epoch": 0.8903785488958991,
+      "grad_norm": 0.5082170155704829,
+      "learning_rate": 1.766840923955105e-05,
+      "loss": 0.3896,
+      "step": 4516
+    },
+    {
+      "epoch": 0.8905757097791798,
+      "grad_norm": 0.542583849270664,
+      "learning_rate": 1.76674142830938e-05,
+      "loss": 0.4499,
+      "step": 4517
+    },
+    {
+      "epoch": 0.8907728706624606,
+      "grad_norm": 0.549649515755227,
+      "learning_rate": 1.7666419142420746e-05,
+      "loss": 0.4205,
+      "step": 4518
+    },
+    {
+      "epoch": 0.8909700315457413,
+      "grad_norm": 0.5575951769960029,
+      "learning_rate": 1.76654238175558e-05,
+      "loss": 0.4465,
+      "step": 4519
+    },
+    {
+      "epoch": 0.8911671924290221,
+      "grad_norm": 0.5453782064529624,
+      "learning_rate": 1.766442830852287e-05,
+      "loss": 0.4769,
+      "step": 4520
+    },
+    {
+      "epoch": 0.8913643533123028,
+      "grad_norm": 0.5101131437540427,
+      "learning_rate": 1.766343261534588e-05,
+      "loss": 0.4017,
+      "step": 4521
+    },
+    {
+      "epoch": 0.8915615141955836,
+      "grad_norm": 0.6136158485523702,
+      "learning_rate": 1.766243673804875e-05,
+      "loss": 0.4098,
+      "step": 4522
+    },
+    {
+      "epoch": 0.8917586750788643,
+      "grad_norm": 0.5292957632786092,
+      "learning_rate": 1.7661440676655407e-05,
+      "loss": 0.4223,
+      "step": 4523
+    },
+    {
+      "epoch": 0.8919558359621451,
+      "grad_norm": 0.5540759577823543,
+      "learning_rate": 1.766044443118978e-05,
+      "loss": 0.4593,
+      "step": 4524
+    },
+    {
+      "epoch": 0.8921529968454258,
+      "grad_norm": 0.5392743345482487,
+      "learning_rate": 1.765944800167581e-05,
+      "loss": 0.4235,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8923501577287066,
+      "grad_norm": 0.6346400393371832,
+      "learning_rate": 1.7658451388137432e-05,
+      "loss": 0.4231,
+      "step": 4526
+    },
+    {
+      "epoch": 0.8925473186119873,
+      "grad_norm": 0.5201806359082028,
+      "learning_rate": 1.7657454590598594e-05,
+      "loss": 0.4375,
+      "step": 4527
+    },
+    {
+      "epoch": 0.8927444794952681,
+      "grad_norm": 0.5175630311254862,
+      "learning_rate": 1.765645760908324e-05,
+      "loss": 0.4408,
+      "step": 4528
+    },
+    {
+      "epoch": 0.892941640378549,
+      "grad_norm": 0.5592078904199146,
+      "learning_rate": 1.7655460443615327e-05,
+      "loss": 0.4714,
+      "step": 4529
+    },
+    {
+      "epoch": 0.8931388012618297,
+      "grad_norm": 0.595148154874751,
+      "learning_rate": 1.7654463094218813e-05,
+      "loss": 0.4265,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8933359621451105,
+      "grad_norm": 0.5367589779182992,
+      "learning_rate": 1.7653465560917656e-05,
+      "loss": 0.4664,
+      "step": 4531
+    },
+    {
+      "epoch": 0.8935331230283912,
+      "grad_norm": 0.5404275654069878,
+      "learning_rate": 1.7652467843735828e-05,
+      "loss": 0.4304,
+      "step": 4532
+    },
+    {
+      "epoch": 0.893730283911672,
+      "grad_norm": 0.5426667005563901,
+      "learning_rate": 1.7651469942697296e-05,
+      "loss": 0.4187,
+      "step": 4533
+    },
+    {
+      "epoch": 0.8939274447949527,
+      "grad_norm": 0.5355183086191119,
+      "learning_rate": 1.7650471857826038e-05,
+      "loss": 0.42,
+      "step": 4534
+    },
+    {
+      "epoch": 0.8941246056782335,
+      "grad_norm": 0.5434987852881384,
+      "learning_rate": 1.7649473589146032e-05,
+      "loss": 0.4129,
+      "step": 4535
+    },
+    {
+      "epoch": 0.8943217665615142,
+      "grad_norm": 0.49571415053957635,
+      "learning_rate": 1.7648475136681265e-05,
+      "loss": 0.4332,
+      "step": 4536
+    },
+    {
+      "epoch": 0.894518927444795,
+      "grad_norm": 0.545085011797932,
+      "learning_rate": 1.7647476500455723e-05,
+      "loss": 0.4722,
+      "step": 4537
+    },
+    {
+      "epoch": 0.8947160883280757,
+      "grad_norm": 0.5458717696446517,
+      "learning_rate": 1.76464776804934e-05,
+      "loss": 0.4437,
+      "step": 4538
+    },
+    {
+      "epoch": 0.8949132492113565,
+      "grad_norm": 0.5198965233821335,
+      "learning_rate": 1.764547867681829e-05,
+      "loss": 0.4335,
+      "step": 4539
+    },
+    {
+      "epoch": 0.8951104100946372,
+      "grad_norm": 0.5098927604754595,
+      "learning_rate": 1.76444794894544e-05,
+      "loss": 0.4281,
+      "step": 4540
+    },
+    {
+      "epoch": 0.895307570977918,
+      "grad_norm": 0.5476572140939739,
+      "learning_rate": 1.7643480118425733e-05,
+      "loss": 0.4575,
+      "step": 4541
+    },
+    {
+      "epoch": 0.8955047318611987,
+      "grad_norm": 0.5506996953424919,
+      "learning_rate": 1.7642480563756305e-05,
+      "loss": 0.4308,
+      "step": 4542
+    },
+    {
+      "epoch": 0.8957018927444795,
+      "grad_norm": 0.5418834197677749,
+      "learning_rate": 1.7641480825470123e-05,
+      "loss": 0.3992,
+      "step": 4543
+    },
+    {
+      "epoch": 0.8958990536277602,
+      "grad_norm": 0.5330741568842563,
+      "learning_rate": 1.764048090359121e-05,
+      "loss": 0.4314,
+      "step": 4544
+    },
+    {
+      "epoch": 0.896096214511041,
+      "grad_norm": 0.5364576977743398,
+      "learning_rate": 1.7639480798143593e-05,
+      "loss": 0.4311,
+      "step": 4545
+    },
+    {
+      "epoch": 0.8962933753943217,
+      "grad_norm": 0.5244339795995854,
+      "learning_rate": 1.7638480509151297e-05,
+      "loss": 0.4576,
+      "step": 4546
+    },
+    {
+      "epoch": 0.8964905362776026,
+      "grad_norm": 0.563203046041122,
+      "learning_rate": 1.7637480036638356e-05,
+      "loss": 0.4828,
+      "step": 4547
+    },
+    {
+      "epoch": 0.8966876971608833,
+      "grad_norm": 0.5070783658444626,
+      "learning_rate": 1.7636479380628806e-05,
+      "loss": 0.4226,
+      "step": 4548
+    },
+    {
+      "epoch": 0.8968848580441641,
+      "grad_norm": 0.5595950137316498,
+      "learning_rate": 1.7635478541146687e-05,
+      "loss": 0.459,
+      "step": 4549
+    },
+    {
+      "epoch": 0.8970820189274448,
+      "grad_norm": 0.5564160601048573,
+      "learning_rate": 1.763447751821605e-05,
+      "loss": 0.4069,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8972791798107256,
+      "grad_norm": 0.5397157852142757,
+      "learning_rate": 1.763347631186094e-05,
+      "loss": 0.4336,
+      "step": 4551
+    },
+    {
+      "epoch": 0.8974763406940063,
+      "grad_norm": 0.5028947071543393,
+      "learning_rate": 1.7632474922105416e-05,
+      "loss": 0.4111,
+      "step": 4552
+    },
+    {
+      "epoch": 0.8976735015772871,
+      "grad_norm": 0.6653704389356556,
+      "learning_rate": 1.7631473348973537e-05,
+      "loss": 0.4817,
+      "step": 4553
+    },
+    {
+      "epoch": 0.8978706624605678,
+      "grad_norm": 0.6954841719181198,
+      "learning_rate": 1.7630471592489366e-05,
+      "loss": 0.4422,
+      "step": 4554
+    },
+    {
+      "epoch": 0.8980678233438486,
+      "grad_norm": 0.5904197333916205,
+      "learning_rate": 1.7629469652676965e-05,
+      "loss": 0.4538,
+      "step": 4555
+    },
+    {
+      "epoch": 0.8982649842271293,
+      "grad_norm": 1.260867911052856,
+      "learning_rate": 1.7628467529560417e-05,
+      "loss": 0.4514,
+      "step": 4556
+    },
+    {
+      "epoch": 0.8984621451104101,
+      "grad_norm": 1.0289033380566197,
+      "learning_rate": 1.762746522316379e-05,
+      "loss": 0.4128,
+      "step": 4557
+    },
+    {
+      "epoch": 0.8986593059936908,
+      "grad_norm": 0.5244592891802363,
+      "learning_rate": 1.762646273351117e-05,
+      "loss": 0.4338,
+      "step": 4558
+    },
+    {
+      "epoch": 0.8988564668769716,
+      "grad_norm": 0.5044945072717523,
+      "learning_rate": 1.7625460060626644e-05,
+      "loss": 0.4117,
+      "step": 4559
+    },
+    {
+      "epoch": 0.8990536277602523,
+      "grad_norm": 0.7264810983583718,
+      "learning_rate": 1.7624457204534292e-05,
+      "loss": 0.4583,
+      "step": 4560
+    },
+    {
+      "epoch": 0.8992507886435331,
+      "grad_norm": 0.5588409715862622,
+      "learning_rate": 1.762345416525822e-05,
+      "loss": 0.4361,
+      "step": 4561
+    },
+    {
+      "epoch": 0.8994479495268138,
+      "grad_norm": 0.4900029305965899,
+      "learning_rate": 1.7622450942822524e-05,
+      "loss": 0.4037,
+      "step": 4562
+    },
+    {
+      "epoch": 0.8996451104100947,
+      "grad_norm": 0.6753821625040557,
+      "learning_rate": 1.7621447537251307e-05,
+      "loss": 0.4286,
+      "step": 4563
+    },
+    {
+      "epoch": 0.8998422712933754,
+      "grad_norm": 0.5928729012070617,
+      "learning_rate": 1.762044394856867e-05,
+      "loss": 0.4391,
+      "step": 4564
+    },
+    {
+      "epoch": 0.9000394321766562,
+      "grad_norm": 0.5403084220268555,
+      "learning_rate": 1.7619440176798733e-05,
+      "loss": 0.4305,
+      "step": 4565
+    },
+    {
+      "epoch": 0.9002365930599369,
+      "grad_norm": 0.5023737487603344,
+      "learning_rate": 1.761843622196561e-05,
+      "loss": 0.432,
+      "step": 4566
+    },
+    {
+      "epoch": 0.9004337539432177,
+      "grad_norm": 0.8214257640360723,
+      "learning_rate": 1.7617432084093424e-05,
+      "loss": 0.4732,
+      "step": 4567
+    },
+    {
+      "epoch": 0.9006309148264984,
+      "grad_norm": 0.6157040111093859,
+      "learning_rate": 1.7616427763206294e-05,
+      "loss": 0.4106,
+      "step": 4568
+    },
+    {
+      "epoch": 0.9008280757097792,
+      "grad_norm": 0.6386520150246892,
+      "learning_rate": 1.7615423259328356e-05,
+      "loss": 0.4317,
+      "step": 4569
+    },
+    {
+      "epoch": 0.9010252365930599,
+      "grad_norm": 0.5233550893573012,
+      "learning_rate": 1.761441857248374e-05,
+      "loss": 0.428,
+      "step": 4570
+    },
+    {
+      "epoch": 0.9012223974763407,
+      "grad_norm": 0.5816027891775581,
+      "learning_rate": 1.7613413702696584e-05,
+      "loss": 0.4624,
+      "step": 4571
+    },
+    {
+      "epoch": 0.9014195583596214,
+      "grad_norm": 0.5226310374487503,
+      "learning_rate": 1.7612408649991037e-05,
+      "loss": 0.4252,
+      "step": 4572
+    },
+    {
+      "epoch": 0.9016167192429022,
+      "grad_norm": 0.7039706825623913,
+      "learning_rate": 1.761140341439124e-05,
+      "loss": 0.4465,
+      "step": 4573
+    },
+    {
+      "epoch": 0.901813880126183,
+      "grad_norm": 0.6335577419820779,
+      "learning_rate": 1.7610397995921348e-05,
+      "loss": 0.4193,
+      "step": 4574
+    },
+    {
+      "epoch": 0.9020110410094637,
+      "grad_norm": 0.5647597040461998,
+      "learning_rate": 1.760939239460551e-05,
+      "loss": 0.4354,
+      "step": 4575
+    },
+    {
+      "epoch": 0.9022082018927445,
+      "grad_norm": 0.5628152685174825,
+      "learning_rate": 1.7608386610467898e-05,
+      "loss": 0.4757,
+      "step": 4576
+    },
+    {
+      "epoch": 0.9024053627760252,
+      "grad_norm": 3.2929032363930917,
+      "learning_rate": 1.7607380643532667e-05,
+      "loss": 0.4282,
+      "step": 4577
+    },
+    {
+      "epoch": 0.902602523659306,
+      "grad_norm": 0.7504633569369734,
+      "learning_rate": 1.7606374493823993e-05,
+      "loss": 0.4495,
+      "step": 4578
+    },
+    {
+      "epoch": 0.9027996845425867,
+      "grad_norm": 0.5656614412611999,
+      "learning_rate": 1.7605368161366043e-05,
+      "loss": 0.4242,
+      "step": 4579
+    },
+    {
+      "epoch": 0.9029968454258676,
+      "grad_norm": 0.5437572557240907,
+      "learning_rate": 1.7604361646183004e-05,
+      "loss": 0.4365,
+      "step": 4580
+    },
+    {
+      "epoch": 0.9031940063091483,
+      "grad_norm": 0.5704886873684303,
+      "learning_rate": 1.760335494829905e-05,
+      "loss": 0.4378,
+      "step": 4581
+    },
+    {
+      "epoch": 0.9033911671924291,
+      "grad_norm": 0.5670944223065693,
+      "learning_rate": 1.7602348067738367e-05,
+      "loss": 0.4222,
+      "step": 4582
+    },
+    {
+      "epoch": 0.9035883280757098,
+      "grad_norm": 0.571465040387628,
+      "learning_rate": 1.760134100452515e-05,
+      "loss": 0.4459,
+      "step": 4583
+    },
+    {
+      "epoch": 0.9037854889589906,
+      "grad_norm": 0.5411774398722019,
+      "learning_rate": 1.7600333758683598e-05,
+      "loss": 0.4373,
+      "step": 4584
+    },
+    {
+      "epoch": 0.9039826498422713,
+      "grad_norm": 0.5164990496084579,
+      "learning_rate": 1.7599326330237906e-05,
+      "loss": 0.3954,
+      "step": 4585
+    },
+    {
+      "epoch": 0.9041798107255521,
+      "grad_norm": 0.5954434300630094,
+      "learning_rate": 1.7598318719212274e-05,
+      "loss": 0.4541,
+      "step": 4586
+    },
+    {
+      "epoch": 0.9043769716088328,
+      "grad_norm": 0.5629588725994905,
+      "learning_rate": 1.7597310925630922e-05,
+      "loss": 0.4696,
+      "step": 4587
+    },
+    {
+      "epoch": 0.9045741324921136,
+      "grad_norm": 1.0533510364008871,
+      "learning_rate": 1.7596302949518054e-05,
+      "loss": 0.4087,
+      "step": 4588
+    },
+    {
+      "epoch": 0.9047712933753943,
+      "grad_norm": 0.5848783608010563,
+      "learning_rate": 1.759529479089789e-05,
+      "loss": 0.4562,
+      "step": 4589
+    },
+    {
+      "epoch": 0.9049684542586751,
+      "grad_norm": 0.5359425519217931,
+      "learning_rate": 1.7594286449794655e-05,
+      "loss": 0.4139,
+      "step": 4590
+    },
+    {
+      "epoch": 0.9051656151419558,
+      "grad_norm": 0.5766734785111527,
+      "learning_rate": 1.759327792623257e-05,
+      "loss": 0.4232,
+      "step": 4591
+    },
+    {
+      "epoch": 0.9053627760252366,
+      "grad_norm": 0.5120340477249785,
+      "learning_rate": 1.759226922023587e-05,
+      "loss": 0.4266,
+      "step": 4592
+    },
+    {
+      "epoch": 0.9055599369085173,
+      "grad_norm": 0.5398891426274068,
+      "learning_rate": 1.7591260331828785e-05,
+      "loss": 0.4179,
+      "step": 4593
+    },
+    {
+      "epoch": 0.9057570977917981,
+      "grad_norm": 0.5184539964235813,
+      "learning_rate": 1.759025126103556e-05,
+      "loss": 0.4125,
+      "step": 4594
+    },
+    {
+      "epoch": 0.9059542586750788,
+      "grad_norm": 0.6029467871180186,
+      "learning_rate": 1.7589242007880435e-05,
+      "loss": 0.4991,
+      "step": 4595
+    },
+    {
+      "epoch": 0.9061514195583596,
+      "grad_norm": 0.5757959829050494,
+      "learning_rate": 1.7588232572387657e-05,
+      "loss": 0.4294,
+      "step": 4596
+    },
+    {
+      "epoch": 0.9063485804416404,
+      "grad_norm": 0.5612525679026711,
+      "learning_rate": 1.7587222954581483e-05,
+      "loss": 0.4779,
+      "step": 4597
+    },
+    {
+      "epoch": 0.9065457413249212,
+      "grad_norm": 0.5047317331482842,
+      "learning_rate": 1.758621315448617e-05,
+      "loss": 0.3938,
+      "step": 4598
+    },
+    {
+      "epoch": 0.9067429022082019,
+      "grad_norm": 0.5667514613061159,
+      "learning_rate": 1.7585203172125972e-05,
+      "loss": 0.4599,
+      "step": 4599
+    },
+    {
+      "epoch": 0.9069400630914827,
+      "grad_norm": 0.5293002056349737,
+      "learning_rate": 1.758419300752516e-05,
+      "loss": 0.4219,
+      "step": 4600
+    },
+    {
+      "epoch": 0.9071372239747634,
+      "grad_norm": 0.5080066799539775,
+      "learning_rate": 1.758318266070801e-05,
+      "loss": 0.3933,
+      "step": 4601
+    },
+    {
+      "epoch": 0.9073343848580442,
+      "grad_norm": 0.5576817270172586,
+      "learning_rate": 1.758217213169878e-05,
+      "loss": 0.4389,
+      "step": 4602
+    },
+    {
+      "epoch": 0.9075315457413249,
+      "grad_norm": 0.5264486271047948,
+      "learning_rate": 1.7581161420521765e-05,
+      "loss": 0.4522,
+      "step": 4603
+    },
+    {
+      "epoch": 0.9077287066246057,
+      "grad_norm": 0.5894471822557419,
+      "learning_rate": 1.758015052720124e-05,
+      "loss": 0.4453,
+      "step": 4604
+    },
+    {
+      "epoch": 0.9079258675078864,
+      "grad_norm": 0.5534225411257014,
+      "learning_rate": 1.7579139451761495e-05,
+      "loss": 0.4532,
+      "step": 4605
+    },
+    {
+      "epoch": 0.9081230283911672,
+      "grad_norm": 0.5369235116741078,
+      "learning_rate": 1.7578128194226823e-05,
+      "loss": 0.4251,
+      "step": 4606
+    },
+    {
+      "epoch": 0.9083201892744479,
+      "grad_norm": 0.4830071277784729,
+      "learning_rate": 1.7577116754621512e-05,
+      "loss": 0.4189,
+      "step": 4607
+    },
+    {
+      "epoch": 0.9085173501577287,
+      "grad_norm": 0.7383285750265494,
+      "learning_rate": 1.7576105132969874e-05,
+      "loss": 0.4759,
+      "step": 4608
+    },
+    {
+      "epoch": 0.9087145110410094,
+      "grad_norm": 0.5395821153111766,
+      "learning_rate": 1.757509332929621e-05,
+      "loss": 0.4595,
+      "step": 4609
+    },
+    {
+      "epoch": 0.9089116719242902,
+      "grad_norm": 0.5971105354676437,
+      "learning_rate": 1.7574081343624827e-05,
+      "loss": 0.4501,
+      "step": 4610
+    },
+    {
+      "epoch": 0.9091088328075709,
+      "grad_norm": 0.5322422724423118,
+      "learning_rate": 1.757306917598004e-05,
+      "loss": 0.4432,
+      "step": 4611
+    },
+    {
+      "epoch": 0.9093059936908517,
+      "grad_norm": 0.6382374433003123,
+      "learning_rate": 1.7572056826386167e-05,
+      "loss": 0.4918,
+      "step": 4612
+    },
+    {
+      "epoch": 0.9095031545741324,
+      "grad_norm": 1.0874753951271274,
+      "learning_rate": 1.7571044294867533e-05,
+      "loss": 0.4519,
+      "step": 4613
+    },
+    {
+      "epoch": 0.9097003154574133,
+      "grad_norm": 1.0844704783301853,
+      "learning_rate": 1.757003158144846e-05,
+      "loss": 0.4352,
+      "step": 4614
+    },
+    {
+      "epoch": 0.909897476340694,
+      "grad_norm": 0.522978838373464,
+      "learning_rate": 1.7569018686153286e-05,
+      "loss": 0.4233,
+      "step": 4615
+    },
+    {
+      "epoch": 0.9100946372239748,
+      "grad_norm": 0.5933248972486103,
+      "learning_rate": 1.756800560900634e-05,
+      "loss": 0.4419,
+      "step": 4616
+    },
+    {
+      "epoch": 0.9102917981072555,
+      "grad_norm": 1.216489202161031,
+      "learning_rate": 1.7566992350031965e-05,
+      "loss": 0.4741,
+      "step": 4617
+    },
+    {
+      "epoch": 0.9104889589905363,
+      "grad_norm": 0.5819684520560912,
+      "learning_rate": 1.7565978909254508e-05,
+      "loss": 0.4278,
+      "step": 4618
+    },
+    {
+      "epoch": 0.910686119873817,
+      "grad_norm": 0.580755399529423,
+      "learning_rate": 1.756496528669831e-05,
+      "loss": 0.4708,
+      "step": 4619
+    },
+    {
+      "epoch": 0.9108832807570978,
+      "grad_norm": 0.5363608983821153,
+      "learning_rate": 1.7563951482387733e-05,
+      "loss": 0.4008,
+      "step": 4620
+    },
+    {
+      "epoch": 0.9110804416403786,
+      "grad_norm": 0.5949632674163584,
+      "learning_rate": 1.7562937496347126e-05,
+      "loss": 0.4663,
+      "step": 4621
+    },
+    {
+      "epoch": 0.9112776025236593,
+      "grad_norm": 0.7340546915226606,
+      "learning_rate": 1.756192332860086e-05,
+      "loss": 0.4453,
+      "step": 4622
+    },
+    {
+      "epoch": 0.9114747634069401,
+      "grad_norm": 0.5539458382888915,
+      "learning_rate": 1.7560908979173294e-05,
+      "loss": 0.4457,
+      "step": 4623
+    },
+    {
+      "epoch": 0.9116719242902208,
+      "grad_norm": 0.5614114939363166,
+      "learning_rate": 1.7559894448088802e-05,
+      "loss": 0.4023,
+      "step": 4624
+    },
+    {
+      "epoch": 0.9118690851735016,
+      "grad_norm": 0.5710583292679061,
+      "learning_rate": 1.7558879735371753e-05,
+      "loss": 0.4688,
+      "step": 4625
+    },
+    {
+      "epoch": 0.9120662460567823,
+      "grad_norm": 0.5550240437903982,
+      "learning_rate": 1.755786484104654e-05,
+      "loss": 0.3906,
+      "step": 4626
+    },
+    {
+      "epoch": 0.9122634069400631,
+      "grad_norm": 0.4837275585777602,
+      "learning_rate": 1.755684976513753e-05,
+      "loss": 0.3919,
+      "step": 4627
+    },
+    {
+      "epoch": 0.9124605678233438,
+      "grad_norm": 0.5844614371573156,
+      "learning_rate": 1.7555834507669124e-05,
+      "loss": 0.4086,
+      "step": 4628
+    },
+    {
+      "epoch": 0.9126577287066246,
+      "grad_norm": 0.5180930399677969,
+      "learning_rate": 1.7554819068665707e-05,
+      "loss": 0.4039,
+      "step": 4629
+    },
+    {
+      "epoch": 0.9128548895899053,
+      "grad_norm": 0.5819268556226904,
+      "learning_rate": 1.7553803448151678e-05,
+      "loss": 0.4102,
+      "step": 4630
+    },
+    {
+      "epoch": 0.9130520504731862,
+      "grad_norm": 0.5158313726358502,
+      "learning_rate": 1.755278764615144e-05,
+      "loss": 0.4266,
+      "step": 4631
+    },
+    {
+      "epoch": 0.9132492113564669,
+      "grad_norm": 0.5930149183798764,
+      "learning_rate": 1.7551771662689393e-05,
+      "loss": 0.4468,
+      "step": 4632
+    },
+    {
+      "epoch": 0.9134463722397477,
+      "grad_norm": 0.5517958419973356,
+      "learning_rate": 1.7550755497789955e-05,
+      "loss": 0.4381,
+      "step": 4633
+    },
+    {
+      "epoch": 0.9136435331230284,
+      "grad_norm": 0.6790482699026756,
+      "learning_rate": 1.754973915147753e-05,
+      "loss": 0.4301,
+      "step": 4634
+    },
+    {
+      "epoch": 0.9138406940063092,
+      "grad_norm": 0.5451328047356113,
+      "learning_rate": 1.7548722623776547e-05,
+      "loss": 0.455,
+      "step": 4635
+    },
+    {
+      "epoch": 0.9140378548895899,
+      "grad_norm": 0.5153803270688035,
+      "learning_rate": 1.754770591471142e-05,
+      "loss": 0.4178,
+      "step": 4636
+    },
+    {
+      "epoch": 0.9142350157728707,
+      "grad_norm": 0.5355618034950106,
+      "learning_rate": 1.7546689024306585e-05,
+      "loss": 0.4328,
+      "step": 4637
+    },
+    {
+      "epoch": 0.9144321766561514,
+      "grad_norm": 0.5793723146542517,
+      "learning_rate": 1.7545671952586464e-05,
+      "loss": 0.4498,
+      "step": 4638
+    },
+    {
+      "epoch": 0.9146293375394322,
+      "grad_norm": 0.49690302528171043,
+      "learning_rate": 1.75446546995755e-05,
+      "loss": 0.4184,
+      "step": 4639
+    },
+    {
+      "epoch": 0.9148264984227129,
+      "grad_norm": 0.5093059262044622,
+      "learning_rate": 1.7543637265298136e-05,
+      "loss": 0.3955,
+      "step": 4640
+    },
+    {
+      "epoch": 0.9150236593059937,
+      "grad_norm": 0.5592098323156065,
+      "learning_rate": 1.7542619649778804e-05,
+      "loss": 0.4376,
+      "step": 4641
+    },
+    {
+      "epoch": 0.9152208201892744,
+      "grad_norm": 0.5525412638322774,
+      "learning_rate": 1.7541601853041963e-05,
+      "loss": 0.4554,
+      "step": 4642
+    },
+    {
+      "epoch": 0.9154179810725552,
+      "grad_norm": 0.5262471397287505,
+      "learning_rate": 1.7540583875112065e-05,
+      "loss": 0.4337,
+      "step": 4643
+    },
+    {
+      "epoch": 0.9156151419558359,
+      "grad_norm": 0.5235757919665939,
+      "learning_rate": 1.753956571601357e-05,
+      "loss": 0.4335,
+      "step": 4644
+    },
+    {
+      "epoch": 0.9158123028391167,
+      "grad_norm": 0.5494919507343168,
+      "learning_rate": 1.7538547375770934e-05,
+      "loss": 0.4198,
+      "step": 4645
+    },
+    {
+      "epoch": 0.9160094637223974,
+      "grad_norm": 0.5155811541332505,
+      "learning_rate": 1.7537528854408625e-05,
+      "loss": 0.4006,
+      "step": 4646
+    },
+    {
+      "epoch": 0.9162066246056783,
+      "grad_norm": 0.619279355812408,
+      "learning_rate": 1.753651015195112e-05,
+      "loss": 0.4524,
+      "step": 4647
+    },
+    {
+      "epoch": 0.916403785488959,
+      "grad_norm": 0.5234709511842155,
+      "learning_rate": 1.7535491268422885e-05,
+      "loss": 0.4365,
+      "step": 4648
+    },
+    {
+      "epoch": 0.9166009463722398,
+      "grad_norm": 0.5760967298774721,
+      "learning_rate": 1.7534472203848402e-05,
+      "loss": 0.421,
+      "step": 4649
+    },
+    {
+      "epoch": 0.9167981072555205,
+      "grad_norm": 0.5744984868953806,
+      "learning_rate": 1.7533452958252164e-05,
+      "loss": 0.4393,
+      "step": 4650
+    },
+    {
+      "epoch": 0.9169952681388013,
+      "grad_norm": 0.5325399090464799,
+      "learning_rate": 1.7532433531658646e-05,
+      "loss": 0.4407,
+      "step": 4651
+    },
+    {
+      "epoch": 0.917192429022082,
+      "grad_norm": 0.5340699591886052,
+      "learning_rate": 1.7531413924092347e-05,
+      "loss": 0.4295,
+      "step": 4652
+    },
+    {
+      "epoch": 0.9173895899053628,
+      "grad_norm": 0.5494041102557055,
+      "learning_rate": 1.7530394135577768e-05,
+      "loss": 0.4489,
+      "step": 4653
+    },
+    {
+      "epoch": 0.9175867507886435,
+      "grad_norm": 0.5182867692860177,
+      "learning_rate": 1.75293741661394e-05,
+      "loss": 0.4204,
+      "step": 4654
+    },
+    {
+      "epoch": 0.9177839116719243,
+      "grad_norm": 0.5728288091111668,
+      "learning_rate": 1.752835401580176e-05,
+      "loss": 0.4238,
+      "step": 4655
+    },
+    {
+      "epoch": 0.917981072555205,
+      "grad_norm": 0.49589426702980033,
+      "learning_rate": 1.752733368458935e-05,
+      "loss": 0.4539,
+      "step": 4656
+    },
+    {
+      "epoch": 0.9181782334384858,
+      "grad_norm": 0.48458956489019156,
+      "learning_rate": 1.7526313172526687e-05,
+      "loss": 0.3936,
+      "step": 4657
+    },
+    {
+      "epoch": 0.9183753943217665,
+      "grad_norm": 0.5274678434361841,
+      "learning_rate": 1.7525292479638286e-05,
+      "loss": 0.4484,
+      "step": 4658
+    },
+    {
+      "epoch": 0.9185725552050473,
+      "grad_norm": 0.5211344047356368,
+      "learning_rate": 1.7524271605948676e-05,
+      "loss": 0.4461,
+      "step": 4659
+    },
+    {
+      "epoch": 0.918769716088328,
+      "grad_norm": 0.511296727944466,
+      "learning_rate": 1.752325055148238e-05,
+      "loss": 0.4185,
+      "step": 4660
+    },
+    {
+      "epoch": 0.9189668769716088,
+      "grad_norm": 0.5647095561867232,
+      "learning_rate": 1.752222931626393e-05,
+      "loss": 0.4473,
+      "step": 4661
+    },
+    {
+      "epoch": 0.9191640378548895,
+      "grad_norm": 0.513631356257729,
+      "learning_rate": 1.7521207900317866e-05,
+      "loss": 0.4403,
+      "step": 4662
+    },
+    {
+      "epoch": 0.9193611987381703,
+      "grad_norm": 0.5155278543678777,
+      "learning_rate": 1.7520186303668722e-05,
+      "loss": 0.4393,
+      "step": 4663
+    },
+    {
+      "epoch": 0.919558359621451,
+      "grad_norm": 0.5019934238237479,
+      "learning_rate": 1.751916452634105e-05,
+      "loss": 0.4006,
+      "step": 4664
+    },
+    {
+      "epoch": 0.9197555205047319,
+      "grad_norm": 0.5133536326390901,
+      "learning_rate": 1.7518142568359395e-05,
+      "loss": 0.4422,
+      "step": 4665
+    },
+    {
+      "epoch": 0.9199526813880127,
+      "grad_norm": 1.1425170395813518,
+      "learning_rate": 1.7517120429748305e-05,
+      "loss": 0.4254,
+      "step": 4666
+    },
+    {
+      "epoch": 0.9201498422712934,
+      "grad_norm": 0.5062808556522904,
+      "learning_rate": 1.751609811053235e-05,
+      "loss": 0.4093,
+      "step": 4667
+    },
+    {
+      "epoch": 0.9203470031545742,
+      "grad_norm": 0.5329727124070021,
+      "learning_rate": 1.7515075610736077e-05,
+      "loss": 0.445,
+      "step": 4668
+    },
+    {
+      "epoch": 0.9205441640378549,
+      "grad_norm": 0.5773802180358756,
+      "learning_rate": 1.751405293038407e-05,
+      "loss": 0.4348,
+      "step": 4669
+    },
+    {
+      "epoch": 0.9207413249211357,
+      "grad_norm": 0.5692775257801391,
+      "learning_rate": 1.7513030069500885e-05,
+      "loss": 0.4427,
+      "step": 4670
+    },
+    {
+      "epoch": 0.9209384858044164,
+      "grad_norm": 0.5036100073808859,
+      "learning_rate": 1.7512007028111103e-05,
+      "loss": 0.4376,
+      "step": 4671
+    },
+    {
+      "epoch": 0.9211356466876972,
+      "grad_norm": 0.5271798471163804,
+      "learning_rate": 1.75109838062393e-05,
+      "loss": 0.4504,
+      "step": 4672
+    },
+    {
+      "epoch": 0.9213328075709779,
+      "grad_norm": 0.5357814969525453,
+      "learning_rate": 1.750996040391007e-05,
+      "loss": 0.4528,
+      "step": 4673
+    },
+    {
+      "epoch": 0.9215299684542587,
+      "grad_norm": 0.5702199988441526,
+      "learning_rate": 1.7508936821147986e-05,
+      "loss": 0.453,
+      "step": 4674
+    },
+    {
+      "epoch": 0.9217271293375394,
+      "grad_norm": 0.5404783434059391,
+      "learning_rate": 1.750791305797765e-05,
+      "loss": 0.4315,
+      "step": 4675
+    },
+    {
+      "epoch": 0.9219242902208202,
+      "grad_norm": 0.5084713263493303,
+      "learning_rate": 1.7506889114423658e-05,
+      "loss": 0.4159,
+      "step": 4676
+    },
+    {
+      "epoch": 0.9221214511041009,
+      "grad_norm": 0.5201149594816759,
+      "learning_rate": 1.750586499051061e-05,
+      "loss": 0.4029,
+      "step": 4677
+    },
+    {
+      "epoch": 0.9223186119873817,
+      "grad_norm": 0.5436352106831215,
+      "learning_rate": 1.750484068626311e-05,
+      "loss": 0.4239,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9225157728706624,
+      "grad_norm": 0.5231124883960365,
+      "learning_rate": 1.7503816201705772e-05,
+      "loss": 0.4557,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9227129337539433,
+      "grad_norm": 0.5377568165510164,
+      "learning_rate": 1.75027915368632e-05,
+      "loss": 0.4332,
+      "step": 4680
+    },
+    {
+      "epoch": 0.922910094637224,
+      "grad_norm": 0.49556159503049685,
+      "learning_rate": 1.7501766691760027e-05,
+      "loss": 0.3759,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9231072555205048,
+      "grad_norm": 0.5366737464738414,
+      "learning_rate": 1.7500741666420863e-05,
+      "loss": 0.4272,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9233044164037855,
+      "grad_norm": 0.5333368284829383,
+      "learning_rate": 1.749971646087034e-05,
+      "loss": 0.4893,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9235015772870663,
+      "grad_norm": 0.5191084944970333,
+      "learning_rate": 1.7498691075133094e-05,
+      "loss": 0.4537,
+      "step": 4684
+    },
+    {
+      "epoch": 0.923698738170347,
+      "grad_norm": 0.4874538317773819,
+      "learning_rate": 1.7497665509233753e-05,
+      "loss": 0.4073,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9238958990536278,
+      "grad_norm": 0.5841148468262191,
+      "learning_rate": 1.7496639763196965e-05,
+      "loss": 0.4568,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9240930599369085,
+      "grad_norm": 0.46497152647935325,
+      "learning_rate": 1.7495613837047362e-05,
+      "loss": 0.3768,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9242902208201893,
+      "grad_norm": 0.6951617732456002,
+      "learning_rate": 1.7494587730809603e-05,
+      "loss": 0.4369,
+      "step": 4688
+    },
+    {
+      "epoch": 0.92448738170347,
+      "grad_norm": 0.5439225603081767,
+      "learning_rate": 1.749356144450834e-05,
+      "loss": 0.4204,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9246845425867508,
+      "grad_norm": 0.5545139274444996,
+      "learning_rate": 1.749253497816823e-05,
+      "loss": 0.4533,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9248817034700315,
+      "grad_norm": 0.5645877878868146,
+      "learning_rate": 1.7491508331813928e-05,
+      "loss": 0.4588,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9250788643533123,
+      "grad_norm": 0.5001014226230943,
+      "learning_rate": 1.7490481505470112e-05,
+      "loss": 0.413,
+      "step": 4692
+    },
+    {
+      "epoch": 0.925276025236593,
+      "grad_norm": 0.5340731127970042,
+      "learning_rate": 1.748945449916144e-05,
+      "loss": 0.4355,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9254731861198738,
+      "grad_norm": 0.4812452465843482,
+      "learning_rate": 1.7488427312912596e-05,
+      "loss": 0.4345,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9256703470031545,
+      "grad_norm": 1.0890928667586521,
+      "learning_rate": 1.7487399946748253e-05,
+      "loss": 0.4734,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9258675078864353,
+      "grad_norm": 0.5505478980835827,
+      "learning_rate": 1.74863724006931e-05,
+      "loss": 0.4808,
+      "step": 4696
+    },
+    {
+      "epoch": 0.926064668769716,
+      "grad_norm": 0.5008122123586376,
+      "learning_rate": 1.7485344674771817e-05,
+      "loss": 0.4225,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9262618296529969,
+      "grad_norm": 0.6702157966102137,
+      "learning_rate": 1.7484316769009105e-05,
+      "loss": 0.4698,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9264589905362776,
+      "grad_norm": 0.5272899289334974,
+      "learning_rate": 1.7483288683429655e-05,
+      "loss": 0.414,
+      "step": 4699
+    },
+    {
+      "epoch": 0.9266561514195584,
+      "grad_norm": 0.5858391308330164,
+      "learning_rate": 1.7482260418058167e-05,
+      "loss": 0.4288,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9268533123028391,
+      "grad_norm": 0.5833034743596007,
+      "learning_rate": 1.7481231972919346e-05,
+      "loss": 0.4879,
+      "step": 4701
+    },
+    {
+      "epoch": 0.9270504731861199,
+      "grad_norm": 0.5320561242621449,
+      "learning_rate": 1.74802033480379e-05,
+      "loss": 0.4144,
+      "step": 4702
+    },
+    {
+      "epoch": 0.9272476340694006,
+      "grad_norm": 0.5387587973540497,
+      "learning_rate": 1.7479174543438547e-05,
+      "loss": 0.4487,
+      "step": 4703
+    },
+    {
+      "epoch": 0.9274447949526814,
+      "grad_norm": 0.5997202359893333,
+      "learning_rate": 1.7478145559146002e-05,
+      "loss": 0.4796,
+      "step": 4704
+    },
+    {
+      "epoch": 0.9276419558359621,
+      "grad_norm": 0.5857148110448572,
+      "learning_rate": 1.747711639518499e-05,
+      "loss": 0.4235,
+      "step": 4705
+    },
+    {
+      "epoch": 0.9278391167192429,
+      "grad_norm": 0.5798656081956509,
+      "learning_rate": 1.7476087051580235e-05,
+      "loss": 0.4518,
+      "step": 4706
+    },
+    {
+      "epoch": 0.9280362776025236,
+      "grad_norm": 0.5831794755793467,
+      "learning_rate": 1.7475057528356466e-05,
+      "loss": 0.4613,
+      "step": 4707
+    },
+    {
+      "epoch": 0.9282334384858044,
+      "grad_norm": 0.5068716865871977,
+      "learning_rate": 1.7474027825538422e-05,
+      "loss": 0.4025,
+      "step": 4708
+    },
+    {
+      "epoch": 0.9284305993690851,
+      "grad_norm": 0.5949150495584267,
+      "learning_rate": 1.747299794315084e-05,
+      "loss": 0.4116,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9286277602523659,
+      "grad_norm": 0.5125219184067067,
+      "learning_rate": 1.7471967881218466e-05,
+      "loss": 0.4338,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9288249211356467,
+      "grad_norm": 0.5270990266912229,
+      "learning_rate": 1.7470937639766042e-05,
+      "loss": 0.4209,
+      "step": 4711
+    },
+    {
+      "epoch": 0.9290220820189274,
+      "grad_norm": 0.5495969435777367,
+      "learning_rate": 1.746990721881833e-05,
+      "loss": 0.4161,
+      "step": 4712
+    },
+    {
+      "epoch": 0.9292192429022083,
+      "grad_norm": 0.519716074091138,
+      "learning_rate": 1.746887661840008e-05,
+      "loss": 0.4226,
+      "step": 4713
+    },
+    {
+      "epoch": 0.929416403785489,
+      "grad_norm": 0.5108308327628213,
+      "learning_rate": 1.7467845838536054e-05,
+      "loss": 0.4335,
+      "step": 4714
+    },
+    {
+      "epoch": 0.9296135646687698,
+      "grad_norm": 4.264832132353777,
+      "learning_rate": 1.746681487925102e-05,
+      "loss": 0.5218,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9298107255520505,
+      "grad_norm": 0.6179886893181556,
+      "learning_rate": 1.746578374056974e-05,
+      "loss": 0.4544,
+      "step": 4716
+    },
+    {
+      "epoch": 0.9300078864353313,
+      "grad_norm": 0.5304906705707955,
+      "learning_rate": 1.7464752422516996e-05,
+      "loss": 0.4232,
+      "step": 4717
+    },
+    {
+      "epoch": 0.930205047318612,
+      "grad_norm": 0.5331325354477906,
+      "learning_rate": 1.7463720925117565e-05,
+      "loss": 0.3829,
+      "step": 4718
+    },
+    {
+      "epoch": 0.9304022082018928,
+      "grad_norm": 0.5765730607350507,
+      "learning_rate": 1.7462689248396228e-05,
+      "loss": 0.4317,
+      "step": 4719
+    },
+    {
+      "epoch": 0.9305993690851735,
+      "grad_norm": 0.5313090247910929,
+      "learning_rate": 1.7461657392377772e-05,
+      "loss": 0.4628,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9307965299684543,
+      "grad_norm": 0.5604970128075755,
+      "learning_rate": 1.746062535708699e-05,
+      "loss": 0.4345,
+      "step": 4721
+    },
+    {
+      "epoch": 0.930993690851735,
+      "grad_norm": 0.6273618716282534,
+      "learning_rate": 1.7459593142548674e-05,
+      "loss": 0.4068,
+      "step": 4722
+    },
+    {
+      "epoch": 0.9311908517350158,
+      "grad_norm": 0.5829766636894154,
+      "learning_rate": 1.7458560748787625e-05,
+      "loss": 0.4577,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9313880126182965,
+      "grad_norm": 0.6025168414566998,
+      "learning_rate": 1.7457528175828648e-05,
+      "loss": 0.4054,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9315851735015773,
+      "grad_norm": 1.0019218230150604,
+      "learning_rate": 1.745649542369655e-05,
+      "loss": 0.4536,
+      "step": 4725
+    },
+    {
+      "epoch": 0.931782334384858,
+      "grad_norm": 0.5862918544220463,
+      "learning_rate": 1.7455462492416148e-05,
+      "loss": 0.4495,
+      "step": 4726
+    },
+    {
+      "epoch": 0.9319794952681388,
+      "grad_norm": 0.5461382110764923,
+      "learning_rate": 1.7454429382012255e-05,
+      "loss": 0.4391,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9321766561514195,
+      "grad_norm": 0.5936092061621204,
+      "learning_rate": 1.745339609250969e-05,
+      "loss": 0.4482,
+      "step": 4728
+    },
+    {
+      "epoch": 0.9323738170347003,
+      "grad_norm": 0.5346698342600564,
+      "learning_rate": 1.7452362623933283e-05,
+      "loss": 0.4358,
+      "step": 4729
+    },
+    {
+      "epoch": 0.932570977917981,
+      "grad_norm": 0.5604545243460433,
+      "learning_rate": 1.7451328976307864e-05,
+      "loss": 0.4072,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9327681388012619,
+      "grad_norm": 0.5624845340732466,
+      "learning_rate": 1.7450295149658265e-05,
+      "loss": 0.4933,
+      "step": 4731
+    },
+    {
+      "epoch": 0.9329652996845426,
+      "grad_norm": 0.5321758225204737,
+      "learning_rate": 1.7449261144009325e-05,
+      "loss": 0.4828,
+      "step": 4732
+    },
+    {
+      "epoch": 0.9331624605678234,
+      "grad_norm": 0.5442798225790718,
+      "learning_rate": 1.744822695938589e-05,
+      "loss": 0.4275,
+      "step": 4733
+    },
+    {
+      "epoch": 0.9333596214511041,
+      "grad_norm": 2.0994490161932124,
+      "learning_rate": 1.74471925958128e-05,
+      "loss": 0.4325,
+      "step": 4734
+    },
+    {
+      "epoch": 0.9335567823343849,
+      "grad_norm": 0.5447286623554133,
+      "learning_rate": 1.744615805331491e-05,
+      "loss": 0.4577,
+      "step": 4735
+    },
+    {
+      "epoch": 0.9337539432176656,
+      "grad_norm": 1.0389460335213794,
+      "learning_rate": 1.744512333191708e-05,
+      "loss": 0.4268,
+      "step": 4736
+    },
+    {
+      "epoch": 0.9339511041009464,
+      "grad_norm": 0.5585776839463066,
+      "learning_rate": 1.7444088431644166e-05,
+      "loss": 0.4436,
+      "step": 4737
+    },
+    {
+      "epoch": 0.9341482649842271,
+      "grad_norm": 0.5415927198264985,
+      "learning_rate": 1.7443053352521032e-05,
+      "loss": 0.4297,
+      "step": 4738
+    },
+    {
+      "epoch": 0.9343454258675079,
+      "grad_norm": 0.7020791964781086,
+      "learning_rate": 1.7442018094572546e-05,
+      "loss": 0.3968,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9345425867507886,
+      "grad_norm": 0.5293642076390404,
+      "learning_rate": 1.7440982657823583e-05,
+      "loss": 0.4414,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9347397476340694,
+      "grad_norm": 0.5699485347614668,
+      "learning_rate": 1.743994704229902e-05,
+      "loss": 0.4452,
+      "step": 4741
+    },
+    {
+      "epoch": 0.9349369085173501,
+      "grad_norm": 0.9714674676297455,
+      "learning_rate": 1.743891124802374e-05,
+      "loss": 0.4315,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9351340694006309,
+      "grad_norm": 0.5220962127597865,
+      "learning_rate": 1.7437875275022622e-05,
+      "loss": 0.4145,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9353312302839116,
+      "grad_norm": 0.538501745627749,
+      "learning_rate": 1.7436839123320566e-05,
+      "loss": 0.46,
+      "step": 4744
+    },
+    {
+      "epoch": 0.9355283911671924,
+      "grad_norm": 0.49639504847696175,
+      "learning_rate": 1.743580279294246e-05,
+      "loss": 0.4082,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9357255520504731,
+      "grad_norm": 0.531129670237936,
+      "learning_rate": 1.74347662839132e-05,
+      "loss": 0.4302,
+      "step": 4746
+    },
+    {
+      "epoch": 0.935922712933754,
+      "grad_norm": 0.6286351478232094,
+      "learning_rate": 1.7433729596257694e-05,
+      "loss": 0.4533,
+      "step": 4747
+    },
+    {
+      "epoch": 0.9361198738170347,
+      "grad_norm": 0.5416558622895243,
+      "learning_rate": 1.743269273000085e-05,
+      "loss": 0.4097,
+      "step": 4748
+    },
+    {
+      "epoch": 0.9363170347003155,
+      "grad_norm": 0.5339455229568808,
+      "learning_rate": 1.7431655685167578e-05,
+      "loss": 0.439,
+      "step": 4749
+    },
+    {
+      "epoch": 0.9365141955835962,
+      "grad_norm": 0.5760351851798424,
+      "learning_rate": 1.743061846178279e-05,
+      "loss": 0.4818,
+      "step": 4750
+    },
+    {
+      "epoch": 0.936711356466877,
+      "grad_norm": 0.5917480235250198,
+      "learning_rate": 1.742958105987141e-05,
+      "loss": 0.4568,
+      "step": 4751
+    },
+    {
+      "epoch": 0.9369085173501577,
+      "grad_norm": 0.8339420837901182,
+      "learning_rate": 1.7428543479458367e-05,
+      "loss": 0.4443,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9371056782334385,
+      "grad_norm": 0.5697309446607353,
+      "learning_rate": 1.7427505720568583e-05,
+      "loss": 0.4565,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9373028391167192,
+      "grad_norm": 0.5415366469073645,
+      "learning_rate": 1.7426467783226992e-05,
+      "loss": 0.4764,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.6019352545618148,
+      "learning_rate": 1.742542966745853e-05,
+      "loss": 0.4492,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9376971608832808,
+      "grad_norm": 0.6685857679022774,
+      "learning_rate": 1.7424391373288142e-05,
+      "loss": 0.4581,
+      "step": 4756
+    },
+    {
+      "epoch": 0.9378943217665615,
+      "grad_norm": 0.5894795119442221,
+      "learning_rate": 1.742335290074077e-05,
+      "loss": 0.4677,
+      "step": 4757
+    },
+    {
+      "epoch": 0.9380914826498423,
+      "grad_norm": 0.5528329269741018,
+      "learning_rate": 1.7422314249841373e-05,
+      "loss": 0.4342,
+      "step": 4758
+    },
+    {
+      "epoch": 0.938288643533123,
+      "grad_norm": 0.5633875242021811,
+      "learning_rate": 1.7421275420614895e-05,
+      "loss": 0.4297,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9384858044164038,
+      "grad_norm": 0.6057160126806879,
+      "learning_rate": 1.7420236413086298e-05,
+      "loss": 0.4469,
+      "step": 4760
+    },
+    {
+      "epoch": 0.9386829652996845,
+      "grad_norm": 0.5062209536970022,
+      "learning_rate": 1.7419197227280545e-05,
+      "loss": 0.413,
+      "step": 4761
+    },
+    {
+      "epoch": 0.9388801261829653,
+      "grad_norm": 0.5947801340129587,
+      "learning_rate": 1.7418157863222608e-05,
+      "loss": 0.4505,
+      "step": 4762
+    },
+    {
+      "epoch": 0.939077287066246,
+      "grad_norm": 0.5473326180648399,
+      "learning_rate": 1.7417118320937452e-05,
+      "loss": 0.4298,
+      "step": 4763
+    },
+    {
+      "epoch": 0.9392744479495269,
+      "grad_norm": 0.5842555110402521,
+      "learning_rate": 1.7416078600450053e-05,
+      "loss": 0.4366,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9394716088328076,
+      "grad_norm": 6.596709214102324,
+      "learning_rate": 1.7415038701785397e-05,
+      "loss": 0.4933,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9396687697160884,
+      "grad_norm": 0.5983220979388622,
+      "learning_rate": 1.741399862496846e-05,
+      "loss": 0.4303,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9398659305993691,
+      "grad_norm": 0.5605426934566005,
+      "learning_rate": 1.741295837002424e-05,
+      "loss": 0.4376,
+      "step": 4767
+    },
+    {
+      "epoch": 0.9400630914826499,
+      "grad_norm": 0.6623850194449121,
+      "learning_rate": 1.7411917936977728e-05,
+      "loss": 0.4891,
+      "step": 4768
+    },
+    {
+      "epoch": 0.9402602523659306,
+      "grad_norm": 0.5643318735416755,
+      "learning_rate": 1.7410877325853914e-05,
+      "loss": 0.4342,
+      "step": 4769
+    },
+    {
+      "epoch": 0.9404574132492114,
+      "grad_norm": 0.510419807668752,
+      "learning_rate": 1.7409836536677804e-05,
+      "loss": 0.4094,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9406545741324921,
+      "grad_norm": 0.5957515521253349,
+      "learning_rate": 1.7408795569474407e-05,
+      "loss": 0.4534,
+      "step": 4771
+    },
+    {
+      "epoch": 0.9408517350157729,
+      "grad_norm": 0.5938740829653067,
+      "learning_rate": 1.7407754424268727e-05,
+      "loss": 0.4209,
+      "step": 4772
+    },
+    {
+      "epoch": 0.9410488958990536,
+      "grad_norm": 0.5430175081242069,
+      "learning_rate": 1.7406713101085782e-05,
+      "loss": 0.4428,
+      "step": 4773
+    },
+    {
+      "epoch": 0.9412460567823344,
+      "grad_norm": 0.6092122034861576,
+      "learning_rate": 1.7405671599950593e-05,
+      "loss": 0.4552,
+      "step": 4774
+    },
+    {
+      "epoch": 0.9414432176656151,
+      "grad_norm": 0.5301372981781962,
+      "learning_rate": 1.7404629920888178e-05,
+      "loss": 0.3949,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9416403785488959,
+      "grad_norm": 0.874324067325636,
+      "learning_rate": 1.7403588063923565e-05,
+      "loss": 0.4171,
+      "step": 4776
+    },
+    {
+      "epoch": 0.9418375394321766,
+      "grad_norm": 0.562779555203244,
+      "learning_rate": 1.7402546029081793e-05,
+      "loss": 0.4508,
+      "step": 4777
+    },
+    {
+      "epoch": 0.9420347003154574,
+      "grad_norm": 0.5751034428164898,
+      "learning_rate": 1.7401503816387886e-05,
+      "loss": 0.4648,
+      "step": 4778
+    },
+    {
+      "epoch": 0.9422318611987381,
+      "grad_norm": 0.4928153903001422,
+      "learning_rate": 1.740046142586689e-05,
+      "loss": 0.4094,
+      "step": 4779
+    },
+    {
+      "epoch": 0.942429022082019,
+      "grad_norm": 0.5593941847892724,
+      "learning_rate": 1.7399418857543848e-05,
+      "loss": 0.4448,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9426261829652997,
+      "grad_norm": 0.5120731327364189,
+      "learning_rate": 1.739837611144381e-05,
+      "loss": 0.4433,
+      "step": 4781
+    },
+    {
+      "epoch": 0.9428233438485805,
+      "grad_norm": 0.5547010617938658,
+      "learning_rate": 1.739733318759183e-05,
+      "loss": 0.4517,
+      "step": 4782
+    },
+    {
+      "epoch": 0.9430205047318612,
+      "grad_norm": 0.5279709568184927,
+      "learning_rate": 1.739629008601296e-05,
+      "loss": 0.4414,
+      "step": 4783
+    },
+    {
+      "epoch": 0.943217665615142,
+      "grad_norm": 0.6546356596831223,
+      "learning_rate": 1.7395246806732266e-05,
+      "loss": 0.502,
+      "step": 4784
+    },
+    {
+      "epoch": 0.9434148264984227,
+      "grad_norm": 0.5551495693288206,
+      "learning_rate": 1.739420334977481e-05,
+      "loss": 0.4212,
+      "step": 4785
+    },
+    {
+      "epoch": 0.9436119873817035,
+      "grad_norm": 0.5092496805523852,
+      "learning_rate": 1.7393159715165668e-05,
+      "loss": 0.405,
+      "step": 4786
+    },
+    {
+      "epoch": 0.9438091482649842,
+      "grad_norm": 0.5118148863889752,
+      "learning_rate": 1.739211590292991e-05,
+      "loss": 0.4185,
+      "step": 4787
+    },
+    {
+      "epoch": 0.944006309148265,
+      "grad_norm": 0.5543188761300978,
+      "learning_rate": 1.739107191309261e-05,
+      "loss": 0.4229,
+      "step": 4788
+    },
+    {
+      "epoch": 0.9442034700315457,
+      "grad_norm": 0.4947381983906698,
+      "learning_rate": 1.7390027745678857e-05,
+      "loss": 0.4347,
+      "step": 4789
+    },
+    {
+      "epoch": 0.9444006309148265,
+      "grad_norm": 0.5333604588136764,
+      "learning_rate": 1.7388983400713736e-05,
+      "loss": 0.4273,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9445977917981072,
+      "grad_norm": 0.48436693257142593,
+      "learning_rate": 1.7387938878222337e-05,
+      "loss": 0.3842,
+      "step": 4791
+    },
+    {
+      "epoch": 0.944794952681388,
+      "grad_norm": 0.48098662178856766,
+      "learning_rate": 1.7386894178229764e-05,
+      "loss": 0.4078,
+      "step": 4792
+    },
+    {
+      "epoch": 0.9449921135646687,
+      "grad_norm": 0.5927563970559201,
+      "learning_rate": 1.7385849300761104e-05,
+      "loss": 0.4479,
+      "step": 4793
+    },
+    {
+      "epoch": 0.9451892744479495,
+      "grad_norm": 0.5292188971068847,
+      "learning_rate": 1.7384804245841468e-05,
+      "loss": 0.4466,
+      "step": 4794
+    },
+    {
+      "epoch": 0.9453864353312302,
+      "grad_norm": 0.5495901269166898,
+      "learning_rate": 1.7383759013495965e-05,
+      "loss": 0.423,
+      "step": 4795
+    },
+    {
+      "epoch": 0.945583596214511,
+      "grad_norm": 0.5124360410567629,
+      "learning_rate": 1.73827136037497e-05,
+      "loss": 0.4669,
+      "step": 4796
+    },
+    {
+      "epoch": 0.9457807570977917,
+      "grad_norm": 0.5728281769441116,
+      "learning_rate": 1.7381668016627798e-05,
+      "loss": 0.4712,
+      "step": 4797
+    },
+    {
+      "epoch": 0.9459779179810726,
+      "grad_norm": 0.48078417126183814,
+      "learning_rate": 1.738062225215538e-05,
+      "loss": 0.4534,
+      "step": 4798
+    },
+    {
+      "epoch": 0.9461750788643533,
+      "grad_norm": 0.4972157768011806,
+      "learning_rate": 1.7379576310357568e-05,
+      "loss": 0.4335,
+      "step": 4799
+    },
+    {
+      "epoch": 0.9463722397476341,
+      "grad_norm": 0.6537133826858609,
+      "learning_rate": 1.7378530191259492e-05,
+      "loss": 0.4604,
+      "step": 4800
+    },
+    {
+      "epoch": 0.9465694006309149,
+      "grad_norm": 0.5206764483089402,
+      "learning_rate": 1.7377483894886285e-05,
+      "loss": 0.4377,
+      "step": 4801
+    },
+    {
+      "epoch": 0.9467665615141956,
+      "grad_norm": 0.5571810897809417,
+      "learning_rate": 1.7376437421263088e-05,
+      "loss": 0.4732,
+      "step": 4802
+    },
+    {
+      "epoch": 0.9469637223974764,
+      "grad_norm": 0.5154485693291944,
+      "learning_rate": 1.737539077041504e-05,
+      "loss": 0.4422,
+      "step": 4803
+    },
+    {
+      "epoch": 0.9471608832807571,
+      "grad_norm": 2.538605874424299,
+      "learning_rate": 1.737434394236729e-05,
+      "loss": 0.4926,
+      "step": 4804
+    },
+    {
+      "epoch": 0.9473580441640379,
+      "grad_norm": 0.5854347379433069,
+      "learning_rate": 1.737329693714499e-05,
+      "loss": 0.4397,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9475552050473186,
+      "grad_norm": 0.5692750771949014,
+      "learning_rate": 1.7372249754773292e-05,
+      "loss": 0.4892,
+      "step": 4806
+    },
+    {
+      "epoch": 0.9477523659305994,
+      "grad_norm": 0.5453924056164228,
+      "learning_rate": 1.7371202395277357e-05,
+      "loss": 0.4488,
+      "step": 4807
+    },
+    {
+      "epoch": 0.9479495268138801,
+      "grad_norm": 0.5329921263116298,
+      "learning_rate": 1.7370154858682347e-05,
+      "loss": 0.4305,
+      "step": 4808
+    },
+    {
+      "epoch": 0.9481466876971609,
+      "grad_norm": 0.5070927581481749,
+      "learning_rate": 1.736910714501343e-05,
+      "loss": 0.3879,
+      "step": 4809
+    },
+    {
+      "epoch": 0.9483438485804416,
+      "grad_norm": 0.49761316318511184,
+      "learning_rate": 1.7368059254295783e-05,
+      "loss": 0.4439,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9485410094637224,
+      "grad_norm": 0.6116580466663731,
+      "learning_rate": 1.736701118655458e-05,
+      "loss": 0.4213,
+      "step": 4811
+    },
+    {
+      "epoch": 0.9487381703470031,
+      "grad_norm": 0.6059773769906507,
+      "learning_rate": 1.7365962941814998e-05,
+      "loss": 0.4447,
+      "step": 4812
+    },
+    {
+      "epoch": 0.948935331230284,
+      "grad_norm": 0.5302599878215019,
+      "learning_rate": 1.7364914520102223e-05,
+      "loss": 0.4219,
+      "step": 4813
+    },
+    {
+      "epoch": 0.9491324921135647,
+      "grad_norm": 0.5346407557209393,
+      "learning_rate": 1.736386592144145e-05,
+      "loss": 0.4452,
+      "step": 4814
+    },
+    {
+      "epoch": 0.9493296529968455,
+      "grad_norm": 2.5973339435009675,
+      "learning_rate": 1.7362817145857866e-05,
+      "loss": 0.5188,
+      "step": 4815
+    },
+    {
+      "epoch": 0.9495268138801262,
+      "grad_norm": 0.6369521233467632,
+      "learning_rate": 1.736176819337667e-05,
+      "loss": 0.4557,
+      "step": 4816
+    },
+    {
+      "epoch": 0.949723974763407,
+      "grad_norm": 0.7229696457122112,
+      "learning_rate": 1.7360719064023067e-05,
+      "loss": 0.4351,
+      "step": 4817
+    },
+    {
+      "epoch": 0.9499211356466877,
+      "grad_norm": 0.5548249000280323,
+      "learning_rate": 1.7359669757822256e-05,
+      "loss": 0.4495,
+      "step": 4818
+    },
+    {
+      "epoch": 0.9501182965299685,
+      "grad_norm": 0.5371920142295181,
+      "learning_rate": 1.7358620274799455e-05,
+      "loss": 0.4468,
+      "step": 4819
+    },
+    {
+      "epoch": 0.9503154574132492,
+      "grad_norm": 0.5768920071321693,
+      "learning_rate": 1.7357570614979878e-05,
+      "loss": 0.4429,
+      "step": 4820
+    },
+    {
+      "epoch": 0.95051261829653,
+      "grad_norm": 0.5491663645524869,
+      "learning_rate": 1.735652077838874e-05,
+      "loss": 0.4645,
+      "step": 4821
+    },
+    {
+      "epoch": 0.9507097791798107,
+      "grad_norm": 0.5993991809124819,
+      "learning_rate": 1.735547076505127e-05,
+      "loss": 0.4071,
+      "step": 4822
+    },
+    {
+      "epoch": 0.9509069400630915,
+      "grad_norm": 0.4880554858272554,
+      "learning_rate": 1.7354420574992686e-05,
+      "loss": 0.4059,
+      "step": 4823
+    },
+    {
+      "epoch": 0.9511041009463722,
+      "grad_norm": 0.903588480396732,
+      "learning_rate": 1.7353370208238226e-05,
+      "loss": 0.4422,
+      "step": 4824
+    },
+    {
+      "epoch": 0.951301261829653,
+      "grad_norm": 0.5459492044215943,
+      "learning_rate": 1.7352319664813126e-05,
+      "loss": 0.4284,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9514984227129337,
+      "grad_norm": 0.531596665432785,
+      "learning_rate": 1.7351268944742626e-05,
+      "loss": 0.4111,
+      "step": 4826
+    },
+    {
+      "epoch": 0.9516955835962145,
+      "grad_norm": 0.5420852235661492,
+      "learning_rate": 1.735021804805197e-05,
+      "loss": 0.4238,
+      "step": 4827
+    },
+    {
+      "epoch": 0.9518927444794952,
+      "grad_norm": 0.5118772611762697,
+      "learning_rate": 1.7349166974766407e-05,
+      "loss": 0.4063,
+      "step": 4828
+    },
+    {
+      "epoch": 0.952089905362776,
+      "grad_norm": 0.6928236531905183,
+      "learning_rate": 1.7348115724911188e-05,
+      "loss": 0.4473,
+      "step": 4829
+    },
+    {
+      "epoch": 0.9522870662460567,
+      "grad_norm": 0.5459811584499307,
+      "learning_rate": 1.734706429851157e-05,
+      "loss": 0.4339,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9524842271293376,
+      "grad_norm": 0.5306998499179901,
+      "learning_rate": 1.7346012695592817e-05,
+      "loss": 0.4496,
+      "step": 4831
+    },
+    {
+      "epoch": 0.9526813880126183,
+      "grad_norm": 0.5488004261374878,
+      "learning_rate": 1.7344960916180192e-05,
+      "loss": 0.4504,
+      "step": 4832
+    },
+    {
+      "epoch": 0.9528785488958991,
+      "grad_norm": 0.5667953843020231,
+      "learning_rate": 1.734390896029897e-05,
+      "loss": 0.4086,
+      "step": 4833
+    },
+    {
+      "epoch": 0.9530757097791798,
+      "grad_norm": 0.5323191653010501,
+      "learning_rate": 1.7342856827974417e-05,
+      "loss": 0.4353,
+      "step": 4834
+    },
+    {
+      "epoch": 0.9532728706624606,
+      "grad_norm": 0.5423134586875568,
+      "learning_rate": 1.7341804519231815e-05,
+      "loss": 0.4411,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9534700315457413,
+      "grad_norm": 0.548006233665663,
+      "learning_rate": 1.734075203409645e-05,
+      "loss": 0.4433,
+      "step": 4836
+    },
+    {
+      "epoch": 0.9536671924290221,
+      "grad_norm": 0.5273485527910695,
+      "learning_rate": 1.7339699372593605e-05,
+      "loss": 0.4673,
+      "step": 4837
+    },
+    {
+      "epoch": 0.9538643533123028,
+      "grad_norm": 0.5406256570631455,
+      "learning_rate": 1.7338646534748572e-05,
+      "loss": 0.4424,
+      "step": 4838
+    },
+    {
+      "epoch": 0.9540615141955836,
+      "grad_norm": 0.5228289639533955,
+      "learning_rate": 1.7337593520586645e-05,
+      "loss": 0.4113,
+      "step": 4839
+    },
+    {
+      "epoch": 0.9542586750788643,
+      "grad_norm": 0.5199695216281296,
+      "learning_rate": 1.7336540330133126e-05,
+      "loss": 0.4334,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9544558359621451,
+      "grad_norm": 0.5157055642134136,
+      "learning_rate": 1.7335486963413318e-05,
+      "loss": 0.4114,
+      "step": 4841
+    },
+    {
+      "epoch": 0.9546529968454258,
+      "grad_norm": 0.5627963986729889,
+      "learning_rate": 1.7334433420452527e-05,
+      "loss": 0.454,
+      "step": 4842
+    },
+    {
+      "epoch": 0.9548501577287066,
+      "grad_norm": 0.5354466461718106,
+      "learning_rate": 1.7333379701276068e-05,
+      "loss": 0.4418,
+      "step": 4843
+    },
+    {
+      "epoch": 0.9550473186119873,
+      "grad_norm": 0.9673410881580022,
+      "learning_rate": 1.7332325805909256e-05,
+      "loss": 0.4306,
+      "step": 4844
+    },
+    {
+      "epoch": 0.9552444794952681,
+      "grad_norm": 5.372837853533415,
+      "learning_rate": 1.733127173437741e-05,
+      "loss": 0.4094,
+      "step": 4845
+    },
+    {
+      "epoch": 0.955441640378549,
+      "grad_norm": 0.6273908422804465,
+      "learning_rate": 1.7330217486705862e-05,
+      "loss": 0.4191,
+      "step": 4846
+    },
+    {
+      "epoch": 0.9556388012618297,
+      "grad_norm": 0.5769927479813597,
+      "learning_rate": 1.732916306291993e-05,
+      "loss": 0.4635,
+      "step": 4847
+    },
+    {
+      "epoch": 0.9558359621451105,
+      "grad_norm": 0.5345159528878386,
+      "learning_rate": 1.7328108463044953e-05,
+      "loss": 0.4228,
+      "step": 4848
+    },
+    {
+      "epoch": 0.9560331230283912,
+      "grad_norm": 0.5196926420397846,
+      "learning_rate": 1.7327053687106273e-05,
+      "loss": 0.3951,
+      "step": 4849
+    },
+    {
+      "epoch": 0.956230283911672,
+      "grad_norm": 0.4976003610694764,
+      "learning_rate": 1.7325998735129227e-05,
+      "loss": 0.3933,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9564274447949527,
+      "grad_norm": 0.5629160034402856,
+      "learning_rate": 1.7324943607139158e-05,
+      "loss": 0.4207,
+      "step": 4851
+    },
+    {
+      "epoch": 0.9566246056782335,
+      "grad_norm": 0.541499287915572,
+      "learning_rate": 1.7323888303161422e-05,
+      "loss": 0.4565,
+      "step": 4852
+    },
+    {
+      "epoch": 0.9568217665615142,
+      "grad_norm": 0.5494402412976808,
+      "learning_rate": 1.732283282322137e-05,
+      "loss": 0.4487,
+      "step": 4853
+    },
+    {
+      "epoch": 0.957018927444795,
+      "grad_norm": 0.5145315890387855,
+      "learning_rate": 1.7321777167344367e-05,
+      "loss": 0.4067,
+      "step": 4854
+    },
+    {
+      "epoch": 0.9572160883280757,
+      "grad_norm": 0.5356564633940517,
+      "learning_rate": 1.732072133555577e-05,
+      "loss": 0.4432,
+      "step": 4855
+    },
+    {
+      "epoch": 0.9574132492113565,
+      "grad_norm": 0.498975459571729,
+      "learning_rate": 1.7319665327880945e-05,
+      "loss": 0.3735,
+      "step": 4856
+    },
+    {
+      "epoch": 0.9576104100946372,
+      "grad_norm": 0.618141773791518,
+      "learning_rate": 1.7318609144345265e-05,
+      "loss": 0.4099,
+      "step": 4857
+    },
+    {
+      "epoch": 0.957807570977918,
+      "grad_norm": 0.5160325448737755,
+      "learning_rate": 1.7317552784974113e-05,
+      "loss": 0.3971,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9580047318611987,
+      "grad_norm": 0.5514233004043473,
+      "learning_rate": 1.7316496249792857e-05,
+      "loss": 0.46,
+      "step": 4859
+    },
+    {
+      "epoch": 0.9582018927444795,
+      "grad_norm": 0.54586722386132,
+      "learning_rate": 1.7315439538826887e-05,
+      "loss": 0.4271,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9583990536277602,
+      "grad_norm": 0.50615085528196,
+      "learning_rate": 1.7314382652101595e-05,
+      "loss": 0.4251,
+      "step": 4861
+    },
+    {
+      "epoch": 0.958596214511041,
+      "grad_norm": 0.5311049525152373,
+      "learning_rate": 1.7313325589642363e-05,
+      "loss": 0.4354,
+      "step": 4862
+    },
+    {
+      "epoch": 0.9587933753943217,
+      "grad_norm": 0.5243846048285586,
+      "learning_rate": 1.7312268351474603e-05,
+      "loss": 0.423,
+      "step": 4863
+    },
+    {
+      "epoch": 0.9589905362776026,
+      "grad_norm": 0.5359735257750494,
+      "learning_rate": 1.73112109376237e-05,
+      "loss": 0.4092,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9591876971608833,
+      "grad_norm": 0.5599828574009664,
+      "learning_rate": 1.7310153348115068e-05,
+      "loss": 0.4465,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9593848580441641,
+      "grad_norm": 0.53367931056808,
+      "learning_rate": 1.7309095582974115e-05,
+      "loss": 0.4485,
+      "step": 4866
+    },
+    {
+      "epoch": 0.9595820189274448,
+      "grad_norm": 0.5485569037489243,
+      "learning_rate": 1.7308037642226258e-05,
+      "loss": 0.4181,
+      "step": 4867
+    },
+    {
+      "epoch": 0.9597791798107256,
+      "grad_norm": 0.5325054492925666,
+      "learning_rate": 1.7306979525896907e-05,
+      "loss": 0.4338,
+      "step": 4868
+    },
+    {
+      "epoch": 0.9599763406940063,
+      "grad_norm": 1.2447432969875094,
+      "learning_rate": 1.730592123401149e-05,
+      "loss": 0.447,
+      "step": 4869
+    },
+    {
+      "epoch": 0.9601735015772871,
+      "grad_norm": 0.5405855805944394,
+      "learning_rate": 1.7304862766595433e-05,
+      "loss": 0.4556,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9603706624605678,
+      "grad_norm": 0.5014468926425727,
+      "learning_rate": 1.7303804123674165e-05,
+      "loss": 0.4127,
+      "step": 4871
+    },
+    {
+      "epoch": 0.9605678233438486,
+      "grad_norm": 0.5242180756855537,
+      "learning_rate": 1.730274530527312e-05,
+      "loss": 0.4488,
+      "step": 4872
+    },
+    {
+      "epoch": 0.9607649842271293,
+      "grad_norm": 0.5590887485992455,
+      "learning_rate": 1.730168631141774e-05,
+      "loss": 0.4408,
+      "step": 4873
+    },
+    {
+      "epoch": 0.9609621451104101,
+      "grad_norm": 0.5590641148729568,
+      "learning_rate": 1.7300627142133466e-05,
+      "loss": 0.4421,
+      "step": 4874
+    },
+    {
+      "epoch": 0.9611593059936908,
+      "grad_norm": 0.5381918633394875,
+      "learning_rate": 1.7299567797445744e-05,
+      "loss": 0.4152,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9613564668769716,
+      "grad_norm": 0.5116503025098527,
+      "learning_rate": 1.729850827738003e-05,
+      "loss": 0.4137,
+      "step": 4876
+    },
+    {
+      "epoch": 0.9615536277602523,
+      "grad_norm": 0.601103697573897,
+      "learning_rate": 1.7297448581961775e-05,
+      "loss": 0.4367,
+      "step": 4877
+    },
+    {
+      "epoch": 0.9617507886435331,
+      "grad_norm": 0.5234000253181185,
+      "learning_rate": 1.7296388711216442e-05,
+      "loss": 0.4264,
+      "step": 4878
+    },
+    {
+      "epoch": 0.9619479495268138,
+      "grad_norm": 0.5489406687628315,
+      "learning_rate": 1.7295328665169495e-05,
+      "loss": 0.4522,
+      "step": 4879
+    },
+    {
+      "epoch": 0.9621451104100947,
+      "grad_norm": 0.5304149080662556,
+      "learning_rate": 1.7294268443846403e-05,
+      "loss": 0.4531,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9623422712933754,
+      "grad_norm": 0.5277630403000341,
+      "learning_rate": 1.7293208047272635e-05,
+      "loss": 0.4233,
+      "step": 4881
+    },
+    {
+      "epoch": 0.9625394321766562,
+      "grad_norm": 0.5216948283800477,
+      "learning_rate": 1.729214747547367e-05,
+      "loss": 0.4634,
+      "step": 4882
+    },
+    {
+      "epoch": 0.9627365930599369,
+      "grad_norm": 0.5128948306308777,
+      "learning_rate": 1.7291086728474992e-05,
+      "loss": 0.3948,
+      "step": 4883
+    },
+    {
+      "epoch": 0.9629337539432177,
+      "grad_norm": 0.4864061144860882,
+      "learning_rate": 1.729002580630208e-05,
+      "loss": 0.4011,
+      "step": 4884
+    },
+    {
+      "epoch": 0.9631309148264984,
+      "grad_norm": 1.3934757331709215,
+      "learning_rate": 1.7288964708980432e-05,
+      "loss": 0.4406,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9633280757097792,
+      "grad_norm": 0.5314428535537528,
+      "learning_rate": 1.7287903436535535e-05,
+      "loss": 0.4282,
+      "step": 4886
+    },
+    {
+      "epoch": 0.9635252365930599,
+      "grad_norm": 0.5113959395669175,
+      "learning_rate": 1.728684198899289e-05,
+      "loss": 0.4459,
+      "step": 4887
+    },
+    {
+      "epoch": 0.9637223974763407,
+      "grad_norm": 0.7377416915879886,
+      "learning_rate": 1.7285780366377998e-05,
+      "loss": 0.4226,
+      "step": 4888
+    },
+    {
+      "epoch": 0.9639195583596214,
+      "grad_norm": 0.625103270758122,
+      "learning_rate": 1.7284718568716362e-05,
+      "loss": 0.4857,
+      "step": 4889
+    },
+    {
+      "epoch": 0.9641167192429022,
+      "grad_norm": 0.49384862341141844,
+      "learning_rate": 1.7283656596033502e-05,
+      "loss": 0.4377,
+      "step": 4890
+    },
+    {
+      "epoch": 0.964313880126183,
+      "grad_norm": 0.8265103239871188,
+      "learning_rate": 1.7282594448354922e-05,
+      "loss": 0.4826,
+      "step": 4891
+    },
+    {
+      "epoch": 0.9645110410094637,
+      "grad_norm": 0.5660829434512461,
+      "learning_rate": 1.728153212570615e-05,
+      "loss": 0.436,
+      "step": 4892
+    },
+    {
+      "epoch": 0.9647082018927445,
+      "grad_norm": 0.49876988837276626,
+      "learning_rate": 1.7280469628112698e-05,
+      "loss": 0.4192,
+      "step": 4893
+    },
+    {
+      "epoch": 0.9649053627760252,
+      "grad_norm": 0.6150615512750844,
+      "learning_rate": 1.7279406955600107e-05,
+      "loss": 0.4016,
+      "step": 4894
+    },
+    {
+      "epoch": 0.965102523659306,
+      "grad_norm": 0.536942351647561,
+      "learning_rate": 1.7278344108193897e-05,
+      "loss": 0.4207,
+      "step": 4895
+    },
+    {
+      "epoch": 0.9652996845425867,
+      "grad_norm": 0.9470023911925746,
+      "learning_rate": 1.7277281085919613e-05,
+      "loss": 0.4274,
+      "step": 4896
+    },
+    {
+      "epoch": 0.9654968454258676,
+      "grad_norm": 0.5235369204502012,
+      "learning_rate": 1.727621788880279e-05,
+      "loss": 0.4244,
+      "step": 4897
+    },
+    {
+      "epoch": 0.9656940063091483,
+      "grad_norm": 0.6071731227037889,
+      "learning_rate": 1.727515451686897e-05,
+      "loss": 0.3893,
+      "step": 4898
+    },
+    {
+      "epoch": 0.9658911671924291,
+      "grad_norm": 0.5903477067942368,
+      "learning_rate": 1.7274090970143705e-05,
+      "loss": 0.4356,
+      "step": 4899
+    },
+    {
+      "epoch": 0.9660883280757098,
+      "grad_norm": 0.5376060698400171,
+      "learning_rate": 1.7273027248652545e-05,
+      "loss": 0.4561,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9662854889589906,
+      "grad_norm": 0.5838617039015437,
+      "learning_rate": 1.727196335242105e-05,
+      "loss": 0.4383,
+      "step": 4901
+    },
+    {
+      "epoch": 0.9664826498422713,
+      "grad_norm": 0.5167788808435415,
+      "learning_rate": 1.727089928147478e-05,
+      "loss": 0.4024,
+      "step": 4902
+    },
+    {
+      "epoch": 0.9666798107255521,
+      "grad_norm": 0.5624366894471787,
+      "learning_rate": 1.72698350358393e-05,
+      "loss": 0.4309,
+      "step": 4903
+    },
+    {
+      "epoch": 0.9668769716088328,
+      "grad_norm": 0.753504301288875,
+      "learning_rate": 1.726877061554018e-05,
+      "loss": 0.3958,
+      "step": 4904
+    },
+    {
+      "epoch": 0.9670741324921136,
+      "grad_norm": 0.5684825023726245,
+      "learning_rate": 1.726770602060299e-05,
+      "loss": 0.4328,
+      "step": 4905
+    },
+    {
+      "epoch": 0.9672712933753943,
+      "grad_norm": 0.6897233391043007,
+      "learning_rate": 1.726664125105331e-05,
+      "loss": 0.4256,
+      "step": 4906
+    },
+    {
+      "epoch": 0.9674684542586751,
+      "grad_norm": 0.5640572708097991,
+      "learning_rate": 1.726557630691672e-05,
+      "loss": 0.4343,
+      "step": 4907
+    },
+    {
+      "epoch": 0.9676656151419558,
+      "grad_norm": 0.5552212931426607,
+      "learning_rate": 1.7264511188218812e-05,
+      "loss": 0.4653,
+      "step": 4908
+    },
+    {
+      "epoch": 0.9678627760252366,
+      "grad_norm": 1.0633900724505037,
+      "learning_rate": 1.726344589498517e-05,
+      "loss": 0.4826,
+      "step": 4909
+    },
+    {
+      "epoch": 0.9680599369085173,
+      "grad_norm": 0.7383588772446893,
+      "learning_rate": 1.7262380427241394e-05,
+      "loss": 0.4684,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9682570977917981,
+      "grad_norm": 0.6826171176254585,
+      "learning_rate": 1.7261314785013078e-05,
+      "loss": 0.4237,
+      "step": 4911
+    },
+    {
+      "epoch": 0.9684542586750788,
+      "grad_norm": 0.6364722722252665,
+      "learning_rate": 1.7260248968325828e-05,
+      "loss": 0.4489,
+      "step": 4912
+    },
+    {
+      "epoch": 0.9686514195583596,
+      "grad_norm": 0.5663790368688596,
+      "learning_rate": 1.7259182977205248e-05,
+      "loss": 0.4189,
+      "step": 4913
+    },
+    {
+      "epoch": 0.9688485804416404,
+      "grad_norm": 0.5420023619747468,
+      "learning_rate": 1.7258116811676956e-05,
+      "loss": 0.4295,
+      "step": 4914
+    },
+    {
+      "epoch": 0.9690457413249212,
+      "grad_norm": 0.5377721001533368,
+      "learning_rate": 1.7257050471766558e-05,
+      "loss": 0.4638,
+      "step": 4915
+    },
+    {
+      "epoch": 0.9692429022082019,
+      "grad_norm": 0.6046459705426991,
+      "learning_rate": 1.7255983957499676e-05,
+      "loss": 0.4886,
+      "step": 4916
+    },
+    {
+      "epoch": 0.9694400630914827,
+      "grad_norm": 0.5236164202200155,
+      "learning_rate": 1.7254917268901942e-05,
+      "loss": 0.4251,
+      "step": 4917
+    },
+    {
+      "epoch": 0.9696372239747634,
+      "grad_norm": 0.5378880369504965,
+      "learning_rate": 1.7253850405998976e-05,
+      "loss": 0.4662,
+      "step": 4918
+    },
+    {
+      "epoch": 0.9698343848580442,
+      "grad_norm": 0.5749445692065384,
+      "learning_rate": 1.7252783368816413e-05,
+      "loss": 0.4855,
+      "step": 4919
+    },
+    {
+      "epoch": 0.9700315457413249,
+      "grad_norm": 0.49857545471871095,
+      "learning_rate": 1.7251716157379887e-05,
+      "loss": 0.4235,
+      "step": 4920
+    },
+    {
+      "epoch": 0.9702287066246057,
+      "grad_norm": 0.5825438945665429,
+      "learning_rate": 1.725064877171504e-05,
+      "loss": 0.4292,
+      "step": 4921
+    },
+    {
+      "epoch": 0.9704258675078864,
+      "grad_norm": 0.7116226202261247,
+      "learning_rate": 1.724958121184752e-05,
+      "loss": 0.4942,
+      "step": 4922
+    },
+    {
+      "epoch": 0.9706230283911672,
+      "grad_norm": 0.5595515787710044,
+      "learning_rate": 1.7248513477802973e-05,
+      "loss": 0.4251,
+      "step": 4923
+    },
+    {
+      "epoch": 0.9708201892744479,
+      "grad_norm": 0.5609266287961348,
+      "learning_rate": 1.724744556960705e-05,
+      "loss": 0.4362,
+      "step": 4924
+    },
+    {
+      "epoch": 0.9710173501577287,
+      "grad_norm": 0.5332114412325697,
+      "learning_rate": 1.7246377487285415e-05,
+      "loss": 0.4222,
+      "step": 4925
+    },
+    {
+      "epoch": 0.9712145110410094,
+      "grad_norm": 0.5579951518622186,
+      "learning_rate": 1.7245309230863723e-05,
+      "loss": 0.4351,
+      "step": 4926
+    },
+    {
+      "epoch": 0.9714116719242902,
+      "grad_norm": 0.5504796489083765,
+      "learning_rate": 1.7244240800367642e-05,
+      "loss": 0.466,
+      "step": 4927
+    },
+    {
+      "epoch": 0.9716088328075709,
+      "grad_norm": 0.5364866603592925,
+      "learning_rate": 1.724317219582284e-05,
+      "loss": 0.4356,
+      "step": 4928
+    },
+    {
+      "epoch": 0.9718059936908517,
+      "grad_norm": 0.5307686124133709,
+      "learning_rate": 1.7242103417255e-05,
+      "loss": 0.3943,
+      "step": 4929
+    },
+    {
+      "epoch": 0.9720031545741324,
+      "grad_norm": 0.535779232578516,
+      "learning_rate": 1.724103446468979e-05,
+      "loss": 0.4479,
+      "step": 4930
+    },
+    {
+      "epoch": 0.9722003154574133,
+      "grad_norm": 0.8527430234546016,
+      "learning_rate": 1.723996533815289e-05,
+      "loss": 0.4271,
+      "step": 4931
+    },
+    {
+      "epoch": 0.972397476340694,
+      "grad_norm": 0.4893815242972822,
+      "learning_rate": 1.723889603767e-05,
+      "loss": 0.4294,
+      "step": 4932
+    },
+    {
+      "epoch": 0.9725946372239748,
+      "grad_norm": 0.5142240975494096,
+      "learning_rate": 1.7237826563266797e-05,
+      "loss": 0.4111,
+      "step": 4933
+    },
+    {
+      "epoch": 0.9727917981072555,
+      "grad_norm": 0.5056050398485274,
+      "learning_rate": 1.7236756914968985e-05,
+      "loss": 0.4459,
+      "step": 4934
+    },
+    {
+      "epoch": 0.9729889589905363,
+      "grad_norm": 0.6875911603497251,
+      "learning_rate": 1.723568709280226e-05,
+      "loss": 0.4393,
+      "step": 4935
+    },
+    {
+      "epoch": 0.973186119873817,
+      "grad_norm": 0.5410138293600987,
+      "learning_rate": 1.7234617096792328e-05,
+      "loss": 0.4337,
+      "step": 4936
+    },
+    {
+      "epoch": 0.9733832807570978,
+      "grad_norm": 0.5406394954943095,
+      "learning_rate": 1.723354692696489e-05,
+      "loss": 0.4787,
+      "step": 4937
+    },
+    {
+      "epoch": 0.9735804416403786,
+      "grad_norm": 0.5039617193695565,
+      "learning_rate": 1.7232476583345667e-05,
+      "loss": 0.4269,
+      "step": 4938
+    },
+    {
+      "epoch": 0.9737776025236593,
+      "grad_norm": 0.5667126358374103,
+      "learning_rate": 1.7231406065960365e-05,
+      "loss": 0.4314,
+      "step": 4939
+    },
+    {
+      "epoch": 0.9739747634069401,
+      "grad_norm": 0.5223138203259416,
+      "learning_rate": 1.723033537483471e-05,
+      "loss": 0.4238,
+      "step": 4940
+    },
+    {
+      "epoch": 0.9741719242902208,
+      "grad_norm": 0.5134992034885483,
+      "learning_rate": 1.722926450999443e-05,
+      "loss": 0.435,
+      "step": 4941
+    },
+    {
+      "epoch": 0.9743690851735016,
+      "grad_norm": 0.5335314712456779,
+      "learning_rate": 1.7228193471465243e-05,
+      "loss": 0.4448,
+      "step": 4942
+    },
+    {
+      "epoch": 0.9745662460567823,
+      "grad_norm": 0.5187220320984581,
+      "learning_rate": 1.722712225927289e-05,
+      "loss": 0.4305,
+      "step": 4943
+    },
+    {
+      "epoch": 0.9747634069400631,
+      "grad_norm": 0.5236993505503822,
+      "learning_rate": 1.7226050873443103e-05,
+      "loss": 0.4324,
+      "step": 4944
+    },
+    {
+      "epoch": 0.9749605678233438,
+      "grad_norm": 0.5501948139908517,
+      "learning_rate": 1.7224979314001623e-05,
+      "loss": 0.4445,
+      "step": 4945
+    },
+    {
+      "epoch": 0.9751577287066246,
+      "grad_norm": 0.5025205140860234,
+      "learning_rate": 1.72239075809742e-05,
+      "loss": 0.4067,
+      "step": 4946
+    },
+    {
+      "epoch": 0.9753548895899053,
+      "grad_norm": 0.509021011008656,
+      "learning_rate": 1.722283567438658e-05,
+      "loss": 0.4127,
+      "step": 4947
+    },
+    {
+      "epoch": 0.9755520504731862,
+      "grad_norm": 0.5169322695492775,
+      "learning_rate": 1.7221763594264513e-05,
+      "loss": 0.4183,
+      "step": 4948
+    },
+    {
+      "epoch": 0.9757492113564669,
+      "grad_norm": 0.5893693443229753,
+      "learning_rate": 1.7220691340633762e-05,
+      "loss": 0.413,
+      "step": 4949
+    },
+    {
+      "epoch": 0.9759463722397477,
+      "grad_norm": 0.5032678433324692,
+      "learning_rate": 1.7219618913520086e-05,
+      "loss": 0.3999,
+      "step": 4950
+    },
+    {
+      "epoch": 0.9761435331230284,
+      "grad_norm": 0.5611832448812477,
+      "learning_rate": 1.7218546312949255e-05,
+      "loss": 0.4208,
+      "step": 4951
+    },
+    {
+      "epoch": 0.9763406940063092,
+      "grad_norm": 0.5402039539918267,
+      "learning_rate": 1.7217473538947032e-05,
+      "loss": 0.4592,
+      "step": 4952
+    },
+    {
+      "epoch": 0.9765378548895899,
+      "grad_norm": 0.5312684422263533,
+      "learning_rate": 1.7216400591539194e-05,
+      "loss": 0.3958,
+      "step": 4953
+    },
+    {
+      "epoch": 0.9767350157728707,
+      "grad_norm": 0.4838095794471189,
+      "learning_rate": 1.7215327470751525e-05,
+      "loss": 0.4012,
+      "step": 4954
+    },
+    {
+      "epoch": 0.9769321766561514,
+      "grad_norm": 0.5459043063215803,
+      "learning_rate": 1.72142541766098e-05,
+      "loss": 0.4889,
+      "step": 4955
+    },
+    {
+      "epoch": 0.9771293375394322,
+      "grad_norm": 0.4762352325018086,
+      "learning_rate": 1.721318070913981e-05,
+      "loss": 0.4194,
+      "step": 4956
+    },
+    {
+      "epoch": 0.9773264984227129,
+      "grad_norm": 1.1041727399927024,
+      "learning_rate": 1.7212107068367343e-05,
+      "loss": 0.4595,
+      "step": 4957
+    },
+    {
+      "epoch": 0.9775236593059937,
+      "grad_norm": 0.5072331377511347,
+      "learning_rate": 1.7211033254318195e-05,
+      "loss": 0.4192,
+      "step": 4958
+    },
+    {
+      "epoch": 0.9777208201892744,
+      "grad_norm": 0.5051945491324831,
+      "learning_rate": 1.720995926701817e-05,
+      "loss": 0.4018,
+      "step": 4959
+    },
+    {
+      "epoch": 0.9779179810725552,
+      "grad_norm": 0.5317945948669217,
+      "learning_rate": 1.7208885106493068e-05,
+      "loss": 0.4245,
+      "step": 4960
+    },
+    {
+      "epoch": 0.9781151419558359,
+      "grad_norm": 0.5003276873104907,
+      "learning_rate": 1.7207810772768692e-05,
+      "loss": 0.4175,
+      "step": 4961
+    },
+    {
+      "epoch": 0.9783123028391167,
+      "grad_norm": 0.5129068918599115,
+      "learning_rate": 1.720673626587086e-05,
+      "loss": 0.4263,
+      "step": 4962
+    },
+    {
+      "epoch": 0.9785094637223974,
+      "grad_norm": 0.49567393969405865,
+      "learning_rate": 1.7205661585825385e-05,
+      "loss": 0.4224,
+      "step": 4963
+    },
+    {
+      "epoch": 0.9787066246056783,
+      "grad_norm": 0.5210701609022403,
+      "learning_rate": 1.7204586732658088e-05,
+      "loss": 0.4054,
+      "step": 4964
+    },
+    {
+      "epoch": 0.978903785488959,
+      "grad_norm": 0.5237810329795967,
+      "learning_rate": 1.720351170639479e-05,
+      "loss": 0.4373,
+      "step": 4965
+    },
+    {
+      "epoch": 0.9791009463722398,
+      "grad_norm": 0.49172980877968686,
+      "learning_rate": 1.7202436507061327e-05,
+      "loss": 0.3981,
+      "step": 4966
+    },
+    {
+      "epoch": 0.9792981072555205,
+      "grad_norm": 0.5033156694120647,
+      "learning_rate": 1.7201361134683522e-05,
+      "loss": 0.405,
+      "step": 4967
+    },
+    {
+      "epoch": 0.9794952681388013,
+      "grad_norm": 0.532232285017258,
+      "learning_rate": 1.720028558928722e-05,
+      "loss": 0.4189,
+      "step": 4968
+    },
+    {
+      "epoch": 0.979692429022082,
+      "grad_norm": 0.5084865188261308,
+      "learning_rate": 1.7199209870898257e-05,
+      "loss": 0.426,
+      "step": 4969
+    },
+    {
+      "epoch": 0.9798895899053628,
+      "grad_norm": 0.5195794855402261,
+      "learning_rate": 1.719813397954248e-05,
+      "loss": 0.4279,
+      "step": 4970
+    },
+    {
+      "epoch": 0.9800867507886435,
+      "grad_norm": 0.5120454170989818,
+      "learning_rate": 1.7197057915245738e-05,
+      "loss": 0.4465,
+      "step": 4971
+    },
+    {
+      "epoch": 0.9802839116719243,
+      "grad_norm": 0.5104835439357177,
+      "learning_rate": 1.7195981678033883e-05,
+      "loss": 0.4551,
+      "step": 4972
+    },
+    {
+      "epoch": 0.980481072555205,
+      "grad_norm": 0.545753139194094,
+      "learning_rate": 1.7194905267932775e-05,
+      "loss": 0.4515,
+      "step": 4973
+    },
+    {
+      "epoch": 0.9806782334384858,
+      "grad_norm": 0.48063719552920575,
+      "learning_rate": 1.719382868496827e-05,
+      "loss": 0.3959,
+      "step": 4974
+    },
+    {
+      "epoch": 0.9808753943217665,
+      "grad_norm": 0.5081444056915589,
+      "learning_rate": 1.7192751929166237e-05,
+      "loss": 0.4267,
+      "step": 4975
+    },
+    {
+      "epoch": 0.9810725552050473,
+      "grad_norm": 0.7889320206894174,
+      "learning_rate": 1.7191675000552552e-05,
+      "loss": 0.4256,
+      "step": 4976
+    },
+    {
+      "epoch": 0.981269716088328,
+      "grad_norm": 0.5161493421048737,
+      "learning_rate": 1.719059789915308e-05,
+      "loss": 0.3944,
+      "step": 4977
+    },
+    {
+      "epoch": 0.9814668769716088,
+      "grad_norm": 0.5526599869469584,
+      "learning_rate": 1.7189520624993706e-05,
+      "loss": 0.4733,
+      "step": 4978
+    },
+    {
+      "epoch": 0.9816640378548895,
+      "grad_norm": 0.5921214088474026,
+      "learning_rate": 1.7188443178100306e-05,
+      "loss": 0.4402,
+      "step": 4979
+    },
+    {
+      "epoch": 0.9818611987381703,
+      "grad_norm": 0.4899387375326017,
+      "learning_rate": 1.7187365558498772e-05,
+      "loss": 0.403,
+      "step": 4980
+    },
+    {
+      "epoch": 0.982058359621451,
+      "grad_norm": 0.6284691568651574,
+      "learning_rate": 1.7186287766214992e-05,
+      "loss": 0.4247,
+      "step": 4981
+    },
+    {
+      "epoch": 0.9822555205047319,
+      "grad_norm": 0.5219769363772543,
+      "learning_rate": 1.7185209801274863e-05,
+      "loss": 0.4378,
+      "step": 4982
+    },
+    {
+      "epoch": 0.9824526813880127,
+      "grad_norm": 0.561058268499637,
+      "learning_rate": 1.718413166370428e-05,
+      "loss": 0.4402,
+      "step": 4983
+    },
+    {
+      "epoch": 0.9826498422712934,
+      "grad_norm": 0.5228751281665158,
+      "learning_rate": 1.7183053353529146e-05,
+      "loss": 0.4342,
+      "step": 4984
+    },
+    {
+      "epoch": 0.9828470031545742,
+      "grad_norm": 0.6359423300026139,
+      "learning_rate": 1.7181974870775374e-05,
+      "loss": 0.4791,
+      "step": 4985
+    },
+    {
+      "epoch": 0.9830441640378549,
+      "grad_norm": 0.5452791966413713,
+      "learning_rate": 1.718089621546887e-05,
+      "loss": 0.4311,
+      "step": 4986
+    },
+    {
+      "epoch": 0.9832413249211357,
+      "grad_norm": 0.5585163320278087,
+      "learning_rate": 1.7179817387635552e-05,
+      "loss": 0.3905,
+      "step": 4987
+    },
+    {
+      "epoch": 0.9834384858044164,
+      "grad_norm": 1.1312602954702684,
+      "learning_rate": 1.7178738387301342e-05,
+      "loss": 0.5146,
+      "step": 4988
+    },
+    {
+      "epoch": 0.9836356466876972,
+      "grad_norm": 0.544619439438417,
+      "learning_rate": 1.7177659214492162e-05,
+      "loss": 0.4722,
+      "step": 4989
+    },
+    {
+      "epoch": 0.9838328075709779,
+      "grad_norm": 0.48850131893298554,
+      "learning_rate": 1.7176579869233935e-05,
+      "loss": 0.387,
+      "step": 4990
+    },
+    {
+      "epoch": 0.9840299684542587,
+      "grad_norm": 0.646046327237621,
+      "learning_rate": 1.71755003515526e-05,
+      "loss": 0.4457,
+      "step": 4991
+    },
+    {
+      "epoch": 0.9842271293375394,
+      "grad_norm": 0.5156158640893732,
+      "learning_rate": 1.717442066147409e-05,
+      "loss": 0.4206,
+      "step": 4992
+    },
+    {
+      "epoch": 0.9844242902208202,
+      "grad_norm": 0.5424189331510925,
+      "learning_rate": 1.7173340799024346e-05,
+      "loss": 0.397,
+      "step": 4993
+    },
+    {
+      "epoch": 0.9846214511041009,
+      "grad_norm": 0.5235819964470497,
+      "learning_rate": 1.7172260764229312e-05,
+      "loss": 0.4467,
+      "step": 4994
+    },
+    {
+      "epoch": 0.9848186119873817,
+      "grad_norm": 0.5909753498690635,
+      "learning_rate": 1.717118055711494e-05,
+      "loss": 0.4398,
+      "step": 4995
+    },
+    {
+      "epoch": 0.9850157728706624,
+      "grad_norm": 0.4987198779431939,
+      "learning_rate": 1.7170100177707177e-05,
+      "loss": 0.4332,
+      "step": 4996
+    },
+    {
+      "epoch": 0.9852129337539433,
+      "grad_norm": 0.5563239550956313,
+      "learning_rate": 1.7169019626031985e-05,
+      "loss": 0.4193,
+      "step": 4997
+    },
+    {
+      "epoch": 0.985410094637224,
+      "grad_norm": 0.542094709088602,
+      "learning_rate": 1.7167938902115323e-05,
+      "loss": 0.4354,
+      "step": 4998
+    },
+    {
+      "epoch": 0.9856072555205048,
+      "grad_norm": 0.6222848829402873,
+      "learning_rate": 1.7166858005983154e-05,
+      "loss": 0.4455,
+      "step": 4999
+    },
+    {
+      "epoch": 0.9858044164037855,
+      "grad_norm": 0.5278021757853874,
+      "learning_rate": 1.7165776937661453e-05,
+      "loss": 0.4299,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9860015772870663,
+      "grad_norm": 0.6100419080088035,
+      "learning_rate": 1.716469569717619e-05,
+      "loss": 0.3916,
+      "step": 5001
+    },
+    {
+      "epoch": 0.986198738170347,
+      "grad_norm": 0.5391831742527743,
+      "learning_rate": 1.716361428455334e-05,
+      "loss": 0.4103,
+      "step": 5002
+    },
+    {
+      "epoch": 0.9863958990536278,
+      "grad_norm": 0.5826789977832045,
+      "learning_rate": 1.7162532699818893e-05,
+      "loss": 0.4617,
+      "step": 5003
+    },
+    {
+      "epoch": 0.9865930599369085,
+      "grad_norm": 0.4977680784543337,
+      "learning_rate": 1.7161450942998827e-05,
+      "loss": 0.4102,
+      "step": 5004
+    },
+    {
+      "epoch": 0.9867902208201893,
+      "grad_norm": 0.5793799937360323,
+      "learning_rate": 1.7160369014119136e-05,
+      "loss": 0.424,
+      "step": 5005
+    },
+    {
+      "epoch": 0.98698738170347,
+      "grad_norm": 0.5113962294204524,
+      "learning_rate": 1.7159286913205813e-05,
+      "loss": 0.4317,
+      "step": 5006
+    },
+    {
+      "epoch": 0.9871845425867508,
+      "grad_norm": 0.5398003533758813,
+      "learning_rate": 1.7158204640284855e-05,
+      "loss": 0.445,
+      "step": 5007
+    },
+    {
+      "epoch": 0.9873817034700315,
+      "grad_norm": 0.5979699074869115,
+      "learning_rate": 1.7157122195382267e-05,
+      "loss": 0.452,
+      "step": 5008
+    },
+    {
+      "epoch": 0.9875788643533123,
+      "grad_norm": 0.5285476936820424,
+      "learning_rate": 1.7156039578524055e-05,
+      "loss": 0.4193,
+      "step": 5009
+    },
+    {
+      "epoch": 0.987776025236593,
+      "grad_norm": 0.5910776224079434,
+      "learning_rate": 1.715495678973623e-05,
+      "loss": 0.4859,
+      "step": 5010
+    },
+    {
+      "epoch": 0.9879731861198738,
+      "grad_norm": 0.5036681780427391,
+      "learning_rate": 1.7153873829044805e-05,
+      "loss": 0.4086,
+      "step": 5011
+    },
+    {
+      "epoch": 0.9881703470031545,
+      "grad_norm": 0.5363578632899204,
+      "learning_rate": 1.7152790696475804e-05,
+      "loss": 0.4109,
+      "step": 5012
+    },
+    {
+      "epoch": 0.9883675078864353,
+      "grad_norm": 0.5058850256040565,
+      "learning_rate": 1.715170739205524e-05,
+      "loss": 0.417,
+      "step": 5013
+    },
+    {
+      "epoch": 0.988564668769716,
+      "grad_norm": 0.5343518165708425,
+      "learning_rate": 1.7150623915809154e-05,
+      "loss": 0.4317,
+      "step": 5014
+    },
+    {
+      "epoch": 0.9887618296529969,
+      "grad_norm": 0.4732197281950188,
+      "learning_rate": 1.7149540267763566e-05,
+      "loss": 0.419,
+      "step": 5015
+    },
+    {
+      "epoch": 0.9889589905362776,
+      "grad_norm": 0.5460018301214419,
+      "learning_rate": 1.7148456447944514e-05,
+      "loss": 0.4371,
+      "step": 5016
+    },
+    {
+      "epoch": 0.9891561514195584,
+      "grad_norm": 0.5228662093888589,
+      "learning_rate": 1.714737245637804e-05,
+      "loss": 0.4348,
+      "step": 5017
+    },
+    {
+      "epoch": 0.9893533123028391,
+      "grad_norm": 0.5136606982937066,
+      "learning_rate": 1.7146288293090187e-05,
+      "loss": 0.4207,
+      "step": 5018
+    },
+    {
+      "epoch": 0.9895504731861199,
+      "grad_norm": 0.49381060992162196,
+      "learning_rate": 1.7145203958107005e-05,
+      "loss": 0.4386,
+      "step": 5019
+    },
+    {
+      "epoch": 0.9897476340694006,
+      "grad_norm": 0.5097224069523834,
+      "learning_rate": 1.714411945145454e-05,
+      "loss": 0.449,
+      "step": 5020
+    },
+    {
+      "epoch": 0.9899447949526814,
+      "grad_norm": 0.48982369331094644,
+      "learning_rate": 1.714303477315886e-05,
+      "loss": 0.4135,
+      "step": 5021
+    },
+    {
+      "epoch": 0.9901419558359621,
+      "grad_norm": 0.5348342417259278,
+      "learning_rate": 1.7141949923246007e-05,
+      "loss": 0.4645,
+      "step": 5022
+    },
+    {
+      "epoch": 0.9903391167192429,
+      "grad_norm": 0.4935854194155088,
+      "learning_rate": 1.7140864901742062e-05,
+      "loss": 0.4156,
+      "step": 5023
+    },
+    {
+      "epoch": 0.9905362776025236,
+      "grad_norm": 0.538994627268955,
+      "learning_rate": 1.7139779708673084e-05,
+      "loss": 0.434,
+      "step": 5024
+    },
+    {
+      "epoch": 0.9907334384858044,
+      "grad_norm": 0.5072487466974747,
+      "learning_rate": 1.7138694344065152e-05,
+      "loss": 0.4265,
+      "step": 5025
+    },
+    {
+      "epoch": 0.9909305993690851,
+      "grad_norm": 14.35107865712939,
+      "learning_rate": 1.7137608807944337e-05,
+      "loss": 0.4277,
+      "step": 5026
+    },
+    {
+      "epoch": 0.9911277602523659,
+      "grad_norm": 0.5905049052128314,
+      "learning_rate": 1.7136523100336725e-05,
+      "loss": 0.4204,
+      "step": 5027
+    },
+    {
+      "epoch": 0.9913249211356467,
+      "grad_norm": 0.5954561128927663,
+      "learning_rate": 1.7135437221268397e-05,
+      "loss": 0.4066,
+      "step": 5028
+    },
+    {
+      "epoch": 0.9915220820189274,
+      "grad_norm": 0.5481450959305048,
+      "learning_rate": 1.7134351170765443e-05,
+      "loss": 0.434,
+      "step": 5029
+    },
+    {
+      "epoch": 0.9917192429022083,
+      "grad_norm": 0.5757944206123269,
+      "learning_rate": 1.7133264948853957e-05,
+      "loss": 0.4157,
+      "step": 5030
+    },
+    {
+      "epoch": 0.991916403785489,
+      "grad_norm": 0.5811091665563136,
+      "learning_rate": 1.7132178555560038e-05,
+      "loss": 0.463,
+      "step": 5031
+    },
+    {
+      "epoch": 0.9921135646687698,
+      "grad_norm": 0.5095692078366557,
+      "learning_rate": 1.7131091990909786e-05,
+      "loss": 0.4403,
+      "step": 5032
+    },
+    {
+      "epoch": 0.9923107255520505,
+      "grad_norm": 0.5285063457801826,
+      "learning_rate": 1.71300052549293e-05,
+      "loss": 0.4018,
+      "step": 5033
+    },
+    {
+      "epoch": 0.9925078864353313,
+      "grad_norm": 0.5743875490503324,
+      "learning_rate": 1.7128918347644704e-05,
+      "loss": 0.463,
+      "step": 5034
+    },
+    {
+      "epoch": 0.992705047318612,
+      "grad_norm": 0.5553066045950534,
+      "learning_rate": 1.7127831269082103e-05,
+      "loss": 0.4358,
+      "step": 5035
+    },
+    {
+      "epoch": 0.9929022082018928,
+      "grad_norm": 0.5201145328150542,
+      "learning_rate": 1.712674401926761e-05,
+      "loss": 0.3973,
+      "step": 5036
+    },
+    {
+      "epoch": 0.9930993690851735,
+      "grad_norm": 0.5308940515290245,
+      "learning_rate": 1.7125656598227357e-05,
+      "loss": 0.4505,
+      "step": 5037
+    },
+    {
+      "epoch": 0.9932965299684543,
+      "grad_norm": 0.5307597719587108,
+      "learning_rate": 1.7124569005987466e-05,
+      "loss": 0.4296,
+      "step": 5038
+    },
+    {
+      "epoch": 0.993493690851735,
+      "grad_norm": 0.5368886391034855,
+      "learning_rate": 1.7123481242574066e-05,
+      "loss": 0.4627,
+      "step": 5039
+    },
+    {
+      "epoch": 0.9936908517350158,
+      "grad_norm": 0.5304804270019259,
+      "learning_rate": 1.7122393308013294e-05,
+      "loss": 0.4513,
+      "step": 5040
+    },
+    {
+      "epoch": 0.9938880126182965,
+      "grad_norm": 0.4837269924708032,
+      "learning_rate": 1.7121305202331284e-05,
+      "loss": 0.4263,
+      "step": 5041
+    },
+    {
+      "epoch": 0.9940851735015773,
+      "grad_norm": 0.538409666606145,
+      "learning_rate": 1.7120216925554185e-05,
+      "loss": 0.4424,
+      "step": 5042
+    },
+    {
+      "epoch": 0.994282334384858,
+      "grad_norm": 0.5107171099135106,
+      "learning_rate": 1.7119128477708137e-05,
+      "loss": 0.4143,
+      "step": 5043
+    },
+    {
+      "epoch": 0.9944794952681388,
+      "grad_norm": 0.5109327464115445,
+      "learning_rate": 1.7118039858819297e-05,
+      "loss": 0.4406,
+      "step": 5044
+    },
+    {
+      "epoch": 0.9946766561514195,
+      "grad_norm": 0.528632568268682,
+      "learning_rate": 1.711695106891382e-05,
+      "loss": 0.4322,
+      "step": 5045
+    },
+    {
+      "epoch": 0.9948738170347003,
+      "grad_norm": 0.5281250909959354,
+      "learning_rate": 1.711586210801786e-05,
+      "loss": 0.4486,
+      "step": 5046
+    },
+    {
+      "epoch": 0.995070977917981,
+      "grad_norm": 0.5636600965669576,
+      "learning_rate": 1.7114772976157578e-05,
+      "loss": 0.4576,
+      "step": 5047
+    },
+    {
+      "epoch": 0.9952681388012619,
+      "grad_norm": 0.5721847302432999,
+      "learning_rate": 1.711368367335915e-05,
+      "loss": 0.4841,
+      "step": 5048
+    },
+    {
+      "epoch": 0.9954652996845426,
+      "grad_norm": 0.4984992031526205,
+      "learning_rate": 1.7112594199648742e-05,
+      "loss": 0.4029,
+      "step": 5049
+    },
+    {
+      "epoch": 0.9956624605678234,
+      "grad_norm": 0.4985204521738928,
+      "learning_rate": 1.7111504555052533e-05,
+      "loss": 0.4062,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9958596214511041,
+      "grad_norm": 0.510291189780098,
+      "learning_rate": 1.7110414739596697e-05,
+      "loss": 0.4291,
+      "step": 5051
+    },
+    {
+      "epoch": 0.9960567823343849,
+      "grad_norm": 0.5643661497979349,
+      "learning_rate": 1.710932475330742e-05,
+      "loss": 0.429,
+      "step": 5052
+    },
+    {
+      "epoch": 0.9962539432176656,
+      "grad_norm": 0.5118534697302934,
+      "learning_rate": 1.7108234596210892e-05,
+      "loss": 0.4399,
+      "step": 5053
+    },
+    {
+      "epoch": 0.9964511041009464,
+      "grad_norm": 0.49932647918668954,
+      "learning_rate": 1.7107144268333307e-05,
+      "loss": 0.383,
+      "step": 5054
+    },
+    {
+      "epoch": 0.9966482649842271,
+      "grad_norm": 0.5207920247222175,
+      "learning_rate": 1.7106053769700855e-05,
+      "loss": 0.4259,
+      "step": 5055
+    },
+    {
+      "epoch": 0.9968454258675079,
+      "grad_norm": 0.5040712672787065,
+      "learning_rate": 1.7104963100339738e-05,
+      "loss": 0.4396,
+      "step": 5056
+    },
+    {
+      "epoch": 0.9970425867507886,
+      "grad_norm": 0.49364985751506657,
+      "learning_rate": 1.7103872260276163e-05,
+      "loss": 0.3975,
+      "step": 5057
+    },
+    {
+      "epoch": 0.9972397476340694,
+      "grad_norm": 0.5562215293332458,
+      "learning_rate": 1.7102781249536333e-05,
+      "loss": 0.4626,
+      "step": 5058
+    },
+    {
+      "epoch": 0.9974369085173501,
+      "grad_norm": 0.5298513969515571,
+      "learning_rate": 1.7101690068146466e-05,
+      "loss": 0.4142,
+      "step": 5059
+    },
+    {
+      "epoch": 0.9976340694006309,
+      "grad_norm": 0.54683487357073,
+      "learning_rate": 1.7100598716132775e-05,
+      "loss": 0.4327,
+      "step": 5060
+    },
+    {
+      "epoch": 0.9978312302839116,
+      "grad_norm": 0.5440040494015383,
+      "learning_rate": 1.7099507193521482e-05,
+      "loss": 0.462,
+      "step": 5061
+    },
+    {
+      "epoch": 0.9980283911671924,
+      "grad_norm": 0.49509844160455657,
+      "learning_rate": 1.709841550033881e-05,
+      "loss": 0.4183,
+      "step": 5062
+    },
+    {
+      "epoch": 0.9982255520504731,
+      "grad_norm": 1.056087374364193,
+      "learning_rate": 1.7097323636610992e-05,
+      "loss": 0.4725,
+      "step": 5063
+    },
+    {
+      "epoch": 0.998422712933754,
+      "grad_norm": 0.5237038370525577,
+      "learning_rate": 1.7096231602364257e-05,
+      "loss": 0.414,
+      "step": 5064
+    },
+    {
+      "epoch": 0.9986198738170347,
+      "grad_norm": 0.5948843667990235,
+      "learning_rate": 1.7095139397624843e-05,
+      "loss": 0.4527,
+      "step": 5065
+    },
+    {
+      "epoch": 0.9988170347003155,
+      "grad_norm": 0.5150733173683878,
+      "learning_rate": 1.7094047022418995e-05,
+      "loss": 0.4904,
+      "step": 5066
+    },
+    {
+      "epoch": 0.9990141955835962,
+      "grad_norm": 0.7227117490491202,
+      "learning_rate": 1.709295447677295e-05,
+      "loss": 0.4206,
+      "step": 5067
+    },
+    {
+      "epoch": 0.999211356466877,
+      "grad_norm": 0.5268898131828004,
+      "learning_rate": 1.7091861760712963e-05,
+      "loss": 0.4402,
+      "step": 5068
+    },
+    {
+      "epoch": 0.9994085173501577,
+      "grad_norm": 0.4895461883001276,
+      "learning_rate": 1.7090768874265285e-05,
+      "loss": 0.43,
+      "step": 5069
+    },
+    {
+      "epoch": 0.9996056782334385,
+      "grad_norm": 0.5224645141727187,
+      "learning_rate": 1.7089675817456175e-05,
+      "loss": 0.4484,
+      "step": 5070
+    },
+    {
+      "epoch": 0.9998028391167192,
+      "grad_norm": 0.5380192178833433,
+      "learning_rate": 1.7088582590311896e-05,
+      "loss": 0.4619,
+      "step": 5071
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.48949137873370546,
+      "learning_rate": 1.708748919285871e-05,
+      "loss": 0.4243,
+      "step": 5072
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.4332016706466675,
+      "eval_runtime": 344.5029,
+      "eval_samples_per_second": 23.599,
+      "eval_steps_per_second": 1.477,
+      "step": 5072
+    },
+    {
+      "epoch": 1.0001971608832807,
+      "grad_norm": 0.5212823866085835,
+      "learning_rate": 1.7086395625122888e-05,
+      "loss": 0.4414,
+      "step": 5073
+    },
+    {
+      "epoch": 1.0001971608832807,
+      "grad_norm": 0.5373287163559608,
+      "learning_rate": 1.7085301887130708e-05,
+      "loss": 0.3227,
+      "step": 5074
+    },
+    {
+      "epoch": 1.0003943217665616,
+      "grad_norm": 0.6038566158503993,
+      "learning_rate": 1.708420797890844e-05,
+      "loss": 0.3653,
+      "step": 5075
+    },
+    {
+      "epoch": 1.0005914826498423,
+      "grad_norm": 0.6164839814322967,
+      "learning_rate": 1.7083113900482374e-05,
+      "loss": 0.3345,
+      "step": 5076
+    },
+    {
+      "epoch": 1.000788643533123,
+      "grad_norm": 0.5898548067122023,
+      "learning_rate": 1.708201965187879e-05,
+      "loss": 0.2978,
+      "step": 5077
+    },
+    {
+      "epoch": 1.0009858044164037,
+      "grad_norm": 0.6086636984908,
+      "learning_rate": 1.708092523312398e-05,
+      "loss": 0.3338,
+      "step": 5078
+    },
+    {
+      "epoch": 1.0011829652996846,
+      "grad_norm": 0.85830944121542,
+      "learning_rate": 1.707983064424424e-05,
+      "loss": 0.3628,
+      "step": 5079
+    },
+    {
+      "epoch": 1.0013801261829653,
+      "grad_norm": 0.7446958830040732,
+      "learning_rate": 1.7078735885265872e-05,
+      "loss": 0.3397,
+      "step": 5080
+    },
+    {
+      "epoch": 1.001577287066246,
+      "grad_norm": 0.5753953072853087,
+      "learning_rate": 1.707764095621517e-05,
+      "loss": 0.3338,
+      "step": 5081
+    },
+    {
+      "epoch": 1.0017744479495267,
+      "grad_norm": 0.5313937087435588,
+      "learning_rate": 1.707654585711844e-05,
+      "loss": 0.3294,
+      "step": 5082
+    },
+    {
+      "epoch": 1.0019716088328077,
+      "grad_norm": 0.5433192534151587,
+      "learning_rate": 1.7075450588002004e-05,
+      "loss": 0.3005,
+      "step": 5083
+    },
+    {
+      "epoch": 1.0021687697160884,
+      "grad_norm": 0.5974671475978584,
+      "learning_rate": 1.7074355148892167e-05,
+      "loss": 0.3355,
+      "step": 5084
+    },
+    {
+      "epoch": 1.002365930599369,
+      "grad_norm": 0.5292036690619668,
+      "learning_rate": 1.707325953981525e-05,
+      "loss": 0.3133,
+      "step": 5085
+    },
+    {
+      "epoch": 1.0025630914826498,
+      "grad_norm": 0.5791264643946715,
+      "learning_rate": 1.707216376079758e-05,
+      "loss": 0.3628,
+      "step": 5086
+    },
+    {
+      "epoch": 1.0027602523659307,
+      "grad_norm": 0.5367457017037465,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.323,
+      "step": 5087
+    },
+    {
+      "epoch": 1.0029574132492114,
+      "grad_norm": 0.5463141350198112,
+      "learning_rate": 1.7069971693045276e-05,
+      "loss": 0.3191,
+      "step": 5088
+    },
+    {
+      "epoch": 1.003154574132492,
+      "grad_norm": 0.5683958910749297,
+      "learning_rate": 1.706887540436331e-05,
+      "loss": 0.3338,
+      "step": 5089
+    },
+    {
+      "epoch": 1.0033517350157728,
+      "grad_norm": 2.555600213008143,
+      "learning_rate": 1.7067778945845923e-05,
+      "loss": 0.3227,
+      "step": 5090
+    },
+    {
+      "epoch": 1.0035488958990537,
+      "grad_norm": 0.602035566632611,
+      "learning_rate": 1.7066682317519453e-05,
+      "loss": 0.3194,
+      "step": 5091
+    },
+    {
+      "epoch": 1.0037460567823344,
+      "grad_norm": 0.5294253122717554,
+      "learning_rate": 1.7065585519410253e-05,
+      "loss": 0.3259,
+      "step": 5092
+    },
+    {
+      "epoch": 1.0039432176656151,
+      "grad_norm": 0.534836832974196,
+      "learning_rate": 1.706448855154467e-05,
+      "loss": 0.3003,
+      "step": 5093
+    },
+    {
+      "epoch": 1.0041403785488958,
+      "grad_norm": 0.5351347670561281,
+      "learning_rate": 1.7063391413949056e-05,
+      "loss": 0.3055,
+      "step": 5094
+    },
+    {
+      "epoch": 1.0043375394321767,
+      "grad_norm": 0.5619292444300296,
+      "learning_rate": 1.7062294106649777e-05,
+      "loss": 0.3276,
+      "step": 5095
+    },
+    {
+      "epoch": 1.0045347003154574,
+      "grad_norm": 0.5124532518746542,
+      "learning_rate": 1.7061196629673198e-05,
+      "loss": 0.3216,
+      "step": 5096
+    },
+    {
+      "epoch": 1.0047318611987381,
+      "grad_norm": 0.5657829767971062,
+      "learning_rate": 1.706009898304568e-05,
+      "loss": 0.3475,
+      "step": 5097
+    },
+    {
+      "epoch": 1.0049290220820188,
+      "grad_norm": 0.5118747834196178,
+      "learning_rate": 1.7059001166793604e-05,
+      "loss": 0.3161,
+      "step": 5098
+    },
+    {
+      "epoch": 1.0051261829652998,
+      "grad_norm": 0.5506568794847809,
+      "learning_rate": 1.7057903180943334e-05,
+      "loss": 0.3096,
+      "step": 5099
+    },
+    {
+      "epoch": 1.0053233438485805,
+      "grad_norm": 0.831367439521069,
+      "learning_rate": 1.7056805025521258e-05,
+      "loss": 0.3331,
+      "step": 5100
+    },
+    {
+      "epoch": 1.0055205047318612,
+      "grad_norm": 1.070362368316711,
+      "learning_rate": 1.705570670055376e-05,
+      "loss": 0.316,
+      "step": 5101
+    },
+    {
+      "epoch": 1.0057176656151419,
+      "grad_norm": 0.5858155470361337,
+      "learning_rate": 1.7054608206067225e-05,
+      "loss": 0.3482,
+      "step": 5102
+    },
+    {
+      "epoch": 1.0059148264984228,
+      "grad_norm": 1.0138659359886917,
+      "learning_rate": 1.705350954208805e-05,
+      "loss": 0.3311,
+      "step": 5103
+    },
+    {
+      "epoch": 1.0061119873817035,
+      "grad_norm": 0.5577957395835643,
+      "learning_rate": 1.705241070864262e-05,
+      "loss": 0.3392,
+      "step": 5104
+    },
+    {
+      "epoch": 1.0063091482649842,
+      "grad_norm": 0.5829503084317307,
+      "learning_rate": 1.7051311705757353e-05,
+      "loss": 0.3484,
+      "step": 5105
+    },
+    {
+      "epoch": 1.0065063091482649,
+      "grad_norm": 0.5629335133052799,
+      "learning_rate": 1.7050212533458637e-05,
+      "loss": 0.3403,
+      "step": 5106
+    },
+    {
+      "epoch": 1.0067034700315458,
+      "grad_norm": 0.545530969036501,
+      "learning_rate": 1.7049113191772892e-05,
+      "loss": 0.3129,
+      "step": 5107
+    },
+    {
+      "epoch": 1.0069006309148265,
+      "grad_norm": 0.5766133922035589,
+      "learning_rate": 1.7048013680726524e-05,
+      "loss": 0.3228,
+      "step": 5108
+    },
+    {
+      "epoch": 1.0070977917981072,
+      "grad_norm": 0.49331230147259286,
+      "learning_rate": 1.7046914000345955e-05,
+      "loss": 0.2823,
+      "step": 5109
+    },
+    {
+      "epoch": 1.007294952681388,
+      "grad_norm": 10.932286242518197,
+      "learning_rate": 1.7045814150657597e-05,
+      "loss": 0.3303,
+      "step": 5110
+    },
+    {
+      "epoch": 1.0074921135646688,
+      "grad_norm": 0.6743010499249699,
+      "learning_rate": 1.704471413168788e-05,
+      "loss": 0.3184,
+      "step": 5111
+    },
+    {
+      "epoch": 1.0076892744479495,
+      "grad_norm": 0.5417814723560809,
+      "learning_rate": 1.7043613943463236e-05,
+      "loss": 0.3313,
+      "step": 5112
+    },
+    {
+      "epoch": 1.0078864353312302,
+      "grad_norm": 0.6255338295712346,
+      "learning_rate": 1.7042513586010096e-05,
+      "loss": 0.3199,
+      "step": 5113
+    },
+    {
+      "epoch": 1.008083596214511,
+      "grad_norm": 0.5321460876030243,
+      "learning_rate": 1.7041413059354893e-05,
+      "loss": 0.3015,
+      "step": 5114
+    },
+    {
+      "epoch": 1.0082807570977919,
+      "grad_norm": 0.5526276788887993,
+      "learning_rate": 1.704031236352407e-05,
+      "loss": 0.3266,
+      "step": 5115
+    },
+    {
+      "epoch": 1.0084779179810726,
+      "grad_norm": 0.6188868104429317,
+      "learning_rate": 1.7039211498544075e-05,
+      "loss": 0.347,
+      "step": 5116
+    },
+    {
+      "epoch": 1.0086750788643533,
+      "grad_norm": 0.5552679635253073,
+      "learning_rate": 1.7038110464441354e-05,
+      "loss": 0.3225,
+      "step": 5117
+    },
+    {
+      "epoch": 1.008872239747634,
+      "grad_norm": 0.525568513553716,
+      "learning_rate": 1.703700926124236e-05,
+      "loss": 0.3194,
+      "step": 5118
+    },
+    {
+      "epoch": 1.0090694006309149,
+      "grad_norm": 0.6134298315197882,
+      "learning_rate": 1.7035907888973556e-05,
+      "loss": 0.3509,
+      "step": 5119
+    },
+    {
+      "epoch": 1.0092665615141956,
+      "grad_norm": 0.5390374770864321,
+      "learning_rate": 1.7034806347661398e-05,
+      "loss": 0.3143,
+      "step": 5120
+    },
+    {
+      "epoch": 1.0094637223974763,
+      "grad_norm": 0.5293103732186882,
+      "learning_rate": 1.703370463733235e-05,
+      "loss": 0.3249,
+      "step": 5121
+    },
+    {
+      "epoch": 1.0096608832807572,
+      "grad_norm": 0.7725038009082341,
+      "learning_rate": 1.7032602758012884e-05,
+      "loss": 0.3411,
+      "step": 5122
+    },
+    {
+      "epoch": 1.009858044164038,
+      "grad_norm": 0.5621329825507172,
+      "learning_rate": 1.703150070972947e-05,
+      "loss": 0.3549,
+      "step": 5123
+    },
+    {
+      "epoch": 1.0100552050473186,
+      "grad_norm": 0.5337193860173397,
+      "learning_rate": 1.7030398492508595e-05,
+      "loss": 0.3117,
+      "step": 5124
+    },
+    {
+      "epoch": 1.0102523659305993,
+      "grad_norm": 0.5648883470891011,
+      "learning_rate": 1.7029296106376732e-05,
+      "loss": 0.3125,
+      "step": 5125
+    },
+    {
+      "epoch": 1.0104495268138802,
+      "grad_norm": 0.5193119110839971,
+      "learning_rate": 1.702819355136037e-05,
+      "loss": 0.3176,
+      "step": 5126
+    },
+    {
+      "epoch": 1.010646687697161,
+      "grad_norm": 1.1538498444708232,
+      "learning_rate": 1.7027090827486e-05,
+      "loss": 0.3035,
+      "step": 5127
+    },
+    {
+      "epoch": 1.0108438485804416,
+      "grad_norm": 0.5536713733939184,
+      "learning_rate": 1.702598793478011e-05,
+      "loss": 0.3113,
+      "step": 5128
+    },
+    {
+      "epoch": 1.0110410094637223,
+      "grad_norm": 0.5121398955428231,
+      "learning_rate": 1.7024884873269206e-05,
+      "loss": 0.2913,
+      "step": 5129
+    },
+    {
+      "epoch": 1.0112381703470033,
+      "grad_norm": 0.7433927419882325,
+      "learning_rate": 1.7023781642979786e-05,
+      "loss": 0.3399,
+      "step": 5130
+    },
+    {
+      "epoch": 1.011435331230284,
+      "grad_norm": 0.5206836608515906,
+      "learning_rate": 1.7022678243938352e-05,
+      "loss": 0.2917,
+      "step": 5131
+    },
+    {
+      "epoch": 1.0116324921135647,
+      "grad_norm": 0.652680038981418,
+      "learning_rate": 1.7021574676171418e-05,
+      "loss": 0.3144,
+      "step": 5132
+    },
+    {
+      "epoch": 1.0118296529968454,
+      "grad_norm": 0.5271509086644123,
+      "learning_rate": 1.70204709397055e-05,
+      "loss": 0.3219,
+      "step": 5133
+    },
+    {
+      "epoch": 1.0120268138801263,
+      "grad_norm": 0.6052127133577494,
+      "learning_rate": 1.7019367034567115e-05,
+      "loss": 0.3163,
+      "step": 5134
+    },
+    {
+      "epoch": 1.012223974763407,
+      "grad_norm": 0.4968211645367436,
+      "learning_rate": 1.7018262960782783e-05,
+      "loss": 0.312,
+      "step": 5135
+    },
+    {
+      "epoch": 1.0124211356466877,
+      "grad_norm": 0.5506646508851154,
+      "learning_rate": 1.701715871837903e-05,
+      "loss": 0.3277,
+      "step": 5136
+    },
+    {
+      "epoch": 1.0126182965299684,
+      "grad_norm": 0.5474780371203971,
+      "learning_rate": 1.7016054307382387e-05,
+      "loss": 0.3104,
+      "step": 5137
+    },
+    {
+      "epoch": 1.0128154574132493,
+      "grad_norm": 0.5652503845654994,
+      "learning_rate": 1.7014949727819395e-05,
+      "loss": 0.3375,
+      "step": 5138
+    },
+    {
+      "epoch": 1.01301261829653,
+      "grad_norm": 0.5762231960940658,
+      "learning_rate": 1.701384497971658e-05,
+      "loss": 0.3515,
+      "step": 5139
+    },
+    {
+      "epoch": 1.0132097791798107,
+      "grad_norm": 0.5555119994103955,
+      "learning_rate": 1.7012740063100495e-05,
+      "loss": 0.304,
+      "step": 5140
+    },
+    {
+      "epoch": 1.0134069400630914,
+      "grad_norm": 0.5638374760070589,
+      "learning_rate": 1.7011634977997683e-05,
+      "loss": 0.3496,
+      "step": 5141
+    },
+    {
+      "epoch": 1.0136041009463723,
+      "grad_norm": 0.5643948965503662,
+      "learning_rate": 1.701052972443469e-05,
+      "loss": 0.3397,
+      "step": 5142
+    },
+    {
+      "epoch": 1.013801261829653,
+      "grad_norm": 0.5135163466842045,
+      "learning_rate": 1.700942430243808e-05,
+      "loss": 0.301,
+      "step": 5143
+    },
+    {
+      "epoch": 1.0139984227129337,
+      "grad_norm": 0.5539748285463232,
+      "learning_rate": 1.7008318712034405e-05,
+      "loss": 0.3244,
+      "step": 5144
+    },
+    {
+      "epoch": 1.0141955835962144,
+      "grad_norm": 0.5740027085216346,
+      "learning_rate": 1.700721295325023e-05,
+      "loss": 0.314,
+      "step": 5145
+    },
+    {
+      "epoch": 1.0143927444794953,
+      "grad_norm": 0.4898966401251555,
+      "learning_rate": 1.7006107026112117e-05,
+      "loss": 0.3106,
+      "step": 5146
+    },
+    {
+      "epoch": 1.014589905362776,
+      "grad_norm": 0.5414399089916463,
+      "learning_rate": 1.7005000930646643e-05,
+      "loss": 0.2973,
+      "step": 5147
+    },
+    {
+      "epoch": 1.0147870662460567,
+      "grad_norm": 0.5389719439899145,
+      "learning_rate": 1.700389466688038e-05,
+      "loss": 0.3285,
+      "step": 5148
+    },
+    {
+      "epoch": 1.0149842271293374,
+      "grad_norm": 0.5180514047808813,
+      "learning_rate": 1.7002788234839908e-05,
+      "loss": 0.3141,
+      "step": 5149
+    },
+    {
+      "epoch": 1.0151813880126184,
+      "grad_norm": 0.5437706206122538,
+      "learning_rate": 1.7001681634551813e-05,
+      "loss": 0.3277,
+      "step": 5150
+    },
+    {
+      "epoch": 1.015378548895899,
+      "grad_norm": 1.396200116042807,
+      "learning_rate": 1.700057486604267e-05,
+      "loss": 0.3558,
+      "step": 5151
+    },
+    {
+      "epoch": 1.0155757097791798,
+      "grad_norm": 0.563327018363544,
+      "learning_rate": 1.6999467929339086e-05,
+      "loss": 0.3421,
+      "step": 5152
+    },
+    {
+      "epoch": 1.0157728706624605,
+      "grad_norm": 0.5184359383889053,
+      "learning_rate": 1.6998360824467644e-05,
+      "loss": 0.3295,
+      "step": 5153
+    },
+    {
+      "epoch": 1.0159700315457414,
+      "grad_norm": 0.5208557657594138,
+      "learning_rate": 1.6997253551454948e-05,
+      "loss": 0.2896,
+      "step": 5154
+    },
+    {
+      "epoch": 1.016167192429022,
+      "grad_norm": 0.5291211490149857,
+      "learning_rate": 1.6996146110327604e-05,
+      "loss": 0.3234,
+      "step": 5155
+    },
+    {
+      "epoch": 1.0163643533123028,
+      "grad_norm": 0.7659645289802868,
+      "learning_rate": 1.699503850111221e-05,
+      "loss": 0.3266,
+      "step": 5156
+    },
+    {
+      "epoch": 1.0165615141955835,
+      "grad_norm": 0.5341642375091761,
+      "learning_rate": 1.699393072383539e-05,
+      "loss": 0.3227,
+      "step": 5157
+    },
+    {
+      "epoch": 1.0167586750788644,
+      "grad_norm": 0.5237990761161321,
+      "learning_rate": 1.6992822778523745e-05,
+      "loss": 0.3156,
+      "step": 5158
+    },
+    {
+      "epoch": 1.0169558359621451,
+      "grad_norm": 0.5556636270933202,
+      "learning_rate": 1.6991714665203905e-05,
+      "loss": 0.3382,
+      "step": 5159
+    },
+    {
+      "epoch": 1.0171529968454258,
+      "grad_norm": 0.5169961857756208,
+      "learning_rate": 1.699060638390249e-05,
+      "loss": 0.3083,
+      "step": 5160
+    },
+    {
+      "epoch": 1.0173501577287065,
+      "grad_norm": 0.577712432873463,
+      "learning_rate": 1.6989497934646128e-05,
+      "loss": 0.3235,
+      "step": 5161
+    },
+    {
+      "epoch": 1.0175473186119874,
+      "grad_norm": 0.48166753712469484,
+      "learning_rate": 1.6988389317461448e-05,
+      "loss": 0.3296,
+      "step": 5162
+    },
+    {
+      "epoch": 1.0177444794952681,
+      "grad_norm": 0.5588468932069051,
+      "learning_rate": 1.6987280532375082e-05,
+      "loss": 0.3205,
+      "step": 5163
+    },
+    {
+      "epoch": 1.0179416403785488,
+      "grad_norm": 0.5022075140943705,
+      "learning_rate": 1.698617157941368e-05,
+      "loss": 0.3183,
+      "step": 5164
+    },
+    {
+      "epoch": 1.0181388012618298,
+      "grad_norm": 0.5824010042694542,
+      "learning_rate": 1.698506245860388e-05,
+      "loss": 0.3339,
+      "step": 5165
+    },
+    {
+      "epoch": 1.0183359621451105,
+      "grad_norm": 0.48671865197581843,
+      "learning_rate": 1.6983953169972333e-05,
+      "loss": 0.3063,
+      "step": 5166
+    },
+    {
+      "epoch": 1.0185331230283912,
+      "grad_norm": 0.5664051734703313,
+      "learning_rate": 1.6982843713545678e-05,
+      "loss": 0.318,
+      "step": 5167
+    },
+    {
+      "epoch": 1.0187302839116719,
+      "grad_norm": 0.5307701667465677,
+      "learning_rate": 1.6981734089350585e-05,
+      "loss": 0.3375,
+      "step": 5168
+    },
+    {
+      "epoch": 1.0189274447949528,
+      "grad_norm": 0.5241759654651865,
+      "learning_rate": 1.698062429741371e-05,
+      "loss": 0.3219,
+      "step": 5169
+    },
+    {
+      "epoch": 1.0191246056782335,
+      "grad_norm": 0.5450182402988996,
+      "learning_rate": 1.697951433776171e-05,
+      "loss": 0.3373,
+      "step": 5170
+    },
+    {
+      "epoch": 1.0193217665615142,
+      "grad_norm": 0.5065193637889227,
+      "learning_rate": 1.6978404210421257e-05,
+      "loss": 0.3224,
+      "step": 5171
+    },
+    {
+      "epoch": 1.0195189274447949,
+      "grad_norm": 0.5440437100960368,
+      "learning_rate": 1.6977293915419025e-05,
+      "loss": 0.346,
+      "step": 5172
+    },
+    {
+      "epoch": 1.0197160883280758,
+      "grad_norm": 0.5167797629051604,
+      "learning_rate": 1.697618345278169e-05,
+      "loss": 0.3114,
+      "step": 5173
+    },
+    {
+      "epoch": 1.0199132492113565,
+      "grad_norm": 0.5242186802344402,
+      "learning_rate": 1.6975072822535924e-05,
+      "loss": 0.3092,
+      "step": 5174
+    },
+    {
+      "epoch": 1.0201104100946372,
+      "grad_norm": 0.5201227294258078,
+      "learning_rate": 1.6973962024708425e-05,
+      "loss": 0.331,
+      "step": 5175
+    },
+    {
+      "epoch": 1.020307570977918,
+      "grad_norm": 0.5072548524687776,
+      "learning_rate": 1.6972851059325866e-05,
+      "loss": 0.3226,
+      "step": 5176
+    },
+    {
+      "epoch": 1.0205047318611988,
+      "grad_norm": 0.5003313380328426,
+      "learning_rate": 1.6971739926414946e-05,
+      "loss": 0.3188,
+      "step": 5177
+    },
+    {
+      "epoch": 1.0207018927444795,
+      "grad_norm": 0.5360098965820386,
+      "learning_rate": 1.6970628626002362e-05,
+      "loss": 0.3379,
+      "step": 5178
+    },
+    {
+      "epoch": 1.0208990536277602,
+      "grad_norm": 0.5006169156718928,
+      "learning_rate": 1.6969517158114807e-05,
+      "loss": 0.33,
+      "step": 5179
+    },
+    {
+      "epoch": 1.021096214511041,
+      "grad_norm": 0.4797157904337746,
+      "learning_rate": 1.6968405522778996e-05,
+      "loss": 0.3035,
+      "step": 5180
+    },
+    {
+      "epoch": 1.0212933753943219,
+      "grad_norm": 0.5189338955937625,
+      "learning_rate": 1.6967293720021628e-05,
+      "loss": 0.3063,
+      "step": 5181
+    },
+    {
+      "epoch": 1.0214905362776026,
+      "grad_norm": 0.5284709213206259,
+      "learning_rate": 1.6966181749869417e-05,
+      "loss": 0.3307,
+      "step": 5182
+    },
+    {
+      "epoch": 1.0216876971608833,
+      "grad_norm": 0.5586108885453471,
+      "learning_rate": 1.6965069612349082e-05,
+      "loss": 0.3354,
+      "step": 5183
+    },
+    {
+      "epoch": 1.021884858044164,
+      "grad_norm": 0.4825892059596201,
+      "learning_rate": 1.6963957307487337e-05,
+      "loss": 0.3116,
+      "step": 5184
+    },
+    {
+      "epoch": 1.0220820189274449,
+      "grad_norm": 0.5626137147948211,
+      "learning_rate": 1.6962844835310912e-05,
+      "loss": 0.3449,
+      "step": 5185
+    },
+    {
+      "epoch": 1.0222791798107256,
+      "grad_norm": 0.4957252119548732,
+      "learning_rate": 1.696173219584653e-05,
+      "loss": 0.3201,
+      "step": 5186
+    },
+    {
+      "epoch": 1.0224763406940063,
+      "grad_norm": 0.5997034402929716,
+      "learning_rate": 1.696061938912093e-05,
+      "loss": 0.3329,
+      "step": 5187
+    },
+    {
+      "epoch": 1.022673501577287,
+      "grad_norm": 0.5074547201736874,
+      "learning_rate": 1.6959506415160838e-05,
+      "loss": 0.3124,
+      "step": 5188
+    },
+    {
+      "epoch": 1.022870662460568,
+      "grad_norm": 0.525820644885496,
+      "learning_rate": 1.6958393273993004e-05,
+      "loss": 0.3149,
+      "step": 5189
+    },
+    {
+      "epoch": 1.0230678233438486,
+      "grad_norm": 0.49149463015492956,
+      "learning_rate": 1.695727996564417e-05,
+      "loss": 0.3033,
+      "step": 5190
+    },
+    {
+      "epoch": 1.0232649842271293,
+      "grad_norm": 0.5812486165845497,
+      "learning_rate": 1.6956166490141076e-05,
+      "loss": 0.3592,
+      "step": 5191
+    },
+    {
+      "epoch": 1.02346214511041,
+      "grad_norm": 0.47448602856705274,
+      "learning_rate": 1.6955052847510486e-05,
+      "loss": 0.298,
+      "step": 5192
+    },
+    {
+      "epoch": 1.023659305993691,
+      "grad_norm": 0.4933808645190031,
+      "learning_rate": 1.6953939037779147e-05,
+      "loss": 0.3174,
+      "step": 5193
+    },
+    {
+      "epoch": 1.0238564668769716,
+      "grad_norm": 0.49672191971801205,
+      "learning_rate": 1.6952825060973826e-05,
+      "loss": 0.3035,
+      "step": 5194
+    },
+    {
+      "epoch": 1.0240536277602523,
+      "grad_norm": 0.5303109710476851,
+      "learning_rate": 1.695171091712128e-05,
+      "loss": 0.3174,
+      "step": 5195
+    },
+    {
+      "epoch": 1.024250788643533,
+      "grad_norm": 0.4983343105685653,
+      "learning_rate": 1.6950596606248283e-05,
+      "loss": 0.325,
+      "step": 5196
+    },
+    {
+      "epoch": 1.024447949526814,
+      "grad_norm": 0.5149354624056497,
+      "learning_rate": 1.6949482128381607e-05,
+      "loss": 0.3172,
+      "step": 5197
+    },
+    {
+      "epoch": 1.0246451104100947,
+      "grad_norm": 0.505149696080138,
+      "learning_rate": 1.694836748354802e-05,
+      "loss": 0.3276,
+      "step": 5198
+    },
+    {
+      "epoch": 1.0248422712933754,
+      "grad_norm": 0.5800181685156677,
+      "learning_rate": 1.6947252671774317e-05,
+      "loss": 0.3669,
+      "step": 5199
+    },
+    {
+      "epoch": 1.025039432176656,
+      "grad_norm": 0.530041440210273,
+      "learning_rate": 1.694613769308727e-05,
+      "loss": 0.3162,
+      "step": 5200
+    },
+    {
+      "epoch": 1.025236593059937,
+      "grad_norm": 0.5328709074168889,
+      "learning_rate": 1.6945022547513672e-05,
+      "loss": 0.3364,
+      "step": 5201
+    },
+    {
+      "epoch": 1.0254337539432177,
+      "grad_norm": 0.5659050716130124,
+      "learning_rate": 1.694390723508031e-05,
+      "loss": 0.352,
+      "step": 5202
+    },
+    {
+      "epoch": 1.0256309148264984,
+      "grad_norm": 0.47684840324515954,
+      "learning_rate": 1.694279175581399e-05,
+      "loss": 0.2927,
+      "step": 5203
+    },
+    {
+      "epoch": 1.025828075709779,
+      "grad_norm": 0.5261599552118007,
+      "learning_rate": 1.6941676109741506e-05,
+      "loss": 0.3474,
+      "step": 5204
+    },
+    {
+      "epoch": 1.02602523659306,
+      "grad_norm": 0.5295012959661953,
+      "learning_rate": 1.694056029688966e-05,
+      "loss": 0.3451,
+      "step": 5205
+    },
+    {
+      "epoch": 1.0262223974763407,
+      "grad_norm": 0.520003663405905,
+      "learning_rate": 1.693944431728527e-05,
+      "loss": 0.3223,
+      "step": 5206
+    },
+    {
+      "epoch": 1.0264195583596214,
+      "grad_norm": 0.4972533544450635,
+      "learning_rate": 1.693832817095514e-05,
+      "loss": 0.329,
+      "step": 5207
+    },
+    {
+      "epoch": 1.0266167192429023,
+      "grad_norm": 0.49743818639590465,
+      "learning_rate": 1.6937211857926087e-05,
+      "loss": 0.3246,
+      "step": 5208
+    },
+    {
+      "epoch": 1.026813880126183,
+      "grad_norm": 0.5108957204240979,
+      "learning_rate": 1.693609537822493e-05,
+      "loss": 0.3604,
+      "step": 5209
+    },
+    {
+      "epoch": 1.0270110410094637,
+      "grad_norm": 0.5027559180142284,
+      "learning_rate": 1.69349787318785e-05,
+      "loss": 0.3127,
+      "step": 5210
+    },
+    {
+      "epoch": 1.0272082018927444,
+      "grad_norm": 0.5060763508870827,
+      "learning_rate": 1.6933861918913617e-05,
+      "loss": 0.3157,
+      "step": 5211
+    },
+    {
+      "epoch": 1.0274053627760253,
+      "grad_norm": 0.5587497322838068,
+      "learning_rate": 1.693274493935712e-05,
+      "loss": 0.3335,
+      "step": 5212
+    },
+    {
+      "epoch": 1.027602523659306,
+      "grad_norm": 0.5328215936145846,
+      "learning_rate": 1.6931627793235845e-05,
+      "loss": 0.3401,
+      "step": 5213
+    },
+    {
+      "epoch": 1.0277996845425867,
+      "grad_norm": 0.4746075756485485,
+      "learning_rate": 1.693051048057663e-05,
+      "loss": 0.2946,
+      "step": 5214
+    },
+    {
+      "epoch": 1.0279968454258674,
+      "grad_norm": 0.5138896173325503,
+      "learning_rate": 1.6929393001406317e-05,
+      "loss": 0.3183,
+      "step": 5215
+    },
+    {
+      "epoch": 1.0281940063091484,
+      "grad_norm": 0.6224304648800318,
+      "learning_rate": 1.6928275355751758e-05,
+      "loss": 0.3385,
+      "step": 5216
+    },
+    {
+      "epoch": 1.028391167192429,
+      "grad_norm": 0.4778598347974904,
+      "learning_rate": 1.69271575436398e-05,
+      "loss": 0.2942,
+      "step": 5217
+    },
+    {
+      "epoch": 1.0285883280757098,
+      "grad_norm": 0.5206657983017704,
+      "learning_rate": 1.6926039565097313e-05,
+      "loss": 0.3232,
+      "step": 5218
+    },
+    {
+      "epoch": 1.0287854889589905,
+      "grad_norm": 0.517735918520925,
+      "learning_rate": 1.6924921420151143e-05,
+      "loss": 0.3415,
+      "step": 5219
+    },
+    {
+      "epoch": 1.0289826498422714,
+      "grad_norm": 0.49436293737722586,
+      "learning_rate": 1.6923803108828155e-05,
+      "loss": 0.2997,
+      "step": 5220
+    },
+    {
+      "epoch": 1.029179810725552,
+      "grad_norm": 0.48751091503344113,
+      "learning_rate": 1.6922684631155226e-05,
+      "loss": 0.3047,
+      "step": 5221
+    },
+    {
+      "epoch": 1.0293769716088328,
+      "grad_norm": 0.520881620751667,
+      "learning_rate": 1.6921565987159226e-05,
+      "loss": 0.335,
+      "step": 5222
+    },
+    {
+      "epoch": 1.0295741324921135,
+      "grad_norm": 0.6347905848505744,
+      "learning_rate": 1.6920447176867022e-05,
+      "loss": 0.2997,
+      "step": 5223
+    },
+    {
+      "epoch": 1.0297712933753944,
+      "grad_norm": 0.5493020000879799,
+      "learning_rate": 1.6919328200305507e-05,
+      "loss": 0.3646,
+      "step": 5224
+    },
+    {
+      "epoch": 1.0299684542586751,
+      "grad_norm": 0.5057286317589824,
+      "learning_rate": 1.691820905750156e-05,
+      "loss": 0.3125,
+      "step": 5225
+    },
+    {
+      "epoch": 1.0301656151419558,
+      "grad_norm": 0.5040769564600079,
+      "learning_rate": 1.691708974848207e-05,
+      "loss": 0.3069,
+      "step": 5226
+    },
+    {
+      "epoch": 1.0303627760252365,
+      "grad_norm": 0.5449234431968593,
+      "learning_rate": 1.6915970273273927e-05,
+      "loss": 0.3322,
+      "step": 5227
+    },
+    {
+      "epoch": 1.0305599369085174,
+      "grad_norm": 0.5763354754529781,
+      "learning_rate": 1.6914850631904027e-05,
+      "loss": 0.3389,
+      "step": 5228
+    },
+    {
+      "epoch": 1.0307570977917981,
+      "grad_norm": 0.51550599843538,
+      "learning_rate": 1.6913730824399274e-05,
+      "loss": 0.3618,
+      "step": 5229
+    },
+    {
+      "epoch": 1.0309542586750788,
+      "grad_norm": 0.5422495461192716,
+      "learning_rate": 1.691261085078657e-05,
+      "loss": 0.328,
+      "step": 5230
+    },
+    {
+      "epoch": 1.0311514195583595,
+      "grad_norm": 0.49409053708036643,
+      "learning_rate": 1.6911490711092824e-05,
+      "loss": 0.3175,
+      "step": 5231
+    },
+    {
+      "epoch": 1.0313485804416405,
+      "grad_norm": 0.5082339134425249,
+      "learning_rate": 1.6910370405344948e-05,
+      "loss": 0.3372,
+      "step": 5232
+    },
+    {
+      "epoch": 1.0315457413249212,
+      "grad_norm": 0.5304668788200992,
+      "learning_rate": 1.6909249933569856e-05,
+      "loss": 0.3237,
+      "step": 5233
+    },
+    {
+      "epoch": 1.0317429022082019,
+      "grad_norm": 0.5133955618334345,
+      "learning_rate": 1.690812929579447e-05,
+      "loss": 0.3391,
+      "step": 5234
+    },
+    {
+      "epoch": 1.0319400630914826,
+      "grad_norm": 0.5657397770422078,
+      "learning_rate": 1.690700849204572e-05,
+      "loss": 0.3315,
+      "step": 5235
+    },
+    {
+      "epoch": 1.0321372239747635,
+      "grad_norm": 0.5017068565390534,
+      "learning_rate": 1.6905887522350522e-05,
+      "loss": 0.3137,
+      "step": 5236
+    },
+    {
+      "epoch": 1.0323343848580442,
+      "grad_norm": 0.49468986516238367,
+      "learning_rate": 1.690476638673582e-05,
+      "loss": 0.3166,
+      "step": 5237
+    },
+    {
+      "epoch": 1.0325315457413249,
+      "grad_norm": 0.4900916730092504,
+      "learning_rate": 1.690364508522854e-05,
+      "loss": 0.3041,
+      "step": 5238
+    },
+    {
+      "epoch": 1.0327287066246056,
+      "grad_norm": 0.4786911698428022,
+      "learning_rate": 1.6902523617855633e-05,
+      "loss": 0.3154,
+      "step": 5239
+    },
+    {
+      "epoch": 1.0329258675078865,
+      "grad_norm": 0.5146263841992783,
+      "learning_rate": 1.6901401984644034e-05,
+      "loss": 0.2831,
+      "step": 5240
+    },
+    {
+      "epoch": 1.0331230283911672,
+      "grad_norm": 0.5189552944117162,
+      "learning_rate": 1.6900280185620697e-05,
+      "loss": 0.31,
+      "step": 5241
+    },
+    {
+      "epoch": 1.033320189274448,
+      "grad_norm": 0.5532426804255108,
+      "learning_rate": 1.689915822081257e-05,
+      "loss": 0.329,
+      "step": 5242
+    },
+    {
+      "epoch": 1.0335173501577286,
+      "grad_norm": 0.8076083652448988,
+      "learning_rate": 1.689803609024661e-05,
+      "loss": 0.3378,
+      "step": 5243
+    },
+    {
+      "epoch": 1.0337145110410095,
+      "grad_norm": 0.5725048421083666,
+      "learning_rate": 1.6896913793949782e-05,
+      "loss": 0.3247,
+      "step": 5244
+    },
+    {
+      "epoch": 1.0339116719242902,
+      "grad_norm": 0.49855659279047376,
+      "learning_rate": 1.6895791331949045e-05,
+      "loss": 0.32,
+      "step": 5245
+    },
+    {
+      "epoch": 1.034108832807571,
+      "grad_norm": 0.6416688834979003,
+      "learning_rate": 1.6894668704271363e-05,
+      "loss": 0.3466,
+      "step": 5246
+    },
+    {
+      "epoch": 1.0343059936908516,
+      "grad_norm": 0.51271995997633,
+      "learning_rate": 1.6893545910943718e-05,
+      "loss": 0.3314,
+      "step": 5247
+    },
+    {
+      "epoch": 1.0345031545741326,
+      "grad_norm": 0.5357629294597502,
+      "learning_rate": 1.689242295199308e-05,
+      "loss": 0.3238,
+      "step": 5248
+    },
+    {
+      "epoch": 1.0347003154574133,
+      "grad_norm": 2.907322545828229,
+      "learning_rate": 1.6891299827446428e-05,
+      "loss": 0.38,
+      "step": 5249
+    },
+    {
+      "epoch": 1.034897476340694,
+      "grad_norm": 0.5492499750569434,
+      "learning_rate": 1.689017653733075e-05,
+      "loss": 0.3136,
+      "step": 5250
+    },
+    {
+      "epoch": 1.0350946372239747,
+      "grad_norm": 0.5411422818363292,
+      "learning_rate": 1.688905308167303e-05,
+      "loss": 0.3282,
+      "step": 5251
+    },
+    {
+      "epoch": 1.0352917981072556,
+      "grad_norm": 0.5143337197907343,
+      "learning_rate": 1.6887929460500264e-05,
+      "loss": 0.2836,
+      "step": 5252
+    },
+    {
+      "epoch": 1.0354889589905363,
+      "grad_norm": 0.5153934469793959,
+      "learning_rate": 1.688680567383945e-05,
+      "loss": 0.3336,
+      "step": 5253
+    },
+    {
+      "epoch": 1.035686119873817,
+      "grad_norm": 0.5296091136769037,
+      "learning_rate": 1.6885681721717575e-05,
+      "loss": 0.3246,
+      "step": 5254
+    },
+    {
+      "epoch": 1.0358832807570977,
+      "grad_norm": 0.5144127492160583,
+      "learning_rate": 1.688455760416166e-05,
+      "loss": 0.3193,
+      "step": 5255
+    },
+    {
+      "epoch": 1.0360804416403786,
+      "grad_norm": 0.5575440600913101,
+      "learning_rate": 1.6883433321198697e-05,
+      "loss": 0.3244,
+      "step": 5256
+    },
+    {
+      "epoch": 1.0362776025236593,
+      "grad_norm": 0.4886892524188935,
+      "learning_rate": 1.688230887285571e-05,
+      "loss": 0.2849,
+      "step": 5257
+    },
+    {
+      "epoch": 1.03647476340694,
+      "grad_norm": 0.5024265088647863,
+      "learning_rate": 1.6881184259159708e-05,
+      "loss": 0.3105,
+      "step": 5258
+    },
+    {
+      "epoch": 1.036671924290221,
+      "grad_norm": 0.5130152621541834,
+      "learning_rate": 1.6880059480137715e-05,
+      "loss": 0.3275,
+      "step": 5259
+    },
+    {
+      "epoch": 1.0368690851735016,
+      "grad_norm": 0.4909264681540603,
+      "learning_rate": 1.687893453581675e-05,
+      "loss": 0.3102,
+      "step": 5260
+    },
+    {
+      "epoch": 1.0370662460567823,
+      "grad_norm": 0.5196024238559768,
+      "learning_rate": 1.6877809426223846e-05,
+      "loss": 0.3231,
+      "step": 5261
+    },
+    {
+      "epoch": 1.037263406940063,
+      "grad_norm": 0.5799374687724274,
+      "learning_rate": 1.6876684151386028e-05,
+      "loss": 0.3441,
+      "step": 5262
+    },
+    {
+      "epoch": 1.037460567823344,
+      "grad_norm": 0.5397818543610164,
+      "learning_rate": 1.687555871133034e-05,
+      "loss": 0.3306,
+      "step": 5263
+    },
+    {
+      "epoch": 1.0376577287066246,
+      "grad_norm": 8.650236468601472,
+      "learning_rate": 1.6874433106083815e-05,
+      "loss": 0.3565,
+      "step": 5264
+    },
+    {
+      "epoch": 1.0378548895899053,
+      "grad_norm": 0.653217018292931,
+      "learning_rate": 1.6873307335673498e-05,
+      "loss": 0.3491,
+      "step": 5265
+    },
+    {
+      "epoch": 1.038052050473186,
+      "grad_norm": 0.5706483442294323,
+      "learning_rate": 1.6872181400126434e-05,
+      "loss": 0.3686,
+      "step": 5266
+    },
+    {
+      "epoch": 1.038249211356467,
+      "grad_norm": 1.8045203865710944,
+      "learning_rate": 1.6871055299469686e-05,
+      "loss": 0.3007,
+      "step": 5267
+    },
+    {
+      "epoch": 1.0384463722397477,
+      "grad_norm": 0.5787568141280821,
+      "learning_rate": 1.6869929033730294e-05,
+      "loss": 0.339,
+      "step": 5268
+    },
+    {
+      "epoch": 1.0386435331230284,
+      "grad_norm": 0.5659159080692836,
+      "learning_rate": 1.6868802602935327e-05,
+      "loss": 0.3399,
+      "step": 5269
+    },
+    {
+      "epoch": 1.038840694006309,
+      "grad_norm": 0.6818154220606805,
+      "learning_rate": 1.6867676007111847e-05,
+      "loss": 0.3008,
+      "step": 5270
+    },
+    {
+      "epoch": 1.03903785488959,
+      "grad_norm": 0.617492098823351,
+      "learning_rate": 1.6866549246286918e-05,
+      "loss": 0.3438,
+      "step": 5271
+    },
+    {
+      "epoch": 1.0392350157728707,
+      "grad_norm": 0.49987762250168705,
+      "learning_rate": 1.6865422320487617e-05,
+      "loss": 0.295,
+      "step": 5272
+    },
+    {
+      "epoch": 1.0394321766561514,
+      "grad_norm": 0.5542454344482823,
+      "learning_rate": 1.6864295229741014e-05,
+      "loss": 0.3195,
+      "step": 5273
+    },
+    {
+      "epoch": 1.039629337539432,
+      "grad_norm": 0.5477043440442139,
+      "learning_rate": 1.686316797407419e-05,
+      "loss": 0.3124,
+      "step": 5274
+    },
+    {
+      "epoch": 1.039826498422713,
+      "grad_norm": 0.541129200628956,
+      "learning_rate": 1.6862040553514227e-05,
+      "loss": 0.3037,
+      "step": 5275
+    },
+    {
+      "epoch": 1.0400236593059937,
+      "grad_norm": 0.5301479545638493,
+      "learning_rate": 1.686091296808822e-05,
+      "loss": 0.3073,
+      "step": 5276
+    },
+    {
+      "epoch": 1.0402208201892744,
+      "grad_norm": 0.5425477118739194,
+      "learning_rate": 1.6859785217823247e-05,
+      "loss": 0.3094,
+      "step": 5277
+    },
+    {
+      "epoch": 1.0404179810725551,
+      "grad_norm": 20.30229094176611,
+      "learning_rate": 1.6858657302746412e-05,
+      "loss": 0.3857,
+      "step": 5278
+    },
+    {
+      "epoch": 1.040615141955836,
+      "grad_norm": 0.5535859868074469,
+      "learning_rate": 1.6857529222884813e-05,
+      "loss": 0.3197,
+      "step": 5279
+    },
+    {
+      "epoch": 1.0408123028391167,
+      "grad_norm": 0.5253753488109291,
+      "learning_rate": 1.685640097826555e-05,
+      "loss": 0.3229,
+      "step": 5280
+    },
+    {
+      "epoch": 1.0410094637223974,
+      "grad_norm": 0.5265804552077402,
+      "learning_rate": 1.6855272568915738e-05,
+      "loss": 0.3349,
+      "step": 5281
+    },
+    {
+      "epoch": 1.0412066246056781,
+      "grad_norm": 0.5239300594186814,
+      "learning_rate": 1.6854143994862476e-05,
+      "loss": 0.3265,
+      "step": 5282
+    },
+    {
+      "epoch": 1.041403785488959,
+      "grad_norm": 0.503539634551038,
+      "learning_rate": 1.685301525613289e-05,
+      "loss": 0.3356,
+      "step": 5283
+    },
+    {
+      "epoch": 1.0416009463722398,
+      "grad_norm": 0.540113881476233,
+      "learning_rate": 1.685188635275409e-05,
+      "loss": 0.3271,
+      "step": 5284
+    },
+    {
+      "epoch": 1.0417981072555205,
+      "grad_norm": 0.5497912101764829,
+      "learning_rate": 1.6850757284753202e-05,
+      "loss": 0.3238,
+      "step": 5285
+    },
+    {
+      "epoch": 1.0419952681388012,
+      "grad_norm": 0.5419609896862225,
+      "learning_rate": 1.6849628052157353e-05,
+      "loss": 0.328,
+      "step": 5286
+    },
+    {
+      "epoch": 1.042192429022082,
+      "grad_norm": 0.5282911482063344,
+      "learning_rate": 1.6848498654993676e-05,
+      "loss": 0.3391,
+      "step": 5287
+    },
+    {
+      "epoch": 1.0423895899053628,
+      "grad_norm": 0.5365488857264754,
+      "learning_rate": 1.6847369093289304e-05,
+      "loss": 0.3443,
+      "step": 5288
+    },
+    {
+      "epoch": 1.0425867507886435,
+      "grad_norm": 0.4884175319952028,
+      "learning_rate": 1.6846239367071372e-05,
+      "loss": 0.33,
+      "step": 5289
+    },
+    {
+      "epoch": 1.0427839116719242,
+      "grad_norm": 0.564983002840711,
+      "learning_rate": 1.684510947636703e-05,
+      "loss": 0.3469,
+      "step": 5290
+    },
+    {
+      "epoch": 1.0429810725552051,
+      "grad_norm": 0.5449133450818839,
+      "learning_rate": 1.684397942120342e-05,
+      "loss": 0.3259,
+      "step": 5291
+    },
+    {
+      "epoch": 1.0431782334384858,
+      "grad_norm": 0.5622289793546967,
+      "learning_rate": 1.684284920160769e-05,
+      "loss": 0.3485,
+      "step": 5292
+    },
+    {
+      "epoch": 1.0433753943217665,
+      "grad_norm": 0.5340801322832177,
+      "learning_rate": 1.6841718817607003e-05,
+      "loss": 0.3358,
+      "step": 5293
+    },
+    {
+      "epoch": 1.0435725552050472,
+      "grad_norm": 0.6083293183480287,
+      "learning_rate": 1.6840588269228507e-05,
+      "loss": 0.3317,
+      "step": 5294
+    },
+    {
+      "epoch": 1.0437697160883281,
+      "grad_norm": 0.4950617147026909,
+      "learning_rate": 1.6839457556499372e-05,
+      "loss": 0.3276,
+      "step": 5295
+    },
+    {
+      "epoch": 1.0439668769716088,
+      "grad_norm": 0.5662216265381399,
+      "learning_rate": 1.6838326679446756e-05,
+      "loss": 0.3515,
+      "step": 5296
+    },
+    {
+      "epoch": 1.0441640378548895,
+      "grad_norm": 0.5275768249043005,
+      "learning_rate": 1.683719563809784e-05,
+      "loss": 0.3291,
+      "step": 5297
+    },
+    {
+      "epoch": 1.0443611987381702,
+      "grad_norm": 0.5363502739274106,
+      "learning_rate": 1.683606443247979e-05,
+      "loss": 0.3348,
+      "step": 5298
+    },
+    {
+      "epoch": 1.0445583596214512,
+      "grad_norm": 0.5434941004906498,
+      "learning_rate": 1.6834933062619788e-05,
+      "loss": 0.3339,
+      "step": 5299
+    },
+    {
+      "epoch": 1.0447555205047319,
+      "grad_norm": 0.49586659679323425,
+      "learning_rate": 1.6833801528545016e-05,
+      "loss": 0.3006,
+      "step": 5300
+    },
+    {
+      "epoch": 1.0449526813880126,
+      "grad_norm": 0.5211848831043154,
+      "learning_rate": 1.6832669830282658e-05,
+      "loss": 0.3197,
+      "step": 5301
+    },
+    {
+      "epoch": 1.0451498422712935,
+      "grad_norm": 0.531819238704122,
+      "learning_rate": 1.6831537967859904e-05,
+      "loss": 0.3438,
+      "step": 5302
+    },
+    {
+      "epoch": 1.0453470031545742,
+      "grad_norm": 0.5489480069934828,
+      "learning_rate": 1.6830405941303948e-05,
+      "loss": 0.3295,
+      "step": 5303
+    },
+    {
+      "epoch": 1.0455441640378549,
+      "grad_norm": 0.5107386270726029,
+      "learning_rate": 1.6829273750641995e-05,
+      "loss": 0.3047,
+      "step": 5304
+    },
+    {
+      "epoch": 1.0457413249211356,
+      "grad_norm": 0.5400652269430585,
+      "learning_rate": 1.6828141395901236e-05,
+      "loss": 0.3339,
+      "step": 5305
+    },
+    {
+      "epoch": 1.0459384858044165,
+      "grad_norm": 0.5346562175161331,
+      "learning_rate": 1.682700887710888e-05,
+      "loss": 0.3171,
+      "step": 5306
+    },
+    {
+      "epoch": 1.0461356466876972,
+      "grad_norm": 0.5453238920962701,
+      "learning_rate": 1.682587619429214e-05,
+      "loss": 0.3333,
+      "step": 5307
+    },
+    {
+      "epoch": 1.046332807570978,
+      "grad_norm": 0.5404341648713517,
+      "learning_rate": 1.6824743347478224e-05,
+      "loss": 0.3207,
+      "step": 5308
+    },
+    {
+      "epoch": 1.0465299684542586,
+      "grad_norm": 0.48716882855128474,
+      "learning_rate": 1.682361033669436e-05,
+      "loss": 0.3093,
+      "step": 5309
+    },
+    {
+      "epoch": 1.0467271293375395,
+      "grad_norm": 0.5182039037513314,
+      "learning_rate": 1.6822477161967757e-05,
+      "loss": 0.3075,
+      "step": 5310
+    },
+    {
+      "epoch": 1.0469242902208202,
+      "grad_norm": 0.5052936292274123,
+      "learning_rate": 1.682134382332565e-05,
+      "loss": 0.3087,
+      "step": 5311
+    },
+    {
+      "epoch": 1.047121451104101,
+      "grad_norm": 0.5089202938636761,
+      "learning_rate": 1.682021032079526e-05,
+      "loss": 0.3342,
+      "step": 5312
+    },
+    {
+      "epoch": 1.0473186119873816,
+      "grad_norm": 0.5487804098811082,
+      "learning_rate": 1.681907665440383e-05,
+      "loss": 0.326,
+      "step": 5313
+    },
+    {
+      "epoch": 1.0475157728706626,
+      "grad_norm": 0.581936936175654,
+      "learning_rate": 1.6817942824178587e-05,
+      "loss": 0.3605,
+      "step": 5314
+    },
+    {
+      "epoch": 1.0477129337539433,
+      "grad_norm": 0.5264515076671745,
+      "learning_rate": 1.681680883014678e-05,
+      "loss": 0.3541,
+      "step": 5315
+    },
+    {
+      "epoch": 1.047910094637224,
+      "grad_norm": 0.49271519940647124,
+      "learning_rate": 1.6815674672335652e-05,
+      "loss": 0.3023,
+      "step": 5316
+    },
+    {
+      "epoch": 1.0481072555205047,
+      "grad_norm": 0.5364492375977938,
+      "learning_rate": 1.681454035077245e-05,
+      "loss": 0.3016,
+      "step": 5317
+    },
+    {
+      "epoch": 1.0483044164037856,
+      "grad_norm": 1.5271525514198456,
+      "learning_rate": 1.6813405865484426e-05,
+      "loss": 0.3644,
+      "step": 5318
+    },
+    {
+      "epoch": 1.0485015772870663,
+      "grad_norm": 0.5513325850528616,
+      "learning_rate": 1.6812271216498842e-05,
+      "loss": 0.3354,
+      "step": 5319
+    },
+    {
+      "epoch": 1.048698738170347,
+      "grad_norm": 0.5098889426381192,
+      "learning_rate": 1.6811136403842955e-05,
+      "loss": 0.3291,
+      "step": 5320
+    },
+    {
+      "epoch": 1.0488958990536277,
+      "grad_norm": 0.5852649954559482,
+      "learning_rate": 1.681000142754403e-05,
+      "loss": 0.2934,
+      "step": 5321
+    },
+    {
+      "epoch": 1.0490930599369086,
+      "grad_norm": 0.5187903351376137,
+      "learning_rate": 1.680886628762934e-05,
+      "loss": 0.29,
+      "step": 5322
+    },
+    {
+      "epoch": 1.0492902208201893,
+      "grad_norm": 0.5150689333720393,
+      "learning_rate": 1.6807730984126155e-05,
+      "loss": 0.3159,
+      "step": 5323
+    },
+    {
+      "epoch": 1.04948738170347,
+      "grad_norm": 0.5865797200646796,
+      "learning_rate": 1.6806595517061744e-05,
+      "loss": 0.3345,
+      "step": 5324
+    },
+    {
+      "epoch": 1.0496845425867507,
+      "grad_norm": 0.5411960758909427,
+      "learning_rate": 1.68054598864634e-05,
+      "loss": 0.3213,
+      "step": 5325
+    },
+    {
+      "epoch": 1.0498817034700316,
+      "grad_norm": 0.5402846419962667,
+      "learning_rate": 1.6804324092358402e-05,
+      "loss": 0.32,
+      "step": 5326
+    },
+    {
+      "epoch": 1.0500788643533123,
+      "grad_norm": 0.5039112744373472,
+      "learning_rate": 1.6803188134774037e-05,
+      "loss": 0.3259,
+      "step": 5327
+    },
+    {
+      "epoch": 1.050276025236593,
+      "grad_norm": 0.5462197388737413,
+      "learning_rate": 1.6802052013737595e-05,
+      "loss": 0.3309,
+      "step": 5328
+    },
+    {
+      "epoch": 1.0504731861198737,
+      "grad_norm": 0.5522764205643538,
+      "learning_rate": 1.680091572927638e-05,
+      "loss": 0.3086,
+      "step": 5329
+    },
+    {
+      "epoch": 1.0506703470031546,
+      "grad_norm": 0.5647445368693557,
+      "learning_rate": 1.6799779281417685e-05,
+      "loss": 0.349,
+      "step": 5330
+    },
+    {
+      "epoch": 1.0508675078864353,
+      "grad_norm": 0.7627307676176236,
+      "learning_rate": 1.679864267018882e-05,
+      "loss": 0.3259,
+      "step": 5331
+    },
+    {
+      "epoch": 1.051064668769716,
+      "grad_norm": 0.5107897902631611,
+      "learning_rate": 1.6797505895617087e-05,
+      "loss": 0.2968,
+      "step": 5332
+    },
+    {
+      "epoch": 1.0512618296529967,
+      "grad_norm": 0.525494347575112,
+      "learning_rate": 1.6796368957729802e-05,
+      "loss": 0.3038,
+      "step": 5333
+    },
+    {
+      "epoch": 1.0514589905362777,
+      "grad_norm": 0.5371290338339879,
+      "learning_rate": 1.679523185655428e-05,
+      "loss": 0.3366,
+      "step": 5334
+    },
+    {
+      "epoch": 1.0516561514195584,
+      "grad_norm": 0.5533840964718997,
+      "learning_rate": 1.6794094592117846e-05,
+      "loss": 0.3443,
+      "step": 5335
+    },
+    {
+      "epoch": 1.051853312302839,
+      "grad_norm": 0.571584471042435,
+      "learning_rate": 1.6792957164447807e-05,
+      "loss": 0.3503,
+      "step": 5336
+    },
+    {
+      "epoch": 1.0520504731861198,
+      "grad_norm": 0.5233336933797986,
+      "learning_rate": 1.6791819573571507e-05,
+      "loss": 0.3158,
+      "step": 5337
+    },
+    {
+      "epoch": 1.0522476340694007,
+      "grad_norm": 0.5494692589886215,
+      "learning_rate": 1.6790681819516275e-05,
+      "loss": 0.3101,
+      "step": 5338
+    },
+    {
+      "epoch": 1.0524447949526814,
+      "grad_norm": 0.5188119339090779,
+      "learning_rate": 1.6789543902309443e-05,
+      "loss": 0.3182,
+      "step": 5339
+    },
+    {
+      "epoch": 1.052641955835962,
+      "grad_norm": 0.4981089705542774,
+      "learning_rate": 1.6788405821978347e-05,
+      "loss": 0.3138,
+      "step": 5340
+    },
+    {
+      "epoch": 1.0528391167192428,
+      "grad_norm": 0.5222423077867616,
+      "learning_rate": 1.6787267578550338e-05,
+      "loss": 0.3465,
+      "step": 5341
+    },
+    {
+      "epoch": 1.0530362776025237,
+      "grad_norm": 0.6072170278417297,
+      "learning_rate": 1.678612917205276e-05,
+      "loss": 0.3759,
+      "step": 5342
+    },
+    {
+      "epoch": 1.0532334384858044,
+      "grad_norm": 0.49978080972584255,
+      "learning_rate": 1.6784990602512962e-05,
+      "loss": 0.3196,
+      "step": 5343
+    },
+    {
+      "epoch": 1.0534305993690851,
+      "grad_norm": 0.5037981316390495,
+      "learning_rate": 1.67838518699583e-05,
+      "loss": 0.3545,
+      "step": 5344
+    },
+    {
+      "epoch": 1.053627760252366,
+      "grad_norm": 0.48805755679759366,
+      "learning_rate": 1.6782712974416136e-05,
+      "loss": 0.3178,
+      "step": 5345
+    },
+    {
+      "epoch": 1.0538249211356467,
+      "grad_norm": 0.528992909703354,
+      "learning_rate": 1.678157391591383e-05,
+      "loss": 0.3609,
+      "step": 5346
+    },
+    {
+      "epoch": 1.0540220820189274,
+      "grad_norm": 0.48711439504089044,
+      "learning_rate": 1.6780434694478748e-05,
+      "loss": 0.3306,
+      "step": 5347
+    },
+    {
+      "epoch": 1.0542192429022081,
+      "grad_norm": 0.5482559735722008,
+      "learning_rate": 1.6779295310138264e-05,
+      "loss": 0.3241,
+      "step": 5348
+    },
+    {
+      "epoch": 1.054416403785489,
+      "grad_norm": 0.526517838176848,
+      "learning_rate": 1.677815576291975e-05,
+      "loss": 0.3523,
+      "step": 5349
+    },
+    {
+      "epoch": 1.0546135646687698,
+      "grad_norm": 0.560425688696832,
+      "learning_rate": 1.6777016052850586e-05,
+      "loss": 0.3306,
+      "step": 5350
+    },
+    {
+      "epoch": 1.0548107255520505,
+      "grad_norm": 0.5204711823514945,
+      "learning_rate": 1.6775876179958154e-05,
+      "loss": 0.31,
+      "step": 5351
+    },
+    {
+      "epoch": 1.0550078864353312,
+      "grad_norm": 0.5230775365579958,
+      "learning_rate": 1.677473614426984e-05,
+      "loss": 0.3317,
+      "step": 5352
+    },
+    {
+      "epoch": 1.055205047318612,
+      "grad_norm": 0.503669553362592,
+      "learning_rate": 1.6773595945813033e-05,
+      "loss": 0.31,
+      "step": 5353
+    },
+    {
+      "epoch": 1.0554022082018928,
+      "grad_norm": 0.5353586843398921,
+      "learning_rate": 1.677245558461513e-05,
+      "loss": 0.34,
+      "step": 5354
+    },
+    {
+      "epoch": 1.0555993690851735,
+      "grad_norm": 0.5027562913026351,
+      "learning_rate": 1.6771315060703525e-05,
+      "loss": 0.3345,
+      "step": 5355
+    },
+    {
+      "epoch": 1.0557965299684542,
+      "grad_norm": 0.5357732118957388,
+      "learning_rate": 1.6770174374105626e-05,
+      "loss": 0.3313,
+      "step": 5356
+    },
+    {
+      "epoch": 1.0559936908517351,
+      "grad_norm": 0.5261857108517175,
+      "learning_rate": 1.6769033524848833e-05,
+      "loss": 0.3115,
+      "step": 5357
+    },
+    {
+      "epoch": 1.0561908517350158,
+      "grad_norm": 0.5009914049381263,
+      "learning_rate": 1.6767892512960565e-05,
+      "loss": 0.3274,
+      "step": 5358
+    },
+    {
+      "epoch": 1.0563880126182965,
+      "grad_norm": 0.5301353617492294,
+      "learning_rate": 1.6766751338468222e-05,
+      "loss": 0.3338,
+      "step": 5359
+    },
+    {
+      "epoch": 1.0565851735015772,
+      "grad_norm": 0.4889759349749068,
+      "learning_rate": 1.6765610001399232e-05,
+      "loss": 0.3063,
+      "step": 5360
+    },
+    {
+      "epoch": 1.0567823343848581,
+      "grad_norm": 0.5153052654541606,
+      "learning_rate": 1.676446850178101e-05,
+      "loss": 0.3212,
+      "step": 5361
+    },
+    {
+      "epoch": 1.0569794952681388,
+      "grad_norm": 0.5432847626107267,
+      "learning_rate": 1.6763326839640993e-05,
+      "loss": 0.3401,
+      "step": 5362
+    },
+    {
+      "epoch": 1.0571766561514195,
+      "grad_norm": 0.5095940207939735,
+      "learning_rate": 1.6762185015006597e-05,
+      "loss": 0.3035,
+      "step": 5363
+    },
+    {
+      "epoch": 1.0573738170347002,
+      "grad_norm": 0.5082084315576471,
+      "learning_rate": 1.676104302790526e-05,
+      "loss": 0.3164,
+      "step": 5364
+    },
+    {
+      "epoch": 1.0575709779179812,
+      "grad_norm": 0.5317861846949568,
+      "learning_rate": 1.675990087836442e-05,
+      "loss": 0.3234,
+      "step": 5365
+    },
+    {
+      "epoch": 1.0577681388012619,
+      "grad_norm": 0.5438204191584657,
+      "learning_rate": 1.6758758566411516e-05,
+      "loss": 0.3276,
+      "step": 5366
+    },
+    {
+      "epoch": 1.0579652996845426,
+      "grad_norm": 0.5378516453459048,
+      "learning_rate": 1.6757616092073993e-05,
+      "loss": 0.3213,
+      "step": 5367
+    },
+    {
+      "epoch": 1.0581624605678233,
+      "grad_norm": 0.5241736593808551,
+      "learning_rate": 1.6756473455379306e-05,
+      "loss": 0.347,
+      "step": 5368
+    },
+    {
+      "epoch": 1.0583596214511042,
+      "grad_norm": 0.5281704964843222,
+      "learning_rate": 1.67553306563549e-05,
+      "loss": 0.3509,
+      "step": 5369
+    },
+    {
+      "epoch": 1.0585567823343849,
+      "grad_norm": 0.519218991265461,
+      "learning_rate": 1.675418769502824e-05,
+      "loss": 0.3438,
+      "step": 5370
+    },
+    {
+      "epoch": 1.0587539432176656,
+      "grad_norm": 0.5056062844302116,
+      "learning_rate": 1.6753044571426777e-05,
+      "loss": 0.3209,
+      "step": 5371
+    },
+    {
+      "epoch": 1.0589511041009463,
+      "grad_norm": 0.547727929795884,
+      "learning_rate": 1.6751901285577986e-05,
+      "loss": 0.3459,
+      "step": 5372
+    },
+    {
+      "epoch": 1.0591482649842272,
+      "grad_norm": 0.5147703062425203,
+      "learning_rate": 1.675075783750932e-05,
+      "loss": 0.309,
+      "step": 5373
+    },
+    {
+      "epoch": 1.059345425867508,
+      "grad_norm": 0.5136539353424504,
+      "learning_rate": 1.6749614227248265e-05,
+      "loss": 0.3356,
+      "step": 5374
+    },
+    {
+      "epoch": 1.0595425867507886,
+      "grad_norm": 0.5065991374201936,
+      "learning_rate": 1.6748470454822295e-05,
+      "loss": 0.3117,
+      "step": 5375
+    },
+    {
+      "epoch": 1.0597397476340693,
+      "grad_norm": 0.4802852456122994,
+      "learning_rate": 1.6747326520258884e-05,
+      "loss": 0.2974,
+      "step": 5376
+    },
+    {
+      "epoch": 1.0599369085173502,
+      "grad_norm": 0.590323007451885,
+      "learning_rate": 1.6746182423585524e-05,
+      "loss": 0.3275,
+      "step": 5377
+    },
+    {
+      "epoch": 1.060134069400631,
+      "grad_norm": 0.5031627981781183,
+      "learning_rate": 1.6745038164829695e-05,
+      "loss": 0.3443,
+      "step": 5378
+    },
+    {
+      "epoch": 1.0603312302839116,
+      "grad_norm": 0.5335220847872871,
+      "learning_rate": 1.6743893744018892e-05,
+      "loss": 0.3522,
+      "step": 5379
+    },
+    {
+      "epoch": 1.0605283911671923,
+      "grad_norm": 0.4886051328326671,
+      "learning_rate": 1.6742749161180614e-05,
+      "loss": 0.3241,
+      "step": 5380
+    },
+    {
+      "epoch": 1.0607255520504733,
+      "grad_norm": 0.5426979534289905,
+      "learning_rate": 1.6741604416342355e-05,
+      "loss": 0.3307,
+      "step": 5381
+    },
+    {
+      "epoch": 1.060922712933754,
+      "grad_norm": 0.4747697116176086,
+      "learning_rate": 1.674045950953162e-05,
+      "loss": 0.3072,
+      "step": 5382
+    },
+    {
+      "epoch": 1.0611198738170347,
+      "grad_norm": 0.4990309113836239,
+      "learning_rate": 1.673931444077592e-05,
+      "loss": 0.3243,
+      "step": 5383
+    },
+    {
+      "epoch": 1.0613170347003154,
+      "grad_norm": 0.5391230910740937,
+      "learning_rate": 1.6738169210102765e-05,
+      "loss": 0.3258,
+      "step": 5384
+    },
+    {
+      "epoch": 1.0615141955835963,
+      "grad_norm": 0.5003732086840169,
+      "learning_rate": 1.6737023817539665e-05,
+      "loss": 0.3081,
+      "step": 5385
+    },
+    {
+      "epoch": 1.061711356466877,
+      "grad_norm": 0.511666602281063,
+      "learning_rate": 1.6735878263114146e-05,
+      "loss": 0.3251,
+      "step": 5386
+    },
+    {
+      "epoch": 1.0619085173501577,
+      "grad_norm": 0.5342493746181992,
+      "learning_rate": 1.673473254685372e-05,
+      "loss": 0.3242,
+      "step": 5387
+    },
+    {
+      "epoch": 1.0621056782334386,
+      "grad_norm": 0.5066446089845047,
+      "learning_rate": 1.6733586668785926e-05,
+      "loss": 0.3415,
+      "step": 5388
+    },
+    {
+      "epoch": 1.0623028391167193,
+      "grad_norm": 0.49933317033206337,
+      "learning_rate": 1.6732440628938293e-05,
+      "loss": 0.3069,
+      "step": 5389
+    },
+    {
+      "epoch": 1.0625,
+      "grad_norm": 0.5395443938451984,
+      "learning_rate": 1.6731294427338344e-05,
+      "loss": 0.3343,
+      "step": 5390
+    },
+    {
+      "epoch": 1.0626971608832807,
+      "grad_norm": 0.5154874242963391,
+      "learning_rate": 1.6730148064013633e-05,
+      "loss": 0.3353,
+      "step": 5391
+    },
+    {
+      "epoch": 1.0628943217665614,
+      "grad_norm": 0.549850655755667,
+      "learning_rate": 1.672900153899169e-05,
+      "loss": 0.3339,
+      "step": 5392
+    },
+    {
+      "epoch": 1.0630914826498423,
+      "grad_norm": 0.5437692223908803,
+      "learning_rate": 1.6727854852300073e-05,
+      "loss": 0.325,
+      "step": 5393
+    },
+    {
+      "epoch": 1.063288643533123,
+      "grad_norm": 0.5319120551752757,
+      "learning_rate": 1.672670800396632e-05,
+      "loss": 0.3333,
+      "step": 5394
+    },
+    {
+      "epoch": 1.0634858044164037,
+      "grad_norm": 0.5888809384452421,
+      "learning_rate": 1.672556099401799e-05,
+      "loss": 0.3289,
+      "step": 5395
+    },
+    {
+      "epoch": 1.0636829652996846,
+      "grad_norm": 0.5174224708917468,
+      "learning_rate": 1.672441382248264e-05,
+      "loss": 0.325,
+      "step": 5396
+    },
+    {
+      "epoch": 1.0638801261829653,
+      "grad_norm": 0.4940106410321626,
+      "learning_rate": 1.6723266489387837e-05,
+      "loss": 0.3277,
+      "step": 5397
+    },
+    {
+      "epoch": 1.064077287066246,
+      "grad_norm": 0.5924938542192701,
+      "learning_rate": 1.672211899476114e-05,
+      "loss": 0.3159,
+      "step": 5398
+    },
+    {
+      "epoch": 1.0642744479495267,
+      "grad_norm": 0.6030406417673295,
+      "learning_rate": 1.672097133863012e-05,
+      "loss": 0.352,
+      "step": 5399
+    },
+    {
+      "epoch": 1.0644716088328077,
+      "grad_norm": 0.46675304029243586,
+      "learning_rate": 1.6719823521022355e-05,
+      "loss": 0.273,
+      "step": 5400
+    },
+    {
+      "epoch": 1.0646687697160884,
+      "grad_norm": 10.313913415266276,
+      "learning_rate": 1.6718675541965413e-05,
+      "loss": 0.3712,
+      "step": 5401
+    },
+    {
+      "epoch": 1.064865930599369,
+      "grad_norm": 0.650331375307367,
+      "learning_rate": 1.6717527401486882e-05,
+      "loss": 0.3173,
+      "step": 5402
+    },
+    {
+      "epoch": 1.0650630914826498,
+      "grad_norm": 0.5516904225558585,
+      "learning_rate": 1.6716379099614348e-05,
+      "loss": 0.3332,
+      "step": 5403
+    },
+    {
+      "epoch": 1.0652602523659307,
+      "grad_norm": 0.5826270309682446,
+      "learning_rate": 1.6715230636375397e-05,
+      "loss": 0.3162,
+      "step": 5404
+    },
+    {
+      "epoch": 1.0654574132492114,
+      "grad_norm": 0.5567258991880055,
+      "learning_rate": 1.6714082011797625e-05,
+      "loss": 0.3091,
+      "step": 5405
+    },
+    {
+      "epoch": 1.065654574132492,
+      "grad_norm": 0.5797543264480705,
+      "learning_rate": 1.6712933225908618e-05,
+      "loss": 0.3638,
+      "step": 5406
+    },
+    {
+      "epoch": 1.0658517350157728,
+      "grad_norm": 0.500686608596155,
+      "learning_rate": 1.6711784278735993e-05,
+      "loss": 0.3183,
+      "step": 5407
+    },
+    {
+      "epoch": 1.0660488958990537,
+      "grad_norm": 0.5482037516123067,
+      "learning_rate": 1.6710635170307336e-05,
+      "loss": 0.3282,
+      "step": 5408
+    },
+    {
+      "epoch": 1.0662460567823344,
+      "grad_norm": 0.5003783186696478,
+      "learning_rate": 1.6709485900650274e-05,
+      "loss": 0.2917,
+      "step": 5409
+    },
+    {
+      "epoch": 1.0664432176656151,
+      "grad_norm": 0.5886317793002753,
+      "learning_rate": 1.6708336469792407e-05,
+      "loss": 0.3297,
+      "step": 5410
+    },
+    {
+      "epoch": 1.0666403785488958,
+      "grad_norm": 0.5466590442325215,
+      "learning_rate": 1.670718687776135e-05,
+      "loss": 0.3549,
+      "step": 5411
+    },
+    {
+      "epoch": 1.0668375394321767,
+      "grad_norm": 0.5540316157067651,
+      "learning_rate": 1.670603712458473e-05,
+      "loss": 0.3215,
+      "step": 5412
+    },
+    {
+      "epoch": 1.0670347003154574,
+      "grad_norm": 0.593181096332669,
+      "learning_rate": 1.670488721029017e-05,
+      "loss": 0.3648,
+      "step": 5413
+    },
+    {
+      "epoch": 1.0672318611987381,
+      "grad_norm": 0.5482843023549779,
+      "learning_rate": 1.6703737134905296e-05,
+      "loss": 0.3269,
+      "step": 5414
+    },
+    {
+      "epoch": 1.0674290220820188,
+      "grad_norm": 0.5958232910254453,
+      "learning_rate": 1.6702586898457737e-05,
+      "loss": 0.3219,
+      "step": 5415
+    },
+    {
+      "epoch": 1.0676261829652998,
+      "grad_norm": 0.5248785667847379,
+      "learning_rate": 1.6701436500975127e-05,
+      "loss": 0.3113,
+      "step": 5416
+    },
+    {
+      "epoch": 1.0678233438485805,
+      "grad_norm": 0.506343180110072,
+      "learning_rate": 1.6700285942485113e-05,
+      "loss": 0.2969,
+      "step": 5417
+    },
+    {
+      "epoch": 1.0680205047318612,
+      "grad_norm": 0.6323082926274546,
+      "learning_rate": 1.669913522301533e-05,
+      "loss": 0.3197,
+      "step": 5418
+    },
+    {
+      "epoch": 1.0682176656151419,
+      "grad_norm": 0.5020570452865531,
+      "learning_rate": 1.669798434259343e-05,
+      "loss": 0.3101,
+      "step": 5419
+    },
+    {
+      "epoch": 1.0684148264984228,
+      "grad_norm": 0.7538065179489385,
+      "learning_rate": 1.669683330124706e-05,
+      "loss": 0.3097,
+      "step": 5420
+    },
+    {
+      "epoch": 1.0686119873817035,
+      "grad_norm": 0.7336379735027907,
+      "learning_rate": 1.669568209900388e-05,
+      "loss": 0.2805,
+      "step": 5421
+    },
+    {
+      "epoch": 1.0688091482649842,
+      "grad_norm": 0.5148666894817214,
+      "learning_rate": 1.6694530735891548e-05,
+      "loss": 0.3298,
+      "step": 5422
+    },
+    {
+      "epoch": 1.0690063091482649,
+      "grad_norm": 0.5938791350921867,
+      "learning_rate": 1.6693379211937717e-05,
+      "loss": 0.3594,
+      "step": 5423
+    },
+    {
+      "epoch": 1.0692034700315458,
+      "grad_norm": 0.5268352695165609,
+      "learning_rate": 1.6692227527170067e-05,
+      "loss": 0.3245,
+      "step": 5424
+    },
+    {
+      "epoch": 1.0694006309148265,
+      "grad_norm": 0.8124267042610954,
+      "learning_rate": 1.6691075681616257e-05,
+      "loss": 0.3672,
+      "step": 5425
+    },
+    {
+      "epoch": 1.0695977917981072,
+      "grad_norm": 0.5680106061709315,
+      "learning_rate": 1.6689923675303967e-05,
+      "loss": 0.3314,
+      "step": 5426
+    },
+    {
+      "epoch": 1.069794952681388,
+      "grad_norm": 0.5607346328108409,
+      "learning_rate": 1.6688771508260876e-05,
+      "loss": 0.3461,
+      "step": 5427
+    },
+    {
+      "epoch": 1.0699921135646688,
+      "grad_norm": 0.5433165552798762,
+      "learning_rate": 1.668761918051466e-05,
+      "loss": 0.3425,
+      "step": 5428
+    },
+    {
+      "epoch": 1.0701892744479495,
+      "grad_norm": 0.5476574461852847,
+      "learning_rate": 1.6686466692093007e-05,
+      "loss": 0.3172,
+      "step": 5429
+    },
+    {
+      "epoch": 1.0703864353312302,
+      "grad_norm": 0.6111021096749791,
+      "learning_rate": 1.6685314043023608e-05,
+      "loss": 0.3577,
+      "step": 5430
+    },
+    {
+      "epoch": 1.0705835962145112,
+      "grad_norm": 0.5552410097887198,
+      "learning_rate": 1.6684161233334157e-05,
+      "loss": 0.3154,
+      "step": 5431
+    },
+    {
+      "epoch": 1.0707807570977919,
+      "grad_norm": 0.5107392149662423,
+      "learning_rate": 1.6683008263052344e-05,
+      "loss": 0.3275,
+      "step": 5432
+    },
+    {
+      "epoch": 1.0709779179810726,
+      "grad_norm": 0.6074720035217149,
+      "learning_rate": 1.6681855132205882e-05,
+      "loss": 0.3597,
+      "step": 5433
+    },
+    {
+      "epoch": 1.0711750788643533,
+      "grad_norm": 0.515248985051321,
+      "learning_rate": 1.6680701840822468e-05,
+      "loss": 0.3278,
+      "step": 5434
+    },
+    {
+      "epoch": 1.071372239747634,
+      "grad_norm": 0.5676731678796564,
+      "learning_rate": 1.667954838892981e-05,
+      "loss": 0.324,
+      "step": 5435
+    },
+    {
+      "epoch": 1.0715694006309149,
+      "grad_norm": 0.5266114021796704,
+      "learning_rate": 1.6678394776555625e-05,
+      "loss": 0.3298,
+      "step": 5436
+    },
+    {
+      "epoch": 1.0717665615141956,
+      "grad_norm": 0.515711098109902,
+      "learning_rate": 1.667724100372763e-05,
+      "loss": 0.3301,
+      "step": 5437
+    },
+    {
+      "epoch": 1.0719637223974763,
+      "grad_norm": 0.5413115348636542,
+      "learning_rate": 1.667608707047354e-05,
+      "loss": 0.329,
+      "step": 5438
+    },
+    {
+      "epoch": 1.0721608832807572,
+      "grad_norm": 0.5133409836464313,
+      "learning_rate": 1.6674932976821078e-05,
+      "loss": 0.3062,
+      "step": 5439
+    },
+    {
+      "epoch": 1.072358044164038,
+      "grad_norm": 0.5514269531982732,
+      "learning_rate": 1.667377872279798e-05,
+      "loss": 0.3224,
+      "step": 5440
+    },
+    {
+      "epoch": 1.0725552050473186,
+      "grad_norm": 0.5381267123295007,
+      "learning_rate": 1.6672624308431977e-05,
+      "loss": 0.3554,
+      "step": 5441
+    },
+    {
+      "epoch": 1.0727523659305993,
+      "grad_norm": 0.5357973892732231,
+      "learning_rate": 1.6671469733750795e-05,
+      "loss": 0.3306,
+      "step": 5442
+    },
+    {
+      "epoch": 1.0729495268138802,
+      "grad_norm": 0.5334845958350081,
+      "learning_rate": 1.6670314998782183e-05,
+      "loss": 0.3522,
+      "step": 5443
+    },
+    {
+      "epoch": 1.073146687697161,
+      "grad_norm": 0.5078507960335631,
+      "learning_rate": 1.6669160103553884e-05,
+      "loss": 0.3192,
+      "step": 5444
+    },
+    {
+      "epoch": 1.0733438485804416,
+      "grad_norm": 0.526273076314691,
+      "learning_rate": 1.666800504809364e-05,
+      "loss": 0.3295,
+      "step": 5445
+    },
+    {
+      "epoch": 1.0735410094637223,
+      "grad_norm": 0.5501746473063156,
+      "learning_rate": 1.6666849832429207e-05,
+      "loss": 0.3559,
+      "step": 5446
+    },
+    {
+      "epoch": 1.0737381703470033,
+      "grad_norm": 0.5417331807201895,
+      "learning_rate": 1.6665694456588335e-05,
+      "loss": 0.3395,
+      "step": 5447
+    },
+    {
+      "epoch": 1.073935331230284,
+      "grad_norm": 0.4892217439505845,
+      "learning_rate": 1.666453892059879e-05,
+      "loss": 0.2989,
+      "step": 5448
+    },
+    {
+      "epoch": 1.0741324921135647,
+      "grad_norm": 0.5427686650470855,
+      "learning_rate": 1.6663383224488323e-05,
+      "loss": 0.3249,
+      "step": 5449
+    },
+    {
+      "epoch": 1.0743296529968454,
+      "grad_norm": 0.5467284567115355,
+      "learning_rate": 1.6662227368284716e-05,
+      "loss": 0.3353,
+      "step": 5450
+    },
+    {
+      "epoch": 1.0745268138801263,
+      "grad_norm": 0.49209796772640735,
+      "learning_rate": 1.6661071352015725e-05,
+      "loss": 0.3205,
+      "step": 5451
+    },
+    {
+      "epoch": 1.074723974763407,
+      "grad_norm": 0.5566384176423259,
+      "learning_rate": 1.6659915175709135e-05,
+      "loss": 0.3221,
+      "step": 5452
+    },
+    {
+      "epoch": 1.0749211356466877,
+      "grad_norm": 0.526499393616949,
+      "learning_rate": 1.665875883939272e-05,
+      "loss": 0.3263,
+      "step": 5453
+    },
+    {
+      "epoch": 1.0751182965299684,
+      "grad_norm": 2.739695761851264,
+      "learning_rate": 1.665760234309426e-05,
+      "loss": 0.3506,
+      "step": 5454
+    },
+    {
+      "epoch": 1.0753154574132493,
+      "grad_norm": 0.6593642459442813,
+      "learning_rate": 1.665644568684154e-05,
+      "loss": 0.3426,
+      "step": 5455
+    },
+    {
+      "epoch": 1.07551261829653,
+      "grad_norm": 0.5151614291220461,
+      "learning_rate": 1.6655288870662354e-05,
+      "loss": 0.328,
+      "step": 5456
+    },
+    {
+      "epoch": 1.0757097791798107,
+      "grad_norm": 0.5714376826042502,
+      "learning_rate": 1.6654131894584494e-05,
+      "loss": 0.318,
+      "step": 5457
+    },
+    {
+      "epoch": 1.0759069400630914,
+      "grad_norm": 0.5326114164538318,
+      "learning_rate": 1.665297475863576e-05,
+      "loss": 0.3268,
+      "step": 5458
+    },
+    {
+      "epoch": 1.0761041009463723,
+      "grad_norm": 0.5318642965648936,
+      "learning_rate": 1.6651817462843944e-05,
+      "loss": 0.3393,
+      "step": 5459
+    },
+    {
+      "epoch": 1.076301261829653,
+      "grad_norm": 0.5208615095763689,
+      "learning_rate": 1.665066000723686e-05,
+      "loss": 0.3048,
+      "step": 5460
+    },
+    {
+      "epoch": 1.0764984227129337,
+      "grad_norm": 0.5112237680335591,
+      "learning_rate": 1.6649502391842313e-05,
+      "loss": 0.3306,
+      "step": 5461
+    },
+    {
+      "epoch": 1.0766955835962144,
+      "grad_norm": 0.5020109988963485,
+      "learning_rate": 1.6648344616688116e-05,
+      "loss": 0.3079,
+      "step": 5462
+    },
+    {
+      "epoch": 1.0768927444794953,
+      "grad_norm": 0.5141839034905135,
+      "learning_rate": 1.664718668180208e-05,
+      "loss": 0.3433,
+      "step": 5463
+    },
+    {
+      "epoch": 1.077089905362776,
+      "grad_norm": 0.522715217511301,
+      "learning_rate": 1.664602858721204e-05,
+      "loss": 0.3273,
+      "step": 5464
+    },
+    {
+      "epoch": 1.0772870662460567,
+      "grad_norm": 0.4873672081469081,
+      "learning_rate": 1.6644870332945807e-05,
+      "loss": 0.3131,
+      "step": 5465
+    },
+    {
+      "epoch": 1.0774842271293374,
+      "grad_norm": 0.5208359970585005,
+      "learning_rate": 1.6643711919031217e-05,
+      "loss": 0.3326,
+      "step": 5466
+    },
+    {
+      "epoch": 1.0776813880126184,
+      "grad_norm": 0.5085175763798949,
+      "learning_rate": 1.6642553345496093e-05,
+      "loss": 0.336,
+      "step": 5467
+    },
+    {
+      "epoch": 1.077878548895899,
+      "grad_norm": 0.5395303777396451,
+      "learning_rate": 1.664139461236828e-05,
+      "loss": 0.3519,
+      "step": 5468
+    },
+    {
+      "epoch": 1.0780757097791798,
+      "grad_norm": 0.4939691741186544,
+      "learning_rate": 1.6640235719675607e-05,
+      "loss": 0.3149,
+      "step": 5469
+    },
+    {
+      "epoch": 1.0782728706624605,
+      "grad_norm": 0.5045822941059132,
+      "learning_rate": 1.663907666744593e-05,
+      "loss": 0.3257,
+      "step": 5470
+    },
+    {
+      "epoch": 1.0784700315457414,
+      "grad_norm": 0.4842189935972799,
+      "learning_rate": 1.6637917455707085e-05,
+      "loss": 0.2941,
+      "step": 5471
+    },
+    {
+      "epoch": 1.078667192429022,
+      "grad_norm": 0.488411947048396,
+      "learning_rate": 1.663675808448693e-05,
+      "loss": 0.3263,
+      "step": 5472
+    },
+    {
+      "epoch": 1.0788643533123028,
+      "grad_norm": 0.5204936040738815,
+      "learning_rate": 1.663559855381332e-05,
+      "loss": 0.337,
+      "step": 5473
+    },
+    {
+      "epoch": 1.0790615141955835,
+      "grad_norm": 0.4907540263785109,
+      "learning_rate": 1.6634438863714108e-05,
+      "loss": 0.3341,
+      "step": 5474
+    },
+    {
+      "epoch": 1.0792586750788644,
+      "grad_norm": 0.5341500420019042,
+      "learning_rate": 1.6633279014217158e-05,
+      "loss": 0.361,
+      "step": 5475
+    },
+    {
+      "epoch": 1.0794558359621451,
+      "grad_norm": 0.566795481318772,
+      "learning_rate": 1.663211900535034e-05,
+      "loss": 0.3565,
+      "step": 5476
+    },
+    {
+      "epoch": 1.0796529968454258,
+      "grad_norm": 0.5104378663095975,
+      "learning_rate": 1.663095883714152e-05,
+      "loss": 0.3172,
+      "step": 5477
+    },
+    {
+      "epoch": 1.0798501577287065,
+      "grad_norm": 0.7836320347832308,
+      "learning_rate": 1.6629798509618575e-05,
+      "loss": 0.3279,
+      "step": 5478
+    },
+    {
+      "epoch": 1.0800473186119874,
+      "grad_norm": 0.4941220351531997,
+      "learning_rate": 1.6628638022809384e-05,
+      "loss": 0.3237,
+      "step": 5479
+    },
+    {
+      "epoch": 1.0802444794952681,
+      "grad_norm": 0.538070953999582,
+      "learning_rate": 1.6627477376741824e-05,
+      "loss": 0.3186,
+      "step": 5480
+    },
+    {
+      "epoch": 1.0804416403785488,
+      "grad_norm": 0.49651925324616386,
+      "learning_rate": 1.6626316571443782e-05,
+      "loss": 0.3328,
+      "step": 5481
+    },
+    {
+      "epoch": 1.0806388012618298,
+      "grad_norm": 0.6699435537682001,
+      "learning_rate": 1.662515560694315e-05,
+      "loss": 0.3353,
+      "step": 5482
+    },
+    {
+      "epoch": 1.0808359621451105,
+      "grad_norm": 0.5278612363315855,
+      "learning_rate": 1.6623994483267823e-05,
+      "loss": 0.3275,
+      "step": 5483
+    },
+    {
+      "epoch": 1.0810331230283912,
+      "grad_norm": 0.5521883665813071,
+      "learning_rate": 1.6622833200445688e-05,
+      "loss": 0.3395,
+      "step": 5484
+    },
+    {
+      "epoch": 1.0812302839116719,
+      "grad_norm": 0.5101814960888886,
+      "learning_rate": 1.6621671758504656e-05,
+      "loss": 0.3229,
+      "step": 5485
+    },
+    {
+      "epoch": 1.0814274447949526,
+      "grad_norm": 0.5082133461105458,
+      "learning_rate": 1.6620510157472626e-05,
+      "loss": 0.3416,
+      "step": 5486
+    },
+    {
+      "epoch": 1.0816246056782335,
+      "grad_norm": 0.5152393878133755,
+      "learning_rate": 1.6619348397377508e-05,
+      "loss": 0.2995,
+      "step": 5487
+    },
+    {
+      "epoch": 1.0818217665615142,
+      "grad_norm": 0.5030381994487839,
+      "learning_rate": 1.6618186478247214e-05,
+      "loss": 0.3219,
+      "step": 5488
+    },
+    {
+      "epoch": 1.0820189274447949,
+      "grad_norm": 0.5033251374070803,
+      "learning_rate": 1.661702440010966e-05,
+      "loss": 0.3278,
+      "step": 5489
+    },
+    {
+      "epoch": 1.0822160883280758,
+      "grad_norm": 0.4858411621778881,
+      "learning_rate": 1.6615862162992765e-05,
+      "loss": 0.321,
+      "step": 5490
+    },
+    {
+      "epoch": 1.0824132492113565,
+      "grad_norm": 0.5299981365610831,
+      "learning_rate": 1.6614699766924457e-05,
+      "loss": 0.334,
+      "step": 5491
+    },
+    {
+      "epoch": 1.0826104100946372,
+      "grad_norm": 0.5398683026062827,
+      "learning_rate": 1.661353721193266e-05,
+      "loss": 0.3679,
+      "step": 5492
+    },
+    {
+      "epoch": 1.082807570977918,
+      "grad_norm": 0.4922123143204083,
+      "learning_rate": 1.6612374498045303e-05,
+      "loss": 0.3252,
+      "step": 5493
+    },
+    {
+      "epoch": 1.0830047318611988,
+      "grad_norm": 0.5402134649137565,
+      "learning_rate": 1.6611211625290328e-05,
+      "loss": 0.3488,
+      "step": 5494
+    },
+    {
+      "epoch": 1.0832018927444795,
+      "grad_norm": 0.49308565622911255,
+      "learning_rate": 1.6610048593695665e-05,
+      "loss": 0.3274,
+      "step": 5495
+    },
+    {
+      "epoch": 1.0833990536277602,
+      "grad_norm": 0.5330415075689533,
+      "learning_rate": 1.6608885403289264e-05,
+      "loss": 0.3416,
+      "step": 5496
+    },
+    {
+      "epoch": 1.083596214511041,
+      "grad_norm": 0.5546315403447396,
+      "learning_rate": 1.6607722054099066e-05,
+      "loss": 0.3428,
+      "step": 5497
+    },
+    {
+      "epoch": 1.0837933753943219,
+      "grad_norm": 0.548153275553841,
+      "learning_rate": 1.6606558546153027e-05,
+      "loss": 0.3274,
+      "step": 5498
+    },
+    {
+      "epoch": 1.0839905362776026,
+      "grad_norm": 0.5741893560590743,
+      "learning_rate": 1.6605394879479102e-05,
+      "loss": 0.3408,
+      "step": 5499
+    },
+    {
+      "epoch": 1.0841876971608833,
+      "grad_norm": 0.563655275701286,
+      "learning_rate": 1.660423105410524e-05,
+      "loss": 0.3535,
+      "step": 5500
+    },
+    {
+      "epoch": 1.084384858044164,
+      "grad_norm": 0.5292670828512388,
+      "learning_rate": 1.6603067070059413e-05,
+      "loss": 0.3451,
+      "step": 5501
+    },
+    {
+      "epoch": 1.0845820189274449,
+      "grad_norm": 0.4943062953455165,
+      "learning_rate": 1.6601902927369577e-05,
+      "loss": 0.305,
+      "step": 5502
+    },
+    {
+      "epoch": 1.0847791798107256,
+      "grad_norm": 0.5684207519421706,
+      "learning_rate": 1.660073862606371e-05,
+      "loss": 0.3544,
+      "step": 5503
+    },
+    {
+      "epoch": 1.0849763406940063,
+      "grad_norm": 0.5269370109869425,
+      "learning_rate": 1.6599574166169783e-05,
+      "loss": 0.3364,
+      "step": 5504
+    },
+    {
+      "epoch": 1.085173501577287,
+      "grad_norm": 0.5930908127100334,
+      "learning_rate": 1.659840954771577e-05,
+      "loss": 0.3293,
+      "step": 5505
+    },
+    {
+      "epoch": 1.085370662460568,
+      "grad_norm": 0.5064001248459025,
+      "learning_rate": 1.6597244770729656e-05,
+      "loss": 0.3364,
+      "step": 5506
+    },
+    {
+      "epoch": 1.0855678233438486,
+      "grad_norm": 0.5583107476093074,
+      "learning_rate": 1.659607983523942e-05,
+      "loss": 0.344,
+      "step": 5507
+    },
+    {
+      "epoch": 1.0857649842271293,
+      "grad_norm": 0.5186990686312586,
+      "learning_rate": 1.6594914741273058e-05,
+      "loss": 0.3203,
+      "step": 5508
+    },
+    {
+      "epoch": 1.08596214511041,
+      "grad_norm": 0.5412209056965988,
+      "learning_rate": 1.6593749488858554e-05,
+      "loss": 0.3514,
+      "step": 5509
+    },
+    {
+      "epoch": 1.086159305993691,
+      "grad_norm": 0.5847946899685581,
+      "learning_rate": 1.6592584078023915e-05,
+      "loss": 0.3498,
+      "step": 5510
+    },
+    {
+      "epoch": 1.0863564668769716,
+      "grad_norm": 0.49519738287188864,
+      "learning_rate": 1.659141850879713e-05,
+      "loss": 0.3065,
+      "step": 5511
+    },
+    {
+      "epoch": 1.0865536277602523,
+      "grad_norm": 0.49468808903487677,
+      "learning_rate": 1.659025278120621e-05,
+      "loss": 0.3129,
+      "step": 5512
+    },
+    {
+      "epoch": 1.086750788643533,
+      "grad_norm": 1.0581729165868015,
+      "learning_rate": 1.6589086895279156e-05,
+      "loss": 0.3673,
+      "step": 5513
+    },
+    {
+      "epoch": 1.086947949526814,
+      "grad_norm": 0.49002563953861666,
+      "learning_rate": 1.6587920851043986e-05,
+      "loss": 0.3223,
+      "step": 5514
+    },
+    {
+      "epoch": 1.0871451104100947,
+      "grad_norm": 0.4905578731396364,
+      "learning_rate": 1.6586754648528712e-05,
+      "loss": 0.3043,
+      "step": 5515
+    },
+    {
+      "epoch": 1.0873422712933754,
+      "grad_norm": 0.667527289823531,
+      "learning_rate": 1.658558828776135e-05,
+      "loss": 0.2788,
+      "step": 5516
+    },
+    {
+      "epoch": 1.087539432176656,
+      "grad_norm": 0.8459394709559949,
+      "learning_rate": 1.6584421768769933e-05,
+      "loss": 0.3415,
+      "step": 5517
+    },
+    {
+      "epoch": 1.087736593059937,
+      "grad_norm": 0.47906959750711475,
+      "learning_rate": 1.6583255091582474e-05,
+      "loss": 0.3405,
+      "step": 5518
+    },
+    {
+      "epoch": 1.0879337539432177,
+      "grad_norm": 0.48428073694586943,
+      "learning_rate": 1.658208825622701e-05,
+      "loss": 0.3322,
+      "step": 5519
+    },
+    {
+      "epoch": 1.0881309148264984,
+      "grad_norm": 0.7968161512793981,
+      "learning_rate": 1.6580921262731582e-05,
+      "loss": 0.3291,
+      "step": 5520
+    },
+    {
+      "epoch": 1.088328075709779,
+      "grad_norm": 0.519690983389764,
+      "learning_rate": 1.6579754111124215e-05,
+      "loss": 0.3248,
+      "step": 5521
+    },
+    {
+      "epoch": 1.08852523659306,
+      "grad_norm": 0.5091809572661937,
+      "learning_rate": 1.6578586801432958e-05,
+      "loss": 0.3133,
+      "step": 5522
+    },
+    {
+      "epoch": 1.0887223974763407,
+      "grad_norm": 0.5549122652560055,
+      "learning_rate": 1.6577419333685855e-05,
+      "loss": 0.3462,
+      "step": 5523
+    },
+    {
+      "epoch": 1.0889195583596214,
+      "grad_norm": 6.326360585378315,
+      "learning_rate": 1.6576251707910955e-05,
+      "loss": 0.3586,
+      "step": 5524
+    },
+    {
+      "epoch": 1.0891167192429023,
+      "grad_norm": 0.594210310368939,
+      "learning_rate": 1.6575083924136313e-05,
+      "loss": 0.2988,
+      "step": 5525
+    },
+    {
+      "epoch": 1.089313880126183,
+      "grad_norm": 0.5653001651310844,
+      "learning_rate": 1.6573915982389986e-05,
+      "loss": 0.3427,
+      "step": 5526
+    },
+    {
+      "epoch": 1.0895110410094637,
+      "grad_norm": 0.6419933562592661,
+      "learning_rate": 1.6572747882700034e-05,
+      "loss": 0.3479,
+      "step": 5527
+    },
+    {
+      "epoch": 1.0897082018927444,
+      "grad_norm": 0.549229419744898,
+      "learning_rate": 1.657157962509452e-05,
+      "loss": 0.3367,
+      "step": 5528
+    },
+    {
+      "epoch": 1.0899053627760251,
+      "grad_norm": 0.5649927832731517,
+      "learning_rate": 1.6570411209601515e-05,
+      "loss": 0.3288,
+      "step": 5529
+    },
+    {
+      "epoch": 1.090102523659306,
+      "grad_norm": 0.5593298618542406,
+      "learning_rate": 1.656924263624909e-05,
+      "loss": 0.3552,
+      "step": 5530
+    },
+    {
+      "epoch": 1.0902996845425867,
+      "grad_norm": 0.5487752889109879,
+      "learning_rate": 1.6568073905065313e-05,
+      "loss": 0.3467,
+      "step": 5531
+    },
+    {
+      "epoch": 1.0904968454258674,
+      "grad_norm": 0.517747474702925,
+      "learning_rate": 1.656690501607828e-05,
+      "loss": 0.3165,
+      "step": 5532
+    },
+    {
+      "epoch": 1.0906940063091484,
+      "grad_norm": 0.5187685507630857,
+      "learning_rate": 1.656573596931606e-05,
+      "loss": 0.3284,
+      "step": 5533
+    },
+    {
+      "epoch": 1.090891167192429,
+      "grad_norm": 0.5501636063381948,
+      "learning_rate": 1.656456676480675e-05,
+      "loss": 0.3071,
+      "step": 5534
+    },
+    {
+      "epoch": 1.0910883280757098,
+      "grad_norm": 0.5456326413445273,
+      "learning_rate": 1.6563397402578432e-05,
+      "loss": 0.3125,
+      "step": 5535
+    },
+    {
+      "epoch": 1.0912854889589905,
+      "grad_norm": 0.5648819649096508,
+      "learning_rate": 1.6562227882659213e-05,
+      "loss": 0.3379,
+      "step": 5536
+    },
+    {
+      "epoch": 1.0914826498422714,
+      "grad_norm": 0.5019198980389612,
+      "learning_rate": 1.656105820507718e-05,
+      "loss": 0.32,
+      "step": 5537
+    },
+    {
+      "epoch": 1.091679810725552,
+      "grad_norm": 0.5549566520131515,
+      "learning_rate": 1.655988836986044e-05,
+      "loss": 0.352,
+      "step": 5538
+    },
+    {
+      "epoch": 1.0918769716088328,
+      "grad_norm": 0.47299201163230054,
+      "learning_rate": 1.65587183770371e-05,
+      "loss": 0.2692,
+      "step": 5539
+    },
+    {
+      "epoch": 1.0920741324921135,
+      "grad_norm": 0.54789761189984,
+      "learning_rate": 1.6557548226635266e-05,
+      "loss": 0.3395,
+      "step": 5540
+    },
+    {
+      "epoch": 1.0922712933753944,
+      "grad_norm": 0.5853498159297642,
+      "learning_rate": 1.655637791868306e-05,
+      "loss": 0.336,
+      "step": 5541
+    },
+    {
+      "epoch": 1.0924684542586751,
+      "grad_norm": 0.630929601289797,
+      "learning_rate": 1.6555207453208596e-05,
+      "loss": 0.3531,
+      "step": 5542
+    },
+    {
+      "epoch": 1.0926656151419558,
+      "grad_norm": 0.5228836259363071,
+      "learning_rate": 1.655403683023999e-05,
+      "loss": 0.3291,
+      "step": 5543
+    },
+    {
+      "epoch": 1.0928627760252365,
+      "grad_norm": 0.5556250366890163,
+      "learning_rate": 1.655286604980537e-05,
+      "loss": 0.3462,
+      "step": 5544
+    },
+    {
+      "epoch": 1.0930599369085174,
+      "grad_norm": 0.5731618356985932,
+      "learning_rate": 1.655169511193287e-05,
+      "loss": 0.3276,
+      "step": 5545
+    },
+    {
+      "epoch": 1.0932570977917981,
+      "grad_norm": 0.5172208442087932,
+      "learning_rate": 1.6550524016650617e-05,
+      "loss": 0.3295,
+      "step": 5546
+    },
+    {
+      "epoch": 1.0934542586750788,
+      "grad_norm": 0.5971863550898441,
+      "learning_rate": 1.6549352763986747e-05,
+      "loss": 0.3635,
+      "step": 5547
+    },
+    {
+      "epoch": 1.0936514195583595,
+      "grad_norm": 0.5043253634738109,
+      "learning_rate": 1.6548181353969407e-05,
+      "loss": 0.3164,
+      "step": 5548
+    },
+    {
+      "epoch": 1.0938485804416405,
+      "grad_norm": 0.5441825059408084,
+      "learning_rate": 1.654700978662674e-05,
+      "loss": 0.3326,
+      "step": 5549
+    },
+    {
+      "epoch": 1.0940457413249212,
+      "grad_norm": 0.5397553446197998,
+      "learning_rate": 1.654583806198688e-05,
+      "loss": 0.3658,
+      "step": 5550
+    },
+    {
+      "epoch": 1.0942429022082019,
+      "grad_norm": 0.5504743804813628,
+      "learning_rate": 1.6544666180077996e-05,
+      "loss": 0.3526,
+      "step": 5551
+    },
+    {
+      "epoch": 1.0944400630914826,
+      "grad_norm": 0.5424844935779093,
+      "learning_rate": 1.6543494140928236e-05,
+      "loss": 0.3307,
+      "step": 5552
+    },
+    {
+      "epoch": 1.0946372239747635,
+      "grad_norm": 0.526220112929215,
+      "learning_rate": 1.654232194456576e-05,
+      "loss": 0.3308,
+      "step": 5553
+    },
+    {
+      "epoch": 1.0948343848580442,
+      "grad_norm": 0.5236749400742969,
+      "learning_rate": 1.6541149591018727e-05,
+      "loss": 0.3453,
+      "step": 5554
+    },
+    {
+      "epoch": 1.0950315457413249,
+      "grad_norm": 0.5266507318689269,
+      "learning_rate": 1.6539977080315313e-05,
+      "loss": 0.3305,
+      "step": 5555
+    },
+    {
+      "epoch": 1.0952287066246056,
+      "grad_norm": 0.5216109610169218,
+      "learning_rate": 1.653880441248368e-05,
+      "loss": 0.3316,
+      "step": 5556
+    },
+    {
+      "epoch": 1.0954258675078865,
+      "grad_norm": 0.5163493651525781,
+      "learning_rate": 1.6537631587552007e-05,
+      "loss": 0.342,
+      "step": 5557
+    },
+    {
+      "epoch": 1.0956230283911672,
+      "grad_norm": 0.5078186202086805,
+      "learning_rate": 1.6536458605548467e-05,
+      "loss": 0.3249,
+      "step": 5558
+    },
+    {
+      "epoch": 1.095820189274448,
+      "grad_norm": 0.5276901659588409,
+      "learning_rate": 1.6535285466501247e-05,
+      "loss": 0.3326,
+      "step": 5559
+    },
+    {
+      "epoch": 1.0960173501577286,
+      "grad_norm": 0.5355919928397173,
+      "learning_rate": 1.653411217043853e-05,
+      "loss": 0.352,
+      "step": 5560
+    },
+    {
+      "epoch": 1.0962145110410095,
+      "grad_norm": 0.4763053400285673,
+      "learning_rate": 1.653293871738851e-05,
+      "loss": 0.2937,
+      "step": 5561
+    },
+    {
+      "epoch": 1.0964116719242902,
+      "grad_norm": 0.5380622063499523,
+      "learning_rate": 1.6531765107379374e-05,
+      "loss": 0.3585,
+      "step": 5562
+    },
+    {
+      "epoch": 1.096608832807571,
+      "grad_norm": 0.5111119725625164,
+      "learning_rate": 1.653059134043932e-05,
+      "loss": 0.3336,
+      "step": 5563
+    },
+    {
+      "epoch": 1.0968059936908516,
+      "grad_norm": 0.49455292690513286,
+      "learning_rate": 1.652941741659655e-05,
+      "loss": 0.3136,
+      "step": 5564
+    },
+    {
+      "epoch": 1.0970031545741326,
+      "grad_norm": 0.6845117374932369,
+      "learning_rate": 1.652824333587927e-05,
+      "loss": 0.332,
+      "step": 5565
+    },
+    {
+      "epoch": 1.0972003154574133,
+      "grad_norm": 0.6523179647836078,
+      "learning_rate": 1.652706909831569e-05,
+      "loss": 0.3441,
+      "step": 5566
+    },
+    {
+      "epoch": 1.097397476340694,
+      "grad_norm": 0.4963174475713189,
+      "learning_rate": 1.6525894703934013e-05,
+      "loss": 0.3077,
+      "step": 5567
+    },
+    {
+      "epoch": 1.0975946372239749,
+      "grad_norm": 0.508004655738834,
+      "learning_rate": 1.6524720152762462e-05,
+      "loss": 0.3201,
+      "step": 5568
+    },
+    {
+      "epoch": 1.0977917981072556,
+      "grad_norm": 0.5984239179107959,
+      "learning_rate": 1.6523545444829254e-05,
+      "loss": 0.3592,
+      "step": 5569
+    },
+    {
+      "epoch": 1.0979889589905363,
+      "grad_norm": 0.5360407905331267,
+      "learning_rate": 1.6522370580162614e-05,
+      "loss": 0.3261,
+      "step": 5570
+    },
+    {
+      "epoch": 1.098186119873817,
+      "grad_norm": 0.5556220479614998,
+      "learning_rate": 1.652119555879077e-05,
+      "loss": 0.3431,
+      "step": 5571
+    },
+    {
+      "epoch": 1.0983832807570977,
+      "grad_norm": 0.5604439658652886,
+      "learning_rate": 1.652002038074195e-05,
+      "loss": 0.3145,
+      "step": 5572
+    },
+    {
+      "epoch": 1.0985804416403786,
+      "grad_norm": 0.5477226089327527,
+      "learning_rate": 1.651884504604439e-05,
+      "loss": 0.3404,
+      "step": 5573
+    },
+    {
+      "epoch": 1.0987776025236593,
+      "grad_norm": 0.5145353614083515,
+      "learning_rate": 1.6517669554726327e-05,
+      "loss": 0.3197,
+      "step": 5574
+    },
+    {
+      "epoch": 1.09897476340694,
+      "grad_norm": 0.5111044242152223,
+      "learning_rate": 1.6516493906816005e-05,
+      "loss": 0.3053,
+      "step": 5575
+    },
+    {
+      "epoch": 1.099171924290221,
+      "grad_norm": 0.5405424477707778,
+      "learning_rate": 1.6515318102341672e-05,
+      "loss": 0.3324,
+      "step": 5576
+    },
+    {
+      "epoch": 1.0993690851735016,
+      "grad_norm": 0.5193363546526818,
+      "learning_rate": 1.651414214133157e-05,
+      "loss": 0.3423,
+      "step": 5577
+    },
+    {
+      "epoch": 1.0995662460567823,
+      "grad_norm": 0.5144317929362779,
+      "learning_rate": 1.651296602381396e-05,
+      "loss": 0.3228,
+      "step": 5578
+    },
+    {
+      "epoch": 1.099763406940063,
+      "grad_norm": 0.5119676049482131,
+      "learning_rate": 1.6511789749817095e-05,
+      "loss": 0.3118,
+      "step": 5579
+    },
+    {
+      "epoch": 1.099960567823344,
+      "grad_norm": 0.5105599242984479,
+      "learning_rate": 1.651061331936924e-05,
+      "loss": 0.3323,
+      "step": 5580
+    },
+    {
+      "epoch": 1.1001577287066246,
+      "grad_norm": 0.5080653593030011,
+      "learning_rate": 1.6509436732498656e-05,
+      "loss": 0.3318,
+      "step": 5581
+    },
+    {
+      "epoch": 1.1003548895899053,
+      "grad_norm": 0.5031898165168965,
+      "learning_rate": 1.650825998923361e-05,
+      "loss": 0.3222,
+      "step": 5582
+    },
+    {
+      "epoch": 1.100552050473186,
+      "grad_norm": 0.5226955063741691,
+      "learning_rate": 1.650708308960238e-05,
+      "loss": 0.355,
+      "step": 5583
+    },
+    {
+      "epoch": 1.100749211356467,
+      "grad_norm": 0.5038434520036276,
+      "learning_rate": 1.6505906033633236e-05,
+      "loss": 0.3391,
+      "step": 5584
+    },
+    {
+      "epoch": 1.1009463722397477,
+      "grad_norm": 0.5059125745144066,
+      "learning_rate": 1.650472882135446e-05,
+      "loss": 0.3039,
+      "step": 5585
+    },
+    {
+      "epoch": 1.1011435331230284,
+      "grad_norm": 0.5040045595649187,
+      "learning_rate": 1.650355145279434e-05,
+      "loss": 0.3477,
+      "step": 5586
+    },
+    {
+      "epoch": 1.101340694006309,
+      "grad_norm": 0.5192804609775385,
+      "learning_rate": 1.6502373927981154e-05,
+      "loss": 0.33,
+      "step": 5587
+    },
+    {
+      "epoch": 1.10153785488959,
+      "grad_norm": 0.5922051341334427,
+      "learning_rate": 1.6501196246943202e-05,
+      "loss": 0.3541,
+      "step": 5588
+    },
+    {
+      "epoch": 1.1017350157728707,
+      "grad_norm": 0.49875642200196624,
+      "learning_rate": 1.650001840970877e-05,
+      "loss": 0.2838,
+      "step": 5589
+    },
+    {
+      "epoch": 1.1019321766561514,
+      "grad_norm": 0.4856484105077422,
+      "learning_rate": 1.6498840416306168e-05,
+      "loss": 0.3083,
+      "step": 5590
+    },
+    {
+      "epoch": 1.102129337539432,
+      "grad_norm": 0.5132652807308573,
+      "learning_rate": 1.6497662266763685e-05,
+      "loss": 0.315,
+      "step": 5591
+    },
+    {
+      "epoch": 1.102326498422713,
+      "grad_norm": 0.5193099983968429,
+      "learning_rate": 1.6496483961109638e-05,
+      "loss": 0.3029,
+      "step": 5592
+    },
+    {
+      "epoch": 1.1025236593059937,
+      "grad_norm": 0.5331749318093294,
+      "learning_rate": 1.649530549937233e-05,
+      "loss": 0.3383,
+      "step": 5593
+    },
+    {
+      "epoch": 1.1027208201892744,
+      "grad_norm": 0.5453618564420977,
+      "learning_rate": 1.6494126881580077e-05,
+      "loss": 0.3239,
+      "step": 5594
+    },
+    {
+      "epoch": 1.1029179810725551,
+      "grad_norm": 0.49856539989644866,
+      "learning_rate": 1.64929481077612e-05,
+      "loss": 0.3423,
+      "step": 5595
+    },
+    {
+      "epoch": 1.103115141955836,
+      "grad_norm": 0.5012478925347682,
+      "learning_rate": 1.649176917794401e-05,
+      "loss": 0.3283,
+      "step": 5596
+    },
+    {
+      "epoch": 1.1033123028391167,
+      "grad_norm": 0.5109791832923377,
+      "learning_rate": 1.649059009215684e-05,
+      "loss": 0.3039,
+      "step": 5597
+    },
+    {
+      "epoch": 1.1035094637223974,
+      "grad_norm": 0.5323841890352232,
+      "learning_rate": 1.6489410850428017e-05,
+      "loss": 0.3441,
+      "step": 5598
+    },
+    {
+      "epoch": 1.1037066246056781,
+      "grad_norm": 0.5666261888064066,
+      "learning_rate": 1.6488231452785867e-05,
+      "loss": 0.3586,
+      "step": 5599
+    },
+    {
+      "epoch": 1.103903785488959,
+      "grad_norm": 0.49589632413833346,
+      "learning_rate": 1.6487051899258738e-05,
+      "loss": 0.3359,
+      "step": 5600
+    },
+    {
+      "epoch": 1.1041009463722398,
+      "grad_norm": 0.522410156012698,
+      "learning_rate": 1.6485872189874962e-05,
+      "loss": 0.3242,
+      "step": 5601
+    },
+    {
+      "epoch": 1.1042981072555205,
+      "grad_norm": 0.5598921164556169,
+      "learning_rate": 1.6484692324662883e-05,
+      "loss": 0.3353,
+      "step": 5602
+    },
+    {
+      "epoch": 1.1044952681388012,
+      "grad_norm": 0.565537446173838,
+      "learning_rate": 1.6483512303650847e-05,
+      "loss": 0.3196,
+      "step": 5603
+    },
+    {
+      "epoch": 1.104692429022082,
+      "grad_norm": 0.5611042309344503,
+      "learning_rate": 1.648233212686721e-05,
+      "loss": 0.3531,
+      "step": 5604
+    },
+    {
+      "epoch": 1.1048895899053628,
+      "grad_norm": 0.6104352248715608,
+      "learning_rate": 1.648115179434032e-05,
+      "loss": 0.3411,
+      "step": 5605
+    },
+    {
+      "epoch": 1.1050867507886435,
+      "grad_norm": 0.5182267968343534,
+      "learning_rate": 1.647997130609854e-05,
+      "loss": 0.3328,
+      "step": 5606
+    },
+    {
+      "epoch": 1.1052839116719242,
+      "grad_norm": 0.6629279110385724,
+      "learning_rate": 1.647879066217023e-05,
+      "loss": 0.3664,
+      "step": 5607
+    },
+    {
+      "epoch": 1.1054810725552051,
+      "grad_norm": 0.48783718160836653,
+      "learning_rate": 1.6477609862583758e-05,
+      "loss": 0.3184,
+      "step": 5608
+    },
+    {
+      "epoch": 1.1056782334384858,
+      "grad_norm": 0.5388589810531627,
+      "learning_rate": 1.6476428907367497e-05,
+      "loss": 0.322,
+      "step": 5609
+    },
+    {
+      "epoch": 1.1058753943217665,
+      "grad_norm": 0.6019162312791431,
+      "learning_rate": 1.647524779654981e-05,
+      "loss": 0.35,
+      "step": 5610
+    },
+    {
+      "epoch": 1.1060725552050474,
+      "grad_norm": 0.5227308598811116,
+      "learning_rate": 1.6474066530159083e-05,
+      "loss": 0.346,
+      "step": 5611
+    },
+    {
+      "epoch": 1.1062697160883281,
+      "grad_norm": 0.5024291222077405,
+      "learning_rate": 1.6472885108223694e-05,
+      "loss": 0.3115,
+      "step": 5612
+    },
+    {
+      "epoch": 1.1064668769716088,
+      "grad_norm": 0.5971378494359267,
+      "learning_rate": 1.647170353077203e-05,
+      "loss": 0.3421,
+      "step": 5613
+    },
+    {
+      "epoch": 1.1066640378548895,
+      "grad_norm": 0.5323688610636911,
+      "learning_rate": 1.647052179783247e-05,
+      "loss": 0.3685,
+      "step": 5614
+    },
+    {
+      "epoch": 1.1068611987381702,
+      "grad_norm": 0.5494630004599759,
+      "learning_rate": 1.646933990943342e-05,
+      "loss": 0.3596,
+      "step": 5615
+    },
+    {
+      "epoch": 1.1070583596214512,
+      "grad_norm": 0.5315790166072238,
+      "learning_rate": 1.6468157865603265e-05,
+      "loss": 0.326,
+      "step": 5616
+    },
+    {
+      "epoch": 1.1072555205047319,
+      "grad_norm": 1.095318117653315,
+      "learning_rate": 1.646697566637041e-05,
+      "loss": 0.3391,
+      "step": 5617
+    },
+    {
+      "epoch": 1.1074526813880126,
+      "grad_norm": 0.4910427373938156,
+      "learning_rate": 1.6465793311763255e-05,
+      "loss": 0.3082,
+      "step": 5618
+    },
+    {
+      "epoch": 1.1076498422712935,
+      "grad_norm": 0.5479397492318889,
+      "learning_rate": 1.646461080181021e-05,
+      "loss": 0.3344,
+      "step": 5619
+    },
+    {
+      "epoch": 1.1078470031545742,
+      "grad_norm": 0.5103248145847307,
+      "learning_rate": 1.6463428136539684e-05,
+      "loss": 0.324,
+      "step": 5620
+    },
+    {
+      "epoch": 1.1080441640378549,
+      "grad_norm": 0.5630855688757884,
+      "learning_rate": 1.6462245315980094e-05,
+      "loss": 0.3275,
+      "step": 5621
+    },
+    {
+      "epoch": 1.1082413249211356,
+      "grad_norm": 0.5263509918106348,
+      "learning_rate": 1.6461062340159853e-05,
+      "loss": 0.3189,
+      "step": 5622
+    },
+    {
+      "epoch": 1.1084384858044165,
+      "grad_norm": 0.6493282607434927,
+      "learning_rate": 1.6459879209107394e-05,
+      "loss": 0.3394,
+      "step": 5623
+    },
+    {
+      "epoch": 1.1086356466876972,
+      "grad_norm": 0.6049538585958488,
+      "learning_rate": 1.6458695922851126e-05,
+      "loss": 0.3308,
+      "step": 5624
+    },
+    {
+      "epoch": 1.108832807570978,
+      "grad_norm": 0.531824713225462,
+      "learning_rate": 1.6457512481419492e-05,
+      "loss": 0.3412,
+      "step": 5625
+    },
+    {
+      "epoch": 1.1090299684542586,
+      "grad_norm": 0.5404382119881224,
+      "learning_rate": 1.6456328884840917e-05,
+      "loss": 0.3427,
+      "step": 5626
+    },
+    {
+      "epoch": 1.1092271293375395,
+      "grad_norm": 0.5405048280095781,
+      "learning_rate": 1.6455145133143843e-05,
+      "loss": 0.3322,
+      "step": 5627
+    },
+    {
+      "epoch": 1.1094242902208202,
+      "grad_norm": 0.5808034796183005,
+      "learning_rate": 1.645396122635671e-05,
+      "loss": 0.3764,
+      "step": 5628
+    },
+    {
+      "epoch": 1.109621451104101,
+      "grad_norm": 0.5655517916406795,
+      "learning_rate": 1.6452777164507957e-05,
+      "loss": 0.3236,
+      "step": 5629
+    },
+    {
+      "epoch": 1.1098186119873816,
+      "grad_norm": 0.5394522259534293,
+      "learning_rate": 1.6451592947626043e-05,
+      "loss": 0.3317,
+      "step": 5630
+    },
+    {
+      "epoch": 1.1100157728706626,
+      "grad_norm": 0.5151963061557067,
+      "learning_rate": 1.6450408575739407e-05,
+      "loss": 0.3467,
+      "step": 5631
+    },
+    {
+      "epoch": 1.1102129337539433,
+      "grad_norm": 0.5279890856885844,
+      "learning_rate": 1.6449224048876512e-05,
+      "loss": 0.3258,
+      "step": 5632
+    },
+    {
+      "epoch": 1.110410094637224,
+      "grad_norm": 0.5779114047767783,
+      "learning_rate": 1.6448039367065816e-05,
+      "loss": 0.3662,
+      "step": 5633
+    },
+    {
+      "epoch": 1.1106072555205047,
+      "grad_norm": 0.5019131557840137,
+      "learning_rate": 1.6446854530335783e-05,
+      "loss": 0.3201,
+      "step": 5634
+    },
+    {
+      "epoch": 1.1108044164037856,
+      "grad_norm": 0.5524102955169621,
+      "learning_rate": 1.6445669538714878e-05,
+      "loss": 0.3678,
+      "step": 5635
+    },
+    {
+      "epoch": 1.1110015772870663,
+      "grad_norm": 0.5274386217024158,
+      "learning_rate": 1.6444484392231574e-05,
+      "loss": 0.3452,
+      "step": 5636
+    },
+    {
+      "epoch": 1.111198738170347,
+      "grad_norm": 0.5065419330922859,
+      "learning_rate": 1.6443299090914336e-05,
+      "loss": 0.3135,
+      "step": 5637
+    },
+    {
+      "epoch": 1.1113958990536277,
+      "grad_norm": 0.5321762740305643,
+      "learning_rate": 1.6442113634791653e-05,
+      "loss": 0.36,
+      "step": 5638
+    },
+    {
+      "epoch": 1.1115930599369086,
+      "grad_norm": 0.4997951806163294,
+      "learning_rate": 1.6440928023892e-05,
+      "loss": 0.3222,
+      "step": 5639
+    },
+    {
+      "epoch": 1.1117902208201893,
+      "grad_norm": 0.4876569661311848,
+      "learning_rate": 1.643974225824387e-05,
+      "loss": 0.3212,
+      "step": 5640
+    },
+    {
+      "epoch": 1.11198738170347,
+      "grad_norm": 0.5216914599856519,
+      "learning_rate": 1.643855633787574e-05,
+      "loss": 0.3373,
+      "step": 5641
+    },
+    {
+      "epoch": 1.1121845425867507,
+      "grad_norm": 0.48155980759723177,
+      "learning_rate": 1.643737026281611e-05,
+      "loss": 0.3003,
+      "step": 5642
+    },
+    {
+      "epoch": 1.1123817034700316,
+      "grad_norm": 0.5168478481124772,
+      "learning_rate": 1.643618403309348e-05,
+      "loss": 0.3435,
+      "step": 5643
+    },
+    {
+      "epoch": 1.1125788643533123,
+      "grad_norm": 0.5105355384691713,
+      "learning_rate": 1.643499764873634e-05,
+      "loss": 0.3098,
+      "step": 5644
+    },
+    {
+      "epoch": 1.112776025236593,
+      "grad_norm": 0.5395714096049974,
+      "learning_rate": 1.6433811109773202e-05,
+      "loss": 0.3495,
+      "step": 5645
+    },
+    {
+      "epoch": 1.1129731861198737,
+      "grad_norm": 0.5042000786890821,
+      "learning_rate": 1.643262441623257e-05,
+      "loss": 0.3231,
+      "step": 5646
+    },
+    {
+      "epoch": 1.1131703470031546,
+      "grad_norm": 0.5278311430339191,
+      "learning_rate": 1.6431437568142956e-05,
+      "loss": 0.3529,
+      "step": 5647
+    },
+    {
+      "epoch": 1.1133675078864353,
+      "grad_norm": 0.5180324474360586,
+      "learning_rate": 1.6430250565532878e-05,
+      "loss": 0.3456,
+      "step": 5648
+    },
+    {
+      "epoch": 1.113564668769716,
+      "grad_norm": 0.5081894704748161,
+      "learning_rate": 1.642906340843085e-05,
+      "loss": 0.3288,
+      "step": 5649
+    },
+    {
+      "epoch": 1.1137618296529967,
+      "grad_norm": 0.6664115588691337,
+      "learning_rate": 1.6427876096865394e-05,
+      "loss": 0.3455,
+      "step": 5650
+    },
+    {
+      "epoch": 1.1139589905362777,
+      "grad_norm": 3.5585025573079028,
+      "learning_rate": 1.642668863086504e-05,
+      "loss": 0.3337,
+      "step": 5651
+    },
+    {
+      "epoch": 1.1141561514195584,
+      "grad_norm": 0.5420249391026032,
+      "learning_rate": 1.642550101045832e-05,
+      "loss": 0.3192,
+      "step": 5652
+    },
+    {
+      "epoch": 1.114353312302839,
+      "grad_norm": 0.5862664575213457,
+      "learning_rate": 1.6424313235673758e-05,
+      "loss": 0.3308,
+      "step": 5653
+    },
+    {
+      "epoch": 1.1145504731861198,
+      "grad_norm": 1.8318481156437365,
+      "learning_rate": 1.6423125306539903e-05,
+      "loss": 0.3252,
+      "step": 5654
+    },
+    {
+      "epoch": 1.1147476340694007,
+      "grad_norm": 0.6180923856298878,
+      "learning_rate": 1.6421937223085284e-05,
+      "loss": 0.3386,
+      "step": 5655
+    },
+    {
+      "epoch": 1.1149447949526814,
+      "grad_norm": 0.5427448619206314,
+      "learning_rate": 1.6420748985338454e-05,
+      "loss": 0.3212,
+      "step": 5656
+    },
+    {
+      "epoch": 1.115141955835962,
+      "grad_norm": 0.5267157193078928,
+      "learning_rate": 1.641956059332796e-05,
+      "loss": 0.3326,
+      "step": 5657
+    },
+    {
+      "epoch": 1.1153391167192428,
+      "grad_norm": 0.5512555264071693,
+      "learning_rate": 1.6418372047082352e-05,
+      "loss": 0.3224,
+      "step": 5658
+    },
+    {
+      "epoch": 1.1155362776025237,
+      "grad_norm": 0.5720422317475261,
+      "learning_rate": 1.6417183346630188e-05,
+      "loss": 0.334,
+      "step": 5659
+    },
+    {
+      "epoch": 1.1157334384858044,
+      "grad_norm": 0.5754561614830481,
+      "learning_rate": 1.6415994492000026e-05,
+      "loss": 0.3141,
+      "step": 5660
+    },
+    {
+      "epoch": 1.1159305993690851,
+      "grad_norm": 0.6375296150804535,
+      "learning_rate": 1.6414805483220426e-05,
+      "loss": 0.3275,
+      "step": 5661
+    },
+    {
+      "epoch": 1.116127760252366,
+      "grad_norm": 0.509049237537517,
+      "learning_rate": 1.641361632031996e-05,
+      "loss": 0.3239,
+      "step": 5662
+    },
+    {
+      "epoch": 1.1163249211356467,
+      "grad_norm": 0.5128636795892693,
+      "learning_rate": 1.64124270033272e-05,
+      "loss": 0.3159,
+      "step": 5663
+    },
+    {
+      "epoch": 1.1165220820189274,
+      "grad_norm": 0.4912807390670567,
+      "learning_rate": 1.6411237532270718e-05,
+      "loss": 0.3144,
+      "step": 5664
+    },
+    {
+      "epoch": 1.1167192429022081,
+      "grad_norm": 0.5397224874275655,
+      "learning_rate": 1.641004790717909e-05,
+      "loss": 0.3313,
+      "step": 5665
+    },
+    {
+      "epoch": 1.1169164037854888,
+      "grad_norm": 0.5253163708246812,
+      "learning_rate": 1.64088581280809e-05,
+      "loss": 0.3363,
+      "step": 5666
+    },
+    {
+      "epoch": 1.1171135646687698,
+      "grad_norm": 0.6246187071629071,
+      "learning_rate": 1.6407668195004726e-05,
+      "loss": 0.3517,
+      "step": 5667
+    },
+    {
+      "epoch": 1.1173107255520505,
+      "grad_norm": 0.510212683182266,
+      "learning_rate": 1.640647810797917e-05,
+      "loss": 0.3423,
+      "step": 5668
+    },
+    {
+      "epoch": 1.1175078864353312,
+      "grad_norm": 0.5433994977386873,
+      "learning_rate": 1.6405287867032816e-05,
+      "loss": 0.3364,
+      "step": 5669
+    },
+    {
+      "epoch": 1.117705047318612,
+      "grad_norm": 0.5114129587959397,
+      "learning_rate": 1.6404097472194264e-05,
+      "loss": 0.3212,
+      "step": 5670
+    },
+    {
+      "epoch": 1.1179022082018928,
+      "grad_norm": 0.6184701890817311,
+      "learning_rate": 1.6402906923492113e-05,
+      "loss": 0.3586,
+      "step": 5671
+    },
+    {
+      "epoch": 1.1180993690851735,
+      "grad_norm": 0.5157737817133912,
+      "learning_rate": 1.6401716220954968e-05,
+      "loss": 0.3466,
+      "step": 5672
+    },
+    {
+      "epoch": 1.1182965299684542,
+      "grad_norm": 0.5239425894073806,
+      "learning_rate": 1.640052536461143e-05,
+      "loss": 0.3367,
+      "step": 5673
+    },
+    {
+      "epoch": 1.1184936908517351,
+      "grad_norm": 0.6216718049365948,
+      "learning_rate": 1.6399334354490123e-05,
+      "loss": 0.3495,
+      "step": 5674
+    },
+    {
+      "epoch": 1.1186908517350158,
+      "grad_norm": 0.5111026021999556,
+      "learning_rate": 1.6398143190619648e-05,
+      "loss": 0.3212,
+      "step": 5675
+    },
+    {
+      "epoch": 1.1188880126182965,
+      "grad_norm": 0.49466574177659056,
+      "learning_rate": 1.6396951873028634e-05,
+      "loss": 0.3287,
+      "step": 5676
+    },
+    {
+      "epoch": 1.1190851735015772,
+      "grad_norm": 0.48459040865141534,
+      "learning_rate": 1.63957604017457e-05,
+      "loss": 0.3209,
+      "step": 5677
+    },
+    {
+      "epoch": 1.1192823343848581,
+      "grad_norm": 0.5002774049961711,
+      "learning_rate": 1.6394568776799472e-05,
+      "loss": 0.3325,
+      "step": 5678
+    },
+    {
+      "epoch": 1.1194794952681388,
+      "grad_norm": 0.49713568658847396,
+      "learning_rate": 1.6393376998218583e-05,
+      "loss": 0.3456,
+      "step": 5679
+    },
+    {
+      "epoch": 1.1196766561514195,
+      "grad_norm": 0.49493207831234715,
+      "learning_rate": 1.6392185066031657e-05,
+      "loss": 0.3127,
+      "step": 5680
+    },
+    {
+      "epoch": 1.1198738170347002,
+      "grad_norm": 0.5362871237694371,
+      "learning_rate": 1.639099298026734e-05,
+      "loss": 0.3451,
+      "step": 5681
+    },
+    {
+      "epoch": 1.1200709779179812,
+      "grad_norm": 0.5294651321602112,
+      "learning_rate": 1.6389800740954268e-05,
+      "loss": 0.362,
+      "step": 5682
+    },
+    {
+      "epoch": 1.1202681388012619,
+      "grad_norm": 0.48735850316187274,
+      "learning_rate": 1.6388608348121088e-05,
+      "loss": 0.341,
+      "step": 5683
+    },
+    {
+      "epoch": 1.1204652996845426,
+      "grad_norm": 0.5248544933902106,
+      "learning_rate": 1.638741580179645e-05,
+      "loss": 0.3176,
+      "step": 5684
+    },
+    {
+      "epoch": 1.1206624605678233,
+      "grad_norm": 0.5439891987403144,
+      "learning_rate": 1.6386223102009e-05,
+      "loss": 0.359,
+      "step": 5685
+    },
+    {
+      "epoch": 1.1208596214511042,
+      "grad_norm": 0.5011856573812568,
+      "learning_rate": 1.6385030248787402e-05,
+      "loss": 0.33,
+      "step": 5686
+    },
+    {
+      "epoch": 1.1210567823343849,
+      "grad_norm": 0.4973353261615954,
+      "learning_rate": 1.6383837242160304e-05,
+      "loss": 0.3253,
+      "step": 5687
+    },
+    {
+      "epoch": 1.1212539432176656,
+      "grad_norm": 0.5271878211430877,
+      "learning_rate": 1.638264408215638e-05,
+      "loss": 0.3273,
+      "step": 5688
+    },
+    {
+      "epoch": 1.1214511041009463,
+      "grad_norm": 0.5007462516405465,
+      "learning_rate": 1.6381450768804293e-05,
+      "loss": 0.3566,
+      "step": 5689
+    },
+    {
+      "epoch": 1.1216482649842272,
+      "grad_norm": 0.5032257739082565,
+      "learning_rate": 1.638025730213271e-05,
+      "loss": 0.3189,
+      "step": 5690
+    },
+    {
+      "epoch": 1.121845425867508,
+      "grad_norm": 0.5057418106825747,
+      "learning_rate": 1.637906368217031e-05,
+      "loss": 0.3296,
+      "step": 5691
+    },
+    {
+      "epoch": 1.1220425867507886,
+      "grad_norm": 0.5220864858440248,
+      "learning_rate": 1.6377869908945763e-05,
+      "loss": 0.3013,
+      "step": 5692
+    },
+    {
+      "epoch": 1.1222397476340693,
+      "grad_norm": 0.49972633540948586,
+      "learning_rate": 1.637667598248776e-05,
+      "loss": 0.3246,
+      "step": 5693
+    },
+    {
+      "epoch": 1.1224369085173502,
+      "grad_norm": 0.48521860052546617,
+      "learning_rate": 1.6375481902824975e-05,
+      "loss": 0.3067,
+      "step": 5694
+    },
+    {
+      "epoch": 1.122634069400631,
+      "grad_norm": 0.5137855325686259,
+      "learning_rate": 1.6374287669986108e-05,
+      "loss": 0.333,
+      "step": 5695
+    },
+    {
+      "epoch": 1.1228312302839116,
+      "grad_norm": 0.5053341438549699,
+      "learning_rate": 1.637309328399985e-05,
+      "loss": 0.3131,
+      "step": 5696
+    },
+    {
+      "epoch": 1.1230283911671923,
+      "grad_norm": 0.4811130427810911,
+      "learning_rate": 1.6371898744894888e-05,
+      "loss": 0.2932,
+      "step": 5697
+    },
+    {
+      "epoch": 1.1232255520504733,
+      "grad_norm": 0.4692455927554586,
+      "learning_rate": 1.6370704052699927e-05,
+      "loss": 0.3081,
+      "step": 5698
+    },
+    {
+      "epoch": 1.123422712933754,
+      "grad_norm": 0.46390772652017165,
+      "learning_rate": 1.6369509207443676e-05,
+      "loss": 0.2891,
+      "step": 5699
+    },
+    {
+      "epoch": 1.1236198738170347,
+      "grad_norm": 0.5387624182052491,
+      "learning_rate": 1.6368314209154836e-05,
+      "loss": 0.357,
+      "step": 5700
+    },
+    {
+      "epoch": 1.1238170347003154,
+      "grad_norm": 0.4959986328725572,
+      "learning_rate": 1.6367119057862115e-05,
+      "loss": 0.3191,
+      "step": 5701
+    },
+    {
+      "epoch": 1.1240141955835963,
+      "grad_norm": 0.5071814974603067,
+      "learning_rate": 1.636592375359423e-05,
+      "loss": 0.3324,
+      "step": 5702
+    },
+    {
+      "epoch": 1.124211356466877,
+      "grad_norm": 0.5104341041644923,
+      "learning_rate": 1.6364728296379904e-05,
+      "loss": 0.3506,
+      "step": 5703
+    },
+    {
+      "epoch": 1.1244085173501577,
+      "grad_norm": 0.5464624679656105,
+      "learning_rate": 1.6363532686247853e-05,
+      "loss": 0.3123,
+      "step": 5704
+    },
+    {
+      "epoch": 1.1246056782334386,
+      "grad_norm": 0.5028129404728894,
+      "learning_rate": 1.636233692322681e-05,
+      "loss": 0.307,
+      "step": 5705
+    },
+    {
+      "epoch": 1.1248028391167193,
+      "grad_norm": 0.6152702419910974,
+      "learning_rate": 1.6361141007345494e-05,
+      "loss": 0.3553,
+      "step": 5706
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 0.48179193603259657,
+      "learning_rate": 1.6359944938632645e-05,
+      "loss": 0.3147,
+      "step": 5707
+    },
+    {
+      "epoch": 1.1251971608832807,
+      "grad_norm": 0.5598824963334975,
+      "learning_rate": 1.6358748717116993e-05,
+      "loss": 0.3402,
+      "step": 5708
+    },
+    {
+      "epoch": 1.1253943217665614,
+      "grad_norm": 0.5571114963157853,
+      "learning_rate": 1.6357552342827284e-05,
+      "loss": 0.3206,
+      "step": 5709
+    },
+    {
+      "epoch": 1.1255914826498423,
+      "grad_norm": 0.4925155006030104,
+      "learning_rate": 1.6356355815792263e-05,
+      "loss": 0.3077,
+      "step": 5710
+    },
+    {
+      "epoch": 1.125788643533123,
+      "grad_norm": 0.6517198961949976,
+      "learning_rate": 1.635515913604067e-05,
+      "loss": 0.3521,
+      "step": 5711
+    },
+    {
+      "epoch": 1.1259858044164037,
+      "grad_norm": 0.5747480536099305,
+      "learning_rate": 1.6353962303601266e-05,
+      "loss": 0.364,
+      "step": 5712
+    },
+    {
+      "epoch": 1.1261829652996846,
+      "grad_norm": 0.5712666247584104,
+      "learning_rate": 1.63527653185028e-05,
+      "loss": 0.3705,
+      "step": 5713
+    },
+    {
+      "epoch": 1.1263801261829653,
+      "grad_norm": 0.5587416233409346,
+      "learning_rate": 1.6351568180774033e-05,
+      "loss": 0.319,
+      "step": 5714
+    },
+    {
+      "epoch": 1.126577287066246,
+      "grad_norm": 0.5358691155957552,
+      "learning_rate": 1.635037089044372e-05,
+      "loss": 0.3336,
+      "step": 5715
+    },
+    {
+      "epoch": 1.1267744479495267,
+      "grad_norm": 0.5829710216948807,
+      "learning_rate": 1.6349173447540634e-05,
+      "loss": 0.3508,
+      "step": 5716
+    },
+    {
+      "epoch": 1.1269716088328074,
+      "grad_norm": 0.5281339705441374,
+      "learning_rate": 1.6347975852093544e-05,
+      "loss": 0.3266,
+      "step": 5717
+    },
+    {
+      "epoch": 1.1271687697160884,
+      "grad_norm": 0.5300750941636183,
+      "learning_rate": 1.6346778104131222e-05,
+      "loss": 0.3367,
+      "step": 5718
+    },
+    {
+      "epoch": 1.127365930599369,
+      "grad_norm": 0.5660626646152135,
+      "learning_rate": 1.634558020368245e-05,
+      "loss": 0.3543,
+      "step": 5719
+    },
+    {
+      "epoch": 1.1275630914826498,
+      "grad_norm": 0.4900690395273459,
+      "learning_rate": 1.6344382150775994e-05,
+      "loss": 0.2959,
+      "step": 5720
+    },
+    {
+      "epoch": 1.1277602523659307,
+      "grad_norm": 0.5929670034437629,
+      "learning_rate": 1.6343183945440656e-05,
+      "loss": 0.3446,
+      "step": 5721
+    },
+    {
+      "epoch": 1.1279574132492114,
+      "grad_norm": 0.5581429670822288,
+      "learning_rate": 1.634198558770521e-05,
+      "loss": 0.3424,
+      "step": 5722
+    },
+    {
+      "epoch": 1.128154574132492,
+      "grad_norm": 0.5554338618967727,
+      "learning_rate": 1.6340787077598456e-05,
+      "loss": 0.3407,
+      "step": 5723
+    },
+    {
+      "epoch": 1.1283517350157728,
+      "grad_norm": 0.4917238000036681,
+      "learning_rate": 1.6339588415149186e-05,
+      "loss": 0.3234,
+      "step": 5724
+    },
+    {
+      "epoch": 1.1285488958990537,
+      "grad_norm": 0.6106081461017013,
+      "learning_rate": 1.63383896003862e-05,
+      "loss": 0.3239,
+      "step": 5725
+    },
+    {
+      "epoch": 1.1287460567823344,
+      "grad_norm": 0.5010753713189542,
+      "learning_rate": 1.6337190633338294e-05,
+      "loss": 0.3191,
+      "step": 5726
+    },
+    {
+      "epoch": 1.1289432176656151,
+      "grad_norm": 0.48674816598759957,
+      "learning_rate": 1.6335991514034283e-05,
+      "loss": 0.2984,
+      "step": 5727
+    },
+    {
+      "epoch": 1.1291403785488958,
+      "grad_norm": 0.505286321715634,
+      "learning_rate": 1.6334792242502978e-05,
+      "loss": 0.3447,
+      "step": 5728
+    },
+    {
+      "epoch": 1.1293375394321767,
+      "grad_norm": 0.5524393690270945,
+      "learning_rate": 1.633359281877318e-05,
+      "loss": 0.3392,
+      "step": 5729
+    },
+    {
+      "epoch": 1.1295347003154574,
+      "grad_norm": 0.49196800593653534,
+      "learning_rate": 1.633239324287372e-05,
+      "loss": 0.3213,
+      "step": 5730
+    },
+    {
+      "epoch": 1.1297318611987381,
+      "grad_norm": 0.5143287994894329,
+      "learning_rate": 1.6331193514833413e-05,
+      "loss": 0.3173,
+      "step": 5731
+    },
+    {
+      "epoch": 1.1299290220820188,
+      "grad_norm": 0.5176908223126655,
+      "learning_rate": 1.632999363468108e-05,
+      "loss": 0.331,
+      "step": 5732
+    },
+    {
+      "epoch": 1.1301261829652998,
+      "grad_norm": 0.5267013800473567,
+      "learning_rate": 1.6328793602445553e-05,
+      "loss": 0.3564,
+      "step": 5733
+    },
+    {
+      "epoch": 1.1303233438485805,
+      "grad_norm": 0.5075643399395108,
+      "learning_rate": 1.6327593418155667e-05,
+      "loss": 0.3161,
+      "step": 5734
+    },
+    {
+      "epoch": 1.1305205047318612,
+      "grad_norm": 0.5262025072548631,
+      "learning_rate": 1.632639308184025e-05,
+      "loss": 0.3345,
+      "step": 5735
+    },
+    {
+      "epoch": 1.1307176656151419,
+      "grad_norm": 0.5292704648957198,
+      "learning_rate": 1.632519259352814e-05,
+      "loss": 0.3492,
+      "step": 5736
+    },
+    {
+      "epoch": 1.1309148264984228,
+      "grad_norm": 0.48829983713727304,
+      "learning_rate": 1.6323991953248192e-05,
+      "loss": 0.3171,
+      "step": 5737
+    },
+    {
+      "epoch": 1.1311119873817035,
+      "grad_norm": 0.4988512975455296,
+      "learning_rate": 1.6322791161029245e-05,
+      "loss": 0.3317,
+      "step": 5738
+    },
+    {
+      "epoch": 1.1313091482649842,
+      "grad_norm": 0.5269442131354157,
+      "learning_rate": 1.6321590216900146e-05,
+      "loss": 0.3308,
+      "step": 5739
+    },
+    {
+      "epoch": 1.1315063091482649,
+      "grad_norm": 0.4902800227785573,
+      "learning_rate": 1.632038912088975e-05,
+      "loss": 0.3061,
+      "step": 5740
+    },
+    {
+      "epoch": 1.1317034700315458,
+      "grad_norm": 0.5247877126830968,
+      "learning_rate": 1.6319187873026917e-05,
+      "loss": 0.3197,
+      "step": 5741
+    },
+    {
+      "epoch": 1.1319006309148265,
+      "grad_norm": 0.5541688059243428,
+      "learning_rate": 1.6317986473340504e-05,
+      "loss": 0.3295,
+      "step": 5742
+    },
+    {
+      "epoch": 1.1320977917981072,
+      "grad_norm": 0.5564999524149586,
+      "learning_rate": 1.631678492185938e-05,
+      "loss": 0.3419,
+      "step": 5743
+    },
+    {
+      "epoch": 1.132294952681388,
+      "grad_norm": 0.4751388780686587,
+      "learning_rate": 1.631558321861241e-05,
+      "loss": 0.3068,
+      "step": 5744
+    },
+    {
+      "epoch": 1.1324921135646688,
+      "grad_norm": 0.5251581221447832,
+      "learning_rate": 1.631438136362847e-05,
+      "loss": 0.3438,
+      "step": 5745
+    },
+    {
+      "epoch": 1.1326892744479495,
+      "grad_norm": 0.49557848543104777,
+      "learning_rate": 1.6313179356936432e-05,
+      "loss": 0.3336,
+      "step": 5746
+    },
+    {
+      "epoch": 1.1328864353312302,
+      "grad_norm": 0.5759205928280032,
+      "learning_rate": 1.631197719856518e-05,
+      "loss": 0.341,
+      "step": 5747
+    },
+    {
+      "epoch": 1.1330835962145112,
+      "grad_norm": 0.5279003018602465,
+      "learning_rate": 1.6310774888543584e-05,
+      "loss": 0.3361,
+      "step": 5748
+    },
+    {
+      "epoch": 1.1332807570977919,
+      "grad_norm": 0.5105221641137494,
+      "learning_rate": 1.6309572426900544e-05,
+      "loss": 0.3463,
+      "step": 5749
+    },
+    {
+      "epoch": 1.1334779179810726,
+      "grad_norm": 0.4979241157777894,
+      "learning_rate": 1.6308369813664945e-05,
+      "loss": 0.3316,
+      "step": 5750
+    },
+    {
+      "epoch": 1.1336750788643533,
+      "grad_norm": 0.480147781148586,
+      "learning_rate": 1.630716704886568e-05,
+      "loss": 0.3278,
+      "step": 5751
+    },
+    {
+      "epoch": 1.133872239747634,
+      "grad_norm": 0.5190158936682682,
+      "learning_rate": 1.630596413253165e-05,
+      "loss": 0.3487,
+      "step": 5752
+    },
+    {
+      "epoch": 1.1340694006309149,
+      "grad_norm": 0.48721620702672924,
+      "learning_rate": 1.6304761064691752e-05,
+      "loss": 0.331,
+      "step": 5753
+    },
+    {
+      "epoch": 1.1342665615141956,
+      "grad_norm": 0.48669878145102674,
+      "learning_rate": 1.6303557845374894e-05,
+      "loss": 0.3157,
+      "step": 5754
+    },
+    {
+      "epoch": 1.1344637223974763,
+      "grad_norm": 0.48762645113611275,
+      "learning_rate": 1.630235447460998e-05,
+      "loss": 0.3414,
+      "step": 5755
+    },
+    {
+      "epoch": 1.1346608832807572,
+      "grad_norm": 0.45531042840793384,
+      "learning_rate": 1.6301150952425925e-05,
+      "loss": 0.3097,
+      "step": 5756
+    },
+    {
+      "epoch": 1.134858044164038,
+      "grad_norm": 0.5085651272774692,
+      "learning_rate": 1.6299947278851644e-05,
+      "loss": 0.333,
+      "step": 5757
+    },
+    {
+      "epoch": 1.1350552050473186,
+      "grad_norm": 0.4893879791327469,
+      "learning_rate": 1.6298743453916057e-05,
+      "loss": 0.3044,
+      "step": 5758
+    },
+    {
+      "epoch": 1.1352523659305993,
+      "grad_norm": 0.5589441744381375,
+      "learning_rate": 1.6297539477648087e-05,
+      "loss": 0.3467,
+      "step": 5759
+    },
+    {
+      "epoch": 1.13544952681388,
+      "grad_norm": 0.4815754124533146,
+      "learning_rate": 1.6296335350076658e-05,
+      "loss": 0.3328,
+      "step": 5760
+    },
+    {
+      "epoch": 1.135646687697161,
+      "grad_norm": 0.4867892480819176,
+      "learning_rate": 1.6295131071230704e-05,
+      "loss": 0.3256,
+      "step": 5761
+    },
+    {
+      "epoch": 1.1358438485804416,
+      "grad_norm": 0.5067944820954798,
+      "learning_rate": 1.6293926641139154e-05,
+      "loss": 0.3323,
+      "step": 5762
+    },
+    {
+      "epoch": 1.1360410094637223,
+      "grad_norm": 0.5113632611165255,
+      "learning_rate": 1.629272205983095e-05,
+      "loss": 0.3311,
+      "step": 5763
+    },
+    {
+      "epoch": 1.1362381703470033,
+      "grad_norm": 0.5560655366878887,
+      "learning_rate": 1.6291517327335027e-05,
+      "loss": 0.3531,
+      "step": 5764
+    },
+    {
+      "epoch": 1.136435331230284,
+      "grad_norm": 0.5079516665115742,
+      "learning_rate": 1.6290312443680335e-05,
+      "loss": 0.3268,
+      "step": 5765
+    },
+    {
+      "epoch": 1.1366324921135647,
+      "grad_norm": 0.49476357370054935,
+      "learning_rate": 1.6289107408895827e-05,
+      "loss": 0.3199,
+      "step": 5766
+    },
+    {
+      "epoch": 1.1368296529968454,
+      "grad_norm": 0.5068561554934843,
+      "learning_rate": 1.6287902223010442e-05,
+      "loss": 0.3093,
+      "step": 5767
+    },
+    {
+      "epoch": 1.1370268138801263,
+      "grad_norm": 0.5241913684270388,
+      "learning_rate": 1.6286696886053146e-05,
+      "loss": 0.3442,
+      "step": 5768
+    },
+    {
+      "epoch": 1.137223974763407,
+      "grad_norm": 0.5303723352436314,
+      "learning_rate": 1.6285491398052892e-05,
+      "loss": 0.3473,
+      "step": 5769
+    },
+    {
+      "epoch": 1.1374211356466877,
+      "grad_norm": 0.5013219330856153,
+      "learning_rate": 1.6284285759038647e-05,
+      "loss": 0.3334,
+      "step": 5770
+    },
+    {
+      "epoch": 1.1376182965299684,
+      "grad_norm": 0.5116295193399402,
+      "learning_rate": 1.6283079969039376e-05,
+      "loss": 0.3348,
+      "step": 5771
+    },
+    {
+      "epoch": 1.1378154574132493,
+      "grad_norm": 0.5072830412189899,
+      "learning_rate": 1.6281874028084048e-05,
+      "loss": 0.3304,
+      "step": 5772
+    },
+    {
+      "epoch": 1.13801261829653,
+      "grad_norm": 0.5969770845470891,
+      "learning_rate": 1.6280667936201638e-05,
+      "loss": 0.3693,
+      "step": 5773
+    },
+    {
+      "epoch": 1.1382097791798107,
+      "grad_norm": 0.49707202160221337,
+      "learning_rate": 1.6279461693421122e-05,
+      "loss": 0.3092,
+      "step": 5774
+    },
+    {
+      "epoch": 1.1384069400630914,
+      "grad_norm": 0.48306958026968283,
+      "learning_rate": 1.6278255299771485e-05,
+      "loss": 0.3133,
+      "step": 5775
+    },
+    {
+      "epoch": 1.1386041009463723,
+      "grad_norm": 0.49732568766268076,
+      "learning_rate": 1.6277048755281706e-05,
+      "loss": 0.3133,
+      "step": 5776
+    },
+    {
+      "epoch": 1.138801261829653,
+      "grad_norm": 0.49440667619738315,
+      "learning_rate": 1.6275842059980777e-05,
+      "loss": 0.3081,
+      "step": 5777
+    },
+    {
+      "epoch": 1.1389984227129337,
+      "grad_norm": 0.49394133306858135,
+      "learning_rate": 1.627463521389769e-05,
+      "loss": 0.3075,
+      "step": 5778
+    },
+    {
+      "epoch": 1.1391955835962144,
+      "grad_norm": 0.5006133223548448,
+      "learning_rate": 1.6273428217061438e-05,
+      "loss": 0.3382,
+      "step": 5779
+    },
+    {
+      "epoch": 1.1393927444794953,
+      "grad_norm": 0.5094072983950105,
+      "learning_rate": 1.627222106950102e-05,
+      "loss": 0.3195,
+      "step": 5780
+    },
+    {
+      "epoch": 1.139589905362776,
+      "grad_norm": 0.4911846217605256,
+      "learning_rate": 1.627101377124544e-05,
+      "loss": 0.3157,
+      "step": 5781
+    },
+    {
+      "epoch": 1.1397870662460567,
+      "grad_norm": 0.5197657629535235,
+      "learning_rate": 1.626980632232371e-05,
+      "loss": 0.3455,
+      "step": 5782
+    },
+    {
+      "epoch": 1.1399842271293374,
+      "grad_norm": 0.5165791218182875,
+      "learning_rate": 1.6268598722764825e-05,
+      "loss": 0.3571,
+      "step": 5783
+    },
+    {
+      "epoch": 1.1401813880126184,
+      "grad_norm": 0.4911136281917946,
+      "learning_rate": 1.6267390972597808e-05,
+      "loss": 0.3258,
+      "step": 5784
+    },
+    {
+      "epoch": 1.140378548895899,
+      "grad_norm": 0.4791334939004954,
+      "learning_rate": 1.626618307185168e-05,
+      "loss": 0.313,
+      "step": 5785
+    },
+    {
+      "epoch": 1.1405757097791798,
+      "grad_norm": 0.507257163220278,
+      "learning_rate": 1.626497502055546e-05,
+      "loss": 0.3348,
+      "step": 5786
+    },
+    {
+      "epoch": 1.1407728706624605,
+      "grad_norm": 0.5095261380886309,
+      "learning_rate": 1.6263766818738163e-05,
+      "loss": 0.3352,
+      "step": 5787
+    },
+    {
+      "epoch": 1.1409700315457414,
+      "grad_norm": 0.489168264070942,
+      "learning_rate": 1.6262558466428827e-05,
+      "loss": 0.3181,
+      "step": 5788
+    },
+    {
+      "epoch": 1.141167192429022,
+      "grad_norm": 0.5758632623492379,
+      "learning_rate": 1.6261349963656482e-05,
+      "loss": 0.3221,
+      "step": 5789
+    },
+    {
+      "epoch": 1.1413643533123028,
+      "grad_norm": 0.5155189277830298,
+      "learning_rate": 1.6260141310450158e-05,
+      "loss": 0.3378,
+      "step": 5790
+    },
+    {
+      "epoch": 1.1415615141955837,
+      "grad_norm": 0.5361821367760298,
+      "learning_rate": 1.6258932506838903e-05,
+      "loss": 0.3548,
+      "step": 5791
+    },
+    {
+      "epoch": 1.1417586750788644,
+      "grad_norm": 0.510073475492786,
+      "learning_rate": 1.6257723552851752e-05,
+      "loss": 0.3259,
+      "step": 5792
+    },
+    {
+      "epoch": 1.1419558359621451,
+      "grad_norm": 0.5181336652283103,
+      "learning_rate": 1.6256514448517753e-05,
+      "loss": 0.3181,
+      "step": 5793
+    },
+    {
+      "epoch": 1.1421529968454258,
+      "grad_norm": 0.5218135564565195,
+      "learning_rate": 1.6255305193865957e-05,
+      "loss": 0.3269,
+      "step": 5794
+    },
+    {
+      "epoch": 1.1423501577287065,
+      "grad_norm": 0.6090159220142208,
+      "learning_rate": 1.6254095788925413e-05,
+      "loss": 0.3487,
+      "step": 5795
+    },
+    {
+      "epoch": 1.1425473186119874,
+      "grad_norm": 0.4918572565910208,
+      "learning_rate": 1.6252886233725186e-05,
+      "loss": 0.3222,
+      "step": 5796
+    },
+    {
+      "epoch": 1.1427444794952681,
+      "grad_norm": 0.5170092942493402,
+      "learning_rate": 1.625167652829433e-05,
+      "loss": 0.3075,
+      "step": 5797
+    },
+    {
+      "epoch": 1.1429416403785488,
+      "grad_norm": 0.5153552400826925,
+      "learning_rate": 1.625046667266191e-05,
+      "loss": 0.3552,
+      "step": 5798
+    },
+    {
+      "epoch": 1.1431388012618298,
+      "grad_norm": 0.5092426572265226,
+      "learning_rate": 1.6249256666856995e-05,
+      "loss": 0.2974,
+      "step": 5799
+    },
+    {
+      "epoch": 1.1433359621451105,
+      "grad_norm": 0.5381814851040011,
+      "learning_rate": 1.6248046510908654e-05,
+      "loss": 0.3186,
+      "step": 5800
+    },
+    {
+      "epoch": 1.1435331230283912,
+      "grad_norm": 0.5393871623205769,
+      "learning_rate": 1.6246836204845967e-05,
+      "loss": 0.3503,
+      "step": 5801
+    },
+    {
+      "epoch": 1.1437302839116719,
+      "grad_norm": 0.5047753301047979,
+      "learning_rate": 1.624562574869801e-05,
+      "loss": 0.3232,
+      "step": 5802
+    },
+    {
+      "epoch": 1.1439274447949526,
+      "grad_norm": 0.5134380411824181,
+      "learning_rate": 1.6244415142493867e-05,
+      "loss": 0.3374,
+      "step": 5803
+    },
+    {
+      "epoch": 1.1441246056782335,
+      "grad_norm": 0.5382530413103754,
+      "learning_rate": 1.6243204386262618e-05,
+      "loss": 0.3472,
+      "step": 5804
+    },
+    {
+      "epoch": 1.1443217665615142,
+      "grad_norm": 0.4860783558028898,
+      "learning_rate": 1.6241993480033353e-05,
+      "loss": 0.2894,
+      "step": 5805
+    },
+    {
+      "epoch": 1.1445189274447949,
+      "grad_norm": 0.5360606741317754,
+      "learning_rate": 1.6240782423835174e-05,
+      "loss": 0.3526,
+      "step": 5806
+    },
+    {
+      "epoch": 1.1447160883280758,
+      "grad_norm": 0.4950368935661035,
+      "learning_rate": 1.6239571217697164e-05,
+      "loss": 0.3282,
+      "step": 5807
+    },
+    {
+      "epoch": 1.1449132492113565,
+      "grad_norm": 0.5007041011557887,
+      "learning_rate": 1.6238359861648438e-05,
+      "loss": 0.3199,
+      "step": 5808
+    },
+    {
+      "epoch": 1.1451104100946372,
+      "grad_norm": 0.4866543775966263,
+      "learning_rate": 1.6237148355718092e-05,
+      "loss": 0.3306,
+      "step": 5809
+    },
+    {
+      "epoch": 1.145307570977918,
+      "grad_norm": 0.4824258516786147,
+      "learning_rate": 1.623593669993523e-05,
+      "loss": 0.3098,
+      "step": 5810
+    },
+    {
+      "epoch": 1.1455047318611988,
+      "grad_norm": 0.4805879120897678,
+      "learning_rate": 1.623472489432897e-05,
+      "loss": 0.3024,
+      "step": 5811
+    },
+    {
+      "epoch": 1.1457018927444795,
+      "grad_norm": 0.5450607872650646,
+      "learning_rate": 1.623351293892842e-05,
+      "loss": 0.3426,
+      "step": 5812
+    },
+    {
+      "epoch": 1.1458990536277602,
+      "grad_norm": 0.49414340389336664,
+      "learning_rate": 1.623230083376271e-05,
+      "loss": 0.3353,
+      "step": 5813
+    },
+    {
+      "epoch": 1.146096214511041,
+      "grad_norm": 0.5993703898368199,
+      "learning_rate": 1.6231088578860946e-05,
+      "loss": 0.3437,
+      "step": 5814
+    },
+    {
+      "epoch": 1.1462933753943219,
+      "grad_norm": 0.5174297786339515,
+      "learning_rate": 1.6229876174252265e-05,
+      "loss": 0.3284,
+      "step": 5815
+    },
+    {
+      "epoch": 1.1464905362776026,
+      "grad_norm": 0.49563315628009424,
+      "learning_rate": 1.6228663619965787e-05,
+      "loss": 0.316,
+      "step": 5816
+    },
+    {
+      "epoch": 1.1466876971608833,
+      "grad_norm": 0.5185380709534432,
+      "learning_rate": 1.6227450916030655e-05,
+      "loss": 0.3324,
+      "step": 5817
+    },
+    {
+      "epoch": 1.146884858044164,
+      "grad_norm": 0.517728568283838,
+      "learning_rate": 1.6226238062476e-05,
+      "loss": 0.3301,
+      "step": 5818
+    },
+    {
+      "epoch": 1.1470820189274449,
+      "grad_norm": 0.5337322717327413,
+      "learning_rate": 1.6225025059330954e-05,
+      "loss": 0.3549,
+      "step": 5819
+    },
+    {
+      "epoch": 1.1472791798107256,
+      "grad_norm": 0.51659644720446,
+      "learning_rate": 1.6223811906624675e-05,
+      "loss": 0.3304,
+      "step": 5820
+    },
+    {
+      "epoch": 1.1474763406940063,
+      "grad_norm": 0.5135534115538601,
+      "learning_rate": 1.6222598604386303e-05,
+      "loss": 0.3253,
+      "step": 5821
+    },
+    {
+      "epoch": 1.147673501577287,
+      "grad_norm": 0.5683240033324098,
+      "learning_rate": 1.6221385152644986e-05,
+      "loss": 0.3719,
+      "step": 5822
+    },
+    {
+      "epoch": 1.147870662460568,
+      "grad_norm": 0.5006723049086741,
+      "learning_rate": 1.622017155142988e-05,
+      "loss": 0.3163,
+      "step": 5823
+    },
+    {
+      "epoch": 1.1480678233438486,
+      "grad_norm": 0.5148405356777013,
+      "learning_rate": 1.6218957800770146e-05,
+      "loss": 0.3548,
+      "step": 5824
+    },
+    {
+      "epoch": 1.1482649842271293,
+      "grad_norm": 0.5750083625058464,
+      "learning_rate": 1.621774390069494e-05,
+      "loss": 0.3561,
+      "step": 5825
+    },
+    {
+      "epoch": 1.14846214511041,
+      "grad_norm": 0.5165474137396464,
+      "learning_rate": 1.621652985123343e-05,
+      "loss": 0.3339,
+      "step": 5826
+    },
+    {
+      "epoch": 1.148659305993691,
+      "grad_norm": 0.5084589221237357,
+      "learning_rate": 1.6215315652414786e-05,
+      "loss": 0.3318,
+      "step": 5827
+    },
+    {
+      "epoch": 1.1488564668769716,
+      "grad_norm": 0.5035248616427002,
+      "learning_rate": 1.6214101304268177e-05,
+      "loss": 0.338,
+      "step": 5828
+    },
+    {
+      "epoch": 1.1490536277602523,
+      "grad_norm": 0.5544001188077334,
+      "learning_rate": 1.621288680682278e-05,
+      "loss": 0.3366,
+      "step": 5829
+    },
+    {
+      "epoch": 1.149250788643533,
+      "grad_norm": 0.5160867252412951,
+      "learning_rate": 1.6211672160107776e-05,
+      "loss": 0.3518,
+      "step": 5830
+    },
+    {
+      "epoch": 1.149447949526814,
+      "grad_norm": 0.5010643439725077,
+      "learning_rate": 1.6210457364152345e-05,
+      "loss": 0.3436,
+      "step": 5831
+    },
+    {
+      "epoch": 1.1496451104100947,
+      "grad_norm": 0.5106652676242482,
+      "learning_rate": 1.6209242418985673e-05,
+      "loss": 0.3226,
+      "step": 5832
+    },
+    {
+      "epoch": 1.1498422712933754,
+      "grad_norm": 0.48330155765068944,
+      "learning_rate": 1.6208027324636956e-05,
+      "loss": 0.3091,
+      "step": 5833
+    },
+    {
+      "epoch": 1.1500394321766563,
+      "grad_norm": 0.5026909229195045,
+      "learning_rate": 1.620681208113538e-05,
+      "loss": 0.313,
+      "step": 5834
+    },
+    {
+      "epoch": 1.150236593059937,
+      "grad_norm": 0.4919091752987136,
+      "learning_rate": 1.6205596688510144e-05,
+      "loss": 0.3118,
+      "step": 5835
+    },
+    {
+      "epoch": 1.1504337539432177,
+      "grad_norm": 0.4987724584414972,
+      "learning_rate": 1.6204381146790452e-05,
+      "loss": 0.3159,
+      "step": 5836
+    },
+    {
+      "epoch": 1.1506309148264984,
+      "grad_norm": 0.47351362341946696,
+      "learning_rate": 1.6203165456005505e-05,
+      "loss": 0.3215,
+      "step": 5837
+    },
+    {
+      "epoch": 1.150828075709779,
+      "grad_norm": 0.5390815160396467,
+      "learning_rate": 1.6201949616184515e-05,
+      "loss": 0.3512,
+      "step": 5838
+    },
+    {
+      "epoch": 1.15102523659306,
+      "grad_norm": 0.4907164796379062,
+      "learning_rate": 1.620073362735669e-05,
+      "loss": 0.3351,
+      "step": 5839
+    },
+    {
+      "epoch": 1.1512223974763407,
+      "grad_norm": 0.5183309334339059,
+      "learning_rate": 1.6199517489551246e-05,
+      "loss": 0.3471,
+      "step": 5840
+    },
+    {
+      "epoch": 1.1514195583596214,
+      "grad_norm": 0.5094113386762072,
+      "learning_rate": 1.61983012027974e-05,
+      "loss": 0.3393,
+      "step": 5841
+    },
+    {
+      "epoch": 1.1516167192429023,
+      "grad_norm": 0.47916372568319,
+      "learning_rate": 1.6197084767124378e-05,
+      "loss": 0.3002,
+      "step": 5842
+    },
+    {
+      "epoch": 1.151813880126183,
+      "grad_norm": 0.5554961928781669,
+      "learning_rate": 1.619586818256141e-05,
+      "loss": 0.3465,
+      "step": 5843
+    },
+    {
+      "epoch": 1.1520110410094637,
+      "grad_norm": 0.4760523166959915,
+      "learning_rate": 1.6194651449137708e-05,
+      "loss": 0.315,
+      "step": 5844
+    },
+    {
+      "epoch": 1.1522082018927444,
+      "grad_norm": 0.49772863471397966,
+      "learning_rate": 1.6193434566882522e-05,
+      "loss": 0.3392,
+      "step": 5845
+    },
+    {
+      "epoch": 1.1524053627760251,
+      "grad_norm": 0.49931835312839523,
+      "learning_rate": 1.6192217535825084e-05,
+      "loss": 0.3259,
+      "step": 5846
+    },
+    {
+      "epoch": 1.152602523659306,
+      "grad_norm": 0.4935523430213534,
+      "learning_rate": 1.619100035599463e-05,
+      "loss": 0.3188,
+      "step": 5847
+    },
+    {
+      "epoch": 1.1527996845425867,
+      "grad_norm": 0.5009652113164703,
+      "learning_rate": 1.618978302742041e-05,
+      "loss": 0.3017,
+      "step": 5848
+    },
+    {
+      "epoch": 1.1529968454258674,
+      "grad_norm": 0.513184184558755,
+      "learning_rate": 1.6188565550131667e-05,
+      "loss": 0.3389,
+      "step": 5849
+    },
+    {
+      "epoch": 1.1531940063091484,
+      "grad_norm": 0.5120199704705086,
+      "learning_rate": 1.6187347924157654e-05,
+      "loss": 0.3272,
+      "step": 5850
+    },
+    {
+      "epoch": 1.153391167192429,
+      "grad_norm": 0.55709606253943,
+      "learning_rate": 1.618613014952762e-05,
+      "loss": 0.3511,
+      "step": 5851
+    },
+    {
+      "epoch": 1.1535883280757098,
+      "grad_norm": 0.528693155299282,
+      "learning_rate": 1.6184912226270833e-05,
+      "loss": 0.3246,
+      "step": 5852
+    },
+    {
+      "epoch": 1.1537854889589905,
+      "grad_norm": 0.4926792899919863,
+      "learning_rate": 1.6183694154416548e-05,
+      "loss": 0.3401,
+      "step": 5853
+    },
+    {
+      "epoch": 1.1539826498422712,
+      "grad_norm": 0.5247479545911596,
+      "learning_rate": 1.618247593399403e-05,
+      "loss": 0.3147,
+      "step": 5854
+    },
+    {
+      "epoch": 1.154179810725552,
+      "grad_norm": 0.5391139938461473,
+      "learning_rate": 1.6181257565032548e-05,
+      "loss": 0.3424,
+      "step": 5855
+    },
+    {
+      "epoch": 1.1543769716088328,
+      "grad_norm": 0.5379238676508531,
+      "learning_rate": 1.6180039047561375e-05,
+      "loss": 0.3489,
+      "step": 5856
+    },
+    {
+      "epoch": 1.1545741324921135,
+      "grad_norm": 0.5433587033942305,
+      "learning_rate": 1.6178820381609793e-05,
+      "loss": 0.3583,
+      "step": 5857
+    },
+    {
+      "epoch": 1.1547712933753944,
+      "grad_norm": 0.5515081493988264,
+      "learning_rate": 1.617760156720707e-05,
+      "loss": 0.3409,
+      "step": 5858
+    },
+    {
+      "epoch": 1.1549684542586751,
+      "grad_norm": 0.5211754285676179,
+      "learning_rate": 1.617638260438249e-05,
+      "loss": 0.3441,
+      "step": 5859
+    },
+    {
+      "epoch": 1.1551656151419558,
+      "grad_norm": 0.5728695227853102,
+      "learning_rate": 1.6175163493165353e-05,
+      "loss": 0.3692,
+      "step": 5860
+    },
+    {
+      "epoch": 1.1553627760252365,
+      "grad_norm": 0.5398536734238296,
+      "learning_rate": 1.6173944233584936e-05,
+      "loss": 0.3333,
+      "step": 5861
+    },
+    {
+      "epoch": 1.1555599369085174,
+      "grad_norm": 0.5261150430536126,
+      "learning_rate": 1.6172724825670537e-05,
+      "loss": 0.3279,
+      "step": 5862
+    },
+    {
+      "epoch": 1.1557570977917981,
+      "grad_norm": 0.5106454983015667,
+      "learning_rate": 1.6171505269451456e-05,
+      "loss": 0.339,
+      "step": 5863
+    },
+    {
+      "epoch": 1.1559542586750788,
+      "grad_norm": 0.5055443785259361,
+      "learning_rate": 1.617028556495699e-05,
+      "loss": 0.2946,
+      "step": 5864
+    },
+    {
+      "epoch": 1.1561514195583595,
+      "grad_norm": 0.5803137026274476,
+      "learning_rate": 1.6169065712216444e-05,
+      "loss": 0.3789,
+      "step": 5865
+    },
+    {
+      "epoch": 1.1563485804416405,
+      "grad_norm": 0.512878015699482,
+      "learning_rate": 1.6167845711259123e-05,
+      "loss": 0.3469,
+      "step": 5866
+    },
+    {
+      "epoch": 1.1565457413249212,
+      "grad_norm": 0.5349976563998219,
+      "learning_rate": 1.6166625562114347e-05,
+      "loss": 0.3428,
+      "step": 5867
+    },
+    {
+      "epoch": 1.1567429022082019,
+      "grad_norm": 0.46278679315128507,
+      "learning_rate": 1.616540526481142e-05,
+      "loss": 0.2879,
+      "step": 5868
+    },
+    {
+      "epoch": 1.1569400630914826,
+      "grad_norm": 0.5228615238771711,
+      "learning_rate": 1.6164184819379673e-05,
+      "loss": 0.33,
+      "step": 5869
+    },
+    {
+      "epoch": 1.1571372239747635,
+      "grad_norm": 0.5253461017754688,
+      "learning_rate": 1.6162964225848416e-05,
+      "loss": 0.3451,
+      "step": 5870
+    },
+    {
+      "epoch": 1.1573343848580442,
+      "grad_norm": 0.5368330566994861,
+      "learning_rate": 1.6161743484246987e-05,
+      "loss": 0.32,
+      "step": 5871
+    },
+    {
+      "epoch": 1.1575315457413249,
+      "grad_norm": 0.5655488658811362,
+      "learning_rate": 1.6160522594604704e-05,
+      "loss": 0.3267,
+      "step": 5872
+    },
+    {
+      "epoch": 1.1577287066246056,
+      "grad_norm": 0.553771815896319,
+      "learning_rate": 1.6159301556950904e-05,
+      "loss": 0.3338,
+      "step": 5873
+    },
+    {
+      "epoch": 1.1579258675078865,
+      "grad_norm": 0.49679032602454654,
+      "learning_rate": 1.6158080371314926e-05,
+      "loss": 0.3207,
+      "step": 5874
+    },
+    {
+      "epoch": 1.1581230283911672,
+      "grad_norm": 0.5110047440222282,
+      "learning_rate": 1.6156859037726108e-05,
+      "loss": 0.3273,
+      "step": 5875
+    },
+    {
+      "epoch": 1.158320189274448,
+      "grad_norm": 0.5164914172279298,
+      "learning_rate": 1.6155637556213793e-05,
+      "loss": 0.3248,
+      "step": 5876
+    },
+    {
+      "epoch": 1.1585173501577288,
+      "grad_norm": 0.5120868579780778,
+      "learning_rate": 1.6154415926807327e-05,
+      "loss": 0.3603,
+      "step": 5877
+    },
+    {
+      "epoch": 1.1587145110410095,
+      "grad_norm": 0.4585506611956392,
+      "learning_rate": 1.6153194149536064e-05,
+      "loss": 0.2974,
+      "step": 5878
+    },
+    {
+      "epoch": 1.1589116719242902,
+      "grad_norm": 0.5616942590634053,
+      "learning_rate": 1.6151972224429356e-05,
+      "loss": 0.3255,
+      "step": 5879
+    },
+    {
+      "epoch": 1.159108832807571,
+      "grad_norm": 0.5276396699788394,
+      "learning_rate": 1.615075015151656e-05,
+      "loss": 0.3336,
+      "step": 5880
+    },
+    {
+      "epoch": 1.1593059936908516,
+      "grad_norm": 0.48785185958924593,
+      "learning_rate": 1.6149527930827043e-05,
+      "loss": 0.3183,
+      "step": 5881
+    },
+    {
+      "epoch": 1.1595031545741326,
+      "grad_norm": 0.5070222632900729,
+      "learning_rate": 1.614830556239016e-05,
+      "loss": 0.3392,
+      "step": 5882
+    },
+    {
+      "epoch": 1.1597003154574133,
+      "grad_norm": 0.5357213010735683,
+      "learning_rate": 1.6147083046235287e-05,
+      "loss": 0.3302,
+      "step": 5883
+    },
+    {
+      "epoch": 1.159897476340694,
+      "grad_norm": 0.492115580680464,
+      "learning_rate": 1.6145860382391792e-05,
+      "loss": 0.3062,
+      "step": 5884
+    },
+    {
+      "epoch": 1.1600946372239749,
+      "grad_norm": 0.4775935270120072,
+      "learning_rate": 1.6144637570889055e-05,
+      "loss": 0.3084,
+      "step": 5885
+    },
+    {
+      "epoch": 1.1602917981072556,
+      "grad_norm": 0.5574502509553525,
+      "learning_rate": 1.6143414611756448e-05,
+      "loss": 0.3762,
+      "step": 5886
+    },
+    {
+      "epoch": 1.1604889589905363,
+      "grad_norm": 0.5324231837542485,
+      "learning_rate": 1.6142191505023362e-05,
+      "loss": 0.3394,
+      "step": 5887
+    },
+    {
+      "epoch": 1.160686119873817,
+      "grad_norm": 0.5540695998548545,
+      "learning_rate": 1.6140968250719177e-05,
+      "loss": 0.3334,
+      "step": 5888
+    },
+    {
+      "epoch": 1.1608832807570977,
+      "grad_norm": 0.5152035583057871,
+      "learning_rate": 1.6139744848873283e-05,
+      "loss": 0.333,
+      "step": 5889
+    },
+    {
+      "epoch": 1.1610804416403786,
+      "grad_norm": 0.5016308910529644,
+      "learning_rate": 1.613852129951508e-05,
+      "loss": 0.328,
+      "step": 5890
+    },
+    {
+      "epoch": 1.1612776025236593,
+      "grad_norm": 0.4671308137477831,
+      "learning_rate": 1.6137297602673955e-05,
+      "loss": 0.3006,
+      "step": 5891
+    },
+    {
+      "epoch": 1.16147476340694,
+      "grad_norm": 0.47717698267201425,
+      "learning_rate": 1.613607375837931e-05,
+      "loss": 0.3085,
+      "step": 5892
+    },
+    {
+      "epoch": 1.161671924290221,
+      "grad_norm": 0.5411199798727852,
+      "learning_rate": 1.6134849766660557e-05,
+      "loss": 0.3512,
+      "step": 5893
+    },
+    {
+      "epoch": 1.1618690851735016,
+      "grad_norm": 0.48072485668659576,
+      "learning_rate": 1.6133625627547096e-05,
+      "loss": 0.2991,
+      "step": 5894
+    },
+    {
+      "epoch": 1.1620662460567823,
+      "grad_norm": 0.49074020021120335,
+      "learning_rate": 1.613240134106834e-05,
+      "loss": 0.3185,
+      "step": 5895
+    },
+    {
+      "epoch": 1.162263406940063,
+      "grad_norm": 0.4742652414752287,
+      "learning_rate": 1.6131176907253703e-05,
+      "loss": 0.3219,
+      "step": 5896
+    },
+    {
+      "epoch": 1.1624605678233437,
+      "grad_norm": 0.4926211188349463,
+      "learning_rate": 1.6129952326132603e-05,
+      "loss": 0.3213,
+      "step": 5897
+    },
+    {
+      "epoch": 1.1626577287066246,
+      "grad_norm": 0.5260995869588211,
+      "learning_rate": 1.6128727597734465e-05,
+      "loss": 0.3266,
+      "step": 5898
+    },
+    {
+      "epoch": 1.1628548895899053,
+      "grad_norm": 0.5005689603012923,
+      "learning_rate": 1.6127502722088703e-05,
+      "loss": 0.3281,
+      "step": 5899
+    },
+    {
+      "epoch": 1.163052050473186,
+      "grad_norm": 0.5286226637548996,
+      "learning_rate": 1.612627769922476e-05,
+      "loss": 0.3383,
+      "step": 5900
+    },
+    {
+      "epoch": 1.163249211356467,
+      "grad_norm": 0.5241554407104515,
+      "learning_rate": 1.612505252917206e-05,
+      "loss": 0.3442,
+      "step": 5901
+    },
+    {
+      "epoch": 1.1634463722397477,
+      "grad_norm": 0.4969893983731258,
+      "learning_rate": 1.6123827211960044e-05,
+      "loss": 0.3355,
+      "step": 5902
+    },
+    {
+      "epoch": 1.1636435331230284,
+      "grad_norm": 0.5176767431545333,
+      "learning_rate": 1.6122601747618144e-05,
+      "loss": 0.3456,
+      "step": 5903
+    },
+    {
+      "epoch": 1.163840694006309,
+      "grad_norm": 0.5541524213460989,
+      "learning_rate": 1.612137613617581e-05,
+      "loss": 0.3082,
+      "step": 5904
+    },
+    {
+      "epoch": 1.16403785488959,
+      "grad_norm": 0.5130645123161176,
+      "learning_rate": 1.612015037766248e-05,
+      "loss": 0.3595,
+      "step": 5905
+    },
+    {
+      "epoch": 1.1642350157728707,
+      "grad_norm": 0.49077914052450783,
+      "learning_rate": 1.611892447210761e-05,
+      "loss": 0.3201,
+      "step": 5906
+    },
+    {
+      "epoch": 1.1644321766561514,
+      "grad_norm": 0.5154065030810983,
+      "learning_rate": 1.6117698419540655e-05,
+      "loss": 0.345,
+      "step": 5907
+    },
+    {
+      "epoch": 1.164629337539432,
+      "grad_norm": 0.5137017205525405,
+      "learning_rate": 1.6116472219991066e-05,
+      "loss": 0.321,
+      "step": 5908
+    },
+    {
+      "epoch": 1.164826498422713,
+      "grad_norm": 0.5185538568944038,
+      "learning_rate": 1.6115245873488308e-05,
+      "loss": 0.3349,
+      "step": 5909
+    },
+    {
+      "epoch": 1.1650236593059937,
+      "grad_norm": 0.47872637622833353,
+      "learning_rate": 1.6114019380061844e-05,
+      "loss": 0.3091,
+      "step": 5910
+    },
+    {
+      "epoch": 1.1652208201892744,
+      "grad_norm": 0.5349706635614739,
+      "learning_rate": 1.6112792739741138e-05,
+      "loss": 0.3416,
+      "step": 5911
+    },
+    {
+      "epoch": 1.1654179810725551,
+      "grad_norm": 0.494058437524682,
+      "learning_rate": 1.6111565952555666e-05,
+      "loss": 0.3411,
+      "step": 5912
+    },
+    {
+      "epoch": 1.165615141955836,
+      "grad_norm": 0.5148890708118689,
+      "learning_rate": 1.6110339018534898e-05,
+      "loss": 0.3379,
+      "step": 5913
+    },
+    {
+      "epoch": 1.1658123028391167,
+      "grad_norm": 0.48546285073114465,
+      "learning_rate": 1.6109111937708317e-05,
+      "loss": 0.3157,
+      "step": 5914
+    },
+    {
+      "epoch": 1.1660094637223974,
+      "grad_norm": 0.5386295226523584,
+      "learning_rate": 1.61078847101054e-05,
+      "loss": 0.3524,
+      "step": 5915
+    },
+    {
+      "epoch": 1.1662066246056781,
+      "grad_norm": 0.5357486231753327,
+      "learning_rate": 1.6106657335755636e-05,
+      "loss": 0.3296,
+      "step": 5916
+    },
+    {
+      "epoch": 1.166403785488959,
+      "grad_norm": 0.5071347852474872,
+      "learning_rate": 1.610542981468851e-05,
+      "loss": 0.2972,
+      "step": 5917
+    },
+    {
+      "epoch": 1.1666009463722398,
+      "grad_norm": 0.5136673972393638,
+      "learning_rate": 1.6104202146933517e-05,
+      "loss": 0.3427,
+      "step": 5918
+    },
+    {
+      "epoch": 1.1667981072555205,
+      "grad_norm": 0.5366052967903738,
+      "learning_rate": 1.6102974332520155e-05,
+      "loss": 0.3348,
+      "step": 5919
+    },
+    {
+      "epoch": 1.1669952681388012,
+      "grad_norm": 0.5142076244004472,
+      "learning_rate": 1.6101746371477915e-05,
+      "loss": 0.3409,
+      "step": 5920
+    },
+    {
+      "epoch": 1.167192429022082,
+      "grad_norm": 0.5156383871361846,
+      "learning_rate": 1.6100518263836305e-05,
+      "loss": 0.3441,
+      "step": 5921
+    },
+    {
+      "epoch": 1.1673895899053628,
+      "grad_norm": 0.48857227681411236,
+      "learning_rate": 1.609929000962483e-05,
+      "loss": 0.3249,
+      "step": 5922
+    },
+    {
+      "epoch": 1.1675867507886435,
+      "grad_norm": 0.5116415565941662,
+      "learning_rate": 1.6098061608873006e-05,
+      "loss": 0.3347,
+      "step": 5923
+    },
+    {
+      "epoch": 1.1677839116719242,
+      "grad_norm": 0.476427280273157,
+      "learning_rate": 1.609683306161034e-05,
+      "loss": 0.2935,
+      "step": 5924
+    },
+    {
+      "epoch": 1.1679810725552051,
+      "grad_norm": 0.5315870178041014,
+      "learning_rate": 1.6095604367866348e-05,
+      "loss": 0.3298,
+      "step": 5925
+    },
+    {
+      "epoch": 1.1681782334384858,
+      "grad_norm": 0.4885155093607642,
+      "learning_rate": 1.6094375527670553e-05,
+      "loss": 0.2914,
+      "step": 5926
+    },
+    {
+      "epoch": 1.1683753943217665,
+      "grad_norm": 0.4958137107083927,
+      "learning_rate": 1.6093146541052472e-05,
+      "loss": 0.3236,
+      "step": 5927
+    },
+    {
+      "epoch": 1.1685725552050474,
+      "grad_norm": 0.5186709145538795,
+      "learning_rate": 1.609191740804165e-05,
+      "loss": 0.3474,
+      "step": 5928
+    },
+    {
+      "epoch": 1.1687697160883281,
+      "grad_norm": 0.5431827154799177,
+      "learning_rate": 1.6090688128667597e-05,
+      "loss": 0.2922,
+      "step": 5929
+    },
+    {
+      "epoch": 1.1689668769716088,
+      "grad_norm": 0.5276177466035217,
+      "learning_rate": 1.608945870295986e-05,
+      "loss": 0.3273,
+      "step": 5930
+    },
+    {
+      "epoch": 1.1691640378548895,
+      "grad_norm": 0.5403429181736961,
+      "learning_rate": 1.6088229130947976e-05,
+      "loss": 0.3573,
+      "step": 5931
+    },
+    {
+      "epoch": 1.1693611987381702,
+      "grad_norm": 0.5682917162052413,
+      "learning_rate": 1.6086999412661483e-05,
+      "loss": 0.3577,
+      "step": 5932
+    },
+    {
+      "epoch": 1.1695583596214512,
+      "grad_norm": 0.5092581015992266,
+      "learning_rate": 1.6085769548129928e-05,
+      "loss": 0.3178,
+      "step": 5933
+    },
+    {
+      "epoch": 1.1697555205047319,
+      "grad_norm": 0.5341988017389373,
+      "learning_rate": 1.6084539537382853e-05,
+      "loss": 0.318,
+      "step": 5934
+    },
+    {
+      "epoch": 1.1699526813880126,
+      "grad_norm": 0.5508242925818101,
+      "learning_rate": 1.6083309380449822e-05,
+      "loss": 0.3399,
+      "step": 5935
+    },
+    {
+      "epoch": 1.1701498422712935,
+      "grad_norm": 0.5516468570148881,
+      "learning_rate": 1.6082079077360382e-05,
+      "loss": 0.3464,
+      "step": 5936
+    },
+    {
+      "epoch": 1.1703470031545742,
+      "grad_norm": 0.5310729616685209,
+      "learning_rate": 1.6080848628144097e-05,
+      "loss": 0.3549,
+      "step": 5937
+    },
+    {
+      "epoch": 1.1705441640378549,
+      "grad_norm": 0.5007231153163549,
+      "learning_rate": 1.6079618032830523e-05,
+      "loss": 0.317,
+      "step": 5938
+    },
+    {
+      "epoch": 1.1707413249211356,
+      "grad_norm": 0.5205324245736465,
+      "learning_rate": 1.6078387291449234e-05,
+      "loss": 0.3343,
+      "step": 5939
+    },
+    {
+      "epoch": 1.1709384858044163,
+      "grad_norm": 0.5293927543610932,
+      "learning_rate": 1.607715640402979e-05,
+      "loss": 0.3364,
+      "step": 5940
+    },
+    {
+      "epoch": 1.1711356466876972,
+      "grad_norm": 0.5437502522427697,
+      "learning_rate": 1.607592537060177e-05,
+      "loss": 0.3625,
+      "step": 5941
+    },
+    {
+      "epoch": 1.171332807570978,
+      "grad_norm": 0.5852873503945257,
+      "learning_rate": 1.6074694191194758e-05,
+      "loss": 0.343,
+      "step": 5942
+    },
+    {
+      "epoch": 1.1715299684542586,
+      "grad_norm": 0.5258888667294329,
+      "learning_rate": 1.607346286583832e-05,
+      "loss": 0.3525,
+      "step": 5943
+    },
+    {
+      "epoch": 1.1717271293375395,
+      "grad_norm": 0.49502127400893137,
+      "learning_rate": 1.6072231394562045e-05,
+      "loss": 0.3429,
+      "step": 5944
+    },
+    {
+      "epoch": 1.1719242902208202,
+      "grad_norm": 0.5340808573650885,
+      "learning_rate": 1.6070999777395522e-05,
+      "loss": 0.3265,
+      "step": 5945
+    },
+    {
+      "epoch": 1.172121451104101,
+      "grad_norm": 0.4913659047348541,
+      "learning_rate": 1.6069768014368344e-05,
+      "loss": 0.3264,
+      "step": 5946
+    },
+    {
+      "epoch": 1.1723186119873816,
+      "grad_norm": 0.5796915968301445,
+      "learning_rate": 1.6068536105510095e-05,
+      "loss": 0.3314,
+      "step": 5947
+    },
+    {
+      "epoch": 1.1725157728706626,
+      "grad_norm": 0.48015072710313483,
+      "learning_rate": 1.606730405085038e-05,
+      "loss": 0.3063,
+      "step": 5948
+    },
+    {
+      "epoch": 1.1727129337539433,
+      "grad_norm": 0.5798304016447527,
+      "learning_rate": 1.60660718504188e-05,
+      "loss": 0.3673,
+      "step": 5949
+    },
+    {
+      "epoch": 1.172910094637224,
+      "grad_norm": 0.49383450156062897,
+      "learning_rate": 1.6064839504244964e-05,
+      "loss": 0.3137,
+      "step": 5950
+    },
+    {
+      "epoch": 1.1731072555205047,
+      "grad_norm": 0.49990802164314824,
+      "learning_rate": 1.6063607012358474e-05,
+      "loss": 0.345,
+      "step": 5951
+    },
+    {
+      "epoch": 1.1733044164037856,
+      "grad_norm": 0.5710511478007725,
+      "learning_rate": 1.6062374374788938e-05,
+      "loss": 0.3489,
+      "step": 5952
+    },
+    {
+      "epoch": 1.1735015772870663,
+      "grad_norm": 0.5406553339299977,
+      "learning_rate": 1.6061141591565977e-05,
+      "loss": 0.3376,
+      "step": 5953
+    },
+    {
+      "epoch": 1.173698738170347,
+      "grad_norm": 0.5266705373762315,
+      "learning_rate": 1.605990866271921e-05,
+      "loss": 0.3169,
+      "step": 5954
+    },
+    {
+      "epoch": 1.1738958990536277,
+      "grad_norm": 0.5290688903751791,
+      "learning_rate": 1.605867558827825e-05,
+      "loss": 0.3314,
+      "step": 5955
+    },
+    {
+      "epoch": 1.1740930599369086,
+      "grad_norm": 0.4981019544107017,
+      "learning_rate": 1.605744236827274e-05,
+      "loss": 0.3499,
+      "step": 5956
+    },
+    {
+      "epoch": 1.1742902208201893,
+      "grad_norm": 0.4966110971975489,
+      "learning_rate": 1.6056209002732293e-05,
+      "loss": 0.3241,
+      "step": 5957
+    },
+    {
+      "epoch": 1.17448738170347,
+      "grad_norm": 0.5075484095401199,
+      "learning_rate": 1.605497549168655e-05,
+      "loss": 0.3164,
+      "step": 5958
+    },
+    {
+      "epoch": 1.1746845425867507,
+      "grad_norm": 0.7450970335942361,
+      "learning_rate": 1.6053741835165146e-05,
+      "loss": 0.3461,
+      "step": 5959
+    },
+    {
+      "epoch": 1.1748817034700316,
+      "grad_norm": 0.5291598106311712,
+      "learning_rate": 1.6052508033197713e-05,
+      "loss": 0.323,
+      "step": 5960
+    },
+    {
+      "epoch": 1.1750788643533123,
+      "grad_norm": 0.5970311083346598,
+      "learning_rate": 1.6051274085813906e-05,
+      "loss": 0.3433,
+      "step": 5961
+    },
+    {
+      "epoch": 1.175276025236593,
+      "grad_norm": 0.5257206942997955,
+      "learning_rate": 1.6050039993043366e-05,
+      "loss": 0.3312,
+      "step": 5962
+    },
+    {
+      "epoch": 1.1754731861198737,
+      "grad_norm": 0.5470667266403514,
+      "learning_rate": 1.604880575491574e-05,
+      "loss": 0.3241,
+      "step": 5963
+    },
+    {
+      "epoch": 1.1756703470031546,
+      "grad_norm": 0.526634277205355,
+      "learning_rate": 1.6047571371460688e-05,
+      "loss": 0.3486,
+      "step": 5964
+    },
+    {
+      "epoch": 1.1758675078864353,
+      "grad_norm": 0.5087896217443573,
+      "learning_rate": 1.6046336842707862e-05,
+      "loss": 0.3373,
+      "step": 5965
+    },
+    {
+      "epoch": 1.176064668769716,
+      "grad_norm": 0.5213596470506094,
+      "learning_rate": 1.6045102168686925e-05,
+      "loss": 0.3399,
+      "step": 5966
+    },
+    {
+      "epoch": 1.1762618296529967,
+      "grad_norm": 0.5025963895764581,
+      "learning_rate": 1.604386734942754e-05,
+      "loss": 0.3309,
+      "step": 5967
+    },
+    {
+      "epoch": 1.1764589905362777,
+      "grad_norm": 0.4947895294362806,
+      "learning_rate": 1.6042632384959377e-05,
+      "loss": 0.3266,
+      "step": 5968
+    },
+    {
+      "epoch": 1.1766561514195584,
+      "grad_norm": 0.5280726977614281,
+      "learning_rate": 1.6041397275312102e-05,
+      "loss": 0.3347,
+      "step": 5969
+    },
+    {
+      "epoch": 1.176853312302839,
+      "grad_norm": 0.49491475836802235,
+      "learning_rate": 1.6040162020515394e-05,
+      "loss": 0.3107,
+      "step": 5970
+    },
+    {
+      "epoch": 1.17705047318612,
+      "grad_norm": 0.50955563379636,
+      "learning_rate": 1.6038926620598924e-05,
+      "loss": 0.3438,
+      "step": 5971
+    },
+    {
+      "epoch": 1.1772476340694007,
+      "grad_norm": 0.47434081410915296,
+      "learning_rate": 1.6037691075592384e-05,
+      "loss": 0.2874,
+      "step": 5972
+    },
+    {
+      "epoch": 1.1774447949526814,
+      "grad_norm": 0.5265992460388663,
+      "learning_rate": 1.6036455385525452e-05,
+      "loss": 0.3138,
+      "step": 5973
+    },
+    {
+      "epoch": 1.177641955835962,
+      "grad_norm": 0.508538165601405,
+      "learning_rate": 1.603521955042782e-05,
+      "loss": 0.3364,
+      "step": 5974
+    },
+    {
+      "epoch": 1.1778391167192428,
+      "grad_norm": 0.536584712113376,
+      "learning_rate": 1.603398357032918e-05,
+      "loss": 0.3425,
+      "step": 5975
+    },
+    {
+      "epoch": 1.1780362776025237,
+      "grad_norm": 0.5637925361320564,
+      "learning_rate": 1.603274744525922e-05,
+      "loss": 0.3751,
+      "step": 5976
+    },
+    {
+      "epoch": 1.1782334384858044,
+      "grad_norm": 0.4828201596184613,
+      "learning_rate": 1.6031511175247648e-05,
+      "loss": 0.3267,
+      "step": 5977
+    },
+    {
+      "epoch": 1.1784305993690851,
+      "grad_norm": 0.5702750284269689,
+      "learning_rate": 1.6030274760324163e-05,
+      "loss": 0.3426,
+      "step": 5978
+    },
+    {
+      "epoch": 1.178627760252366,
+      "grad_norm": 0.5166744599492986,
+      "learning_rate": 1.602903820051847e-05,
+      "loss": 0.3443,
+      "step": 5979
+    },
+    {
+      "epoch": 1.1788249211356467,
+      "grad_norm": 0.4553996700869784,
+      "learning_rate": 1.602780149586028e-05,
+      "loss": 0.2921,
+      "step": 5980
+    },
+    {
+      "epoch": 1.1790220820189274,
+      "grad_norm": 0.5395408232911948,
+      "learning_rate": 1.60265646463793e-05,
+      "loss": 0.3487,
+      "step": 5981
+    },
+    {
+      "epoch": 1.1792192429022081,
+      "grad_norm": 0.5246852833105909,
+      "learning_rate": 1.6025327652105256e-05,
+      "loss": 0.353,
+      "step": 5982
+    },
+    {
+      "epoch": 1.1794164037854888,
+      "grad_norm": 0.5504345490202539,
+      "learning_rate": 1.6024090513067864e-05,
+      "loss": 0.3337,
+      "step": 5983
+    },
+    {
+      "epoch": 1.1796135646687698,
+      "grad_norm": 0.5273314886798296,
+      "learning_rate": 1.6022853229296844e-05,
+      "loss": 0.3613,
+      "step": 5984
+    },
+    {
+      "epoch": 1.1798107255520505,
+      "grad_norm": 0.4976467528520897,
+      "learning_rate": 1.6021615800821923e-05,
+      "loss": 0.3119,
+      "step": 5985
+    },
+    {
+      "epoch": 1.1800078864353312,
+      "grad_norm": 0.544586740515642,
+      "learning_rate": 1.6020378227672834e-05,
+      "loss": 0.3593,
+      "step": 5986
+    },
+    {
+      "epoch": 1.180205047318612,
+      "grad_norm": 0.5288888595664981,
+      "learning_rate": 1.6019140509879312e-05,
+      "loss": 0.365,
+      "step": 5987
+    },
+    {
+      "epoch": 1.1804022082018928,
+      "grad_norm": 0.5499369482970153,
+      "learning_rate": 1.601790264747109e-05,
+      "loss": 0.3673,
+      "step": 5988
+    },
+    {
+      "epoch": 1.1805993690851735,
+      "grad_norm": 0.5132871954718244,
+      "learning_rate": 1.6016664640477912e-05,
+      "loss": 0.349,
+      "step": 5989
+    },
+    {
+      "epoch": 1.1807965299684542,
+      "grad_norm": 0.5107401573623064,
+      "learning_rate": 1.601542648892952e-05,
+      "loss": 0.3328,
+      "step": 5990
+    },
+    {
+      "epoch": 1.1809936908517351,
+      "grad_norm": 0.5003005634953831,
+      "learning_rate": 1.6014188192855667e-05,
+      "loss": 0.3325,
+      "step": 5991
+    },
+    {
+      "epoch": 1.1811908517350158,
+      "grad_norm": 0.5419236213545058,
+      "learning_rate": 1.6012949752286093e-05,
+      "loss": 0.3567,
+      "step": 5992
+    },
+    {
+      "epoch": 1.1813880126182965,
+      "grad_norm": 0.46447030510821447,
+      "learning_rate": 1.6011711167250563e-05,
+      "loss": 0.3133,
+      "step": 5993
+    },
+    {
+      "epoch": 1.1815851735015772,
+      "grad_norm": 0.5146898936963578,
+      "learning_rate": 1.6010472437778827e-05,
+      "loss": 0.3507,
+      "step": 5994
+    },
+    {
+      "epoch": 1.1817823343848581,
+      "grad_norm": 1.9545816777357674,
+      "learning_rate": 1.6009233563900654e-05,
+      "loss": 0.3232,
+      "step": 5995
+    },
+    {
+      "epoch": 1.1819794952681388,
+      "grad_norm": 0.504150566251979,
+      "learning_rate": 1.6007994545645807e-05,
+      "loss": 0.3385,
+      "step": 5996
+    },
+    {
+      "epoch": 1.1821766561514195,
+      "grad_norm": 0.5178105405608192,
+      "learning_rate": 1.600675538304405e-05,
+      "loss": 0.3198,
+      "step": 5997
+    },
+    {
+      "epoch": 1.1823738170347002,
+      "grad_norm": 0.49033729293788897,
+      "learning_rate": 1.600551607612516e-05,
+      "loss": 0.3074,
+      "step": 5998
+    },
+    {
+      "epoch": 1.1825709779179812,
+      "grad_norm": 0.5216261093702694,
+      "learning_rate": 1.6004276624918906e-05,
+      "loss": 0.3524,
+      "step": 5999
+    },
+    {
+      "epoch": 1.1827681388012619,
+      "grad_norm": 0.48871872582944514,
+      "learning_rate": 1.600303702945507e-05,
+      "loss": 0.3333,
+      "step": 6000
+    },
+    {
+      "epoch": 1.1829652996845426,
+      "grad_norm": 0.5425430954135214,
+      "learning_rate": 1.600179728976344e-05,
+      "loss": 0.3351,
+      "step": 6001
+    },
+    {
+      "epoch": 1.1831624605678233,
+      "grad_norm": 0.4997582084309298,
+      "learning_rate": 1.6000557405873793e-05,
+      "loss": 0.3281,
+      "step": 6002
+    },
+    {
+      "epoch": 1.1833596214511042,
+      "grad_norm": 0.5468290480793452,
+      "learning_rate": 1.5999317377815927e-05,
+      "loss": 0.3504,
+      "step": 6003
+    },
+    {
+      "epoch": 1.1835567823343849,
+      "grad_norm": 0.5176645046959065,
+      "learning_rate": 1.5998077205619625e-05,
+      "loss": 0.3362,
+      "step": 6004
+    },
+    {
+      "epoch": 1.1837539432176656,
+      "grad_norm": 0.5013566738277708,
+      "learning_rate": 1.599683688931469e-05,
+      "loss": 0.3276,
+      "step": 6005
+    },
+    {
+      "epoch": 1.1839511041009463,
+      "grad_norm": 0.5694727653608741,
+      "learning_rate": 1.599559642893092e-05,
+      "loss": 0.346,
+      "step": 6006
+    },
+    {
+      "epoch": 1.1841482649842272,
+      "grad_norm": 0.49787309953663916,
+      "learning_rate": 1.5994355824498118e-05,
+      "loss": 0.3237,
+      "step": 6007
+    },
+    {
+      "epoch": 1.184345425867508,
+      "grad_norm": 0.5217262967923424,
+      "learning_rate": 1.5993115076046085e-05,
+      "loss": 0.3272,
+      "step": 6008
+    },
+    {
+      "epoch": 1.1845425867507886,
+      "grad_norm": 0.5171039380183391,
+      "learning_rate": 1.5991874183604638e-05,
+      "loss": 0.3409,
+      "step": 6009
+    },
+    {
+      "epoch": 1.1847397476340693,
+      "grad_norm": 0.5173593200967992,
+      "learning_rate": 1.5990633147203595e-05,
+      "loss": 0.3439,
+      "step": 6010
+    },
+    {
+      "epoch": 1.1849369085173502,
+      "grad_norm": 0.5600380235112447,
+      "learning_rate": 1.598939196687276e-05,
+      "loss": 0.3539,
+      "step": 6011
+    },
+    {
+      "epoch": 1.185134069400631,
+      "grad_norm": 0.47322318486676124,
+      "learning_rate": 1.5988150642641963e-05,
+      "loss": 0.3241,
+      "step": 6012
+    },
+    {
+      "epoch": 1.1853312302839116,
+      "grad_norm": 0.509209862620946,
+      "learning_rate": 1.598690917454102e-05,
+      "loss": 0.3384,
+      "step": 6013
+    },
+    {
+      "epoch": 1.1855283911671926,
+      "grad_norm": 0.5071917580583539,
+      "learning_rate": 1.598566756259977e-05,
+      "loss": 0.3306,
+      "step": 6014
+    },
+    {
+      "epoch": 1.1857255520504733,
+      "grad_norm": 0.5285407284823473,
+      "learning_rate": 1.598442580684803e-05,
+      "loss": 0.353,
+      "step": 6015
+    },
+    {
+      "epoch": 1.185922712933754,
+      "grad_norm": 0.5370829605446722,
+      "learning_rate": 1.598318390731564e-05,
+      "loss": 0.3423,
+      "step": 6016
+    },
+    {
+      "epoch": 1.1861198738170347,
+      "grad_norm": 0.4874835855626257,
+      "learning_rate": 1.5981941864032444e-05,
+      "loss": 0.3312,
+      "step": 6017
+    },
+    {
+      "epoch": 1.1863170347003154,
+      "grad_norm": 0.5441187518003243,
+      "learning_rate": 1.5980699677028276e-05,
+      "loss": 0.3475,
+      "step": 6018
+    },
+    {
+      "epoch": 1.1865141955835963,
+      "grad_norm": 0.5291193191006429,
+      "learning_rate": 1.597945734633298e-05,
+      "loss": 0.3461,
+      "step": 6019
+    },
+    {
+      "epoch": 1.186711356466877,
+      "grad_norm": 0.5254879886979995,
+      "learning_rate": 1.5978214871976408e-05,
+      "loss": 0.3455,
+      "step": 6020
+    },
+    {
+      "epoch": 1.1869085173501577,
+      "grad_norm": 0.5377765608865902,
+      "learning_rate": 1.597697225398841e-05,
+      "loss": 0.3352,
+      "step": 6021
+    },
+    {
+      "epoch": 1.1871056782334386,
+      "grad_norm": 0.5072825222986831,
+      "learning_rate": 1.5975729492398836e-05,
+      "loss": 0.3208,
+      "step": 6022
+    },
+    {
+      "epoch": 1.1873028391167193,
+      "grad_norm": 0.47380472273260626,
+      "learning_rate": 1.5974486587237554e-05,
+      "loss": 0.3119,
+      "step": 6023
+    },
+    {
+      "epoch": 1.1875,
+      "grad_norm": 0.5284173062047662,
+      "learning_rate": 1.5973243538534416e-05,
+      "loss": 0.348,
+      "step": 6024
+    },
+    {
+      "epoch": 1.1876971608832807,
+      "grad_norm": 0.532090130405166,
+      "learning_rate": 1.5972000346319296e-05,
+      "loss": 0.3584,
+      "step": 6025
+    },
+    {
+      "epoch": 1.1878943217665614,
+      "grad_norm": 0.5219678872457383,
+      "learning_rate": 1.5970757010622056e-05,
+      "loss": 0.3243,
+      "step": 6026
+    },
+    {
+      "epoch": 1.1880914826498423,
+      "grad_norm": 0.49984607033393075,
+      "learning_rate": 1.596951353147257e-05,
+      "loss": 0.3215,
+      "step": 6027
+    },
+    {
+      "epoch": 1.188288643533123,
+      "grad_norm": 0.49207605863384773,
+      "learning_rate": 1.5968269908900714e-05,
+      "loss": 0.2912,
+      "step": 6028
+    },
+    {
+      "epoch": 1.1884858044164037,
+      "grad_norm": 0.5613782227268997,
+      "learning_rate": 1.596702614293637e-05,
+      "loss": 0.3553,
+      "step": 6029
+    },
+    {
+      "epoch": 1.1886829652996846,
+      "grad_norm": 0.5250585921793455,
+      "learning_rate": 1.5965782233609416e-05,
+      "loss": 0.3738,
+      "step": 6030
+    },
+    {
+      "epoch": 1.1888801261829653,
+      "grad_norm": 0.5008108248618431,
+      "learning_rate": 1.5964538180949738e-05,
+      "loss": 0.3319,
+      "step": 6031
+    },
+    {
+      "epoch": 1.189077287066246,
+      "grad_norm": 0.5245356584082338,
+      "learning_rate": 1.596329398498723e-05,
+      "loss": 0.3239,
+      "step": 6032
+    },
+    {
+      "epoch": 1.1892744479495267,
+      "grad_norm": 0.5355385935858578,
+      "learning_rate": 1.5962049645751778e-05,
+      "loss": 0.339,
+      "step": 6033
+    },
+    {
+      "epoch": 1.1894716088328074,
+      "grad_norm": 0.6132239805769646,
+      "learning_rate": 1.5960805163273287e-05,
+      "loss": 0.365,
+      "step": 6034
+    },
+    {
+      "epoch": 1.1896687697160884,
+      "grad_norm": 9.838847324850864,
+      "learning_rate": 1.5959560537581646e-05,
+      "loss": 0.343,
+      "step": 6035
+    },
+    {
+      "epoch": 1.189865930599369,
+      "grad_norm": 0.5645786328771389,
+      "learning_rate": 1.5958315768706767e-05,
+      "loss": 0.3648,
+      "step": 6036
+    },
+    {
+      "epoch": 1.1900630914826498,
+      "grad_norm": 0.48867840037373045,
+      "learning_rate": 1.5957070856678553e-05,
+      "loss": 0.3233,
+      "step": 6037
+    },
+    {
+      "epoch": 1.1902602523659307,
+      "grad_norm": 0.529456568702109,
+      "learning_rate": 1.5955825801526918e-05,
+      "loss": 0.327,
+      "step": 6038
+    },
+    {
+      "epoch": 1.1904574132492114,
+      "grad_norm": 1.176277930833411,
+      "learning_rate": 1.5954580603281768e-05,
+      "loss": 0.3511,
+      "step": 6039
+    },
+    {
+      "epoch": 1.190654574132492,
+      "grad_norm": 0.5138247235703703,
+      "learning_rate": 1.5953335261973024e-05,
+      "loss": 0.3266,
+      "step": 6040
+    },
+    {
+      "epoch": 1.1908517350157728,
+      "grad_norm": 0.504680208228039,
+      "learning_rate": 1.5952089777630604e-05,
+      "loss": 0.3379,
+      "step": 6041
+    },
+    {
+      "epoch": 1.1910488958990537,
+      "grad_norm": 0.516549288250157,
+      "learning_rate": 1.5950844150284438e-05,
+      "loss": 0.3391,
+      "step": 6042
+    },
+    {
+      "epoch": 1.1912460567823344,
+      "grad_norm": 0.524867491432488,
+      "learning_rate": 1.5949598379964447e-05,
+      "loss": 0.3345,
+      "step": 6043
+    },
+    {
+      "epoch": 1.1914432176656151,
+      "grad_norm": 0.5311974639533232,
+      "learning_rate": 1.594835246670056e-05,
+      "loss": 0.3475,
+      "step": 6044
+    },
+    {
+      "epoch": 1.1916403785488958,
+      "grad_norm": 0.4916105988107756,
+      "learning_rate": 1.5947106410522722e-05,
+      "loss": 0.318,
+      "step": 6045
+    },
+    {
+      "epoch": 1.1918375394321767,
+      "grad_norm": 0.5324212670370398,
+      "learning_rate": 1.594586021146086e-05,
+      "loss": 0.3236,
+      "step": 6046
+    },
+    {
+      "epoch": 1.1920347003154574,
+      "grad_norm": 0.4964775871993211,
+      "learning_rate": 1.594461386954492e-05,
+      "loss": 0.3247,
+      "step": 6047
+    },
+    {
+      "epoch": 1.1922318611987381,
+      "grad_norm": 0.5471553857731185,
+      "learning_rate": 1.5943367384804842e-05,
+      "loss": 0.3463,
+      "step": 6048
+    },
+    {
+      "epoch": 1.1924290220820188,
+      "grad_norm": 0.7831995811222979,
+      "learning_rate": 1.5942120757270578e-05,
+      "loss": 0.3071,
+      "step": 6049
+    },
+    {
+      "epoch": 1.1926261829652998,
+      "grad_norm": 0.5010080635510805,
+      "learning_rate": 1.5940873986972078e-05,
+      "loss": 0.3082,
+      "step": 6050
+    },
+    {
+      "epoch": 1.1928233438485805,
+      "grad_norm": 0.49204619020050877,
+      "learning_rate": 1.5939627073939298e-05,
+      "loss": 0.325,
+      "step": 6051
+    },
+    {
+      "epoch": 1.1930205047318612,
+      "grad_norm": 0.5694016315454971,
+      "learning_rate": 1.593838001820219e-05,
+      "loss": 0.355,
+      "step": 6052
+    },
+    {
+      "epoch": 1.1932176656151419,
+      "grad_norm": 0.5103103854682324,
+      "learning_rate": 1.5937132819790722e-05,
+      "loss": 0.336,
+      "step": 6053
+    },
+    {
+      "epoch": 1.1934148264984228,
+      "grad_norm": 0.5971611394525508,
+      "learning_rate": 1.593588547873486e-05,
+      "loss": 0.2987,
+      "step": 6054
+    },
+    {
+      "epoch": 1.1936119873817035,
+      "grad_norm": 0.5324309908139868,
+      "learning_rate": 1.593463799506456e-05,
+      "loss": 0.3381,
+      "step": 6055
+    },
+    {
+      "epoch": 1.1938091482649842,
+      "grad_norm": 0.5513616850381257,
+      "learning_rate": 1.593339036880981e-05,
+      "loss": 0.3436,
+      "step": 6056
+    },
+    {
+      "epoch": 1.1940063091482649,
+      "grad_norm": 0.48370871862770526,
+      "learning_rate": 1.5932142600000577e-05,
+      "loss": 0.3131,
+      "step": 6057
+    },
+    {
+      "epoch": 1.1942034700315458,
+      "grad_norm": 0.5723593799005773,
+      "learning_rate": 1.5930894688666843e-05,
+      "loss": 0.3526,
+      "step": 6058
+    },
+    {
+      "epoch": 1.1944006309148265,
+      "grad_norm": 0.5244334828806275,
+      "learning_rate": 1.5929646634838583e-05,
+      "loss": 0.3277,
+      "step": 6059
+    },
+    {
+      "epoch": 1.1945977917981072,
+      "grad_norm": 0.5360911703031235,
+      "learning_rate": 1.5928398438545792e-05,
+      "loss": 0.3359,
+      "step": 6060
+    },
+    {
+      "epoch": 1.194794952681388,
+      "grad_norm": 0.5103692154019288,
+      "learning_rate": 1.5927150099818454e-05,
+      "loss": 0.3397,
+      "step": 6061
+    },
+    {
+      "epoch": 1.1949921135646688,
+      "grad_norm": 0.4920913803322767,
+      "learning_rate": 1.592590161868656e-05,
+      "loss": 0.2967,
+      "step": 6062
+    },
+    {
+      "epoch": 1.1951892744479495,
+      "grad_norm": 0.5338396158035961,
+      "learning_rate": 1.5924652995180106e-05,
+      "loss": 0.3377,
+      "step": 6063
+    },
+    {
+      "epoch": 1.1953864353312302,
+      "grad_norm": 0.4978817940429984,
+      "learning_rate": 1.5923404229329097e-05,
+      "loss": 0.3148,
+      "step": 6064
+    },
+    {
+      "epoch": 1.1955835962145112,
+      "grad_norm": 0.5108248246371543,
+      "learning_rate": 1.5922155321163528e-05,
+      "loss": 0.344,
+      "step": 6065
+    },
+    {
+      "epoch": 1.1957807570977919,
+      "grad_norm": 0.4962965845967901,
+      "learning_rate": 1.592090627071341e-05,
+      "loss": 0.32,
+      "step": 6066
+    },
+    {
+      "epoch": 1.1959779179810726,
+      "grad_norm": 0.5153462208898505,
+      "learning_rate": 1.591965707800875e-05,
+      "loss": 0.3368,
+      "step": 6067
+    },
+    {
+      "epoch": 1.1961750788643533,
+      "grad_norm": 0.5214640418478306,
+      "learning_rate": 1.5918407743079564e-05,
+      "loss": 0.3297,
+      "step": 6068
+    },
+    {
+      "epoch": 1.196372239747634,
+      "grad_norm": 0.5677250498709867,
+      "learning_rate": 1.5917158265955863e-05,
+      "loss": 0.3743,
+      "step": 6069
+    },
+    {
+      "epoch": 1.1965694006309149,
+      "grad_norm": 0.5268238165290028,
+      "learning_rate": 1.591590864666767e-05,
+      "loss": 0.3654,
+      "step": 6070
+    },
+    {
+      "epoch": 1.1967665615141956,
+      "grad_norm": 0.4961093876386208,
+      "learning_rate": 1.5914658885245006e-05,
+      "loss": 0.3319,
+      "step": 6071
+    },
+    {
+      "epoch": 1.1969637223974763,
+      "grad_norm": 0.5169884357870341,
+      "learning_rate": 1.5913408981717902e-05,
+      "loss": 0.355,
+      "step": 6072
+    },
+    {
+      "epoch": 1.1971608832807572,
+      "grad_norm": 0.4863348861419478,
+      "learning_rate": 1.5912158936116383e-05,
+      "loss": 0.3118,
+      "step": 6073
+    },
+    {
+      "epoch": 1.197358044164038,
+      "grad_norm": 0.49535517239582816,
+      "learning_rate": 1.5910908748470485e-05,
+      "loss": 0.2871,
+      "step": 6074
+    },
+    {
+      "epoch": 1.1975552050473186,
+      "grad_norm": 0.5012578788415777,
+      "learning_rate": 1.5909658418810246e-05,
+      "loss": 0.3124,
+      "step": 6075
+    },
+    {
+      "epoch": 1.1977523659305993,
+      "grad_norm": 6.581609483160163,
+      "learning_rate": 1.5908407947165704e-05,
+      "loss": 0.3924,
+      "step": 6076
+    },
+    {
+      "epoch": 1.19794952681388,
+      "grad_norm": 0.5416411980138617,
+      "learning_rate": 1.59071573335669e-05,
+      "loss": 0.3265,
+      "step": 6077
+    },
+    {
+      "epoch": 1.198146687697161,
+      "grad_norm": 5.137867502750501,
+      "learning_rate": 1.5905906578043892e-05,
+      "loss": 0.4025,
+      "step": 6078
+    },
+    {
+      "epoch": 1.1983438485804416,
+      "grad_norm": 0.498758781908228,
+      "learning_rate": 1.5904655680626712e-05,
+      "loss": 0.3361,
+      "step": 6079
+    },
+    {
+      "epoch": 1.1985410094637223,
+      "grad_norm": 0.5129283262777451,
+      "learning_rate": 1.590340464134543e-05,
+      "loss": 0.3302,
+      "step": 6080
+    },
+    {
+      "epoch": 1.1987381703470033,
+      "grad_norm": 0.7337508073690954,
+      "learning_rate": 1.5902153460230097e-05,
+      "loss": 0.3584,
+      "step": 6081
+    },
+    {
+      "epoch": 1.198935331230284,
+      "grad_norm": 0.4927136722041788,
+      "learning_rate": 1.5900902137310777e-05,
+      "loss": 0.315,
+      "step": 6082
+    },
+    {
+      "epoch": 1.1991324921135647,
+      "grad_norm": 0.4999312067571768,
+      "learning_rate": 1.5899650672617526e-05,
+      "loss": 0.3218,
+      "step": 6083
+    },
+    {
+      "epoch": 1.1993296529968454,
+      "grad_norm": 0.5108041586913794,
+      "learning_rate": 1.589839906618042e-05,
+      "loss": 0.3266,
+      "step": 6084
+    },
+    {
+      "epoch": 1.1995268138801263,
+      "grad_norm": 0.5344740208260175,
+      "learning_rate": 1.5897147318029524e-05,
+      "loss": 0.3587,
+      "step": 6085
+    },
+    {
+      "epoch": 1.199723974763407,
+      "grad_norm": 0.5888513085931902,
+      "learning_rate": 1.5895895428194915e-05,
+      "loss": 0.3455,
+      "step": 6086
+    },
+    {
+      "epoch": 1.1999211356466877,
+      "grad_norm": 0.4926690283535484,
+      "learning_rate": 1.5894643396706674e-05,
+      "loss": 0.3181,
+      "step": 6087
+    },
+    {
+      "epoch": 1.2001182965299684,
+      "grad_norm": 0.5047531232097675,
+      "learning_rate": 1.5893391223594873e-05,
+      "loss": 0.3396,
+      "step": 6088
+    },
+    {
+      "epoch": 1.2003154574132493,
+      "grad_norm": 0.5194620816609076,
+      "learning_rate": 1.5892138908889606e-05,
+      "loss": 0.3498,
+      "step": 6089
+    },
+    {
+      "epoch": 1.20051261829653,
+      "grad_norm": 0.49106900692891253,
+      "learning_rate": 1.589088645262096e-05,
+      "loss": 0.3156,
+      "step": 6090
+    },
+    {
+      "epoch": 1.2007097791798107,
+      "grad_norm": 0.4816616653634188,
+      "learning_rate": 1.5889633854819014e-05,
+      "loss": 0.3269,
+      "step": 6091
+    },
+    {
+      "epoch": 1.2009069400630914,
+      "grad_norm": 0.5018051104959026,
+      "learning_rate": 1.5888381115513878e-05,
+      "loss": 0.3261,
+      "step": 6092
+    },
+    {
+      "epoch": 1.2011041009463723,
+      "grad_norm": 0.5243143133854937,
+      "learning_rate": 1.5887128234735638e-05,
+      "loss": 0.3356,
+      "step": 6093
+    },
+    {
+      "epoch": 1.201301261829653,
+      "grad_norm": 0.5816921886358895,
+      "learning_rate": 1.5885875212514408e-05,
+      "loss": 0.3673,
+      "step": 6094
+    },
+    {
+      "epoch": 1.2014984227129337,
+      "grad_norm": 0.4886736205149519,
+      "learning_rate": 1.5884622048880283e-05,
+      "loss": 0.348,
+      "step": 6095
+    },
+    {
+      "epoch": 1.2016955835962144,
+      "grad_norm": 0.5352170470566625,
+      "learning_rate": 1.5883368743863376e-05,
+      "loss": 0.3531,
+      "step": 6096
+    },
+    {
+      "epoch": 1.2018927444794953,
+      "grad_norm": 0.5070353571291337,
+      "learning_rate": 1.5882115297493793e-05,
+      "loss": 0.3074,
+      "step": 6097
+    },
+    {
+      "epoch": 1.202089905362776,
+      "grad_norm": 0.5299404136107287,
+      "learning_rate": 1.588086170980166e-05,
+      "loss": 0.3548,
+      "step": 6098
+    },
+    {
+      "epoch": 1.2022870662460567,
+      "grad_norm": 0.5010029776145218,
+      "learning_rate": 1.5879607980817084e-05,
+      "loss": 0.3322,
+      "step": 6099
+    },
+    {
+      "epoch": 1.2024842271293374,
+      "grad_norm": 0.5443027119113688,
+      "learning_rate": 1.5878354110570188e-05,
+      "loss": 0.3445,
+      "step": 6100
+    },
+    {
+      "epoch": 1.2026813880126184,
+      "grad_norm": 0.5116498162874465,
+      "learning_rate": 1.5877100099091106e-05,
+      "loss": 0.35,
+      "step": 6101
+    },
+    {
+      "epoch": 1.202878548895899,
+      "grad_norm": 0.4990846323105568,
+      "learning_rate": 1.587584594640996e-05,
+      "loss": 0.303,
+      "step": 6102
+    },
+    {
+      "epoch": 1.2030757097791798,
+      "grad_norm": 0.6083340219487477,
+      "learning_rate": 1.5874591652556887e-05,
+      "loss": 0.3466,
+      "step": 6103
+    },
+    {
+      "epoch": 1.2032728706624605,
+      "grad_norm": 0.5699359859289749,
+      "learning_rate": 1.5873337217562012e-05,
+      "loss": 0.3492,
+      "step": 6104
+    },
+    {
+      "epoch": 1.2034700315457414,
+      "grad_norm": 0.5305845146338529,
+      "learning_rate": 1.5872082641455484e-05,
+      "loss": 0.3311,
+      "step": 6105
+    },
+    {
+      "epoch": 1.203667192429022,
+      "grad_norm": 0.4987574811623951,
+      "learning_rate": 1.5870827924267442e-05,
+      "loss": 0.3127,
+      "step": 6106
+    },
+    {
+      "epoch": 1.2038643533123028,
+      "grad_norm": 0.5542197601383188,
+      "learning_rate": 1.586957306602803e-05,
+      "loss": 0.3711,
+      "step": 6107
+    },
+    {
+      "epoch": 1.2040615141955837,
+      "grad_norm": 0.5296385507003382,
+      "learning_rate": 1.58683180667674e-05,
+      "loss": 0.3322,
+      "step": 6108
+    },
+    {
+      "epoch": 1.2042586750788644,
+      "grad_norm": 0.510771246311118,
+      "learning_rate": 1.5867062926515702e-05,
+      "loss": 0.3478,
+      "step": 6109
+    },
+    {
+      "epoch": 1.2044558359621451,
+      "grad_norm": 0.5083219009195116,
+      "learning_rate": 1.586580764530309e-05,
+      "loss": 0.3283,
+      "step": 6110
+    },
+    {
+      "epoch": 1.2046529968454258,
+      "grad_norm": 0.4984543211011751,
+      "learning_rate": 1.586455222315973e-05,
+      "loss": 0.3161,
+      "step": 6111
+    },
+    {
+      "epoch": 1.2048501577287065,
+      "grad_norm": 0.48449825361455534,
+      "learning_rate": 1.5863296660115778e-05,
+      "loss": 0.3235,
+      "step": 6112
+    },
+    {
+      "epoch": 1.2050473186119874,
+      "grad_norm": 0.4880911697740563,
+      "learning_rate": 1.58620409562014e-05,
+      "loss": 0.3221,
+      "step": 6113
+    },
+    {
+      "epoch": 1.2052444794952681,
+      "grad_norm": 0.6718849269017811,
+      "learning_rate": 1.586078511144677e-05,
+      "loss": 0.3891,
+      "step": 6114
+    },
+    {
+      "epoch": 1.2054416403785488,
+      "grad_norm": 0.536419239014414,
+      "learning_rate": 1.5859529125882058e-05,
+      "loss": 0.3603,
+      "step": 6115
+    },
+    {
+      "epoch": 1.2056388012618298,
+      "grad_norm": 0.5329763858491461,
+      "learning_rate": 1.585827299953744e-05,
+      "loss": 0.3488,
+      "step": 6116
+    },
+    {
+      "epoch": 1.2058359621451105,
+      "grad_norm": 1.3571017677348047,
+      "learning_rate": 1.5857016732443096e-05,
+      "loss": 0.331,
+      "step": 6117
+    },
+    {
+      "epoch": 1.2060331230283912,
+      "grad_norm": 0.5105153466144287,
+      "learning_rate": 1.5855760324629204e-05,
+      "loss": 0.324,
+      "step": 6118
+    },
+    {
+      "epoch": 1.2062302839116719,
+      "grad_norm": 0.4980780054716261,
+      "learning_rate": 1.585450377612596e-05,
+      "loss": 0.3108,
+      "step": 6119
+    },
+    {
+      "epoch": 1.2064274447949526,
+      "grad_norm": 0.969040129147523,
+      "learning_rate": 1.5853247086963546e-05,
+      "loss": 0.3388,
+      "step": 6120
+    },
+    {
+      "epoch": 1.2066246056782335,
+      "grad_norm": 0.5442833624434943,
+      "learning_rate": 1.585199025717216e-05,
+      "loss": 0.3501,
+      "step": 6121
+    },
+    {
+      "epoch": 1.2068217665615142,
+      "grad_norm": 0.50130789731431,
+      "learning_rate": 1.585073328678199e-05,
+      "loss": 0.33,
+      "step": 6122
+    },
+    {
+      "epoch": 1.2070189274447949,
+      "grad_norm": 0.5068279641638377,
+      "learning_rate": 1.5849476175823242e-05,
+      "loss": 0.3371,
+      "step": 6123
+    },
+    {
+      "epoch": 1.2072160883280758,
+      "grad_norm": 0.567218532987645,
+      "learning_rate": 1.584821892432612e-05,
+      "loss": 0.3313,
+      "step": 6124
+    },
+    {
+      "epoch": 1.2074132492113565,
+      "grad_norm": 0.5501597914721346,
+      "learning_rate": 1.5846961532320833e-05,
+      "loss": 0.3363,
+      "step": 6125
+    },
+    {
+      "epoch": 1.2076104100946372,
+      "grad_norm": 0.5130749844399497,
+      "learning_rate": 1.584570399983758e-05,
+      "loss": 0.3331,
+      "step": 6126
+    },
+    {
+      "epoch": 1.207807570977918,
+      "grad_norm": 0.7105268873576728,
+      "learning_rate": 1.5844446326906585e-05,
+      "loss": 0.3289,
+      "step": 6127
+    },
+    {
+      "epoch": 1.2080047318611988,
+      "grad_norm": 0.49590774482060557,
+      "learning_rate": 1.5843188513558056e-05,
+      "loss": 0.3159,
+      "step": 6128
+    },
+    {
+      "epoch": 1.2082018927444795,
+      "grad_norm": 0.522580528542394,
+      "learning_rate": 1.5841930559822222e-05,
+      "loss": 0.3258,
+      "step": 6129
+    },
+    {
+      "epoch": 1.2083990536277602,
+      "grad_norm": 0.5232730029770233,
+      "learning_rate": 1.58406724657293e-05,
+      "loss": 0.3242,
+      "step": 6130
+    },
+    {
+      "epoch": 1.208596214511041,
+      "grad_norm": 0.5564685376916361,
+      "learning_rate": 1.583941423130952e-05,
+      "loss": 0.3414,
+      "step": 6131
+    },
+    {
+      "epoch": 1.2087933753943219,
+      "grad_norm": 0.5155603754900965,
+      "learning_rate": 1.583815585659311e-05,
+      "loss": 0.3329,
+      "step": 6132
+    },
+    {
+      "epoch": 1.2089905362776026,
+      "grad_norm": 0.49929691738974885,
+      "learning_rate": 1.58368973416103e-05,
+      "loss": 0.3302,
+      "step": 6133
+    },
+    {
+      "epoch": 1.2091876971608833,
+      "grad_norm": 0.5307058011919017,
+      "learning_rate": 1.5835638686391338e-05,
+      "loss": 0.3232,
+      "step": 6134
+    },
+    {
+      "epoch": 1.209384858044164,
+      "grad_norm": 0.5116249657384987,
+      "learning_rate": 1.583437989096645e-05,
+      "loss": 0.3149,
+      "step": 6135
+    },
+    {
+      "epoch": 1.2095820189274449,
+      "grad_norm": 0.4958052917998363,
+      "learning_rate": 1.5833120955365894e-05,
+      "loss": 0.345,
+      "step": 6136
+    },
+    {
+      "epoch": 1.2097791798107256,
+      "grad_norm": 0.5772963968620416,
+      "learning_rate": 1.5831861879619904e-05,
+      "loss": 0.3658,
+      "step": 6137
+    },
+    {
+      "epoch": 1.2099763406940063,
+      "grad_norm": 0.5176663312841757,
+      "learning_rate": 1.5830602663758737e-05,
+      "loss": 0.3321,
+      "step": 6138
+    },
+    {
+      "epoch": 1.210173501577287,
+      "grad_norm": 0.5572969524001481,
+      "learning_rate": 1.582934330781265e-05,
+      "loss": 0.3646,
+      "step": 6139
+    },
+    {
+      "epoch": 1.210370662460568,
+      "grad_norm": 0.5149376440705687,
+      "learning_rate": 1.582808381181189e-05,
+      "loss": 0.3233,
+      "step": 6140
+    },
+    {
+      "epoch": 1.2105678233438486,
+      "grad_norm": 0.5060377765127014,
+      "learning_rate": 1.5826824175786724e-05,
+      "loss": 0.3241,
+      "step": 6141
+    },
+    {
+      "epoch": 1.2107649842271293,
+      "grad_norm": 0.5075915990283572,
+      "learning_rate": 1.5825564399767416e-05,
+      "loss": 0.3315,
+      "step": 6142
+    },
+    {
+      "epoch": 1.21096214511041,
+      "grad_norm": 0.48101366074112933,
+      "learning_rate": 1.5824304483784234e-05,
+      "loss": 0.3271,
+      "step": 6143
+    },
+    {
+      "epoch": 1.211159305993691,
+      "grad_norm": 0.5006150490244125,
+      "learning_rate": 1.5823044427867446e-05,
+      "loss": 0.3208,
+      "step": 6144
+    },
+    {
+      "epoch": 1.2113564668769716,
+      "grad_norm": 0.5083292731196384,
+      "learning_rate": 1.582178423204732e-05,
+      "loss": 0.3332,
+      "step": 6145
+    },
+    {
+      "epoch": 1.2115536277602523,
+      "grad_norm": 0.5431538616132132,
+      "learning_rate": 1.5820523896354146e-05,
+      "loss": 0.3188,
+      "step": 6146
+    },
+    {
+      "epoch": 1.211750788643533,
+      "grad_norm": 0.4949725843395628,
+      "learning_rate": 1.5819263420818198e-05,
+      "loss": 0.3279,
+      "step": 6147
+    },
+    {
+      "epoch": 1.211947949526814,
+      "grad_norm": 0.5283861058082052,
+      "learning_rate": 1.5818002805469758e-05,
+      "loss": 0.3466,
+      "step": 6148
+    },
+    {
+      "epoch": 1.2121451104100947,
+      "grad_norm": 0.49009428195008536,
+      "learning_rate": 1.581674205033912e-05,
+      "loss": 0.316,
+      "step": 6149
+    },
+    {
+      "epoch": 1.2123422712933754,
+      "grad_norm": 0.527558337367723,
+      "learning_rate": 1.5815481155456566e-05,
+      "loss": 0.3217,
+      "step": 6150
+    },
+    {
+      "epoch": 1.2125394321766563,
+      "grad_norm": 0.46986929303135894,
+      "learning_rate": 1.581422012085239e-05,
+      "loss": 0.3353,
+      "step": 6151
+    },
+    {
+      "epoch": 1.212736593059937,
+      "grad_norm": 0.5248790360545358,
+      "learning_rate": 1.5812958946556897e-05,
+      "loss": 0.3471,
+      "step": 6152
+    },
+    {
+      "epoch": 1.2129337539432177,
+      "grad_norm": 0.4914294810645838,
+      "learning_rate": 1.581169763260039e-05,
+      "loss": 0.3302,
+      "step": 6153
+    },
+    {
+      "epoch": 1.2131309148264984,
+      "grad_norm": 0.4763550637519196,
+      "learning_rate": 1.5810436179013158e-05,
+      "loss": 0.3128,
+      "step": 6154
+    },
+    {
+      "epoch": 1.213328075709779,
+      "grad_norm": 0.5141786955182022,
+      "learning_rate": 1.5809174585825523e-05,
+      "loss": 0.3347,
+      "step": 6155
+    },
+    {
+      "epoch": 1.21352523659306,
+      "grad_norm": 0.5181612279002368,
+      "learning_rate": 1.5807912853067787e-05,
+      "loss": 0.3764,
+      "step": 6156
+    },
+    {
+      "epoch": 1.2137223974763407,
+      "grad_norm": 0.5063765486686969,
+      "learning_rate": 1.5806650980770273e-05,
+      "loss": 0.3455,
+      "step": 6157
+    },
+    {
+      "epoch": 1.2139195583596214,
+      "grad_norm": 0.5010394541454587,
+      "learning_rate": 1.5805388968963286e-05,
+      "loss": 0.3283,
+      "step": 6158
+    },
+    {
+      "epoch": 1.2141167192429023,
+      "grad_norm": 0.4758924748836648,
+      "learning_rate": 1.5804126817677158e-05,
+      "loss": 0.31,
+      "step": 6159
+    },
+    {
+      "epoch": 1.214313880126183,
+      "grad_norm": 0.5025450657121269,
+      "learning_rate": 1.580286452694221e-05,
+      "loss": 0.3045,
+      "step": 6160
+    },
+    {
+      "epoch": 1.2145110410094637,
+      "grad_norm": 0.5146546499338246,
+      "learning_rate": 1.5801602096788768e-05,
+      "loss": 0.3355,
+      "step": 6161
+    },
+    {
+      "epoch": 1.2147082018927444,
+      "grad_norm": 0.5384444969117445,
+      "learning_rate": 1.5800339527247163e-05,
+      "loss": 0.3499,
+      "step": 6162
+    },
+    {
+      "epoch": 1.2149053627760251,
+      "grad_norm": 0.4940594458584491,
+      "learning_rate": 1.579907681834773e-05,
+      "loss": 0.3296,
+      "step": 6163
+    },
+    {
+      "epoch": 1.215102523659306,
+      "grad_norm": 0.5343762136162648,
+      "learning_rate": 1.579781397012081e-05,
+      "loss": 0.3352,
+      "step": 6164
+    },
+    {
+      "epoch": 1.2152996845425867,
+      "grad_norm": 0.48752315188028517,
+      "learning_rate": 1.5796550982596732e-05,
+      "loss": 0.3282,
+      "step": 6165
+    },
+    {
+      "epoch": 1.2154968454258674,
+      "grad_norm": 0.5547331237218581,
+      "learning_rate": 1.5795287855805853e-05,
+      "loss": 0.3829,
+      "step": 6166
+    },
+    {
+      "epoch": 1.2156940063091484,
+      "grad_norm": 0.5004995021041893,
+      "learning_rate": 1.5794024589778518e-05,
+      "loss": 0.328,
+      "step": 6167
+    },
+    {
+      "epoch": 1.215891167192429,
+      "grad_norm": 0.5175275340376262,
+      "learning_rate": 1.5792761184545076e-05,
+      "loss": 0.3307,
+      "step": 6168
+    },
+    {
+      "epoch": 1.2160883280757098,
+      "grad_norm": 0.5027473643529223,
+      "learning_rate": 1.579149764013588e-05,
+      "loss": 0.3395,
+      "step": 6169
+    },
+    {
+      "epoch": 1.2162854889589905,
+      "grad_norm": 0.4817141906134851,
+      "learning_rate": 1.579023395658129e-05,
+      "loss": 0.3249,
+      "step": 6170
+    },
+    {
+      "epoch": 1.2164826498422712,
+      "grad_norm": 0.46188472835340966,
+      "learning_rate": 1.578897013391167e-05,
+      "loss": 0.3147,
+      "step": 6171
+    },
+    {
+      "epoch": 1.216679810725552,
+      "grad_norm": 0.4988989117087451,
+      "learning_rate": 1.5787706172157374e-05,
+      "loss": 0.33,
+      "step": 6172
+    },
+    {
+      "epoch": 1.2168769716088328,
+      "grad_norm": 0.49979162808926375,
+      "learning_rate": 1.578644207134878e-05,
+      "loss": 0.335,
+      "step": 6173
+    },
+    {
+      "epoch": 1.2170741324921135,
+      "grad_norm": 9.87115802371154,
+      "learning_rate": 1.5785177831516255e-05,
+      "loss": 0.5038,
+      "step": 6174
+    },
+    {
+      "epoch": 1.2172712933753944,
+      "grad_norm": 0.49379194281460687,
+      "learning_rate": 1.5783913452690174e-05,
+      "loss": 0.3263,
+      "step": 6175
+    },
+    {
+      "epoch": 1.2174684542586751,
+      "grad_norm": 0.7283072054648132,
+      "learning_rate": 1.5782648934900915e-05,
+      "loss": 0.3288,
+      "step": 6176
+    },
+    {
+      "epoch": 1.2176656151419558,
+      "grad_norm": 0.5063345599140339,
+      "learning_rate": 1.5781384278178858e-05,
+      "loss": 0.3322,
+      "step": 6177
+    },
+    {
+      "epoch": 1.2178627760252365,
+      "grad_norm": 0.49525686855704687,
+      "learning_rate": 1.578011948255439e-05,
+      "loss": 0.3412,
+      "step": 6178
+    },
+    {
+      "epoch": 1.2180599369085174,
+      "grad_norm": 0.4937155808164859,
+      "learning_rate": 1.5778854548057893e-05,
+      "loss": 0.3404,
+      "step": 6179
+    },
+    {
+      "epoch": 1.2182570977917981,
+      "grad_norm": 0.5601096943034672,
+      "learning_rate": 1.5777589474719764e-05,
+      "loss": 0.3301,
+      "step": 6180
+    },
+    {
+      "epoch": 1.2184542586750788,
+      "grad_norm": 0.5134943557442895,
+      "learning_rate": 1.5776324262570394e-05,
+      "loss": 0.3036,
+      "step": 6181
+    },
+    {
+      "epoch": 1.2186514195583595,
+      "grad_norm": 0.5234618403194057,
+      "learning_rate": 1.577505891164018e-05,
+      "loss": 0.3487,
+      "step": 6182
+    },
+    {
+      "epoch": 1.2188485804416405,
+      "grad_norm": 0.5060468910052316,
+      "learning_rate": 1.5773793421959528e-05,
+      "loss": 0.3045,
+      "step": 6183
+    },
+    {
+      "epoch": 1.2190457413249212,
+      "grad_norm": 0.5385344505438445,
+      "learning_rate": 1.577252779355884e-05,
+      "loss": 0.3471,
+      "step": 6184
+    },
+    {
+      "epoch": 1.2192429022082019,
+      "grad_norm": 0.47439335012121125,
+      "learning_rate": 1.577126202646852e-05,
+      "loss": 0.3137,
+      "step": 6185
+    },
+    {
+      "epoch": 1.2194400630914826,
+      "grad_norm": 0.5634484011687884,
+      "learning_rate": 1.5769996120718985e-05,
+      "loss": 0.3504,
+      "step": 6186
+    },
+    {
+      "epoch": 1.2196372239747635,
+      "grad_norm": 0.4783159791866919,
+      "learning_rate": 1.5768730076340646e-05,
+      "loss": 0.3238,
+      "step": 6187
+    },
+    {
+      "epoch": 1.2198343848580442,
+      "grad_norm": 0.5282172252080681,
+      "learning_rate": 1.5767463893363925e-05,
+      "loss": 0.324,
+      "step": 6188
+    },
+    {
+      "epoch": 1.2200315457413249,
+      "grad_norm": 0.508650741550587,
+      "learning_rate": 1.5766197571819234e-05,
+      "loss": 0.3108,
+      "step": 6189
+    },
+    {
+      "epoch": 1.2202287066246056,
+      "grad_norm": 0.47948864235134103,
+      "learning_rate": 1.5764931111737005e-05,
+      "loss": 0.335,
+      "step": 6190
+    },
+    {
+      "epoch": 1.2204258675078865,
+      "grad_norm": 0.5276091720414406,
+      "learning_rate": 1.576366451314766e-05,
+      "loss": 0.3546,
+      "step": 6191
+    },
+    {
+      "epoch": 1.2206230283911672,
+      "grad_norm": 0.48975202677127483,
+      "learning_rate": 1.576239777608164e-05,
+      "loss": 0.3195,
+      "step": 6192
+    },
+    {
+      "epoch": 1.220820189274448,
+      "grad_norm": 0.5115397169665277,
+      "learning_rate": 1.576113090056937e-05,
+      "loss": 0.3552,
+      "step": 6193
+    },
+    {
+      "epoch": 1.2210173501577288,
+      "grad_norm": 0.5544373081682192,
+      "learning_rate": 1.575986388664129e-05,
+      "loss": 0.3388,
+      "step": 6194
+    },
+    {
+      "epoch": 1.2212145110410095,
+      "grad_norm": 0.4789487893265251,
+      "learning_rate": 1.5758596734327842e-05,
+      "loss": 0.3049,
+      "step": 6195
+    },
+    {
+      "epoch": 1.2214116719242902,
+      "grad_norm": 0.5085580816350456,
+      "learning_rate": 1.5757329443659468e-05,
+      "loss": 0.3337,
+      "step": 6196
+    },
+    {
+      "epoch": 1.221608832807571,
+      "grad_norm": 0.5018198548590921,
+      "learning_rate": 1.5756062014666622e-05,
+      "loss": 0.3602,
+      "step": 6197
+    },
+    {
+      "epoch": 1.2218059936908516,
+      "grad_norm": 0.5459993872154436,
+      "learning_rate": 1.5754794447379747e-05,
+      "loss": 0.3734,
+      "step": 6198
+    },
+    {
+      "epoch": 1.2220031545741326,
+      "grad_norm": 0.5069803136116705,
+      "learning_rate": 1.5753526741829302e-05,
+      "loss": 0.3292,
+      "step": 6199
+    },
+    {
+      "epoch": 1.2222003154574133,
+      "grad_norm": 0.5349842354961301,
+      "learning_rate": 1.5752258898045747e-05,
+      "loss": 0.3591,
+      "step": 6200
+    },
+    {
+      "epoch": 1.222397476340694,
+      "grad_norm": 0.5113741033129703,
+      "learning_rate": 1.5750990916059537e-05,
+      "loss": 0.3354,
+      "step": 6201
+    },
+    {
+      "epoch": 1.2225946372239749,
+      "grad_norm": 0.4764417943474872,
+      "learning_rate": 1.5749722795901142e-05,
+      "loss": 0.2776,
+      "step": 6202
+    },
+    {
+      "epoch": 1.2227917981072556,
+      "grad_norm": 0.4852541319749403,
+      "learning_rate": 1.574845453760102e-05,
+      "loss": 0.2877,
+      "step": 6203
+    },
+    {
+      "epoch": 1.2229889589905363,
+      "grad_norm": 0.5011535867185861,
+      "learning_rate": 1.5747186141189654e-05,
+      "loss": 0.3188,
+      "step": 6204
+    },
+    {
+      "epoch": 1.223186119873817,
+      "grad_norm": 0.5585191080381091,
+      "learning_rate": 1.574591760669751e-05,
+      "loss": 0.3278,
+      "step": 6205
+    },
+    {
+      "epoch": 1.2233832807570977,
+      "grad_norm": 0.5095163688965774,
+      "learning_rate": 1.574464893415507e-05,
+      "loss": 0.3388,
+      "step": 6206
+    },
+    {
+      "epoch": 1.2235804416403786,
+      "grad_norm": 0.489944289171812,
+      "learning_rate": 1.5743380123592815e-05,
+      "loss": 0.3298,
+      "step": 6207
+    },
+    {
+      "epoch": 1.2237776025236593,
+      "grad_norm": 0.56552053195053,
+      "learning_rate": 1.5742111175041222e-05,
+      "loss": 0.3271,
+      "step": 6208
+    },
+    {
+      "epoch": 1.22397476340694,
+      "grad_norm": 0.4975929434887488,
+      "learning_rate": 1.5740842088530788e-05,
+      "loss": 0.3171,
+      "step": 6209
+    },
+    {
+      "epoch": 1.224171924290221,
+      "grad_norm": 0.5635718664701436,
+      "learning_rate": 1.5739572864091995e-05,
+      "loss": 0.333,
+      "step": 6210
+    },
+    {
+      "epoch": 1.2243690851735016,
+      "grad_norm": 0.5466750621448705,
+      "learning_rate": 1.573830350175535e-05,
+      "loss": 0.3226,
+      "step": 6211
+    },
+    {
+      "epoch": 1.2245662460567823,
+      "grad_norm": 0.5179765647986949,
+      "learning_rate": 1.5737034001551336e-05,
+      "loss": 0.3451,
+      "step": 6212
+    },
+    {
+      "epoch": 1.224763406940063,
+      "grad_norm": 0.505297026337151,
+      "learning_rate": 1.573576436351046e-05,
+      "loss": 0.3263,
+      "step": 6213
+    },
+    {
+      "epoch": 1.2249605678233437,
+      "grad_norm": 0.5323927892854735,
+      "learning_rate": 1.573449458766323e-05,
+      "loss": 0.3348,
+      "step": 6214
+    },
+    {
+      "epoch": 1.2251577287066246,
+      "grad_norm": 0.4585385421095369,
+      "learning_rate": 1.573322467404015e-05,
+      "loss": 0.2889,
+      "step": 6215
+    },
+    {
+      "epoch": 1.2253548895899053,
+      "grad_norm": 0.5133915738969325,
+      "learning_rate": 1.573195462267173e-05,
+      "loss": 0.3354,
+      "step": 6216
+    },
+    {
+      "epoch": 1.225552050473186,
+      "grad_norm": 0.4976376614042412,
+      "learning_rate": 1.573068443358848e-05,
+      "loss": 0.3261,
+      "step": 6217
+    },
+    {
+      "epoch": 1.225749211356467,
+      "grad_norm": 0.5204545809416458,
+      "learning_rate": 1.572941410682092e-05,
+      "loss": 0.3152,
+      "step": 6218
+    },
+    {
+      "epoch": 1.2259463722397477,
+      "grad_norm": 0.46008522086558107,
+      "learning_rate": 1.572814364239958e-05,
+      "loss": 0.3049,
+      "step": 6219
+    },
+    {
+      "epoch": 1.2261435331230284,
+      "grad_norm": 0.5085240798538043,
+      "learning_rate": 1.572687304035497e-05,
+      "loss": 0.3392,
+      "step": 6220
+    },
+    {
+      "epoch": 1.226340694006309,
+      "grad_norm": 0.49075420527666574,
+      "learning_rate": 1.5725602300717628e-05,
+      "loss": 0.3238,
+      "step": 6221
+    },
+    {
+      "epoch": 1.22653785488959,
+      "grad_norm": 0.5009588499072523,
+      "learning_rate": 1.5724331423518076e-05,
+      "loss": 0.3588,
+      "step": 6222
+    },
+    {
+      "epoch": 1.2267350157728707,
+      "grad_norm": 0.5004985108490898,
+      "learning_rate": 1.572306040878685e-05,
+      "loss": 0.3413,
+      "step": 6223
+    },
+    {
+      "epoch": 1.2269321766561514,
+      "grad_norm": 0.5036837263941046,
+      "learning_rate": 1.5721789256554495e-05,
+      "loss": 0.3472,
+      "step": 6224
+    },
+    {
+      "epoch": 1.227129337539432,
+      "grad_norm": 0.7266521472928784,
+      "learning_rate": 1.5720517966851544e-05,
+      "loss": 0.3553,
+      "step": 6225
+    },
+    {
+      "epoch": 1.227326498422713,
+      "grad_norm": 0.5013043340465552,
+      "learning_rate": 1.5719246539708536e-05,
+      "loss": 0.3436,
+      "step": 6226
+    },
+    {
+      "epoch": 1.2275236593059937,
+      "grad_norm": 0.4940987339117502,
+      "learning_rate": 1.571797497515603e-05,
+      "loss": 0.3127,
+      "step": 6227
+    },
+    {
+      "epoch": 1.2277208201892744,
+      "grad_norm": 0.5237341159455894,
+      "learning_rate": 1.5716703273224568e-05,
+      "loss": 0.3173,
+      "step": 6228
+    },
+    {
+      "epoch": 1.2279179810725551,
+      "grad_norm": 0.5284887213486391,
+      "learning_rate": 1.5715431433944706e-05,
+      "loss": 0.3326,
+      "step": 6229
+    },
+    {
+      "epoch": 1.228115141955836,
+      "grad_norm": 0.5224757623724461,
+      "learning_rate": 1.5714159457347007e-05,
+      "loss": 0.3254,
+      "step": 6230
+    },
+    {
+      "epoch": 1.2283123028391167,
+      "grad_norm": 0.49876895032485197,
+      "learning_rate": 1.571288734346202e-05,
+      "loss": 0.3423,
+      "step": 6231
+    },
+    {
+      "epoch": 1.2285094637223974,
+      "grad_norm": 0.512553101690978,
+      "learning_rate": 1.5711615092320315e-05,
+      "loss": 0.3414,
+      "step": 6232
+    },
+    {
+      "epoch": 1.2287066246056781,
+      "grad_norm": 0.5014118944270054,
+      "learning_rate": 1.571034270395246e-05,
+      "loss": 0.3183,
+      "step": 6233
+    },
+    {
+      "epoch": 1.228903785488959,
+      "grad_norm": 1.4423480845194543,
+      "learning_rate": 1.570907017838902e-05,
+      "loss": 0.3328,
+      "step": 6234
+    },
+    {
+      "epoch": 1.2291009463722398,
+      "grad_norm": 0.4737177745880879,
+      "learning_rate": 1.5707797515660574e-05,
+      "loss": 0.3193,
+      "step": 6235
+    },
+    {
+      "epoch": 1.2292981072555205,
+      "grad_norm": 0.5036458325808413,
+      "learning_rate": 1.5706524715797693e-05,
+      "loss": 0.3438,
+      "step": 6236
+    },
+    {
+      "epoch": 1.2294952681388012,
+      "grad_norm": 0.5174813077151698,
+      "learning_rate": 1.5705251778830962e-05,
+      "loss": 0.3515,
+      "step": 6237
+    },
+    {
+      "epoch": 1.229692429022082,
+      "grad_norm": 0.4799244634840848,
+      "learning_rate": 1.5703978704790962e-05,
+      "loss": 0.3149,
+      "step": 6238
+    },
+    {
+      "epoch": 1.2298895899053628,
+      "grad_norm": 0.5581993797302612,
+      "learning_rate": 1.5702705493708283e-05,
+      "loss": 0.3553,
+      "step": 6239
+    },
+    {
+      "epoch": 1.2300867507886435,
+      "grad_norm": 0.5064827009129059,
+      "learning_rate": 1.5701432145613508e-05,
+      "loss": 0.3266,
+      "step": 6240
+    },
+    {
+      "epoch": 1.2302839116719242,
+      "grad_norm": 0.6670134919739141,
+      "learning_rate": 1.5700158660537235e-05,
+      "loss": 0.3515,
+      "step": 6241
+    },
+    {
+      "epoch": 1.2304810725552051,
+      "grad_norm": 0.510108854937946,
+      "learning_rate": 1.569888503851006e-05,
+      "loss": 0.3321,
+      "step": 6242
+    },
+    {
+      "epoch": 1.2306782334384858,
+      "grad_norm": 0.5046667999599455,
+      "learning_rate": 1.5697611279562584e-05,
+      "loss": 0.3339,
+      "step": 6243
+    },
+    {
+      "epoch": 1.2308753943217665,
+      "grad_norm": 0.49224188852519396,
+      "learning_rate": 1.5696337383725412e-05,
+      "loss": 0.3119,
+      "step": 6244
+    },
+    {
+      "epoch": 1.2310725552050474,
+      "grad_norm": 0.5172616601262686,
+      "learning_rate": 1.569506335102914e-05,
+      "loss": 0.3443,
+      "step": 6245
+    },
+    {
+      "epoch": 1.2312697160883281,
+      "grad_norm": 0.5191142722710178,
+      "learning_rate": 1.569378918150439e-05,
+      "loss": 0.356,
+      "step": 6246
+    },
+    {
+      "epoch": 1.2314668769716088,
+      "grad_norm": 0.5412878379897871,
+      "learning_rate": 1.5692514875181767e-05,
+      "loss": 0.3716,
+      "step": 6247
+    },
+    {
+      "epoch": 1.2316640378548895,
+      "grad_norm": 0.5465050490209235,
+      "learning_rate": 1.5691240432091892e-05,
+      "loss": 0.3478,
+      "step": 6248
+    },
+    {
+      "epoch": 1.2318611987381702,
+      "grad_norm": 0.4939130381768039,
+      "learning_rate": 1.5689965852265383e-05,
+      "loss": 0.334,
+      "step": 6249
+    },
+    {
+      "epoch": 1.2320583596214512,
+      "grad_norm": 0.5058870629567691,
+      "learning_rate": 1.568869113573286e-05,
+      "loss": 0.336,
+      "step": 6250
+    },
+    {
+      "epoch": 1.2322555205047319,
+      "grad_norm": 0.48700312555624475,
+      "learning_rate": 1.568741628252495e-05,
+      "loss": 0.329,
+      "step": 6251
+    },
+    {
+      "epoch": 1.2324526813880126,
+      "grad_norm": 0.5093770682043642,
+      "learning_rate": 1.5686141292672287e-05,
+      "loss": 0.3498,
+      "step": 6252
+    },
+    {
+      "epoch": 1.2326498422712935,
+      "grad_norm": 0.49687858919236133,
+      "learning_rate": 1.56848661662055e-05,
+      "loss": 0.3263,
+      "step": 6253
+    },
+    {
+      "epoch": 1.2328470031545742,
+      "grad_norm": 0.5077643833980352,
+      "learning_rate": 1.5683590903155222e-05,
+      "loss": 0.3385,
+      "step": 6254
+    },
+    {
+      "epoch": 1.2330441640378549,
+      "grad_norm": 0.5185258515611184,
+      "learning_rate": 1.56823155035521e-05,
+      "loss": 0.3456,
+      "step": 6255
+    },
+    {
+      "epoch": 1.2332413249211356,
+      "grad_norm": 0.5278963932263402,
+      "learning_rate": 1.5681039967426773e-05,
+      "loss": 0.3555,
+      "step": 6256
+    },
+    {
+      "epoch": 1.2334384858044163,
+      "grad_norm": 0.4941475259778851,
+      "learning_rate": 1.5679764294809882e-05,
+      "loss": 0.3451,
+      "step": 6257
+    },
+    {
+      "epoch": 1.2336356466876972,
+      "grad_norm": 0.4917220521203785,
+      "learning_rate": 1.567848848573208e-05,
+      "loss": 0.3115,
+      "step": 6258
+    },
+    {
+      "epoch": 1.233832807570978,
+      "grad_norm": 0.4949766275099023,
+      "learning_rate": 1.567721254022402e-05,
+      "loss": 0.3181,
+      "step": 6259
+    },
+    {
+      "epoch": 1.2340299684542586,
+      "grad_norm": 0.5273568098480352,
+      "learning_rate": 1.5675936458316357e-05,
+      "loss": 0.3399,
+      "step": 6260
+    },
+    {
+      "epoch": 1.2342271293375395,
+      "grad_norm": 0.4886485870140973,
+      "learning_rate": 1.567466024003975e-05,
+      "loss": 0.3564,
+      "step": 6261
+    },
+    {
+      "epoch": 1.2344242902208202,
+      "grad_norm": 0.5030459426717067,
+      "learning_rate": 1.567338388542486e-05,
+      "loss": 0.3335,
+      "step": 6262
+    },
+    {
+      "epoch": 1.234621451104101,
+      "grad_norm": 0.4845019722082114,
+      "learning_rate": 1.5672107394502357e-05,
+      "loss": 0.3083,
+      "step": 6263
+    },
+    {
+      "epoch": 1.2348186119873816,
+      "grad_norm": 0.49034401686565704,
+      "learning_rate": 1.56708307673029e-05,
+      "loss": 0.3406,
+      "step": 6264
+    },
+    {
+      "epoch": 1.2350157728706626,
+      "grad_norm": 0.8790381187259672,
+      "learning_rate": 1.5669554003857172e-05,
+      "loss": 0.3443,
+      "step": 6265
+    },
+    {
+      "epoch": 1.2352129337539433,
+      "grad_norm": 0.4998894425749621,
+      "learning_rate": 1.566827710419584e-05,
+      "loss": 0.335,
+      "step": 6266
+    },
+    {
+      "epoch": 1.235410094637224,
+      "grad_norm": 0.532021790282657,
+      "learning_rate": 1.566700006834959e-05,
+      "loss": 0.3623,
+      "step": 6267
+    },
+    {
+      "epoch": 1.2356072555205047,
+      "grad_norm": 0.5388159940953764,
+      "learning_rate": 1.5665722896349098e-05,
+      "loss": 0.3463,
+      "step": 6268
+    },
+    {
+      "epoch": 1.2358044164037856,
+      "grad_norm": 0.5157145162260288,
+      "learning_rate": 1.566444558822505e-05,
+      "loss": 0.3421,
+      "step": 6269
+    },
+    {
+      "epoch": 1.2360015772870663,
+      "grad_norm": 0.5605990079950705,
+      "learning_rate": 1.5663168144008136e-05,
+      "loss": 0.3561,
+      "step": 6270
+    },
+    {
+      "epoch": 1.236198738170347,
+      "grad_norm": 0.4715933715273965,
+      "learning_rate": 1.5661890563729045e-05,
+      "loss": 0.3115,
+      "step": 6271
+    },
+    {
+      "epoch": 1.2363958990536277,
+      "grad_norm": 0.5133180282336455,
+      "learning_rate": 1.5660612847418476e-05,
+      "loss": 0.3462,
+      "step": 6272
+    },
+    {
+      "epoch": 1.2365930599369086,
+      "grad_norm": 0.5181709969632247,
+      "learning_rate": 1.5659334995107124e-05,
+      "loss": 0.349,
+      "step": 6273
+    },
+    {
+      "epoch": 1.2367902208201893,
+      "grad_norm": 0.5196223017058347,
+      "learning_rate": 1.565805700682569e-05,
+      "loss": 0.3526,
+      "step": 6274
+    },
+    {
+      "epoch": 1.23698738170347,
+      "grad_norm": 0.5375416524081914,
+      "learning_rate": 1.565677888260488e-05,
+      "loss": 0.3509,
+      "step": 6275
+    },
+    {
+      "epoch": 1.2371845425867507,
+      "grad_norm": 0.4661403315888071,
+      "learning_rate": 1.5655500622475405e-05,
+      "loss": 0.2998,
+      "step": 6276
+    },
+    {
+      "epoch": 1.2373817034700316,
+      "grad_norm": 0.5026678272938364,
+      "learning_rate": 1.565422222646797e-05,
+      "loss": 0.316,
+      "step": 6277
+    },
+    {
+      "epoch": 1.2375788643533123,
+      "grad_norm": 0.5370007180074455,
+      "learning_rate": 1.5652943694613293e-05,
+      "loss": 0.3638,
+      "step": 6278
+    },
+    {
+      "epoch": 1.237776025236593,
+      "grad_norm": 0.5161124298507306,
+      "learning_rate": 1.5651665026942094e-05,
+      "loss": 0.3346,
+      "step": 6279
+    },
+    {
+      "epoch": 1.2379731861198737,
+      "grad_norm": 0.5652741095921482,
+      "learning_rate": 1.565038622348509e-05,
+      "loss": 0.3288,
+      "step": 6280
+    },
+    {
+      "epoch": 1.2381703470031546,
+      "grad_norm": 0.9272850647527554,
+      "learning_rate": 1.5649107284273007e-05,
+      "loss": 0.3186,
+      "step": 6281
+    },
+    {
+      "epoch": 1.2383675078864353,
+      "grad_norm": 0.5570913267865101,
+      "learning_rate": 1.5647828209336572e-05,
+      "loss": 0.3537,
+      "step": 6282
+    },
+    {
+      "epoch": 1.238564668769716,
+      "grad_norm": 0.5082360498168839,
+      "learning_rate": 1.5646548998706514e-05,
+      "loss": 0.3099,
+      "step": 6283
+    },
+    {
+      "epoch": 1.2387618296529967,
+      "grad_norm": 0.5205218960702347,
+      "learning_rate": 1.5645269652413574e-05,
+      "loss": 0.306,
+      "step": 6284
+    },
+    {
+      "epoch": 1.2389589905362777,
+      "grad_norm": 0.504848365836123,
+      "learning_rate": 1.564399017048848e-05,
+      "loss": 0.3249,
+      "step": 6285
+    },
+    {
+      "epoch": 1.2391561514195584,
+      "grad_norm": 0.5448082615998162,
+      "learning_rate": 1.5642710552961982e-05,
+      "loss": 0.3271,
+      "step": 6286
+    },
+    {
+      "epoch": 1.239353312302839,
+      "grad_norm": 0.5213986756373774,
+      "learning_rate": 1.564143079986481e-05,
+      "loss": 0.3245,
+      "step": 6287
+    },
+    {
+      "epoch": 1.23955047318612,
+      "grad_norm": 0.5069938937165306,
+      "learning_rate": 1.564015091122773e-05,
+      "loss": 0.3336,
+      "step": 6288
+    },
+    {
+      "epoch": 1.2397476340694007,
+      "grad_norm": 0.5053084261882574,
+      "learning_rate": 1.5638870887081476e-05,
+      "loss": 0.3217,
+      "step": 6289
+    },
+    {
+      "epoch": 1.2399447949526814,
+      "grad_norm": 0.521104683492062,
+      "learning_rate": 1.5637590727456808e-05,
+      "loss": 0.3168,
+      "step": 6290
+    },
+    {
+      "epoch": 1.240141955835962,
+      "grad_norm": 0.5467959912507718,
+      "learning_rate": 1.5636310432384487e-05,
+      "loss": 0.3569,
+      "step": 6291
+    },
+    {
+      "epoch": 1.2403391167192428,
+      "grad_norm": 0.5102807325723169,
+      "learning_rate": 1.5635030001895267e-05,
+      "loss": 0.3149,
+      "step": 6292
+    },
+    {
+      "epoch": 1.2405362776025237,
+      "grad_norm": 1.3045505295401632,
+      "learning_rate": 1.5633749436019913e-05,
+      "loss": 0.3757,
+      "step": 6293
+    },
+    {
+      "epoch": 1.2407334384858044,
+      "grad_norm": 0.5444503336415735,
+      "learning_rate": 1.5632468734789192e-05,
+      "loss": 0.3235,
+      "step": 6294
+    },
+    {
+      "epoch": 1.2409305993690851,
+      "grad_norm": 0.5048607240462756,
+      "learning_rate": 1.563118789823387e-05,
+      "loss": 0.3353,
+      "step": 6295
+    },
+    {
+      "epoch": 1.241127760252366,
+      "grad_norm": 0.5667148895534923,
+      "learning_rate": 1.562990692638473e-05,
+      "loss": 0.3819,
+      "step": 6296
+    },
+    {
+      "epoch": 1.2413249211356467,
+      "grad_norm": 0.5446964393578572,
+      "learning_rate": 1.562862581927254e-05,
+      "loss": 0.3701,
+      "step": 6297
+    },
+    {
+      "epoch": 1.2415220820189274,
+      "grad_norm": 0.5257025409305759,
+      "learning_rate": 1.5627344576928085e-05,
+      "loss": 0.3344,
+      "step": 6298
+    },
+    {
+      "epoch": 1.2417192429022081,
+      "grad_norm": 0.5256923271112383,
+      "learning_rate": 1.5626063199382138e-05,
+      "loss": 0.3574,
+      "step": 6299
+    },
+    {
+      "epoch": 1.2419164037854888,
+      "grad_norm": 0.5316599865695444,
+      "learning_rate": 1.5624781686665498e-05,
+      "loss": 0.3574,
+      "step": 6300
+    },
+    {
+      "epoch": 1.2421135646687698,
+      "grad_norm": 0.5313857364398521,
+      "learning_rate": 1.5623500038808946e-05,
+      "loss": 0.3563,
+      "step": 6301
+    },
+    {
+      "epoch": 1.2423107255520505,
+      "grad_norm": 0.5039351482023608,
+      "learning_rate": 1.5622218255843276e-05,
+      "loss": 0.3031,
+      "step": 6302
+    },
+    {
+      "epoch": 1.2425078864353312,
+      "grad_norm": 0.5072016335647022,
+      "learning_rate": 1.5620936337799287e-05,
+      "loss": 0.3515,
+      "step": 6303
+    },
+    {
+      "epoch": 1.242705047318612,
+      "grad_norm": 0.5672328567093293,
+      "learning_rate": 1.5619654284707773e-05,
+      "loss": 0.314,
+      "step": 6304
+    },
+    {
+      "epoch": 1.2429022082018928,
+      "grad_norm": 0.5399431684544427,
+      "learning_rate": 1.5618372096599547e-05,
+      "loss": 0.3492,
+      "step": 6305
+    },
+    {
+      "epoch": 1.2430993690851735,
+      "grad_norm": 0.49668681766651646,
+      "learning_rate": 1.56170897735054e-05,
+      "loss": 0.3435,
+      "step": 6306
+    },
+    {
+      "epoch": 1.2432965299684542,
+      "grad_norm": 0.5134554384784643,
+      "learning_rate": 1.561580731545615e-05,
+      "loss": 0.3542,
+      "step": 6307
+    },
+    {
+      "epoch": 1.2434936908517351,
+      "grad_norm": 0.507277410037156,
+      "learning_rate": 1.5614524722482604e-05,
+      "loss": 0.3333,
+      "step": 6308
+    },
+    {
+      "epoch": 1.2436908517350158,
+      "grad_norm": 0.5301659849549519,
+      "learning_rate": 1.561324199461558e-05,
+      "loss": 0.3351,
+      "step": 6309
+    },
+    {
+      "epoch": 1.2438880126182965,
+      "grad_norm": 0.4915261347852841,
+      "learning_rate": 1.56119591318859e-05,
+      "loss": 0.3319,
+      "step": 6310
+    },
+    {
+      "epoch": 1.2440851735015772,
+      "grad_norm": 0.5361265111719739,
+      "learning_rate": 1.561067613432438e-05,
+      "loss": 0.3189,
+      "step": 6311
+    },
+    {
+      "epoch": 1.2442823343848581,
+      "grad_norm": 0.5259903970751432,
+      "learning_rate": 1.560939300196185e-05,
+      "loss": 0.3497,
+      "step": 6312
+    },
+    {
+      "epoch": 1.2444794952681388,
+      "grad_norm": 0.5066141181133125,
+      "learning_rate": 1.5608109734829134e-05,
+      "loss": 0.3335,
+      "step": 6313
+    },
+    {
+      "epoch": 1.2446766561514195,
+      "grad_norm": 0.5640167051107872,
+      "learning_rate": 1.5606826332957066e-05,
+      "loss": 0.3371,
+      "step": 6314
+    },
+    {
+      "epoch": 1.2448738170347002,
+      "grad_norm": 0.4993378182654541,
+      "learning_rate": 1.560554279637648e-05,
+      "loss": 0.3393,
+      "step": 6315
+    },
+    {
+      "epoch": 1.2450709779179812,
+      "grad_norm": 0.4864313789939513,
+      "learning_rate": 1.560425912511822e-05,
+      "loss": 0.3251,
+      "step": 6316
+    },
+    {
+      "epoch": 1.2452681388012619,
+      "grad_norm": 0.4908528558657437,
+      "learning_rate": 1.5602975319213115e-05,
+      "loss": 0.3377,
+      "step": 6317
+    },
+    {
+      "epoch": 1.2454652996845426,
+      "grad_norm": 0.5386582667468415,
+      "learning_rate": 1.5601691378692014e-05,
+      "loss": 0.3517,
+      "step": 6318
+    },
+    {
+      "epoch": 1.2456624605678233,
+      "grad_norm": 0.5109315855247596,
+      "learning_rate": 1.5600407303585773e-05,
+      "loss": 0.3195,
+      "step": 6319
+    },
+    {
+      "epoch": 1.2458596214511042,
+      "grad_norm": 0.5286420872063581,
+      "learning_rate": 1.559912309392523e-05,
+      "loss": 0.3531,
+      "step": 6320
+    },
+    {
+      "epoch": 1.2460567823343849,
+      "grad_norm": 0.5019801643055133,
+      "learning_rate": 1.559783874974125e-05,
+      "loss": 0.3223,
+      "step": 6321
+    },
+    {
+      "epoch": 1.2462539432176656,
+      "grad_norm": 0.5106613080428457,
+      "learning_rate": 1.559655427106468e-05,
+      "loss": 0.3308,
+      "step": 6322
+    },
+    {
+      "epoch": 1.2464511041009463,
+      "grad_norm": 0.5460596574557645,
+      "learning_rate": 1.5595269657926396e-05,
+      "loss": 0.3453,
+      "step": 6323
+    },
+    {
+      "epoch": 1.2466482649842272,
+      "grad_norm": 0.5260622334928508,
+      "learning_rate": 1.559398491035725e-05,
+      "loss": 0.3253,
+      "step": 6324
+    },
+    {
+      "epoch": 1.246845425867508,
+      "grad_norm": 0.4969801726967249,
+      "learning_rate": 1.5592700028388107e-05,
+      "loss": 0.3253,
+      "step": 6325
+    },
+    {
+      "epoch": 1.2470425867507886,
+      "grad_norm": 0.4980689481986874,
+      "learning_rate": 1.5591415012049846e-05,
+      "loss": 0.3236,
+      "step": 6326
+    },
+    {
+      "epoch": 1.2472397476340693,
+      "grad_norm": 0.49763176731095426,
+      "learning_rate": 1.5590129861373335e-05,
+      "loss": 0.3378,
+      "step": 6327
+    },
+    {
+      "epoch": 1.2474369085173502,
+      "grad_norm": 0.5169444490816916,
+      "learning_rate": 1.5588844576389454e-05,
+      "loss": 0.3312,
+      "step": 6328
+    },
+    {
+      "epoch": 1.247634069400631,
+      "grad_norm": 0.5109026656737135,
+      "learning_rate": 1.5587559157129078e-05,
+      "loss": 0.333,
+      "step": 6329
+    },
+    {
+      "epoch": 1.2478312302839116,
+      "grad_norm": 0.5055755741491358,
+      "learning_rate": 1.5586273603623098e-05,
+      "loss": 0.3435,
+      "step": 6330
+    },
+    {
+      "epoch": 1.2480283911671926,
+      "grad_norm": 0.5005392793433827,
+      "learning_rate": 1.5584987915902393e-05,
+      "loss": 0.3399,
+      "step": 6331
+    },
+    {
+      "epoch": 1.2482255520504733,
+      "grad_norm": 0.4977560846580881,
+      "learning_rate": 1.5583702093997855e-05,
+      "loss": 0.3299,
+      "step": 6332
+    },
+    {
+      "epoch": 1.248422712933754,
+      "grad_norm": 0.49705232631811735,
+      "learning_rate": 1.558241613794038e-05,
+      "loss": 0.3119,
+      "step": 6333
+    },
+    {
+      "epoch": 1.2486198738170347,
+      "grad_norm": 0.5240628686982806,
+      "learning_rate": 1.5581130047760865e-05,
+      "loss": 0.3357,
+      "step": 6334
+    },
+    {
+      "epoch": 1.2488170347003154,
+      "grad_norm": 0.5317216595261723,
+      "learning_rate": 1.55798438234902e-05,
+      "loss": 0.3368,
+      "step": 6335
+    },
+    {
+      "epoch": 1.2490141955835963,
+      "grad_norm": 0.5388823858020947,
+      "learning_rate": 1.5578557465159296e-05,
+      "loss": 0.3451,
+      "step": 6336
+    },
+    {
+      "epoch": 1.249211356466877,
+      "grad_norm": 0.5020216428038514,
+      "learning_rate": 1.5577270972799056e-05,
+      "loss": 0.3266,
+      "step": 6337
+    },
+    {
+      "epoch": 1.2494085173501577,
+      "grad_norm": 0.5251732240656299,
+      "learning_rate": 1.5575984346440393e-05,
+      "loss": 0.3255,
+      "step": 6338
+    },
+    {
+      "epoch": 1.2496056782334386,
+      "grad_norm": 0.5065709033953211,
+      "learning_rate": 1.5574697586114213e-05,
+      "loss": 0.3441,
+      "step": 6339
+    },
+    {
+      "epoch": 1.2498028391167193,
+      "grad_norm": 0.49243721616902913,
+      "learning_rate": 1.5573410691851432e-05,
+      "loss": 0.3347,
+      "step": 6340
+    },
+    {
+      "epoch": 1.2498028391167193,
+      "eval_loss": 0.43939557671546936,
+      "eval_runtime": 344.8004,
+      "eval_samples_per_second": 23.579,
+      "eval_steps_per_second": 1.476,
+      "step": 6340
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5101061503167154,
+      "learning_rate": 1.5572123663682975e-05,
+      "loss": 0.3102,
+      "step": 6341
+    },
+    {
+      "epoch": 1.2501971608832807,
+      "grad_norm": 0.4808666726594073,
+      "learning_rate": 1.5570836501639754e-05,
+      "loss": 0.3435,
+      "step": 6342
+    },
+    {
+      "epoch": 1.2503943217665614,
+      "grad_norm": 0.4999834817361975,
+      "learning_rate": 1.5569549205752707e-05,
+      "loss": 0.3381,
+      "step": 6343
+    },
+    {
+      "epoch": 1.2505914826498423,
+      "grad_norm": 0.49204094898528117,
+      "learning_rate": 1.556826177605275e-05,
+      "loss": 0.3274,
+      "step": 6344
+    },
+    {
+      "epoch": 1.250788643533123,
+      "grad_norm": 0.4949146499447072,
+      "learning_rate": 1.556697421257082e-05,
+      "loss": 0.3592,
+      "step": 6345
+    },
+    {
+      "epoch": 1.2509858044164037,
+      "grad_norm": 0.47762676196393705,
+      "learning_rate": 1.556568651533785e-05,
+      "loss": 0.322,
+      "step": 6346
+    },
+    {
+      "epoch": 1.2511829652996846,
+      "grad_norm": 0.5191724256820243,
+      "learning_rate": 1.5564398684384787e-05,
+      "loss": 0.3442,
+      "step": 6347
+    },
+    {
+      "epoch": 1.2513801261829653,
+      "grad_norm": 0.4787762214934278,
+      "learning_rate": 1.5563110719742558e-05,
+      "loss": 0.3285,
+      "step": 6348
+    },
+    {
+      "epoch": 1.251577287066246,
+      "grad_norm": 0.48201229976584514,
+      "learning_rate": 1.5561822621442114e-05,
+      "loss": 0.326,
+      "step": 6349
+    },
+    {
+      "epoch": 1.2517744479495267,
+      "grad_norm": 0.46392186774839445,
+      "learning_rate": 1.5560534389514407e-05,
+      "loss": 0.3004,
+      "step": 6350
+    },
+    {
+      "epoch": 1.2519716088328074,
+      "grad_norm": 0.4815291317559283,
+      "learning_rate": 1.555924602399038e-05,
+      "loss": 0.307,
+      "step": 6351
+    },
+    {
+      "epoch": 1.2521687697160884,
+      "grad_norm": 0.8785203455669748,
+      "learning_rate": 1.5557957524900993e-05,
+      "loss": 0.3367,
+      "step": 6352
+    },
+    {
+      "epoch": 1.252365930599369,
+      "grad_norm": 0.48203047338077637,
+      "learning_rate": 1.5556668892277197e-05,
+      "loss": 0.3262,
+      "step": 6353
+    },
+    {
+      "epoch": 1.2525630914826498,
+      "grad_norm": 0.5133633911448251,
+      "learning_rate": 1.555538012614996e-05,
+      "loss": 0.3317,
+      "step": 6354
+    },
+    {
+      "epoch": 1.2527602523659307,
+      "grad_norm": 0.5341940509642474,
+      "learning_rate": 1.555409122655024e-05,
+      "loss": 0.3572,
+      "step": 6355
+    },
+    {
+      "epoch": 1.2529574132492114,
+      "grad_norm": 0.482176622994259,
+      "learning_rate": 1.5552802193509003e-05,
+      "loss": 0.3147,
+      "step": 6356
+    },
+    {
+      "epoch": 1.253154574132492,
+      "grad_norm": 0.49374641797023383,
+      "learning_rate": 1.5551513027057225e-05,
+      "loss": 0.3481,
+      "step": 6357
+    },
+    {
+      "epoch": 1.2533517350157728,
+      "grad_norm": 0.46859298112273634,
+      "learning_rate": 1.5550223727225875e-05,
+      "loss": 0.3099,
+      "step": 6358
+    },
+    {
+      "epoch": 1.2535488958990535,
+      "grad_norm": 0.4986948734759226,
+      "learning_rate": 1.554893429404593e-05,
+      "loss": 0.3068,
+      "step": 6359
+    },
+    {
+      "epoch": 1.2537460567823344,
+      "grad_norm": 0.49945204691498807,
+      "learning_rate": 1.5547644727548373e-05,
+      "loss": 0.3256,
+      "step": 6360
+    },
+    {
+      "epoch": 1.2539432176656151,
+      "grad_norm": 0.5162525662741562,
+      "learning_rate": 1.554635502776418e-05,
+      "loss": 0.3326,
+      "step": 6361
+    },
+    {
+      "epoch": 1.2541403785488958,
+      "grad_norm": 0.534638355896996,
+      "learning_rate": 1.554506519472434e-05,
+      "loss": 0.3336,
+      "step": 6362
+    },
+    {
+      "epoch": 1.2543375394321767,
+      "grad_norm": 0.5029286598031089,
+      "learning_rate": 1.5543775228459846e-05,
+      "loss": 0.3516,
+      "step": 6363
+    },
+    {
+      "epoch": 1.2545347003154574,
+      "grad_norm": 0.5261189540493615,
+      "learning_rate": 1.554248512900169e-05,
+      "loss": 0.328,
+      "step": 6364
+    },
+    {
+      "epoch": 1.2547318611987381,
+      "grad_norm": 0.8234311010835397,
+      "learning_rate": 1.5541194896380863e-05,
+      "loss": 0.3409,
+      "step": 6365
+    },
+    {
+      "epoch": 1.254929022082019,
+      "grad_norm": 0.5154543512907336,
+      "learning_rate": 1.5539904530628365e-05,
+      "loss": 0.3431,
+      "step": 6366
+    },
+    {
+      "epoch": 1.2551261829652998,
+      "grad_norm": 0.478219086021865,
+      "learning_rate": 1.55386140317752e-05,
+      "loss": 0.3156,
+      "step": 6367
+    },
+    {
+      "epoch": 1.2553233438485805,
+      "grad_norm": 0.49818756625431365,
+      "learning_rate": 1.5537323399852373e-05,
+      "loss": 0.3157,
+      "step": 6368
+    },
+    {
+      "epoch": 1.2555205047318612,
+      "grad_norm": 0.47906698592620167,
+      "learning_rate": 1.5536032634890892e-05,
+      "loss": 0.3299,
+      "step": 6369
+    },
+    {
+      "epoch": 1.2557176656151419,
+      "grad_norm": 0.5062243330914792,
+      "learning_rate": 1.553474173692177e-05,
+      "loss": 0.3484,
+      "step": 6370
+    },
+    {
+      "epoch": 1.2559148264984228,
+      "grad_norm": 0.5307599529911612,
+      "learning_rate": 1.5533450705976018e-05,
+      "loss": 0.3385,
+      "step": 6371
+    },
+    {
+      "epoch": 1.2561119873817035,
+      "grad_norm": 0.573572643755291,
+      "learning_rate": 1.553215954208466e-05,
+      "loss": 0.357,
+      "step": 6372
+    },
+    {
+      "epoch": 1.2563091482649842,
+      "grad_norm": 0.5968608411513199,
+      "learning_rate": 1.5530868245278708e-05,
+      "loss": 0.3489,
+      "step": 6373
+    },
+    {
+      "epoch": 1.256506309148265,
+      "grad_norm": 0.58115296279192,
+      "learning_rate": 1.55295768155892e-05,
+      "loss": 0.3522,
+      "step": 6374
+    },
+    {
+      "epoch": 1.2567034700315458,
+      "grad_norm": 0.49820315714499325,
+      "learning_rate": 1.5528285253047153e-05,
+      "loss": 0.3303,
+      "step": 6375
+    },
+    {
+      "epoch": 1.2569006309148265,
+      "grad_norm": 0.5089032372495764,
+      "learning_rate": 1.55269935576836e-05,
+      "loss": 0.344,
+      "step": 6376
+    },
+    {
+      "epoch": 1.2570977917981072,
+      "grad_norm": 0.5371768444531454,
+      "learning_rate": 1.5525701729529578e-05,
+      "loss": 0.3675,
+      "step": 6377
+    },
+    {
+      "epoch": 1.257294952681388,
+      "grad_norm": 0.5002102537095656,
+      "learning_rate": 1.552440976861612e-05,
+      "loss": 0.3248,
+      "step": 6378
+    },
+    {
+      "epoch": 1.2574921135646688,
+      "grad_norm": 0.5018100903078823,
+      "learning_rate": 1.5523117674974267e-05,
+      "loss": 0.3217,
+      "step": 6379
+    },
+    {
+      "epoch": 1.2576892744479495,
+      "grad_norm": 0.5336700648969083,
+      "learning_rate": 1.5521825448635066e-05,
+      "loss": 0.3441,
+      "step": 6380
+    },
+    {
+      "epoch": 1.2578864353312302,
+      "grad_norm": 0.48938739876612136,
+      "learning_rate": 1.5520533089629562e-05,
+      "loss": 0.328,
+      "step": 6381
+    },
+    {
+      "epoch": 1.2580835962145112,
+      "grad_norm": 0.49246018639601746,
+      "learning_rate": 1.5519240597988806e-05,
+      "loss": 0.3418,
+      "step": 6382
+    },
+    {
+      "epoch": 1.2582807570977919,
+      "grad_norm": 0.49124354097385314,
+      "learning_rate": 1.551794797374385e-05,
+      "loss": 0.3294,
+      "step": 6383
+    },
+    {
+      "epoch": 1.2584779179810726,
+      "grad_norm": 0.5583036731814169,
+      "learning_rate": 1.5516655216925748e-05,
+      "loss": 0.3576,
+      "step": 6384
+    },
+    {
+      "epoch": 1.2586750788643533,
+      "grad_norm": 0.5097818791868225,
+      "learning_rate": 1.5515362327565564e-05,
+      "loss": 0.3372,
+      "step": 6385
+    },
+    {
+      "epoch": 1.258872239747634,
+      "grad_norm": 0.5176295462370966,
+      "learning_rate": 1.5514069305694356e-05,
+      "loss": 0.3537,
+      "step": 6386
+    },
+    {
+      "epoch": 1.2590694006309149,
+      "grad_norm": 0.5797974728770534,
+      "learning_rate": 1.5512776151343198e-05,
+      "loss": 0.3595,
+      "step": 6387
+    },
+    {
+      "epoch": 1.2592665615141956,
+      "grad_norm": 0.6262421558004105,
+      "learning_rate": 1.5511482864543147e-05,
+      "loss": 0.3253,
+      "step": 6388
+    },
+    {
+      "epoch": 1.2594637223974763,
+      "grad_norm": 0.5273042130470622,
+      "learning_rate": 1.5510189445325284e-05,
+      "loss": 0.3221,
+      "step": 6389
+    },
+    {
+      "epoch": 1.2596608832807572,
+      "grad_norm": 0.5732125441797801,
+      "learning_rate": 1.5508895893720685e-05,
+      "loss": 0.3532,
+      "step": 6390
+    },
+    {
+      "epoch": 1.259858044164038,
+      "grad_norm": 0.5275542533558232,
+      "learning_rate": 1.550760220976042e-05,
+      "loss": 0.356,
+      "step": 6391
+    },
+    {
+      "epoch": 1.2600552050473186,
+      "grad_norm": 0.4857065591782998,
+      "learning_rate": 1.5506308393475582e-05,
+      "loss": 0.3308,
+      "step": 6392
+    },
+    {
+      "epoch": 1.2602523659305993,
+      "grad_norm": 0.5099413329942116,
+      "learning_rate": 1.550501444489725e-05,
+      "loss": 0.3362,
+      "step": 6393
+    },
+    {
+      "epoch": 1.26044952681388,
+      "grad_norm": 0.5056333274989698,
+      "learning_rate": 1.5503720364056512e-05,
+      "loss": 0.3489,
+      "step": 6394
+    },
+    {
+      "epoch": 1.260646687697161,
+      "grad_norm": 0.4774413513695629,
+      "learning_rate": 1.550242615098446e-05,
+      "loss": 0.3304,
+      "step": 6395
+    },
+    {
+      "epoch": 1.2608438485804416,
+      "grad_norm": 0.4843542395552344,
+      "learning_rate": 1.5501131805712188e-05,
+      "loss": 0.3047,
+      "step": 6396
+    },
+    {
+      "epoch": 1.2610410094637223,
+      "grad_norm": 0.4843638404403003,
+      "learning_rate": 1.549983732827079e-05,
+      "loss": 0.319,
+      "step": 6397
+    },
+    {
+      "epoch": 1.2612381703470033,
+      "grad_norm": 0.4783957926301324,
+      "learning_rate": 1.5498542718691378e-05,
+      "loss": 0.2986,
+      "step": 6398
+    },
+    {
+      "epoch": 1.261435331230284,
+      "grad_norm": 0.5171370998231292,
+      "learning_rate": 1.5497247977005047e-05,
+      "loss": 0.3288,
+      "step": 6399
+    },
+    {
+      "epoch": 1.2616324921135647,
+      "grad_norm": 0.5587928424642727,
+      "learning_rate": 1.5495953103242908e-05,
+      "loss": 0.3409,
+      "step": 6400
+    },
+    {
+      "epoch": 1.2618296529968454,
+      "grad_norm": 0.5069107019620349,
+      "learning_rate": 1.549465809743607e-05,
+      "loss": 0.3489,
+      "step": 6401
+    },
+    {
+      "epoch": 1.262026813880126,
+      "grad_norm": 0.5048904621722605,
+      "learning_rate": 1.5493362959615646e-05,
+      "loss": 0.3363,
+      "step": 6402
+    },
+    {
+      "epoch": 1.262223974763407,
+      "grad_norm": 0.4918922172890412,
+      "learning_rate": 1.549206768981275e-05,
+      "loss": 0.3215,
+      "step": 6403
+    },
+    {
+      "epoch": 1.2624211356466877,
+      "grad_norm": 0.5373254295839985,
+      "learning_rate": 1.5490772288058508e-05,
+      "loss": 0.3351,
+      "step": 6404
+    },
+    {
+      "epoch": 1.2626182965299684,
+      "grad_norm": 0.5194441534825306,
+      "learning_rate": 1.5489476754384035e-05,
+      "loss": 0.3395,
+      "step": 6405
+    },
+    {
+      "epoch": 1.2628154574132493,
+      "grad_norm": 0.49136904473331555,
+      "learning_rate": 1.5488181088820468e-05,
+      "loss": 0.3144,
+      "step": 6406
+    },
+    {
+      "epoch": 1.26301261829653,
+      "grad_norm": 0.5815486766369652,
+      "learning_rate": 1.548688529139893e-05,
+      "loss": 0.3718,
+      "step": 6407
+    },
+    {
+      "epoch": 1.2632097791798107,
+      "grad_norm": 0.49216042299624607,
+      "learning_rate": 1.5485589362150552e-05,
+      "loss": 0.3384,
+      "step": 6408
+    },
+    {
+      "epoch": 1.2634069400630916,
+      "grad_norm": 0.49257478050805037,
+      "learning_rate": 1.5484293301106475e-05,
+      "loss": 0.3161,
+      "step": 6409
+    },
+    {
+      "epoch": 1.2636041009463723,
+      "grad_norm": 0.5111418677883345,
+      "learning_rate": 1.5482997108297834e-05,
+      "loss": 0.3159,
+      "step": 6410
+    },
+    {
+      "epoch": 1.263801261829653,
+      "grad_norm": 0.48719719419588625,
+      "learning_rate": 1.5481700783755772e-05,
+      "loss": 0.31,
+      "step": 6411
+    },
+    {
+      "epoch": 1.2639984227129337,
+      "grad_norm": 0.6071472420270517,
+      "learning_rate": 1.548040432751143e-05,
+      "loss": 0.3906,
+      "step": 6412
+    },
+    {
+      "epoch": 1.2641955835962144,
+      "grad_norm": 0.5109344409789744,
+      "learning_rate": 1.5479107739595967e-05,
+      "loss": 0.3505,
+      "step": 6413
+    },
+    {
+      "epoch": 1.2643927444794953,
+      "grad_norm": 0.5163073576362812,
+      "learning_rate": 1.5477811020040525e-05,
+      "loss": 0.3396,
+      "step": 6414
+    },
+    {
+      "epoch": 1.264589905362776,
+      "grad_norm": 0.5159856675009115,
+      "learning_rate": 1.5476514168876264e-05,
+      "loss": 0.3458,
+      "step": 6415
+    },
+    {
+      "epoch": 1.2647870662460567,
+      "grad_norm": 0.5208996581565274,
+      "learning_rate": 1.5475217186134335e-05,
+      "loss": 0.3575,
+      "step": 6416
+    },
+    {
+      "epoch": 1.2649842271293377,
+      "grad_norm": 0.6501488219759475,
+      "learning_rate": 1.5473920071845906e-05,
+      "loss": 0.3406,
+      "step": 6417
+    },
+    {
+      "epoch": 1.2651813880126184,
+      "grad_norm": 0.5045419985244192,
+      "learning_rate": 1.5472622826042144e-05,
+      "loss": 0.3216,
+      "step": 6418
+    },
+    {
+      "epoch": 1.265378548895899,
+      "grad_norm": 0.5083528685308321,
+      "learning_rate": 1.5471325448754207e-05,
+      "loss": 0.3485,
+      "step": 6419
+    },
+    {
+      "epoch": 1.2655757097791798,
+      "grad_norm": 0.4756731932644996,
+      "learning_rate": 1.547002794001327e-05,
+      "loss": 0.3249,
+      "step": 6420
+    },
+    {
+      "epoch": 1.2657728706624605,
+      "grad_norm": 0.5179143417797316,
+      "learning_rate": 1.546873029985051e-05,
+      "loss": 0.3634,
+      "step": 6421
+    },
+    {
+      "epoch": 1.2659700315457414,
+      "grad_norm": 0.5165453574064824,
+      "learning_rate": 1.54674325282971e-05,
+      "loss": 0.3135,
+      "step": 6422
+    },
+    {
+      "epoch": 1.266167192429022,
+      "grad_norm": 0.4897247296838715,
+      "learning_rate": 1.5466134625384216e-05,
+      "loss": 0.3222,
+      "step": 6423
+    },
+    {
+      "epoch": 1.2663643533123028,
+      "grad_norm": 0.5067798904225712,
+      "learning_rate": 1.546483659114305e-05,
+      "loss": 0.3574,
+      "step": 6424
+    },
+    {
+      "epoch": 1.2665615141955837,
+      "grad_norm": 0.4586524801850368,
+      "learning_rate": 1.5463538425604782e-05,
+      "loss": 0.3044,
+      "step": 6425
+    },
+    {
+      "epoch": 1.2667586750788644,
+      "grad_norm": 0.48557837904062384,
+      "learning_rate": 1.5462240128800604e-05,
+      "loss": 0.3395,
+      "step": 6426
+    },
+    {
+      "epoch": 1.2669558359621451,
+      "grad_norm": 0.4999102108308707,
+      "learning_rate": 1.5460941700761706e-05,
+      "loss": 0.3213,
+      "step": 6427
+    },
+    {
+      "epoch": 1.2671529968454258,
+      "grad_norm": 0.5132500026273736,
+      "learning_rate": 1.545964314151929e-05,
+      "loss": 0.3422,
+      "step": 6428
+    },
+    {
+      "epoch": 1.2673501577287065,
+      "grad_norm": 0.4913356576984297,
+      "learning_rate": 1.545834445110455e-05,
+      "loss": 0.3383,
+      "step": 6429
+    },
+    {
+      "epoch": 1.2675473186119874,
+      "grad_norm": 0.5025975016108518,
+      "learning_rate": 1.5457045629548687e-05,
+      "loss": 0.3397,
+      "step": 6430
+    },
+    {
+      "epoch": 1.2677444794952681,
+      "grad_norm": 0.4826877165006056,
+      "learning_rate": 1.545574667688291e-05,
+      "loss": 0.3151,
+      "step": 6431
+    },
+    {
+      "epoch": 1.2679416403785488,
+      "grad_norm": 0.4772272594998397,
+      "learning_rate": 1.5454447593138424e-05,
+      "loss": 0.3349,
+      "step": 6432
+    },
+    {
+      "epoch": 1.2681388012618298,
+      "grad_norm": 2.53849935570129,
+      "learning_rate": 1.5453148378346444e-05,
+      "loss": 0.3515,
+      "step": 6433
+    },
+    {
+      "epoch": 1.2683359621451105,
+      "grad_norm": 0.5128594643909096,
+      "learning_rate": 1.5451849032538185e-05,
+      "loss": 0.3419,
+      "step": 6434
+    },
+    {
+      "epoch": 1.2685331230283912,
+      "grad_norm": 0.493655568948996,
+      "learning_rate": 1.5450549555744857e-05,
+      "loss": 0.3182,
+      "step": 6435
+    },
+    {
+      "epoch": 1.2687302839116719,
+      "grad_norm": 0.4969357386584924,
+      "learning_rate": 1.5449249947997687e-05,
+      "loss": 0.3429,
+      "step": 6436
+    },
+    {
+      "epoch": 1.2689274447949526,
+      "grad_norm": 0.5468340570267272,
+      "learning_rate": 1.5447950209327905e-05,
+      "loss": 0.3493,
+      "step": 6437
+    },
+    {
+      "epoch": 1.2691246056782335,
+      "grad_norm": 0.5418093845032692,
+      "learning_rate": 1.5446650339766723e-05,
+      "loss": 0.3476,
+      "step": 6438
+    },
+    {
+      "epoch": 1.2693217665615142,
+      "grad_norm": 0.6775784568233131,
+      "learning_rate": 1.544535033934539e-05,
+      "loss": 0.3739,
+      "step": 6439
+    },
+    {
+      "epoch": 1.2695189274447949,
+      "grad_norm": 0.48385467083683753,
+      "learning_rate": 1.5444050208095124e-05,
+      "loss": 0.3195,
+      "step": 6440
+    },
+    {
+      "epoch": 1.2697160883280758,
+      "grad_norm": 0.493918698352294,
+      "learning_rate": 1.544274994604717e-05,
+      "loss": 0.3302,
+      "step": 6441
+    },
+    {
+      "epoch": 1.2699132492113565,
+      "grad_norm": 0.5530138835909129,
+      "learning_rate": 1.5441449553232764e-05,
+      "loss": 0.3495,
+      "step": 6442
+    },
+    {
+      "epoch": 1.2701104100946372,
+      "grad_norm": 0.49621433371123225,
+      "learning_rate": 1.544014902968315e-05,
+      "loss": 0.3333,
+      "step": 6443
+    },
+    {
+      "epoch": 1.270307570977918,
+      "grad_norm": 0.49721837630990706,
+      "learning_rate": 1.5438848375429576e-05,
+      "loss": 0.3283,
+      "step": 6444
+    },
+    {
+      "epoch": 1.2705047318611986,
+      "grad_norm": 0.5533506511752488,
+      "learning_rate": 1.5437547590503288e-05,
+      "loss": 0.385,
+      "step": 6445
+    },
+    {
+      "epoch": 1.2707018927444795,
+      "grad_norm": 0.5283600379328754,
+      "learning_rate": 1.5436246674935543e-05,
+      "loss": 0.3434,
+      "step": 6446
+    },
+    {
+      "epoch": 1.2708990536277602,
+      "grad_norm": 0.5013523816913494,
+      "learning_rate": 1.5434945628757595e-05,
+      "loss": 0.338,
+      "step": 6447
+    },
+    {
+      "epoch": 1.271096214511041,
+      "grad_norm": 0.4905233287935434,
+      "learning_rate": 1.54336444520007e-05,
+      "loss": 0.3287,
+      "step": 6448
+    },
+    {
+      "epoch": 1.2712933753943219,
+      "grad_norm": 0.5437498180898698,
+      "learning_rate": 1.5432343144696117e-05,
+      "loss": 0.3708,
+      "step": 6449
+    },
+    {
+      "epoch": 1.2714905362776026,
+      "grad_norm": 0.48421826633903103,
+      "learning_rate": 1.543104170687512e-05,
+      "loss": 0.3125,
+      "step": 6450
+    },
+    {
+      "epoch": 1.2716876971608833,
+      "grad_norm": 0.4793437101259274,
+      "learning_rate": 1.542974013856897e-05,
+      "loss": 0.3341,
+      "step": 6451
+    },
+    {
+      "epoch": 1.271884858044164,
+      "grad_norm": 19.93851850988607,
+      "learning_rate": 1.5428438439808942e-05,
+      "loss": 0.6517,
+      "step": 6452
+    },
+    {
+      "epoch": 1.2720820189274447,
+      "grad_norm": 0.5087943803471097,
+      "learning_rate": 1.542713661062631e-05,
+      "loss": 0.3086,
+      "step": 6453
+    },
+    {
+      "epoch": 1.2722791798107256,
+      "grad_norm": 0.47671813979964683,
+      "learning_rate": 1.542583465105235e-05,
+      "loss": 0.3208,
+      "step": 6454
+    },
+    {
+      "epoch": 1.2724763406940063,
+      "grad_norm": 0.5236254976524284,
+      "learning_rate": 1.542453256111834e-05,
+      "loss": 0.3477,
+      "step": 6455
+    },
+    {
+      "epoch": 1.272673501577287,
+      "grad_norm": 0.49008247823279333,
+      "learning_rate": 1.5423230340855572e-05,
+      "loss": 0.3354,
+      "step": 6456
+    },
+    {
+      "epoch": 1.272870662460568,
+      "grad_norm": 0.48169835947954537,
+      "learning_rate": 1.5421927990295325e-05,
+      "loss": 0.3369,
+      "step": 6457
+    },
+    {
+      "epoch": 1.2730678233438486,
+      "grad_norm": 0.5245321000278956,
+      "learning_rate": 1.5420625509468892e-05,
+      "loss": 0.3466,
+      "step": 6458
+    },
+    {
+      "epoch": 1.2732649842271293,
+      "grad_norm": 0.6897513827097659,
+      "learning_rate": 1.5419322898407562e-05,
+      "loss": 0.3384,
+      "step": 6459
+    },
+    {
+      "epoch": 1.2734621451104102,
+      "grad_norm": 0.5001285523422395,
+      "learning_rate": 1.541802015714264e-05,
+      "loss": 0.3512,
+      "step": 6460
+    },
+    {
+      "epoch": 1.273659305993691,
+      "grad_norm": 0.5336358617640525,
+      "learning_rate": 1.5416717285705417e-05,
+      "loss": 0.3448,
+      "step": 6461
+    },
+    {
+      "epoch": 1.2738564668769716,
+      "grad_norm": 0.5466334074547992,
+      "learning_rate": 1.5415414284127207e-05,
+      "loss": 0.3506,
+      "step": 6462
+    },
+    {
+      "epoch": 1.2740536277602523,
+      "grad_norm": 0.526166079775761,
+      "learning_rate": 1.54141111524393e-05,
+      "loss": 0.3268,
+      "step": 6463
+    },
+    {
+      "epoch": 1.274250788643533,
+      "grad_norm": 0.47399584750409807,
+      "learning_rate": 1.5412807890673015e-05,
+      "loss": 0.3079,
+      "step": 6464
+    },
+    {
+      "epoch": 1.274447949526814,
+      "grad_norm": 0.505741406230065,
+      "learning_rate": 1.541150449885966e-05,
+      "loss": 0.3296,
+      "step": 6465
+    },
+    {
+      "epoch": 1.2746451104100947,
+      "grad_norm": 0.5386050119365994,
+      "learning_rate": 1.5410200977030553e-05,
+      "loss": 0.3652,
+      "step": 6466
+    },
+    {
+      "epoch": 1.2748422712933754,
+      "grad_norm": 0.47587895585088,
+      "learning_rate": 1.5408897325217012e-05,
+      "loss": 0.3239,
+      "step": 6467
+    },
+    {
+      "epoch": 1.2750394321766563,
+      "grad_norm": 0.504126225675362,
+      "learning_rate": 1.5407593543450358e-05,
+      "loss": 0.3362,
+      "step": 6468
+    },
+    {
+      "epoch": 1.275236593059937,
+      "grad_norm": 0.485501366242255,
+      "learning_rate": 1.540628963176191e-05,
+      "loss": 0.3153,
+      "step": 6469
+    },
+    {
+      "epoch": 1.2754337539432177,
+      "grad_norm": 0.49475043227065524,
+      "learning_rate": 1.5404985590183e-05,
+      "loss": 0.3079,
+      "step": 6470
+    },
+    {
+      "epoch": 1.2756309148264984,
+      "grad_norm": 0.4753091395271322,
+      "learning_rate": 1.5403681418744962e-05,
+      "loss": 0.3128,
+      "step": 6471
+    },
+    {
+      "epoch": 1.275828075709779,
+      "grad_norm": 0.5027088286339224,
+      "learning_rate": 1.5402377117479127e-05,
+      "loss": 0.3438,
+      "step": 6472
+    },
+    {
+      "epoch": 1.27602523659306,
+      "grad_norm": 0.4918844457862931,
+      "learning_rate": 1.5401072686416826e-05,
+      "loss": 0.3433,
+      "step": 6473
+    },
+    {
+      "epoch": 1.2762223974763407,
+      "grad_norm": 0.5104026241302757,
+      "learning_rate": 1.539976812558941e-05,
+      "loss": 0.3238,
+      "step": 6474
+    },
+    {
+      "epoch": 1.2764195583596214,
+      "grad_norm": 0.5129479397583762,
+      "learning_rate": 1.539846343502821e-05,
+      "loss": 0.3521,
+      "step": 6475
+    },
+    {
+      "epoch": 1.2766167192429023,
+      "grad_norm": 0.5917550936700113,
+      "learning_rate": 1.5397158614764584e-05,
+      "loss": 0.3882,
+      "step": 6476
+    },
+    {
+      "epoch": 1.276813880126183,
+      "grad_norm": 0.5595087268305293,
+      "learning_rate": 1.5395853664829876e-05,
+      "loss": 0.3397,
+      "step": 6477
+    },
+    {
+      "epoch": 1.2770110410094637,
+      "grad_norm": 0.4600650571043877,
+      "learning_rate": 1.5394548585255437e-05,
+      "loss": 0.3025,
+      "step": 6478
+    },
+    {
+      "epoch": 1.2772082018927444,
+      "grad_norm": 0.49890829885179133,
+      "learning_rate": 1.5393243376072625e-05,
+      "loss": 0.3205,
+      "step": 6479
+    },
+    {
+      "epoch": 1.2774053627760251,
+      "grad_norm": 0.4890558758581133,
+      "learning_rate": 1.5391938037312795e-05,
+      "loss": 0.3315,
+      "step": 6480
+    },
+    {
+      "epoch": 1.277602523659306,
+      "grad_norm": 0.7150885015408796,
+      "learning_rate": 1.5390632569007314e-05,
+      "loss": 0.3116,
+      "step": 6481
+    },
+    {
+      "epoch": 1.2777996845425867,
+      "grad_norm": 0.5271705955690947,
+      "learning_rate": 1.5389326971187543e-05,
+      "loss": 0.3453,
+      "step": 6482
+    },
+    {
+      "epoch": 1.2779968454258674,
+      "grad_norm": 0.508576676488475,
+      "learning_rate": 1.538802124388485e-05,
+      "loss": 0.3424,
+      "step": 6483
+    },
+    {
+      "epoch": 1.2781940063091484,
+      "grad_norm": 0.5597504943231402,
+      "learning_rate": 1.538671538713061e-05,
+      "loss": 0.3741,
+      "step": 6484
+    },
+    {
+      "epoch": 1.278391167192429,
+      "grad_norm": 0.5144168371604326,
+      "learning_rate": 1.5385409400956196e-05,
+      "loss": 0.3434,
+      "step": 6485
+    },
+    {
+      "epoch": 1.2785883280757098,
+      "grad_norm": 0.5022134043659201,
+      "learning_rate": 1.538410328539298e-05,
+      "loss": 0.3229,
+      "step": 6486
+    },
+    {
+      "epoch": 1.2787854889589905,
+      "grad_norm": 0.5129040379585406,
+      "learning_rate": 1.5382797040472352e-05,
+      "loss": 0.326,
+      "step": 6487
+    },
+    {
+      "epoch": 1.2789826498422712,
+      "grad_norm": 0.49329203335900706,
+      "learning_rate": 1.538149066622569e-05,
+      "loss": 0.3395,
+      "step": 6488
+    },
+    {
+      "epoch": 1.279179810725552,
+      "grad_norm": 0.5055724718310393,
+      "learning_rate": 1.538018416268438e-05,
+      "loss": 0.3472,
+      "step": 6489
+    },
+    {
+      "epoch": 1.2793769716088328,
+      "grad_norm": 0.5258954355075676,
+      "learning_rate": 1.537887752987981e-05,
+      "loss": 0.3538,
+      "step": 6490
+    },
+    {
+      "epoch": 1.2795741324921135,
+      "grad_norm": 0.542150664111908,
+      "learning_rate": 1.5377570767843377e-05,
+      "loss": 0.3396,
+      "step": 6491
+    },
+    {
+      "epoch": 1.2797712933753944,
+      "grad_norm": 0.47748206124296466,
+      "learning_rate": 1.5376263876606475e-05,
+      "loss": 0.3288,
+      "step": 6492
+    },
+    {
+      "epoch": 1.2799684542586751,
+      "grad_norm": 3.411379492061442,
+      "learning_rate": 1.5374956856200504e-05,
+      "loss": 0.33,
+      "step": 6493
+    },
+    {
+      "epoch": 1.2801656151419558,
+      "grad_norm": 0.5132738135149726,
+      "learning_rate": 1.537364970665687e-05,
+      "loss": 0.3298,
+      "step": 6494
+    },
+    {
+      "epoch": 1.2803627760252365,
+      "grad_norm": 1.1343123502825927,
+      "learning_rate": 1.537234242800697e-05,
+      "loss": 0.336,
+      "step": 6495
+    },
+    {
+      "epoch": 1.2805599369085172,
+      "grad_norm": 0.518109994618307,
+      "learning_rate": 1.537103502028222e-05,
+      "loss": 0.3338,
+      "step": 6496
+    },
+    {
+      "epoch": 1.2807570977917981,
+      "grad_norm": 0.5142434953565715,
+      "learning_rate": 1.5369727483514026e-05,
+      "loss": 0.3498,
+      "step": 6497
+    },
+    {
+      "epoch": 1.2809542586750788,
+      "grad_norm": 0.4857243954146504,
+      "learning_rate": 1.536841981773381e-05,
+      "loss": 0.3212,
+      "step": 6498
+    },
+    {
+      "epoch": 1.2811514195583595,
+      "grad_norm": 0.5583875828356074,
+      "learning_rate": 1.5367112022972977e-05,
+      "loss": 0.3449,
+      "step": 6499
+    },
+    {
+      "epoch": 1.2813485804416405,
+      "grad_norm": 0.7616833629256355,
+      "learning_rate": 1.536580409926296e-05,
+      "loss": 0.3047,
+      "step": 6500
+    },
+    {
+      "epoch": 1.2815457413249212,
+      "grad_norm": 0.5422560861681756,
+      "learning_rate": 1.5364496046635175e-05,
+      "loss": 0.3336,
+      "step": 6501
+    },
+    {
+      "epoch": 1.2817429022082019,
+      "grad_norm": 0.5234629592906149,
+      "learning_rate": 1.5363187865121058e-05,
+      "loss": 0.3207,
+      "step": 6502
+    },
+    {
+      "epoch": 1.2819400630914828,
+      "grad_norm": 0.5788807821840986,
+      "learning_rate": 1.5361879554752027e-05,
+      "loss": 0.3518,
+      "step": 6503
+    },
+    {
+      "epoch": 1.2821372239747635,
+      "grad_norm": 0.560152361635895,
+      "learning_rate": 1.536057111555953e-05,
+      "loss": 0.3671,
+      "step": 6504
+    },
+    {
+      "epoch": 1.2823343848580442,
+      "grad_norm": 0.573760389051277,
+      "learning_rate": 1.5359262547574986e-05,
+      "loss": 0.3296,
+      "step": 6505
+    },
+    {
+      "epoch": 1.2825315457413249,
+      "grad_norm": 0.639852449462523,
+      "learning_rate": 1.535795385082985e-05,
+      "loss": 0.3648,
+      "step": 6506
+    },
+    {
+      "epoch": 1.2827287066246056,
+      "grad_norm": 0.5684843971461204,
+      "learning_rate": 1.5356645025355556e-05,
+      "loss": 0.3628,
+      "step": 6507
+    },
+    {
+      "epoch": 1.2829258675078865,
+      "grad_norm": 0.48085580385372134,
+      "learning_rate": 1.535533607118355e-05,
+      "loss": 0.3272,
+      "step": 6508
+    },
+    {
+      "epoch": 1.2831230283911672,
+      "grad_norm": 0.5728460211410128,
+      "learning_rate": 1.5354026988345284e-05,
+      "loss": 0.36,
+      "step": 6509
+    },
+    {
+      "epoch": 1.283320189274448,
+      "grad_norm": 0.5181581225517601,
+      "learning_rate": 1.5352717776872208e-05,
+      "loss": 0.3219,
+      "step": 6510
+    },
+    {
+      "epoch": 1.2835173501577288,
+      "grad_norm": 0.5081886354156057,
+      "learning_rate": 1.5351408436795777e-05,
+      "loss": 0.3369,
+      "step": 6511
+    },
+    {
+      "epoch": 1.2837145110410095,
+      "grad_norm": 0.5230355727747609,
+      "learning_rate": 1.535009896814745e-05,
+      "loss": 0.3244,
+      "step": 6512
+    },
+    {
+      "epoch": 1.2839116719242902,
+      "grad_norm": 0.5375516631231971,
+      "learning_rate": 1.5348789370958687e-05,
+      "loss": 0.3419,
+      "step": 6513
+    },
+    {
+      "epoch": 1.284108832807571,
+      "grad_norm": 0.5618743671777963,
+      "learning_rate": 1.534747964526095e-05,
+      "loss": 0.3321,
+      "step": 6514
+    },
+    {
+      "epoch": 1.2843059936908516,
+      "grad_norm": 0.5648952032025825,
+      "learning_rate": 1.5346169791085707e-05,
+      "loss": 0.3586,
+      "step": 6515
+    },
+    {
+      "epoch": 1.2845031545741326,
+      "grad_norm": 0.4704499047660257,
+      "learning_rate": 1.534485980846443e-05,
+      "loss": 0.2988,
+      "step": 6516
+    },
+    {
+      "epoch": 1.2847003154574133,
+      "grad_norm": 0.5115491062943802,
+      "learning_rate": 1.5343549697428596e-05,
+      "loss": 0.3401,
+      "step": 6517
+    },
+    {
+      "epoch": 1.284897476340694,
+      "grad_norm": 0.5183202372840393,
+      "learning_rate": 1.5342239458009675e-05,
+      "loss": 0.3253,
+      "step": 6518
+    },
+    {
+      "epoch": 1.2850946372239749,
+      "grad_norm": 0.48683435924195767,
+      "learning_rate": 1.5340929090239146e-05,
+      "loss": 0.3006,
+      "step": 6519
+    },
+    {
+      "epoch": 1.2852917981072556,
+      "grad_norm": 0.5138313161943242,
+      "learning_rate": 1.5339618594148497e-05,
+      "loss": 0.3063,
+      "step": 6520
+    },
+    {
+      "epoch": 1.2854889589905363,
+      "grad_norm": 0.5017471245997583,
+      "learning_rate": 1.533830796976921e-05,
+      "loss": 0.3297,
+      "step": 6521
+    },
+    {
+      "epoch": 1.285686119873817,
+      "grad_norm": 0.5123547738441006,
+      "learning_rate": 1.5336997217132777e-05,
+      "loss": 0.3525,
+      "step": 6522
+    },
+    {
+      "epoch": 1.2858832807570977,
+      "grad_norm": 0.5338495423729164,
+      "learning_rate": 1.533568633627069e-05,
+      "loss": 0.3295,
+      "step": 6523
+    },
+    {
+      "epoch": 1.2860804416403786,
+      "grad_norm": 0.5013934896990097,
+      "learning_rate": 1.5334375327214437e-05,
+      "loss": 0.3459,
+      "step": 6524
+    },
+    {
+      "epoch": 1.2862776025236593,
+      "grad_norm": 0.4948643141041335,
+      "learning_rate": 1.5333064189995523e-05,
+      "loss": 0.3553,
+      "step": 6525
+    },
+    {
+      "epoch": 1.28647476340694,
+      "grad_norm": 0.46557296850936236,
+      "learning_rate": 1.5331752924645448e-05,
+      "loss": 0.3048,
+      "step": 6526
+    },
+    {
+      "epoch": 1.286671924290221,
+      "grad_norm": 0.48584258034002087,
+      "learning_rate": 1.5330441531195714e-05,
+      "loss": 0.3127,
+      "step": 6527
+    },
+    {
+      "epoch": 1.2868690851735016,
+      "grad_norm": 0.5002880962180022,
+      "learning_rate": 1.532913000967783e-05,
+      "loss": 0.3415,
+      "step": 6528
+    },
+    {
+      "epoch": 1.2870662460567823,
+      "grad_norm": 0.5028923628602395,
+      "learning_rate": 1.5327818360123307e-05,
+      "loss": 0.3148,
+      "step": 6529
+    },
+    {
+      "epoch": 1.287263406940063,
+      "grad_norm": 0.5195591290608911,
+      "learning_rate": 1.532650658256366e-05,
+      "loss": 0.3567,
+      "step": 6530
+    },
+    {
+      "epoch": 1.2874605678233437,
+      "grad_norm": 0.5004401636204268,
+      "learning_rate": 1.5325194677030396e-05,
+      "loss": 0.3399,
+      "step": 6531
+    },
+    {
+      "epoch": 1.2876577287066246,
+      "grad_norm": 0.4867335475638141,
+      "learning_rate": 1.5323882643555045e-05,
+      "loss": 0.3228,
+      "step": 6532
+    },
+    {
+      "epoch": 1.2878548895899053,
+      "grad_norm": 0.501274068173939,
+      "learning_rate": 1.5322570482169127e-05,
+      "loss": 0.3034,
+      "step": 6533
+    },
+    {
+      "epoch": 1.288052050473186,
+      "grad_norm": 0.502185959674175,
+      "learning_rate": 1.5321258192904165e-05,
+      "loss": 0.3134,
+      "step": 6534
+    },
+    {
+      "epoch": 1.288249211356467,
+      "grad_norm": 0.51496969492382,
+      "learning_rate": 1.531994577579169e-05,
+      "loss": 0.3469,
+      "step": 6535
+    },
+    {
+      "epoch": 1.2884463722397477,
+      "grad_norm": 0.4872788476858814,
+      "learning_rate": 1.5318633230863237e-05,
+      "loss": 0.3189,
+      "step": 6536
+    },
+    {
+      "epoch": 1.2886435331230284,
+      "grad_norm": 0.5229951752622122,
+      "learning_rate": 1.5317320558150336e-05,
+      "loss": 0.3392,
+      "step": 6537
+    },
+    {
+      "epoch": 1.288840694006309,
+      "grad_norm": 0.4772007251776008,
+      "learning_rate": 1.5316007757684523e-05,
+      "loss": 0.3089,
+      "step": 6538
+    },
+    {
+      "epoch": 1.2890378548895898,
+      "grad_norm": 0.5003182201377973,
+      "learning_rate": 1.5314694829497344e-05,
+      "loss": 0.3164,
+      "step": 6539
+    },
+    {
+      "epoch": 1.2892350157728707,
+      "grad_norm": 0.5065544170909857,
+      "learning_rate": 1.5313381773620344e-05,
+      "loss": 0.3202,
+      "step": 6540
+    },
+    {
+      "epoch": 1.2894321766561514,
+      "grad_norm": 0.5077740483878833,
+      "learning_rate": 1.5312068590085067e-05,
+      "loss": 0.343,
+      "step": 6541
+    },
+    {
+      "epoch": 1.289629337539432,
+      "grad_norm": 0.5577531098567273,
+      "learning_rate": 1.5310755278923067e-05,
+      "loss": 0.3386,
+      "step": 6542
+    },
+    {
+      "epoch": 1.289826498422713,
+      "grad_norm": 10.547858415373208,
+      "learning_rate": 1.530944184016589e-05,
+      "loss": 0.3561,
+      "step": 6543
+    },
+    {
+      "epoch": 1.2900236593059937,
+      "grad_norm": 0.5345161964355895,
+      "learning_rate": 1.53081282738451e-05,
+      "loss": 0.3251,
+      "step": 6544
+    },
+    {
+      "epoch": 1.2902208201892744,
+      "grad_norm": 0.5104260494177608,
+      "learning_rate": 1.5306814579992254e-05,
+      "loss": 0.3355,
+      "step": 6545
+    },
+    {
+      "epoch": 1.2904179810725553,
+      "grad_norm": 0.46675498748982025,
+      "learning_rate": 1.530550075863891e-05,
+      "loss": 0.3111,
+      "step": 6546
+    },
+    {
+      "epoch": 1.290615141955836,
+      "grad_norm": 0.5293919854328544,
+      "learning_rate": 1.5304186809816644e-05,
+      "loss": 0.3247,
+      "step": 6547
+    },
+    {
+      "epoch": 1.2908123028391167,
+      "grad_norm": 0.4883341570443513,
+      "learning_rate": 1.5302872733557013e-05,
+      "loss": 0.316,
+      "step": 6548
+    },
+    {
+      "epoch": 1.2910094637223974,
+      "grad_norm": 0.5129224662462487,
+      "learning_rate": 1.53015585298916e-05,
+      "loss": 0.3536,
+      "step": 6549
+    },
+    {
+      "epoch": 1.2912066246056781,
+      "grad_norm": 0.4954761184679624,
+      "learning_rate": 1.5300244198851965e-05,
+      "loss": 0.3269,
+      "step": 6550
+    },
+    {
+      "epoch": 1.291403785488959,
+      "grad_norm": 0.5132713779944943,
+      "learning_rate": 1.5298929740469707e-05,
+      "loss": 0.34,
+      "step": 6551
+    },
+    {
+      "epoch": 1.2916009463722398,
+      "grad_norm": 0.5025532639181867,
+      "learning_rate": 1.5297615154776384e-05,
+      "loss": 0.3253,
+      "step": 6552
+    },
+    {
+      "epoch": 1.2917981072555205,
+      "grad_norm": 0.5058178309161923,
+      "learning_rate": 1.5296300441803594e-05,
+      "loss": 0.3464,
+      "step": 6553
+    },
+    {
+      "epoch": 1.2919952681388014,
+      "grad_norm": 0.5248801782387191,
+      "learning_rate": 1.5294985601582922e-05,
+      "loss": 0.3535,
+      "step": 6554
+    },
+    {
+      "epoch": 1.292192429022082,
+      "grad_norm": 0.4613729865477732,
+      "learning_rate": 1.5293670634145955e-05,
+      "loss": 0.2844,
+      "step": 6555
+    },
+    {
+      "epoch": 1.2923895899053628,
+      "grad_norm": 0.548846735245273,
+      "learning_rate": 1.529235553952429e-05,
+      "loss": 0.3593,
+      "step": 6556
+    },
+    {
+      "epoch": 1.2925867507886435,
+      "grad_norm": 0.48961397369557463,
+      "learning_rate": 1.5291040317749522e-05,
+      "loss": 0.3248,
+      "step": 6557
+    },
+    {
+      "epoch": 1.2927839116719242,
+      "grad_norm": 0.5242798395592415,
+      "learning_rate": 1.528972496885325e-05,
+      "loss": 0.3569,
+      "step": 6558
+    },
+    {
+      "epoch": 1.2929810725552051,
+      "grad_norm": 0.5183526269970178,
+      "learning_rate": 1.5288409492867075e-05,
+      "loss": 0.3333,
+      "step": 6559
+    },
+    {
+      "epoch": 1.2931782334384858,
+      "grad_norm": 0.6706130182709,
+      "learning_rate": 1.52870938898226e-05,
+      "loss": 0.3381,
+      "step": 6560
+    },
+    {
+      "epoch": 1.2933753943217665,
+      "grad_norm": 0.5081309223342054,
+      "learning_rate": 1.528577815975144e-05,
+      "loss": 0.3075,
+      "step": 6561
+    },
+    {
+      "epoch": 1.2935725552050474,
+      "grad_norm": 0.5513296961101606,
+      "learning_rate": 1.5284462302685203e-05,
+      "loss": 0.3199,
+      "step": 6562
+    },
+    {
+      "epoch": 1.2937697160883281,
+      "grad_norm": 0.49894267015353083,
+      "learning_rate": 1.52831463186555e-05,
+      "loss": 0.3275,
+      "step": 6563
+    },
+    {
+      "epoch": 1.2939668769716088,
+      "grad_norm": 0.5327112678308222,
+      "learning_rate": 1.5281830207693955e-05,
+      "loss": 0.3446,
+      "step": 6564
+    },
+    {
+      "epoch": 1.2941640378548895,
+      "grad_norm": 0.5119589336578062,
+      "learning_rate": 1.5280513969832185e-05,
+      "loss": 0.3238,
+      "step": 6565
+    },
+    {
+      "epoch": 1.2943611987381702,
+      "grad_norm": 0.5181686124023328,
+      "learning_rate": 1.5279197605101814e-05,
+      "loss": 0.327,
+      "step": 6566
+    },
+    {
+      "epoch": 1.2945583596214512,
+      "grad_norm": 0.5017388093095246,
+      "learning_rate": 1.527788111353447e-05,
+      "loss": 0.3306,
+      "step": 6567
+    },
+    {
+      "epoch": 1.2947555205047319,
+      "grad_norm": 0.5001792806641484,
+      "learning_rate": 1.5276564495161787e-05,
+      "loss": 0.3197,
+      "step": 6568
+    },
+    {
+      "epoch": 1.2949526813880126,
+      "grad_norm": 1.4540496805328393,
+      "learning_rate": 1.5275247750015383e-05,
+      "loss": 0.3723,
+      "step": 6569
+    },
+    {
+      "epoch": 1.2951498422712935,
+      "grad_norm": 1.8774355679525034,
+      "learning_rate": 1.5273930878126912e-05,
+      "loss": 0.3558,
+      "step": 6570
+    },
+    {
+      "epoch": 1.2953470031545742,
+      "grad_norm": 0.6034476548951655,
+      "learning_rate": 1.5272613879528e-05,
+      "loss": 0.3327,
+      "step": 6571
+    },
+    {
+      "epoch": 1.2955441640378549,
+      "grad_norm": 0.6074067971559209,
+      "learning_rate": 1.5271296754250296e-05,
+      "loss": 0.3475,
+      "step": 6572
+    },
+    {
+      "epoch": 1.2957413249211356,
+      "grad_norm": 2.176593999697865,
+      "learning_rate": 1.526997950232544e-05,
+      "loss": 0.388,
+      "step": 6573
+    },
+    {
+      "epoch": 1.2959384858044163,
+      "grad_norm": 0.6536726906942081,
+      "learning_rate": 1.5268662123785084e-05,
+      "loss": 0.3288,
+      "step": 6574
+    },
+    {
+      "epoch": 1.2961356466876972,
+      "grad_norm": 0.5254118788592566,
+      "learning_rate": 1.5267344618660876e-05,
+      "loss": 0.3427,
+      "step": 6575
+    },
+    {
+      "epoch": 1.296332807570978,
+      "grad_norm": 0.504643420107229,
+      "learning_rate": 1.526602698698447e-05,
+      "loss": 0.3277,
+      "step": 6576
+    },
+    {
+      "epoch": 1.2965299684542586,
+      "grad_norm": 0.5190602422795867,
+      "learning_rate": 1.5264709228787534e-05,
+      "loss": 0.3404,
+      "step": 6577
+    },
+    {
+      "epoch": 1.2967271293375395,
+      "grad_norm": 0.525866200522297,
+      "learning_rate": 1.5263391344101713e-05,
+      "loss": 0.351,
+      "step": 6578
+    },
+    {
+      "epoch": 1.2969242902208202,
+      "grad_norm": 0.5358394825870619,
+      "learning_rate": 1.5262073332958677e-05,
+      "loss": 0.3478,
+      "step": 6579
+    },
+    {
+      "epoch": 1.297121451104101,
+      "grad_norm": 0.5301335549732624,
+      "learning_rate": 1.526075519539009e-05,
+      "loss": 0.3253,
+      "step": 6580
+    },
+    {
+      "epoch": 1.2973186119873816,
+      "grad_norm": 0.5239831490523945,
+      "learning_rate": 1.5259436931427624e-05,
+      "loss": 0.3308,
+      "step": 6581
+    },
+    {
+      "epoch": 1.2975157728706623,
+      "grad_norm": 0.5013826442278009,
+      "learning_rate": 1.525811854110295e-05,
+      "loss": 0.3492,
+      "step": 6582
+    },
+    {
+      "epoch": 1.2977129337539433,
+      "grad_norm": 0.5146221951802675,
+      "learning_rate": 1.5256800024447744e-05,
+      "loss": 0.3288,
+      "step": 6583
+    },
+    {
+      "epoch": 1.297910094637224,
+      "grad_norm": 0.5480449934066463,
+      "learning_rate": 1.5255481381493686e-05,
+      "loss": 0.3557,
+      "step": 6584
+    },
+    {
+      "epoch": 1.2981072555205047,
+      "grad_norm": 0.5334935940934137,
+      "learning_rate": 1.5254162612272451e-05,
+      "loss": 0.3467,
+      "step": 6585
+    },
+    {
+      "epoch": 1.2983044164037856,
+      "grad_norm": 0.5358329007175838,
+      "learning_rate": 1.5252843716815733e-05,
+      "loss": 0.3814,
+      "step": 6586
+    },
+    {
+      "epoch": 1.2985015772870663,
+      "grad_norm": 0.49941905926425306,
+      "learning_rate": 1.5251524695155214e-05,
+      "loss": 0.3412,
+      "step": 6587
+    },
+    {
+      "epoch": 1.298698738170347,
+      "grad_norm": 0.5883299960109943,
+      "learning_rate": 1.525020554732258e-05,
+      "loss": 0.3469,
+      "step": 6588
+    },
+    {
+      "epoch": 1.2988958990536277,
+      "grad_norm": 0.4752126077673626,
+      "learning_rate": 1.5248886273349537e-05,
+      "loss": 0.3147,
+      "step": 6589
+    },
+    {
+      "epoch": 1.2990930599369084,
+      "grad_norm": 0.5009771354117454,
+      "learning_rate": 1.524756687326777e-05,
+      "loss": 0.3461,
+      "step": 6590
+    },
+    {
+      "epoch": 1.2992902208201893,
+      "grad_norm": 0.5678355964259746,
+      "learning_rate": 1.5246247347108984e-05,
+      "loss": 0.3774,
+      "step": 6591
+    },
+    {
+      "epoch": 1.29948738170347,
+      "grad_norm": 0.4622894199811546,
+      "learning_rate": 1.524492769490488e-05,
+      "loss": 0.3359,
+      "step": 6592
+    },
+    {
+      "epoch": 1.2996845425867507,
+      "grad_norm": 0.5169518693086831,
+      "learning_rate": 1.5243607916687167e-05,
+      "loss": 0.3513,
+      "step": 6593
+    },
+    {
+      "epoch": 1.2998817034700316,
+      "grad_norm": 0.49140954893508976,
+      "learning_rate": 1.524228801248755e-05,
+      "loss": 0.3222,
+      "step": 6594
+    },
+    {
+      "epoch": 1.3000788643533123,
+      "grad_norm": 0.5198162659011691,
+      "learning_rate": 1.5240967982337738e-05,
+      "loss": 0.3309,
+      "step": 6595
+    },
+    {
+      "epoch": 1.300276025236593,
+      "grad_norm": 0.5186960022124844,
+      "learning_rate": 1.5239647826269455e-05,
+      "loss": 0.3305,
+      "step": 6596
+    },
+    {
+      "epoch": 1.300473186119874,
+      "grad_norm": 0.5377070970944036,
+      "learning_rate": 1.5238327544314409e-05,
+      "loss": 0.3584,
+      "step": 6597
+    },
+    {
+      "epoch": 1.3006703470031546,
+      "grad_norm": 0.5046133132932712,
+      "learning_rate": 1.5237007136504329e-05,
+      "loss": 0.3403,
+      "step": 6598
+    },
+    {
+      "epoch": 1.3008675078864353,
+      "grad_norm": 0.5454800240840171,
+      "learning_rate": 1.5235686602870932e-05,
+      "loss": 0.377,
+      "step": 6599
+    },
+    {
+      "epoch": 1.301064668769716,
+      "grad_norm": 0.5025344180773049,
+      "learning_rate": 1.5234365943445953e-05,
+      "loss": 0.3432,
+      "step": 6600
+    },
+    {
+      "epoch": 1.3012618296529967,
+      "grad_norm": 0.5276225149052198,
+      "learning_rate": 1.523304515826111e-05,
+      "loss": 0.3527,
+      "step": 6601
+    },
+    {
+      "epoch": 1.3014589905362777,
+      "grad_norm": 0.47724462165996756,
+      "learning_rate": 1.5231724247348148e-05,
+      "loss": 0.3126,
+      "step": 6602
+    },
+    {
+      "epoch": 1.3016561514195584,
+      "grad_norm": 0.5567156598290902,
+      "learning_rate": 1.5230403210738796e-05,
+      "loss": 0.3633,
+      "step": 6603
+    },
+    {
+      "epoch": 1.301853312302839,
+      "grad_norm": 0.48550253059271353,
+      "learning_rate": 1.5229082048464796e-05,
+      "loss": 0.3226,
+      "step": 6604
+    },
+    {
+      "epoch": 1.30205047318612,
+      "grad_norm": 0.5343562201937075,
+      "learning_rate": 1.5227760760557887e-05,
+      "loss": 0.3421,
+      "step": 6605
+    },
+    {
+      "epoch": 1.3022476340694007,
+      "grad_norm": 0.5152095426808443,
+      "learning_rate": 1.522643934704982e-05,
+      "loss": 0.3454,
+      "step": 6606
+    },
+    {
+      "epoch": 1.3024447949526814,
+      "grad_norm": 0.5411901174169776,
+      "learning_rate": 1.5225117807972334e-05,
+      "loss": 0.3494,
+      "step": 6607
+    },
+    {
+      "epoch": 1.302641955835962,
+      "grad_norm": 0.5079018703629262,
+      "learning_rate": 1.5223796143357188e-05,
+      "loss": 0.3364,
+      "step": 6608
+    },
+    {
+      "epoch": 1.3028391167192428,
+      "grad_norm": 0.5422309368847433,
+      "learning_rate": 1.522247435323613e-05,
+      "loss": 0.332,
+      "step": 6609
+    },
+    {
+      "epoch": 1.3030362776025237,
+      "grad_norm": 1.8926296586911842,
+      "learning_rate": 1.5221152437640922e-05,
+      "loss": 0.321,
+      "step": 6610
+    },
+    {
+      "epoch": 1.3032334384858044,
+      "grad_norm": 0.5764351786305123,
+      "learning_rate": 1.5219830396603321e-05,
+      "loss": 0.3629,
+      "step": 6611
+    },
+    {
+      "epoch": 1.3034305993690851,
+      "grad_norm": 0.5013264006168145,
+      "learning_rate": 1.5218508230155093e-05,
+      "loss": 0.3306,
+      "step": 6612
+    },
+    {
+      "epoch": 1.303627760252366,
+      "grad_norm": 0.5367421957046887,
+      "learning_rate": 1.5217185938328003e-05,
+      "loss": 0.3283,
+      "step": 6613
+    },
+    {
+      "epoch": 1.3038249211356467,
+      "grad_norm": 0.6051753693199979,
+      "learning_rate": 1.5215863521153817e-05,
+      "loss": 0.3365,
+      "step": 6614
+    },
+    {
+      "epoch": 1.3040220820189274,
+      "grad_norm": 0.5121455196545641,
+      "learning_rate": 1.521454097866431e-05,
+      "loss": 0.3465,
+      "step": 6615
+    },
+    {
+      "epoch": 1.3042192429022081,
+      "grad_norm": 0.5204250557148496,
+      "learning_rate": 1.5213218310891256e-05,
+      "loss": 0.343,
+      "step": 6616
+    },
+    {
+      "epoch": 1.3044164037854888,
+      "grad_norm": 0.5312764234356772,
+      "learning_rate": 1.5211895517866437e-05,
+      "loss": 0.3386,
+      "step": 6617
+    },
+    {
+      "epoch": 1.3046135646687698,
+      "grad_norm": 0.5071761949128407,
+      "learning_rate": 1.5210572599621626e-05,
+      "loss": 0.3328,
+      "step": 6618
+    },
+    {
+      "epoch": 1.3048107255520505,
+      "grad_norm": 0.5433017323113483,
+      "learning_rate": 1.5209249556188619e-05,
+      "loss": 0.3632,
+      "step": 6619
+    },
+    {
+      "epoch": 1.3050078864353312,
+      "grad_norm": 0.568002908855061,
+      "learning_rate": 1.520792638759919e-05,
+      "loss": 0.3341,
+      "step": 6620
+    },
+    {
+      "epoch": 1.305205047318612,
+      "grad_norm": 0.4905457370725845,
+      "learning_rate": 1.520660309388514e-05,
+      "loss": 0.3361,
+      "step": 6621
+    },
+    {
+      "epoch": 1.3054022082018928,
+      "grad_norm": 0.5619377123694016,
+      "learning_rate": 1.5205279675078255e-05,
+      "loss": 0.3574,
+      "step": 6622
+    },
+    {
+      "epoch": 1.3055993690851735,
+      "grad_norm": 0.5451812237029258,
+      "learning_rate": 1.5203956131210333e-05,
+      "loss": 0.3632,
+      "step": 6623
+    },
+    {
+      "epoch": 1.3057965299684542,
+      "grad_norm": 0.541134699754437,
+      "learning_rate": 1.5202632462313178e-05,
+      "loss": 0.35,
+      "step": 6624
+    },
+    {
+      "epoch": 1.305993690851735,
+      "grad_norm": 0.5465872753368111,
+      "learning_rate": 1.5201308668418588e-05,
+      "loss": 0.3686,
+      "step": 6625
+    },
+    {
+      "epoch": 1.3061908517350158,
+      "grad_norm": 0.48991274685024205,
+      "learning_rate": 1.5199984749558367e-05,
+      "loss": 0.3211,
+      "step": 6626
+    },
+    {
+      "epoch": 1.3063880126182965,
+      "grad_norm": 0.4938660731028014,
+      "learning_rate": 1.5198660705764326e-05,
+      "loss": 0.3393,
+      "step": 6627
+    },
+    {
+      "epoch": 1.3065851735015772,
+      "grad_norm": 0.5656182414746819,
+      "learning_rate": 1.5197336537068275e-05,
+      "loss": 0.3565,
+      "step": 6628
+    },
+    {
+      "epoch": 1.3067823343848581,
+      "grad_norm": 0.48654324405735194,
+      "learning_rate": 1.5196012243502027e-05,
+      "loss": 0.3061,
+      "step": 6629
+    },
+    {
+      "epoch": 1.3069794952681388,
+      "grad_norm": 0.5696072738061521,
+      "learning_rate": 1.5194687825097401e-05,
+      "loss": 0.3624,
+      "step": 6630
+    },
+    {
+      "epoch": 1.3071766561514195,
+      "grad_norm": 0.4992301751223368,
+      "learning_rate": 1.5193363281886217e-05,
+      "loss": 0.3332,
+      "step": 6631
+    },
+    {
+      "epoch": 1.3073738170347002,
+      "grad_norm": 0.47480549774075814,
+      "learning_rate": 1.5192038613900297e-05,
+      "loss": 0.33,
+      "step": 6632
+    },
+    {
+      "epoch": 1.307570977917981,
+      "grad_norm": 0.621679280561631,
+      "learning_rate": 1.519071382117147e-05,
+      "loss": 0.3277,
+      "step": 6633
+    },
+    {
+      "epoch": 1.3077681388012619,
+      "grad_norm": 0.5098237493871496,
+      "learning_rate": 1.5189388903731562e-05,
+      "loss": 0.347,
+      "step": 6634
+    },
+    {
+      "epoch": 1.3079652996845426,
+      "grad_norm": 0.4863659684564988,
+      "learning_rate": 1.5188063861612405e-05,
+      "loss": 0.3305,
+      "step": 6635
+    },
+    {
+      "epoch": 1.3081624605678233,
+      "grad_norm": 0.5124706812541354,
+      "learning_rate": 1.518673869484584e-05,
+      "loss": 0.3275,
+      "step": 6636
+    },
+    {
+      "epoch": 1.3083596214511042,
+      "grad_norm": 0.5190003537835063,
+      "learning_rate": 1.5185413403463698e-05,
+      "loss": 0.3355,
+      "step": 6637
+    },
+    {
+      "epoch": 1.3085567823343849,
+      "grad_norm": 1.1096517702373996,
+      "learning_rate": 1.5184087987497824e-05,
+      "loss": 0.3525,
+      "step": 6638
+    },
+    {
+      "epoch": 1.3087539432176656,
+      "grad_norm": 0.531753230713042,
+      "learning_rate": 1.5182762446980061e-05,
+      "loss": 0.3487,
+      "step": 6639
+    },
+    {
+      "epoch": 1.3089511041009465,
+      "grad_norm": 0.5226931181499541,
+      "learning_rate": 1.5181436781942258e-05,
+      "loss": 0.336,
+      "step": 6640
+    },
+    {
+      "epoch": 1.3091482649842272,
+      "grad_norm": 0.49468338224537034,
+      "learning_rate": 1.5180110992416262e-05,
+      "loss": 0.3089,
+      "step": 6641
+    },
+    {
+      "epoch": 1.309345425867508,
+      "grad_norm": 0.47660945364337853,
+      "learning_rate": 1.5178785078433928e-05,
+      "loss": 0.3207,
+      "step": 6642
+    },
+    {
+      "epoch": 1.3095425867507886,
+      "grad_norm": 0.5172184894772541,
+      "learning_rate": 1.5177459040027114e-05,
+      "loss": 0.3366,
+      "step": 6643
+    },
+    {
+      "epoch": 1.3097397476340693,
+      "grad_norm": 0.570608404247494,
+      "learning_rate": 1.5176132877227674e-05,
+      "loss": 0.3571,
+      "step": 6644
+    },
+    {
+      "epoch": 1.3099369085173502,
+      "grad_norm": 0.5231899305828399,
+      "learning_rate": 1.5174806590067475e-05,
+      "loss": 0.3496,
+      "step": 6645
+    },
+    {
+      "epoch": 1.310134069400631,
+      "grad_norm": 0.499514906187556,
+      "learning_rate": 1.517348017857838e-05,
+      "loss": 0.3125,
+      "step": 6646
+    },
+    {
+      "epoch": 1.3103312302839116,
+      "grad_norm": 0.49103990204237496,
+      "learning_rate": 1.517215364279226e-05,
+      "loss": 0.3127,
+      "step": 6647
+    },
+    {
+      "epoch": 1.3105283911671926,
+      "grad_norm": 0.47212331175692046,
+      "learning_rate": 1.517082698274098e-05,
+      "loss": 0.3166,
+      "step": 6648
+    },
+    {
+      "epoch": 1.3107255520504733,
+      "grad_norm": 0.5661729386565737,
+      "learning_rate": 1.5169500198456417e-05,
+      "loss": 0.376,
+      "step": 6649
+    },
+    {
+      "epoch": 1.310922712933754,
+      "grad_norm": 0.5266195177462345,
+      "learning_rate": 1.5168173289970453e-05,
+      "loss": 0.3619,
+      "step": 6650
+    },
+    {
+      "epoch": 1.3111198738170347,
+      "grad_norm": 0.4968849869857472,
+      "learning_rate": 1.5166846257314961e-05,
+      "loss": 0.3527,
+      "step": 6651
+    },
+    {
+      "epoch": 1.3113170347003154,
+      "grad_norm": 0.5036423318919169,
+      "learning_rate": 1.5165519100521828e-05,
+      "loss": 0.339,
+      "step": 6652
+    },
+    {
+      "epoch": 1.3115141955835963,
+      "grad_norm": 0.8033202698895086,
+      "learning_rate": 1.5164191819622937e-05,
+      "loss": 0.3572,
+      "step": 6653
+    },
+    {
+      "epoch": 1.311711356466877,
+      "grad_norm": 0.5025110332228436,
+      "learning_rate": 1.516286441465018e-05,
+      "loss": 0.335,
+      "step": 6654
+    },
+    {
+      "epoch": 1.3119085173501577,
+      "grad_norm": 0.5067929824788047,
+      "learning_rate": 1.5161536885635451e-05,
+      "loss": 0.3255,
+      "step": 6655
+    },
+    {
+      "epoch": 1.3121056782334386,
+      "grad_norm": 0.6200278707751725,
+      "learning_rate": 1.5160209232610637e-05,
+      "loss": 0.3556,
+      "step": 6656
+    },
+    {
+      "epoch": 1.3123028391167193,
+      "grad_norm": 0.5316032164983252,
+      "learning_rate": 1.5158881455607643e-05,
+      "loss": 0.3387,
+      "step": 6657
+    },
+    {
+      "epoch": 1.3125,
+      "grad_norm": 0.5213969466823882,
+      "learning_rate": 1.5157553554658367e-05,
+      "loss": 0.3442,
+      "step": 6658
+    },
+    {
+      "epoch": 1.3126971608832807,
+      "grad_norm": 0.5401001295546158,
+      "learning_rate": 1.5156225529794713e-05,
+      "loss": 0.3456,
+      "step": 6659
+    },
+    {
+      "epoch": 1.3128943217665614,
+      "grad_norm": 0.5668921110797275,
+      "learning_rate": 1.5154897381048588e-05,
+      "loss": 0.3473,
+      "step": 6660
+    },
+    {
+      "epoch": 1.3130914826498423,
+      "grad_norm": 0.5275183994588266,
+      "learning_rate": 1.5153569108451905e-05,
+      "loss": 0.3245,
+      "step": 6661
+    },
+    {
+      "epoch": 1.313288643533123,
+      "grad_norm": 0.5126597475316212,
+      "learning_rate": 1.5152240712036573e-05,
+      "loss": 0.3375,
+      "step": 6662
+    },
+    {
+      "epoch": 1.3134858044164037,
+      "grad_norm": 0.49094805836830735,
+      "learning_rate": 1.5150912191834504e-05,
+      "loss": 0.3237,
+      "step": 6663
+    },
+    {
+      "epoch": 1.3136829652996846,
+      "grad_norm": 0.5276645393537767,
+      "learning_rate": 1.5149583547877629e-05,
+      "loss": 0.3586,
+      "step": 6664
+    },
+    {
+      "epoch": 1.3138801261829653,
+      "grad_norm": 0.5293512949651924,
+      "learning_rate": 1.5148254780197856e-05,
+      "loss": 0.3064,
+      "step": 6665
+    },
+    {
+      "epoch": 1.314077287066246,
+      "grad_norm": 0.5109488720261187,
+      "learning_rate": 1.514692588882712e-05,
+      "loss": 0.3164,
+      "step": 6666
+    },
+    {
+      "epoch": 1.3142744479495267,
+      "grad_norm": 0.49826567908107544,
+      "learning_rate": 1.5145596873797342e-05,
+      "loss": 0.3158,
+      "step": 6667
+    },
+    {
+      "epoch": 1.3144716088328074,
+      "grad_norm": 0.5250546099243462,
+      "learning_rate": 1.5144267735140459e-05,
+      "loss": 0.3415,
+      "step": 6668
+    },
+    {
+      "epoch": 1.3146687697160884,
+      "grad_norm": 0.4974244732092123,
+      "learning_rate": 1.5142938472888395e-05,
+      "loss": 0.3153,
+      "step": 6669
+    },
+    {
+      "epoch": 1.314865930599369,
+      "grad_norm": 0.5030333271298699,
+      "learning_rate": 1.5141609087073099e-05,
+      "loss": 0.3212,
+      "step": 6670
+    },
+    {
+      "epoch": 1.3150630914826498,
+      "grad_norm": 0.49516039944032675,
+      "learning_rate": 1.51402795777265e-05,
+      "loss": 0.3373,
+      "step": 6671
+    },
+    {
+      "epoch": 1.3152602523659307,
+      "grad_norm": 0.5251022630682348,
+      "learning_rate": 1.5138949944880547e-05,
+      "loss": 0.3594,
+      "step": 6672
+    },
+    {
+      "epoch": 1.3154574132492114,
+      "grad_norm": 0.5080937428606618,
+      "learning_rate": 1.5137620188567183e-05,
+      "loss": 0.2982,
+      "step": 6673
+    },
+    {
+      "epoch": 1.315654574132492,
+      "grad_norm": 0.5762031753271423,
+      "learning_rate": 1.5136290308818355e-05,
+      "loss": 0.365,
+      "step": 6674
+    },
+    {
+      "epoch": 1.3158517350157728,
+      "grad_norm": 0.492333953850378,
+      "learning_rate": 1.5134960305666017e-05,
+      "loss": 0.3197,
+      "step": 6675
+    },
+    {
+      "epoch": 1.3160488958990535,
+      "grad_norm": 0.5353009146536675,
+      "learning_rate": 1.5133630179142124e-05,
+      "loss": 0.3206,
+      "step": 6676
+    },
+    {
+      "epoch": 1.3162460567823344,
+      "grad_norm": 0.5193275884525621,
+      "learning_rate": 1.5132299929278631e-05,
+      "loss": 0.3569,
+      "step": 6677
+    },
+    {
+      "epoch": 1.3164432176656151,
+      "grad_norm": 0.5170594306368025,
+      "learning_rate": 1.5130969556107498e-05,
+      "loss": 0.3609,
+      "step": 6678
+    },
+    {
+      "epoch": 1.3166403785488958,
+      "grad_norm": 0.5376643454719803,
+      "learning_rate": 1.512963905966069e-05,
+      "loss": 0.332,
+      "step": 6679
+    },
+    {
+      "epoch": 1.3168375394321767,
+      "grad_norm": 0.5133927845944496,
+      "learning_rate": 1.5128308439970174e-05,
+      "loss": 0.3372,
+      "step": 6680
+    },
+    {
+      "epoch": 1.3170347003154574,
+      "grad_norm": 0.5399002043459604,
+      "learning_rate": 1.5126977697067915e-05,
+      "loss": 0.363,
+      "step": 6681
+    },
+    {
+      "epoch": 1.3172318611987381,
+      "grad_norm": 0.49610768530525307,
+      "learning_rate": 1.5125646830985892e-05,
+      "loss": 0.3136,
+      "step": 6682
+    },
+    {
+      "epoch": 1.317429022082019,
+      "grad_norm": 0.5473955010074304,
+      "learning_rate": 1.5124315841756072e-05,
+      "loss": 0.355,
+      "step": 6683
+    },
+    {
+      "epoch": 1.3176261829652998,
+      "grad_norm": 0.578461631477163,
+      "learning_rate": 1.5122984729410437e-05,
+      "loss": 0.3719,
+      "step": 6684
+    },
+    {
+      "epoch": 1.3178233438485805,
+      "grad_norm": 0.5463357652437992,
+      "learning_rate": 1.5121653493980973e-05,
+      "loss": 0.3432,
+      "step": 6685
+    },
+    {
+      "epoch": 1.3180205047318612,
+      "grad_norm": 0.4772451928962519,
+      "learning_rate": 1.5120322135499654e-05,
+      "loss": 0.3071,
+      "step": 6686
+    },
+    {
+      "epoch": 1.3182176656151419,
+      "grad_norm": 0.5767867603367968,
+      "learning_rate": 1.511899065399848e-05,
+      "loss": 0.3512,
+      "step": 6687
+    },
+    {
+      "epoch": 1.3184148264984228,
+      "grad_norm": 0.5041565865933484,
+      "learning_rate": 1.5117659049509425e-05,
+      "loss": 0.3437,
+      "step": 6688
+    },
+    {
+      "epoch": 1.3186119873817035,
+      "grad_norm": 0.5042883160416222,
+      "learning_rate": 1.5116327322064497e-05,
+      "loss": 0.3271,
+      "step": 6689
+    },
+    {
+      "epoch": 1.3188091482649842,
+      "grad_norm": 0.4844327881363672,
+      "learning_rate": 1.5114995471695679e-05,
+      "loss": 0.2951,
+      "step": 6690
+    },
+    {
+      "epoch": 1.319006309148265,
+      "grad_norm": 0.5160775384323053,
+      "learning_rate": 1.5113663498434979e-05,
+      "loss": 0.3308,
+      "step": 6691
+    },
+    {
+      "epoch": 1.3192034700315458,
+      "grad_norm": 0.5182153228216186,
+      "learning_rate": 1.5112331402314393e-05,
+      "loss": 0.3311,
+      "step": 6692
+    },
+    {
+      "epoch": 1.3194006309148265,
+      "grad_norm": 0.5509813748947842,
+      "learning_rate": 1.5110999183365933e-05,
+      "loss": 0.3603,
+      "step": 6693
+    },
+    {
+      "epoch": 1.3195977917981072,
+      "grad_norm": 0.4984381584603614,
+      "learning_rate": 1.5109666841621597e-05,
+      "loss": 0.3316,
+      "step": 6694
+    },
+    {
+      "epoch": 1.319794952681388,
+      "grad_norm": 0.5204935976874747,
+      "learning_rate": 1.51083343771134e-05,
+      "loss": 0.3161,
+      "step": 6695
+    },
+    {
+      "epoch": 1.3199921135646688,
+      "grad_norm": 0.536865132939174,
+      "learning_rate": 1.510700178987336e-05,
+      "loss": 0.3332,
+      "step": 6696
+    },
+    {
+      "epoch": 1.3201892744479495,
+      "grad_norm": 0.520088029730267,
+      "learning_rate": 1.5105669079933486e-05,
+      "loss": 0.336,
+      "step": 6697
+    },
+    {
+      "epoch": 1.3203864353312302,
+      "grad_norm": 0.898006104693384,
+      "learning_rate": 1.5104336247325803e-05,
+      "loss": 0.3888,
+      "step": 6698
+    },
+    {
+      "epoch": 1.3205835962145112,
+      "grad_norm": 0.49085502589272256,
+      "learning_rate": 1.510300329208233e-05,
+      "loss": 0.3267,
+      "step": 6699
+    },
+    {
+      "epoch": 1.3207807570977919,
+      "grad_norm": 0.5621876865017514,
+      "learning_rate": 1.5101670214235094e-05,
+      "loss": 0.3831,
+      "step": 6700
+    },
+    {
+      "epoch": 1.3209779179810726,
+      "grad_norm": 0.5512732326331895,
+      "learning_rate": 1.5100337013816122e-05,
+      "loss": 0.3478,
+      "step": 6701
+    },
+    {
+      "epoch": 1.3211750788643533,
+      "grad_norm": 0.7789605507183492,
+      "learning_rate": 1.5099003690857448e-05,
+      "loss": 0.3209,
+      "step": 6702
+    },
+    {
+      "epoch": 1.321372239747634,
+      "grad_norm": 0.5419622274274699,
+      "learning_rate": 1.50976702453911e-05,
+      "loss": 0.3443,
+      "step": 6703
+    },
+    {
+      "epoch": 1.3215694006309149,
+      "grad_norm": 0.5100055997526229,
+      "learning_rate": 1.5096336677449125e-05,
+      "loss": 0.3405,
+      "step": 6704
+    },
+    {
+      "epoch": 1.3217665615141956,
+      "grad_norm": 0.4976601575607139,
+      "learning_rate": 1.5095002987063549e-05,
+      "loss": 0.3209,
+      "step": 6705
+    },
+    {
+      "epoch": 1.3219637223974763,
+      "grad_norm": 0.5566156732462322,
+      "learning_rate": 1.509366917426643e-05,
+      "loss": 0.3416,
+      "step": 6706
+    },
+    {
+      "epoch": 1.3221608832807572,
+      "grad_norm": 0.5069545882513634,
+      "learning_rate": 1.5092335239089803e-05,
+      "loss": 0.3394,
+      "step": 6707
+    },
+    {
+      "epoch": 1.322358044164038,
+      "grad_norm": 0.5033024334874232,
+      "learning_rate": 1.5091001181565725e-05,
+      "loss": 0.3424,
+      "step": 6708
+    },
+    {
+      "epoch": 1.3225552050473186,
+      "grad_norm": 0.49579326452908207,
+      "learning_rate": 1.5089667001726243e-05,
+      "loss": 0.3147,
+      "step": 6709
+    },
+    {
+      "epoch": 1.3227523659305993,
+      "grad_norm": 0.5271145089705676,
+      "learning_rate": 1.5088332699603412e-05,
+      "loss": 0.3593,
+      "step": 6710
+    },
+    {
+      "epoch": 1.32294952681388,
+      "grad_norm": 0.5320580774064494,
+      "learning_rate": 1.508699827522929e-05,
+      "loss": 0.3364,
+      "step": 6711
+    },
+    {
+      "epoch": 1.323146687697161,
+      "grad_norm": 0.5104256370446169,
+      "learning_rate": 1.5085663728635935e-05,
+      "loss": 0.333,
+      "step": 6712
+    },
+    {
+      "epoch": 1.3233438485804416,
+      "grad_norm": 0.49891079466541427,
+      "learning_rate": 1.5084329059855419e-05,
+      "loss": 0.3334,
+      "step": 6713
+    },
+    {
+      "epoch": 1.3235410094637223,
+      "grad_norm": 0.5733949567223368,
+      "learning_rate": 1.5082994268919798e-05,
+      "loss": 0.3157,
+      "step": 6714
+    },
+    {
+      "epoch": 1.3237381703470033,
+      "grad_norm": 0.5435038605202684,
+      "learning_rate": 1.508165935586115e-05,
+      "loss": 0.3233,
+      "step": 6715
+    },
+    {
+      "epoch": 1.323935331230284,
+      "grad_norm": 0.5673725835719423,
+      "learning_rate": 1.5080324320711542e-05,
+      "loss": 0.3399,
+      "step": 6716
+    },
+    {
+      "epoch": 1.3241324921135647,
+      "grad_norm": 0.46434225808628016,
+      "learning_rate": 1.507898916350305e-05,
+      "loss": 0.315,
+      "step": 6717
+    },
+    {
+      "epoch": 1.3243296529968454,
+      "grad_norm": 0.5335227211463376,
+      "learning_rate": 1.5077653884267753e-05,
+      "loss": 0.3509,
+      "step": 6718
+    },
+    {
+      "epoch": 1.324526813880126,
+      "grad_norm": 0.5194316243511954,
+      "learning_rate": 1.5076318483037736e-05,
+      "loss": 0.3524,
+      "step": 6719
+    },
+    {
+      "epoch": 1.324723974763407,
+      "grad_norm": 0.5237285016193873,
+      "learning_rate": 1.5074982959845077e-05,
+      "loss": 0.3842,
+      "step": 6720
+    },
+    {
+      "epoch": 1.3249211356466877,
+      "grad_norm": 0.512608848459545,
+      "learning_rate": 1.5073647314721867e-05,
+      "loss": 0.3554,
+      "step": 6721
+    },
+    {
+      "epoch": 1.3251182965299684,
+      "grad_norm": 22.837779278990155,
+      "learning_rate": 1.5072311547700194e-05,
+      "loss": 0.3395,
+      "step": 6722
+    },
+    {
+      "epoch": 1.3253154574132493,
+      "grad_norm": 0.5584335319441984,
+      "learning_rate": 1.507097565881215e-05,
+      "loss": 0.3746,
+      "step": 6723
+    },
+    {
+      "epoch": 1.32551261829653,
+      "grad_norm": 0.4725105964351633,
+      "learning_rate": 1.5069639648089833e-05,
+      "loss": 0.3079,
+      "step": 6724
+    },
+    {
+      "epoch": 1.3257097791798107,
+      "grad_norm": 0.48649112789508087,
+      "learning_rate": 1.506830351556534e-05,
+      "loss": 0.3243,
+      "step": 6725
+    },
+    {
+      "epoch": 1.3259069400630916,
+      "grad_norm": 0.5047703578036561,
+      "learning_rate": 1.5066967261270775e-05,
+      "loss": 0.352,
+      "step": 6726
+    },
+    {
+      "epoch": 1.3261041009463723,
+      "grad_norm": 0.5240622470462101,
+      "learning_rate": 1.506563088523824e-05,
+      "loss": 0.3676,
+      "step": 6727
+    },
+    {
+      "epoch": 1.326301261829653,
+      "grad_norm": 0.5168279521689652,
+      "learning_rate": 1.5064294387499844e-05,
+      "loss": 0.3542,
+      "step": 6728
+    },
+    {
+      "epoch": 1.3264984227129337,
+      "grad_norm": 0.4816304700981381,
+      "learning_rate": 1.5062957768087698e-05,
+      "loss": 0.3053,
+      "step": 6729
+    },
+    {
+      "epoch": 1.3266955835962144,
+      "grad_norm": 0.606385343673032,
+      "learning_rate": 1.5061621027033914e-05,
+      "loss": 0.3315,
+      "step": 6730
+    },
+    {
+      "epoch": 1.3268927444794953,
+      "grad_norm": 0.49813137918087436,
+      "learning_rate": 1.5060284164370606e-05,
+      "loss": 0.3501,
+      "step": 6731
+    },
+    {
+      "epoch": 1.327089905362776,
+      "grad_norm": 0.5003849761342135,
+      "learning_rate": 1.5058947180129902e-05,
+      "loss": 0.3308,
+      "step": 6732
+    },
+    {
+      "epoch": 1.3272870662460567,
+      "grad_norm": 0.5092842617270525,
+      "learning_rate": 1.5057610074343911e-05,
+      "loss": 0.3215,
+      "step": 6733
+    },
+    {
+      "epoch": 1.3274842271293377,
+      "grad_norm": 0.5426283767618892,
+      "learning_rate": 1.505627284704477e-05,
+      "loss": 0.3455,
+      "step": 6734
+    },
+    {
+      "epoch": 1.3276813880126184,
+      "grad_norm": 0.5093218204427572,
+      "learning_rate": 1.50549354982646e-05,
+      "loss": 0.341,
+      "step": 6735
+    },
+    {
+      "epoch": 1.327878548895899,
+      "grad_norm": 0.47254370160551584,
+      "learning_rate": 1.5053598028035534e-05,
+      "loss": 0.3253,
+      "step": 6736
+    },
+    {
+      "epoch": 1.3280757097791798,
+      "grad_norm": 0.5063195573320013,
+      "learning_rate": 1.5052260436389708e-05,
+      "loss": 0.3158,
+      "step": 6737
+    },
+    {
+      "epoch": 1.3282728706624605,
+      "grad_norm": 0.5253016137313395,
+      "learning_rate": 1.5050922723359254e-05,
+      "loss": 0.3387,
+      "step": 6738
+    },
+    {
+      "epoch": 1.3284700315457414,
+      "grad_norm": 0.4894300336691798,
+      "learning_rate": 1.5049584888976311e-05,
+      "loss": 0.3385,
+      "step": 6739
+    },
+    {
+      "epoch": 1.328667192429022,
+      "grad_norm": 0.5590119104561938,
+      "learning_rate": 1.504824693327303e-05,
+      "loss": 0.3397,
+      "step": 6740
+    },
+    {
+      "epoch": 1.3288643533123028,
+      "grad_norm": 0.5350527773958544,
+      "learning_rate": 1.504690885628155e-05,
+      "loss": 0.3326,
+      "step": 6741
+    },
+    {
+      "epoch": 1.3290615141955837,
+      "grad_norm": 0.5138921218572929,
+      "learning_rate": 1.5045570658034022e-05,
+      "loss": 0.3447,
+      "step": 6742
+    },
+    {
+      "epoch": 1.3292586750788644,
+      "grad_norm": 1.730892234036529,
+      "learning_rate": 1.504423233856259e-05,
+      "loss": 0.373,
+      "step": 6743
+    },
+    {
+      "epoch": 1.3294558359621451,
+      "grad_norm": 0.5491132539355397,
+      "learning_rate": 1.5042893897899417e-05,
+      "loss": 0.3453,
+      "step": 6744
+    },
+    {
+      "epoch": 1.3296529968454258,
+      "grad_norm": 0.539322312311082,
+      "learning_rate": 1.5041555336076661e-05,
+      "loss": 0.3468,
+      "step": 6745
+    },
+    {
+      "epoch": 1.3298501577287065,
+      "grad_norm": 0.5125742645152108,
+      "learning_rate": 1.5040216653126471e-05,
+      "loss": 0.3266,
+      "step": 6746
+    },
+    {
+      "epoch": 1.3300473186119874,
+      "grad_norm": 0.5010768816584499,
+      "learning_rate": 1.5038877849081023e-05,
+      "loss": 0.325,
+      "step": 6747
+    },
+    {
+      "epoch": 1.3302444794952681,
+      "grad_norm": 0.5237464923768348,
+      "learning_rate": 1.5037538923972474e-05,
+      "loss": 0.3392,
+      "step": 6748
+    },
+    {
+      "epoch": 1.3304416403785488,
+      "grad_norm": 0.5259667674407368,
+      "learning_rate": 1.5036199877832997e-05,
+      "loss": 0.3405,
+      "step": 6749
+    },
+    {
+      "epoch": 1.3306388012618298,
+      "grad_norm": 0.5653193705799677,
+      "learning_rate": 1.503486071069476e-05,
+      "loss": 0.3687,
+      "step": 6750
+    },
+    {
+      "epoch": 1.3308359621451105,
+      "grad_norm": 0.5180820594611166,
+      "learning_rate": 1.5033521422589943e-05,
+      "loss": 0.3785,
+      "step": 6751
+    },
+    {
+      "epoch": 1.3310331230283912,
+      "grad_norm": 0.5264455832857787,
+      "learning_rate": 1.5032182013550719e-05,
+      "loss": 0.349,
+      "step": 6752
+    },
+    {
+      "epoch": 1.3312302839116719,
+      "grad_norm": 0.5131681170313048,
+      "learning_rate": 1.5030842483609268e-05,
+      "loss": 0.3309,
+      "step": 6753
+    },
+    {
+      "epoch": 1.3314274447949526,
+      "grad_norm": 0.5006781906744829,
+      "learning_rate": 1.5029502832797775e-05,
+      "loss": 0.3106,
+      "step": 6754
+    },
+    {
+      "epoch": 1.3316246056782335,
+      "grad_norm": 0.4732747904769726,
+      "learning_rate": 1.5028163061148432e-05,
+      "loss": 0.3355,
+      "step": 6755
+    },
+    {
+      "epoch": 1.3318217665615142,
+      "grad_norm": 0.4995838593733347,
+      "learning_rate": 1.5026823168693414e-05,
+      "loss": 0.3322,
+      "step": 6756
+    },
+    {
+      "epoch": 1.3320189274447949,
+      "grad_norm": 0.478174941952485,
+      "learning_rate": 1.5025483155464926e-05,
+      "loss": 0.3073,
+      "step": 6757
+    },
+    {
+      "epoch": 1.3322160883280758,
+      "grad_norm": 0.484122961464807,
+      "learning_rate": 1.5024143021495157e-05,
+      "loss": 0.3413,
+      "step": 6758
+    },
+    {
+      "epoch": 1.3324132492113565,
+      "grad_norm": 0.5951585544576415,
+      "learning_rate": 1.5022802766816306e-05,
+      "loss": 0.3635,
+      "step": 6759
+    },
+    {
+      "epoch": 1.3326104100946372,
+      "grad_norm": 0.5340216464990926,
+      "learning_rate": 1.5021462391460576e-05,
+      "loss": 0.3552,
+      "step": 6760
+    },
+    {
+      "epoch": 1.332807570977918,
+      "grad_norm": 0.5645843694433802,
+      "learning_rate": 1.5020121895460165e-05,
+      "loss": 0.3593,
+      "step": 6761
+    },
+    {
+      "epoch": 1.3330047318611986,
+      "grad_norm": 0.5728974673232404,
+      "learning_rate": 1.5018781278847286e-05,
+      "loss": 0.3529,
+      "step": 6762
+    },
+    {
+      "epoch": 1.3332018927444795,
+      "grad_norm": 0.5256692890594539,
+      "learning_rate": 1.501744054165414e-05,
+      "loss": 0.3387,
+      "step": 6763
+    },
+    {
+      "epoch": 1.3333990536277602,
+      "grad_norm": 0.5674819000439734,
+      "learning_rate": 1.501609968391295e-05,
+      "loss": 0.3509,
+      "step": 6764
+    },
+    {
+      "epoch": 1.333596214511041,
+      "grad_norm": 0.4986446571718649,
+      "learning_rate": 1.5014758705655922e-05,
+      "loss": 0.3239,
+      "step": 6765
+    },
+    {
+      "epoch": 1.3337933753943219,
+      "grad_norm": 0.4896630219969831,
+      "learning_rate": 1.5013417606915279e-05,
+      "loss": 0.3204,
+      "step": 6766
+    },
+    {
+      "epoch": 1.3339905362776026,
+      "grad_norm": 0.5177912836591282,
+      "learning_rate": 1.501207638772324e-05,
+      "loss": 0.3386,
+      "step": 6767
+    },
+    {
+      "epoch": 1.3341876971608833,
+      "grad_norm": 0.48153562676597106,
+      "learning_rate": 1.5010735048112031e-05,
+      "loss": 0.3056,
+      "step": 6768
+    },
+    {
+      "epoch": 1.334384858044164,
+      "grad_norm": 0.499695810551327,
+      "learning_rate": 1.5009393588113876e-05,
+      "loss": 0.2962,
+      "step": 6769
+    },
+    {
+      "epoch": 1.3345820189274447,
+      "grad_norm": 0.48456873034066944,
+      "learning_rate": 1.5008052007761009e-05,
+      "loss": 0.3056,
+      "step": 6770
+    },
+    {
+      "epoch": 1.3347791798107256,
+      "grad_norm": 0.5138152037859967,
+      "learning_rate": 1.5006710307085656e-05,
+      "loss": 0.3473,
+      "step": 6771
+    },
+    {
+      "epoch": 1.3349763406940063,
+      "grad_norm": 0.4989689701123164,
+      "learning_rate": 1.5005368486120058e-05,
+      "loss": 0.3527,
+      "step": 6772
+    },
+    {
+      "epoch": 1.335173501577287,
+      "grad_norm": 0.5240843932016397,
+      "learning_rate": 1.5004026544896448e-05,
+      "loss": 0.3289,
+      "step": 6773
+    },
+    {
+      "epoch": 1.335370662460568,
+      "grad_norm": 0.51061194084439,
+      "learning_rate": 1.5002684483447074e-05,
+      "loss": 0.3582,
+      "step": 6774
+    },
+    {
+      "epoch": 1.3355678233438486,
+      "grad_norm": 0.5005446884388874,
+      "learning_rate": 1.5001342301804176e-05,
+      "loss": 0.3458,
+      "step": 6775
+    },
+    {
+      "epoch": 1.3357649842271293,
+      "grad_norm": 0.5140258358304517,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.3606,
+      "step": 6776
+    },
+    {
+      "epoch": 1.3359621451104102,
+      "grad_norm": 0.49447960825964604,
+      "learning_rate": 1.49986575780668e-05,
+      "loss": 0.3337,
+      "step": 6777
+    },
+    {
+      "epoch": 1.336159305993691,
+      "grad_norm": 0.4734352743858005,
+      "learning_rate": 1.4997315036036826e-05,
+      "loss": 0.3288,
+      "step": 6778
+    },
+    {
+      "epoch": 1.3363564668769716,
+      "grad_norm": 0.5153787348270493,
+      "learning_rate": 1.4995972373942334e-05,
+      "loss": 0.3512,
+      "step": 6779
+    },
+    {
+      "epoch": 1.3365536277602523,
+      "grad_norm": 0.5614500712850554,
+      "learning_rate": 1.4994629591815579e-05,
+      "loss": 0.3581,
+      "step": 6780
+    },
+    {
+      "epoch": 1.336750788643533,
+      "grad_norm": 0.5086243455682851,
+      "learning_rate": 1.4993286689688831e-05,
+      "loss": 0.3368,
+      "step": 6781
+    },
+    {
+      "epoch": 1.336947949526814,
+      "grad_norm": 0.5109131318818986,
+      "learning_rate": 1.4991943667594344e-05,
+      "loss": 0.3473,
+      "step": 6782
+    },
+    {
+      "epoch": 1.3371451104100947,
+      "grad_norm": 0.48373369562864504,
+      "learning_rate": 1.4990600525564394e-05,
+      "loss": 0.3171,
+      "step": 6783
+    },
+    {
+      "epoch": 1.3373422712933754,
+      "grad_norm": 0.48077824595792773,
+      "learning_rate": 1.4989257263631246e-05,
+      "loss": 0.3112,
+      "step": 6784
+    },
+    {
+      "epoch": 1.3375394321766563,
+      "grad_norm": 0.5403123400327008,
+      "learning_rate": 1.4987913881827177e-05,
+      "loss": 0.3509,
+      "step": 6785
+    },
+    {
+      "epoch": 1.337736593059937,
+      "grad_norm": 0.5283987542462503,
+      "learning_rate": 1.4986570380184454e-05,
+      "loss": 0.3369,
+      "step": 6786
+    },
+    {
+      "epoch": 1.3379337539432177,
+      "grad_norm": 0.5135808090920458,
+      "learning_rate": 1.4985226758735368e-05,
+      "loss": 0.3436,
+      "step": 6787
+    },
+    {
+      "epoch": 1.3381309148264984,
+      "grad_norm": 0.5241808952900016,
+      "learning_rate": 1.498388301751219e-05,
+      "loss": 0.329,
+      "step": 6788
+    },
+    {
+      "epoch": 1.338328075709779,
+      "grad_norm": 0.539221498149948,
+      "learning_rate": 1.4982539156547214e-05,
+      "loss": 0.3565,
+      "step": 6789
+    },
+    {
+      "epoch": 1.33852523659306,
+      "grad_norm": 0.5312950545419894,
+      "learning_rate": 1.4981195175872718e-05,
+      "loss": 0.3392,
+      "step": 6790
+    },
+    {
+      "epoch": 1.3387223974763407,
+      "grad_norm": 0.5526638195715327,
+      "learning_rate": 1.4979851075521e-05,
+      "loss": 0.3518,
+      "step": 6791
+    },
+    {
+      "epoch": 1.3389195583596214,
+      "grad_norm": 0.5416794319715723,
+      "learning_rate": 1.4978506855524348e-05,
+      "loss": 0.3632,
+      "step": 6792
+    },
+    {
+      "epoch": 1.3391167192429023,
+      "grad_norm": 0.5235504870365736,
+      "learning_rate": 1.497716251591506e-05,
+      "loss": 0.316,
+      "step": 6793
+    },
+    {
+      "epoch": 1.339313880126183,
+      "grad_norm": 0.5043048732799276,
+      "learning_rate": 1.4975818056725433e-05,
+      "loss": 0.3434,
+      "step": 6794
+    },
+    {
+      "epoch": 1.3395110410094637,
+      "grad_norm": 0.49798925055448356,
+      "learning_rate": 1.4974473477987771e-05,
+      "loss": 0.3184,
+      "step": 6795
+    },
+    {
+      "epoch": 1.3397082018927444,
+      "grad_norm": 0.4979629803712915,
+      "learning_rate": 1.4973128779734381e-05,
+      "loss": 0.321,
+      "step": 6796
+    },
+    {
+      "epoch": 1.3399053627760251,
+      "grad_norm": 0.5040873484662636,
+      "learning_rate": 1.4971783961997561e-05,
+      "loss": 0.3426,
+      "step": 6797
+    },
+    {
+      "epoch": 1.340102523659306,
+      "grad_norm": 0.4974631337369193,
+      "learning_rate": 1.4970439024809634e-05,
+      "loss": 0.3271,
+      "step": 6798
+    },
+    {
+      "epoch": 1.3402996845425867,
+      "grad_norm": 0.5323248192377603,
+      "learning_rate": 1.49690939682029e-05,
+      "loss": 0.3537,
+      "step": 6799
+    },
+    {
+      "epoch": 1.3404968454258674,
+      "grad_norm": 0.48171766055599236,
+      "learning_rate": 1.4967748792209689e-05,
+      "loss": 0.3173,
+      "step": 6800
+    },
+    {
+      "epoch": 1.3406940063091484,
+      "grad_norm": 0.510686613417193,
+      "learning_rate": 1.4966403496862304e-05,
+      "loss": 0.3195,
+      "step": 6801
+    },
+    {
+      "epoch": 1.340891167192429,
+      "grad_norm": 0.48429649890064824,
+      "learning_rate": 1.4965058082193084e-05,
+      "loss": 0.3141,
+      "step": 6802
+    },
+    {
+      "epoch": 1.3410883280757098,
+      "grad_norm": 0.49594807930521567,
+      "learning_rate": 1.496371254823434e-05,
+      "loss": 0.321,
+      "step": 6803
+    },
+    {
+      "epoch": 1.3412854889589905,
+      "grad_norm": 0.526673612495566,
+      "learning_rate": 1.496236689501841e-05,
+      "loss": 0.3415,
+      "step": 6804
+    },
+    {
+      "epoch": 1.3414826498422712,
+      "grad_norm": 0.45404096769055313,
+      "learning_rate": 1.4961021122577613e-05,
+      "loss": 0.2986,
+      "step": 6805
+    },
+    {
+      "epoch": 1.341679810725552,
+      "grad_norm": 0.5806061745177113,
+      "learning_rate": 1.495967523094429e-05,
+      "loss": 0.3928,
+      "step": 6806
+    },
+    {
+      "epoch": 1.3418769716088328,
+      "grad_norm": 0.470337630001986,
+      "learning_rate": 1.4958329220150778e-05,
+      "loss": 0.3148,
+      "step": 6807
+    },
+    {
+      "epoch": 1.3420741324921135,
+      "grad_norm": 0.47772699425478293,
+      "learning_rate": 1.4956983090229413e-05,
+      "loss": 0.3203,
+      "step": 6808
+    },
+    {
+      "epoch": 1.3422712933753944,
+      "grad_norm": 0.4886302060088961,
+      "learning_rate": 1.4955636841212538e-05,
+      "loss": 0.3518,
+      "step": 6809
+    },
+    {
+      "epoch": 1.3424684542586751,
+      "grad_norm": 0.48102229230186133,
+      "learning_rate": 1.4954290473132495e-05,
+      "loss": 0.3198,
+      "step": 6810
+    },
+    {
+      "epoch": 1.3426656151419558,
+      "grad_norm": 0.48930648987210024,
+      "learning_rate": 1.4952943986021633e-05,
+      "loss": 0.3305,
+      "step": 6811
+    },
+    {
+      "epoch": 1.3428627760252365,
+      "grad_norm": 0.540876771967004,
+      "learning_rate": 1.4951597379912306e-05,
+      "loss": 0.3504,
+      "step": 6812
+    },
+    {
+      "epoch": 1.3430599369085172,
+      "grad_norm": 0.5164837203011601,
+      "learning_rate": 1.4950250654836862e-05,
+      "loss": 0.3657,
+      "step": 6813
+    },
+    {
+      "epoch": 1.3432570977917981,
+      "grad_norm": 0.4476636874195238,
+      "learning_rate": 1.4948903810827662e-05,
+      "loss": 0.295,
+      "step": 6814
+    },
+    {
+      "epoch": 1.3434542586750788,
+      "grad_norm": 0.5015700179009981,
+      "learning_rate": 1.4947556847917062e-05,
+      "loss": 0.3495,
+      "step": 6815
+    },
+    {
+      "epoch": 1.3436514195583595,
+      "grad_norm": 0.4883735796558397,
+      "learning_rate": 1.4946209766137422e-05,
+      "loss": 0.3262,
+      "step": 6816
+    },
+    {
+      "epoch": 1.3438485804416405,
+      "grad_norm": 0.4960996354859811,
+      "learning_rate": 1.4944862565521113e-05,
+      "loss": 0.3266,
+      "step": 6817
+    },
+    {
+      "epoch": 1.3440457413249212,
+      "grad_norm": 0.5111574948370762,
+      "learning_rate": 1.4943515246100498e-05,
+      "loss": 0.3534,
+      "step": 6818
+    },
+    {
+      "epoch": 1.3442429022082019,
+      "grad_norm": 0.4806601835647626,
+      "learning_rate": 1.4942167807907945e-05,
+      "loss": 0.3156,
+      "step": 6819
+    },
+    {
+      "epoch": 1.3444400630914828,
+      "grad_norm": 0.4771804770811557,
+      "learning_rate": 1.494082025097583e-05,
+      "loss": 0.3411,
+      "step": 6820
+    },
+    {
+      "epoch": 1.3446372239747635,
+      "grad_norm": 0.5010047370612712,
+      "learning_rate": 1.4939472575336535e-05,
+      "loss": 0.3332,
+      "step": 6821
+    },
+    {
+      "epoch": 1.3448343848580442,
+      "grad_norm": 0.5128977303636474,
+      "learning_rate": 1.4938124781022429e-05,
+      "loss": 0.3502,
+      "step": 6822
+    },
+    {
+      "epoch": 1.3450315457413249,
+      "grad_norm": 0.5115752389432223,
+      "learning_rate": 1.4936776868065904e-05,
+      "loss": 0.3251,
+      "step": 6823
+    },
+    {
+      "epoch": 1.3452287066246056,
+      "grad_norm": 0.5115575705230232,
+      "learning_rate": 1.4935428836499333e-05,
+      "loss": 0.3251,
+      "step": 6824
+    },
+    {
+      "epoch": 1.3454258675078865,
+      "grad_norm": 0.45496251393031,
+      "learning_rate": 1.4934080686355112e-05,
+      "loss": 0.3212,
+      "step": 6825
+    },
+    {
+      "epoch": 1.3456230283911672,
+      "grad_norm": 0.4683704737332005,
+      "learning_rate": 1.4932732417665627e-05,
+      "loss": 0.314,
+      "step": 6826
+    },
+    {
+      "epoch": 1.345820189274448,
+      "grad_norm": 0.5058243078911916,
+      "learning_rate": 1.4931384030463276e-05,
+      "loss": 0.3547,
+      "step": 6827
+    },
+    {
+      "epoch": 1.3460173501577288,
+      "grad_norm": 0.49135394576091385,
+      "learning_rate": 1.4930035524780455e-05,
+      "loss": 0.3351,
+      "step": 6828
+    },
+    {
+      "epoch": 1.3462145110410095,
+      "grad_norm": 0.49176472785254427,
+      "learning_rate": 1.4928686900649557e-05,
+      "loss": 0.3467,
+      "step": 6829
+    },
+    {
+      "epoch": 1.3464116719242902,
+      "grad_norm": 0.4877950490063829,
+      "learning_rate": 1.4927338158102988e-05,
+      "loss": 0.32,
+      "step": 6830
+    },
+    {
+      "epoch": 1.346608832807571,
+      "grad_norm": 0.5644001631439873,
+      "learning_rate": 1.4925989297173148e-05,
+      "loss": 0.3492,
+      "step": 6831
+    },
+    {
+      "epoch": 1.3468059936908516,
+      "grad_norm": 0.5182639423714361,
+      "learning_rate": 1.4924640317892457e-05,
+      "loss": 0.3616,
+      "step": 6832
+    },
+    {
+      "epoch": 1.3470031545741326,
+      "grad_norm": 0.4911569508385157,
+      "learning_rate": 1.4923291220293307e-05,
+      "loss": 0.3187,
+      "step": 6833
+    },
+    {
+      "epoch": 1.3472003154574133,
+      "grad_norm": 0.4939068383078795,
+      "learning_rate": 1.4921942004408126e-05,
+      "loss": 0.3352,
+      "step": 6834
+    },
+    {
+      "epoch": 1.347397476340694,
+      "grad_norm": 0.48064147010076375,
+      "learning_rate": 1.4920592670269323e-05,
+      "loss": 0.3334,
+      "step": 6835
+    },
+    {
+      "epoch": 1.3475946372239749,
+      "grad_norm": 0.5150465597535828,
+      "learning_rate": 1.4919243217909318e-05,
+      "loss": 0.3525,
+      "step": 6836
+    },
+    {
+      "epoch": 1.3477917981072556,
+      "grad_norm": 0.48198241690638255,
+      "learning_rate": 1.4917893647360538e-05,
+      "loss": 0.3215,
+      "step": 6837
+    },
+    {
+      "epoch": 1.3479889589905363,
+      "grad_norm": 0.49192236847221166,
+      "learning_rate": 1.4916543958655396e-05,
+      "loss": 0.3446,
+      "step": 6838
+    },
+    {
+      "epoch": 1.348186119873817,
+      "grad_norm": 0.5380105418206355,
+      "learning_rate": 1.491519415182633e-05,
+      "loss": 0.3455,
+      "step": 6839
+    },
+    {
+      "epoch": 1.3483832807570977,
+      "grad_norm": 0.5360035265768491,
+      "learning_rate": 1.4913844226905767e-05,
+      "loss": 0.3578,
+      "step": 6840
+    },
+    {
+      "epoch": 1.3485804416403786,
+      "grad_norm": 0.6845647174250372,
+      "learning_rate": 1.4912494183926139e-05,
+      "loss": 0.3396,
+      "step": 6841
+    },
+    {
+      "epoch": 1.3487776025236593,
+      "grad_norm": 0.5269738365614332,
+      "learning_rate": 1.4911144022919879e-05,
+      "loss": 0.3559,
+      "step": 6842
+    },
+    {
+      "epoch": 1.34897476340694,
+      "grad_norm": 0.4813494392754745,
+      "learning_rate": 1.4909793743919432e-05,
+      "loss": 0.3228,
+      "step": 6843
+    },
+    {
+      "epoch": 1.349171924290221,
+      "grad_norm": 0.4816948637640496,
+      "learning_rate": 1.4908443346957235e-05,
+      "loss": 0.3154,
+      "step": 6844
+    },
+    {
+      "epoch": 1.3493690851735016,
+      "grad_norm": 0.45620872496423787,
+      "learning_rate": 1.4907092832065734e-05,
+      "loss": 0.3105,
+      "step": 6845
+    },
+    {
+      "epoch": 1.3495662460567823,
+      "grad_norm": 0.5146288021545241,
+      "learning_rate": 1.4905742199277376e-05,
+      "loss": 0.3563,
+      "step": 6846
+    },
+    {
+      "epoch": 1.349763406940063,
+      "grad_norm": 0.508730043660222,
+      "learning_rate": 1.4904391448624612e-05,
+      "loss": 0.3473,
+      "step": 6847
+    },
+    {
+      "epoch": 1.3499605678233437,
+      "grad_norm": 0.5656947915688717,
+      "learning_rate": 1.4903040580139891e-05,
+      "loss": 0.3603,
+      "step": 6848
+    },
+    {
+      "epoch": 1.3501577287066246,
+      "grad_norm": 0.48859833009143055,
+      "learning_rate": 1.4901689593855677e-05,
+      "loss": 0.3243,
+      "step": 6849
+    },
+    {
+      "epoch": 1.3503548895899053,
+      "grad_norm": 0.5710007684390376,
+      "learning_rate": 1.4900338489804418e-05,
+      "loss": 0.3194,
+      "step": 6850
+    },
+    {
+      "epoch": 1.350552050473186,
+      "grad_norm": 0.5249828286967833,
+      "learning_rate": 1.4898987268018586e-05,
+      "loss": 0.3633,
+      "step": 6851
+    },
+    {
+      "epoch": 1.350749211356467,
+      "grad_norm": 0.5414339406348158,
+      "learning_rate": 1.4897635928530634e-05,
+      "loss": 0.3512,
+      "step": 6852
+    },
+    {
+      "epoch": 1.3509463722397477,
+      "grad_norm": 0.44873869097621316,
+      "learning_rate": 1.4896284471373038e-05,
+      "loss": 0.3052,
+      "step": 6853
+    },
+    {
+      "epoch": 1.3511435331230284,
+      "grad_norm": 0.4672888541955871,
+      "learning_rate": 1.4894932896578262e-05,
+      "loss": 0.3198,
+      "step": 6854
+    },
+    {
+      "epoch": 1.351340694006309,
+      "grad_norm": 0.5477287513630034,
+      "learning_rate": 1.4893581204178785e-05,
+      "loss": 0.3337,
+      "step": 6855
+    },
+    {
+      "epoch": 1.3515378548895898,
+      "grad_norm": 0.47013217105100197,
+      "learning_rate": 1.4892229394207076e-05,
+      "loss": 0.3484,
+      "step": 6856
+    },
+    {
+      "epoch": 1.3517350157728707,
+      "grad_norm": 0.47570179900580484,
+      "learning_rate": 1.4890877466695617e-05,
+      "loss": 0.3142,
+      "step": 6857
+    },
+    {
+      "epoch": 1.3519321766561514,
+      "grad_norm": 0.5040215090763913,
+      "learning_rate": 1.488952542167689e-05,
+      "loss": 0.3213,
+      "step": 6858
+    },
+    {
+      "epoch": 1.352129337539432,
+      "grad_norm": 0.5374781827924892,
+      "learning_rate": 1.4888173259183375e-05,
+      "loss": 0.3559,
+      "step": 6859
+    },
+    {
+      "epoch": 1.352326498422713,
+      "grad_norm": 0.4893306035975573,
+      "learning_rate": 1.4886820979247561e-05,
+      "loss": 0.3199,
+      "step": 6860
+    },
+    {
+      "epoch": 1.3525236593059937,
+      "grad_norm": 0.47307143181191913,
+      "learning_rate": 1.4885468581901939e-05,
+      "loss": 0.3081,
+      "step": 6861
+    },
+    {
+      "epoch": 1.3527208201892744,
+      "grad_norm": 0.49226139486267406,
+      "learning_rate": 1.4884116067178997e-05,
+      "loss": 0.3359,
+      "step": 6862
+    },
+    {
+      "epoch": 1.3529179810725553,
+      "grad_norm": 2.9054703966108226,
+      "learning_rate": 1.4882763435111236e-05,
+      "loss": 0.3751,
+      "step": 6863
+    },
+    {
+      "epoch": 1.353115141955836,
+      "grad_norm": 0.5783559792239383,
+      "learning_rate": 1.4881410685731152e-05,
+      "loss": 0.341,
+      "step": 6864
+    },
+    {
+      "epoch": 1.3533123028391167,
+      "grad_norm": 0.5284375900313468,
+      "learning_rate": 1.4880057819071244e-05,
+      "loss": 0.3455,
+      "step": 6865
+    },
+    {
+      "epoch": 1.3535094637223974,
+      "grad_norm": 0.47608340152473533,
+      "learning_rate": 1.4878704835164018e-05,
+      "loss": 0.2983,
+      "step": 6866
+    },
+    {
+      "epoch": 1.3537066246056781,
+      "grad_norm": 0.506845850224768,
+      "learning_rate": 1.4877351734041976e-05,
+      "loss": 0.3592,
+      "step": 6867
+    },
+    {
+      "epoch": 1.353903785488959,
+      "grad_norm": 0.4935101468881354,
+      "learning_rate": 1.4875998515737635e-05,
+      "loss": 0.3066,
+      "step": 6868
+    },
+    {
+      "epoch": 1.3541009463722398,
+      "grad_norm": 0.5059246656166723,
+      "learning_rate": 1.48746451802835e-05,
+      "loss": 0.3418,
+      "step": 6869
+    },
+    {
+      "epoch": 1.3542981072555205,
+      "grad_norm": 0.46928991796641456,
+      "learning_rate": 1.4873291727712094e-05,
+      "loss": 0.3037,
+      "step": 6870
+    },
+    {
+      "epoch": 1.3544952681388014,
+      "grad_norm": 0.5073410603661683,
+      "learning_rate": 1.4871938158055926e-05,
+      "loss": 0.3429,
+      "step": 6871
+    },
+    {
+      "epoch": 1.354692429022082,
+      "grad_norm": 0.5676528958205695,
+      "learning_rate": 1.487058447134752e-05,
+      "loss": 0.386,
+      "step": 6872
+    },
+    {
+      "epoch": 1.3548895899053628,
+      "grad_norm": 0.5449613542540548,
+      "learning_rate": 1.4869230667619399e-05,
+      "loss": 0.3391,
+      "step": 6873
+    },
+    {
+      "epoch": 1.3550867507886435,
+      "grad_norm": 0.5309796727926993,
+      "learning_rate": 1.4867876746904093e-05,
+      "loss": 0.3191,
+      "step": 6874
+    },
+    {
+      "epoch": 1.3552839116719242,
+      "grad_norm": 0.48139917834477924,
+      "learning_rate": 1.4866522709234125e-05,
+      "loss": 0.3207,
+      "step": 6875
+    },
+    {
+      "epoch": 1.3554810725552051,
+      "grad_norm": 0.4741924876595785,
+      "learning_rate": 1.4865168554642033e-05,
+      "loss": 0.3279,
+      "step": 6876
+    },
+    {
+      "epoch": 1.3556782334384858,
+      "grad_norm": 0.48900065703542145,
+      "learning_rate": 1.4863814283160348e-05,
+      "loss": 0.3399,
+      "step": 6877
+    },
+    {
+      "epoch": 1.3558753943217665,
+      "grad_norm": 0.5735243814672437,
+      "learning_rate": 1.4862459894821606e-05,
+      "loss": 0.3556,
+      "step": 6878
+    },
+    {
+      "epoch": 1.3560725552050474,
+      "grad_norm": 0.5398399252571137,
+      "learning_rate": 1.486110538965835e-05,
+      "loss": 0.384,
+      "step": 6879
+    },
+    {
+      "epoch": 1.3562697160883281,
+      "grad_norm": 0.5117636275917532,
+      "learning_rate": 1.4859750767703122e-05,
+      "loss": 0.3438,
+      "step": 6880
+    },
+    {
+      "epoch": 1.3564668769716088,
+      "grad_norm": 0.5051003334328982,
+      "learning_rate": 1.4858396028988472e-05,
+      "loss": 0.3428,
+      "step": 6881
+    },
+    {
+      "epoch": 1.3566640378548895,
+      "grad_norm": 0.5441154566116012,
+      "learning_rate": 1.4857041173546941e-05,
+      "loss": 0.3412,
+      "step": 6882
+    },
+    {
+      "epoch": 1.3568611987381702,
+      "grad_norm": 0.4798420407498326,
+      "learning_rate": 1.4855686201411086e-05,
+      "loss": 0.3448,
+      "step": 6883
+    },
+    {
+      "epoch": 1.3570583596214512,
+      "grad_norm": 0.5141849025940272,
+      "learning_rate": 1.485433111261346e-05,
+      "loss": 0.3407,
+      "step": 6884
+    },
+    {
+      "epoch": 1.3572555205047319,
+      "grad_norm": 0.534717445086196,
+      "learning_rate": 1.4852975907186618e-05,
+      "loss": 0.3731,
+      "step": 6885
+    },
+    {
+      "epoch": 1.3574526813880126,
+      "grad_norm": 0.5435535748807395,
+      "learning_rate": 1.4851620585163123e-05,
+      "loss": 0.3712,
+      "step": 6886
+    },
+    {
+      "epoch": 1.3576498422712935,
+      "grad_norm": 0.4963966514211321,
+      "learning_rate": 1.4850265146575535e-05,
+      "loss": 0.3203,
+      "step": 6887
+    },
+    {
+      "epoch": 1.3578470031545742,
+      "grad_norm": 0.5097541899694029,
+      "learning_rate": 1.4848909591456421e-05,
+      "loss": 0.3511,
+      "step": 6888
+    },
+    {
+      "epoch": 1.3580441640378549,
+      "grad_norm": 0.5022870460738562,
+      "learning_rate": 1.4847553919838353e-05,
+      "loss": 0.3459,
+      "step": 6889
+    },
+    {
+      "epoch": 1.3582413249211356,
+      "grad_norm": 0.5010019389495004,
+      "learning_rate": 1.4846198131753894e-05,
+      "loss": 0.349,
+      "step": 6890
+    },
+    {
+      "epoch": 1.3584384858044163,
+      "grad_norm": 0.4838676445079221,
+      "learning_rate": 1.4844842227235628e-05,
+      "loss": 0.3237,
+      "step": 6891
+    },
+    {
+      "epoch": 1.3586356466876972,
+      "grad_norm": 0.49788296424059325,
+      "learning_rate": 1.4843486206316122e-05,
+      "loss": 0.3319,
+      "step": 6892
+    },
+    {
+      "epoch": 1.358832807570978,
+      "grad_norm": 0.5331292005137582,
+      "learning_rate": 1.4842130069027957e-05,
+      "loss": 0.3446,
+      "step": 6893
+    },
+    {
+      "epoch": 1.3590299684542586,
+      "grad_norm": 0.5009763516051616,
+      "learning_rate": 1.4840773815403722e-05,
+      "loss": 0.3255,
+      "step": 6894
+    },
+    {
+      "epoch": 1.3592271293375395,
+      "grad_norm": 0.5154972070799596,
+      "learning_rate": 1.4839417445475995e-05,
+      "loss": 0.3382,
+      "step": 6895
+    },
+    {
+      "epoch": 1.3594242902208202,
+      "grad_norm": 0.522383735377928,
+      "learning_rate": 1.483806095927737e-05,
+      "loss": 0.3481,
+      "step": 6896
+    },
+    {
+      "epoch": 1.359621451104101,
+      "grad_norm": 0.5040201789415579,
+      "learning_rate": 1.4836704356840428e-05,
+      "loss": 0.3329,
+      "step": 6897
+    },
+    {
+      "epoch": 1.3598186119873816,
+      "grad_norm": 0.525037747965827,
+      "learning_rate": 1.4835347638197777e-05,
+      "loss": 0.3359,
+      "step": 6898
+    },
+    {
+      "epoch": 1.3600157728706623,
+      "grad_norm": 0.5051300920201072,
+      "learning_rate": 1.4833990803381997e-05,
+      "loss": 0.3467,
+      "step": 6899
+    },
+    {
+      "epoch": 1.3602129337539433,
+      "grad_norm": 0.4999739246930199,
+      "learning_rate": 1.4832633852425702e-05,
+      "loss": 0.361,
+      "step": 6900
+    },
+    {
+      "epoch": 1.360410094637224,
+      "grad_norm": 0.5107030410105877,
+      "learning_rate": 1.4831276785361484e-05,
+      "loss": 0.3556,
+      "step": 6901
+    },
+    {
+      "epoch": 1.3606072555205047,
+      "grad_norm": 0.4974505550953817,
+      "learning_rate": 1.4829919602221949e-05,
+      "loss": 0.3282,
+      "step": 6902
+    },
+    {
+      "epoch": 1.3608044164037856,
+      "grad_norm": 0.48064179340063234,
+      "learning_rate": 1.4828562303039708e-05,
+      "loss": 0.3416,
+      "step": 6903
+    },
+    {
+      "epoch": 1.3610015772870663,
+      "grad_norm": 0.5127276668204793,
+      "learning_rate": 1.4827204887847369e-05,
+      "loss": 0.3347,
+      "step": 6904
+    },
+    {
+      "epoch": 1.361198738170347,
+      "grad_norm": 0.4780273243210127,
+      "learning_rate": 1.4825847356677546e-05,
+      "loss": 0.304,
+      "step": 6905
+    },
+    {
+      "epoch": 1.3613958990536277,
+      "grad_norm": 0.4887295391580163,
+      "learning_rate": 1.4824489709562854e-05,
+      "loss": 0.3327,
+      "step": 6906
+    },
+    {
+      "epoch": 1.3615930599369084,
+      "grad_norm": 0.48670211288915743,
+      "learning_rate": 1.4823131946535912e-05,
+      "loss": 0.3355,
+      "step": 6907
+    },
+    {
+      "epoch": 1.3617902208201893,
+      "grad_norm": 0.5294737580364534,
+      "learning_rate": 1.4821774067629338e-05,
+      "loss": 0.3631,
+      "step": 6908
+    },
+    {
+      "epoch": 1.36198738170347,
+      "grad_norm": 0.4803323354930984,
+      "learning_rate": 1.482041607287576e-05,
+      "loss": 0.3238,
+      "step": 6909
+    },
+    {
+      "epoch": 1.3621845425867507,
+      "grad_norm": 0.5172113196727169,
+      "learning_rate": 1.4819057962307805e-05,
+      "loss": 0.3276,
+      "step": 6910
+    },
+    {
+      "epoch": 1.3623817034700316,
+      "grad_norm": 0.5122817965816258,
+      "learning_rate": 1.4817699735958103e-05,
+      "loss": 0.34,
+      "step": 6911
+    },
+    {
+      "epoch": 1.3625788643533123,
+      "grad_norm": 0.500184273669358,
+      "learning_rate": 1.4816341393859283e-05,
+      "loss": 0.3579,
+      "step": 6912
+    },
+    {
+      "epoch": 1.362776025236593,
+      "grad_norm": 0.49730830810220844,
+      "learning_rate": 1.4814982936043984e-05,
+      "loss": 0.3338,
+      "step": 6913
+    },
+    {
+      "epoch": 1.362973186119874,
+      "grad_norm": 0.4523508533879372,
+      "learning_rate": 1.481362436254484e-05,
+      "loss": 0.3122,
+      "step": 6914
+    },
+    {
+      "epoch": 1.3631703470031546,
+      "grad_norm": 0.5318999450446498,
+      "learning_rate": 1.4812265673394496e-05,
+      "loss": 0.3442,
+      "step": 6915
+    },
+    {
+      "epoch": 1.3633675078864353,
+      "grad_norm": 0.4991165752969515,
+      "learning_rate": 1.4810906868625595e-05,
+      "loss": 0.3403,
+      "step": 6916
+    },
+    {
+      "epoch": 1.363564668769716,
+      "grad_norm": 0.5238776570531024,
+      "learning_rate": 1.4809547948270782e-05,
+      "loss": 0.358,
+      "step": 6917
+    },
+    {
+      "epoch": 1.3637618296529967,
+      "grad_norm": 0.4971815188764086,
+      "learning_rate": 1.4808188912362705e-05,
+      "loss": 0.3149,
+      "step": 6918
+    },
+    {
+      "epoch": 1.3639589905362777,
+      "grad_norm": 0.5124620131345615,
+      "learning_rate": 1.4806829760934018e-05,
+      "loss": 0.3232,
+      "step": 6919
+    },
+    {
+      "epoch": 1.3641561514195584,
+      "grad_norm": 0.5178627263514879,
+      "learning_rate": 1.4805470494017373e-05,
+      "loss": 0.3513,
+      "step": 6920
+    },
+    {
+      "epoch": 1.364353312302839,
+      "grad_norm": 0.52380799206788,
+      "learning_rate": 1.4804111111645434e-05,
+      "loss": 0.3487,
+      "step": 6921
+    },
+    {
+      "epoch": 1.36455047318612,
+      "grad_norm": 0.48203134434437056,
+      "learning_rate": 1.4802751613850853e-05,
+      "loss": 0.3213,
+      "step": 6922
+    },
+    {
+      "epoch": 1.3647476340694007,
+      "grad_norm": 0.5102651067337406,
+      "learning_rate": 1.4801392000666297e-05,
+      "loss": 0.3266,
+      "step": 6923
+    },
+    {
+      "epoch": 1.3649447949526814,
+      "grad_norm": 0.5414007491135933,
+      "learning_rate": 1.4800032272124432e-05,
+      "loss": 0.3818,
+      "step": 6924
+    },
+    {
+      "epoch": 1.365141955835962,
+      "grad_norm": 0.5237634060391902,
+      "learning_rate": 1.4798672428257928e-05,
+      "loss": 0.3349,
+      "step": 6925
+    },
+    {
+      "epoch": 1.3653391167192428,
+      "grad_norm": 0.5011403516477672,
+      "learning_rate": 1.4797312469099454e-05,
+      "loss": 0.3213,
+      "step": 6926
+    },
+    {
+      "epoch": 1.3655362776025237,
+      "grad_norm": 0.5571042867990981,
+      "learning_rate": 1.4795952394681682e-05,
+      "loss": 0.3691,
+      "step": 6927
+    },
+    {
+      "epoch": 1.3657334384858044,
+      "grad_norm": 0.48652049099214323,
+      "learning_rate": 1.4794592205037295e-05,
+      "loss": 0.3387,
+      "step": 6928
+    },
+    {
+      "epoch": 1.3659305993690851,
+      "grad_norm": 0.4823341354472324,
+      "learning_rate": 1.4793231900198968e-05,
+      "loss": 0.3235,
+      "step": 6929
+    },
+    {
+      "epoch": 1.366127760252366,
+      "grad_norm": 0.5152844218945443,
+      "learning_rate": 1.4791871480199385e-05,
+      "loss": 0.3405,
+      "step": 6930
+    },
+    {
+      "epoch": 1.3663249211356467,
+      "grad_norm": 0.4901566411635077,
+      "learning_rate": 1.479051094507123e-05,
+      "loss": 0.328,
+      "step": 6931
+    },
+    {
+      "epoch": 1.3665220820189274,
+      "grad_norm": 0.5243962870706206,
+      "learning_rate": 1.4789150294847192e-05,
+      "loss": 0.3436,
+      "step": 6932
+    },
+    {
+      "epoch": 1.3667192429022081,
+      "grad_norm": 0.5194816070602105,
+      "learning_rate": 1.4787789529559961e-05,
+      "loss": 0.3309,
+      "step": 6933
+    },
+    {
+      "epoch": 1.3669164037854888,
+      "grad_norm": 0.5123859611810547,
+      "learning_rate": 1.4786428649242232e-05,
+      "loss": 0.3367,
+      "step": 6934
+    },
+    {
+      "epoch": 1.3671135646687698,
+      "grad_norm": 0.4896479386075604,
+      "learning_rate": 1.4785067653926701e-05,
+      "loss": 0.3204,
+      "step": 6935
+    },
+    {
+      "epoch": 1.3673107255520505,
+      "grad_norm": 0.49790804021992413,
+      "learning_rate": 1.4783706543646066e-05,
+      "loss": 0.3368,
+      "step": 6936
+    },
+    {
+      "epoch": 1.3675078864353312,
+      "grad_norm": 0.5019017261167511,
+      "learning_rate": 1.4782345318433025e-05,
+      "loss": 0.3278,
+      "step": 6937
+    },
+    {
+      "epoch": 1.367705047318612,
+      "grad_norm": 0.4904028947701245,
+      "learning_rate": 1.478098397832029e-05,
+      "loss": 0.3129,
+      "step": 6938
+    },
+    {
+      "epoch": 1.3679022082018928,
+      "grad_norm": 0.5321594647922132,
+      "learning_rate": 1.4779622523340562e-05,
+      "loss": 0.3425,
+      "step": 6939
+    },
+    {
+      "epoch": 1.3680993690851735,
+      "grad_norm": 0.4986035776438812,
+      "learning_rate": 1.477826095352656e-05,
+      "loss": 0.3367,
+      "step": 6940
+    },
+    {
+      "epoch": 1.3682965299684542,
+      "grad_norm": 0.5246630991127984,
+      "learning_rate": 1.4776899268910985e-05,
+      "loss": 0.3542,
+      "step": 6941
+    },
+    {
+      "epoch": 1.368493690851735,
+      "grad_norm": 0.4648451353245271,
+      "learning_rate": 1.477553746952656e-05,
+      "loss": 0.3196,
+      "step": 6942
+    },
+    {
+      "epoch": 1.3686908517350158,
+      "grad_norm": 0.546723814002127,
+      "learning_rate": 1.4774175555406e-05,
+      "loss": 0.3466,
+      "step": 6943
+    },
+    {
+      "epoch": 1.3688880126182965,
+      "grad_norm": 0.5018774391569213,
+      "learning_rate": 1.477281352658203e-05,
+      "loss": 0.3239,
+      "step": 6944
+    },
+    {
+      "epoch": 1.3690851735015772,
+      "grad_norm": 0.4926341199579351,
+      "learning_rate": 1.4771451383087373e-05,
+      "loss": 0.3356,
+      "step": 6945
+    },
+    {
+      "epoch": 1.3692823343848581,
+      "grad_norm": 0.5338603435373444,
+      "learning_rate": 1.477008912495475e-05,
+      "loss": 0.3629,
+      "step": 6946
+    },
+    {
+      "epoch": 1.3694794952681388,
+      "grad_norm": 0.46986959001400663,
+      "learning_rate": 1.4768726752216898e-05,
+      "loss": 0.3107,
+      "step": 6947
+    },
+    {
+      "epoch": 1.3696766561514195,
+      "grad_norm": 0.5335668009350745,
+      "learning_rate": 1.4767364264906542e-05,
+      "loss": 0.3481,
+      "step": 6948
+    },
+    {
+      "epoch": 1.3698738170347002,
+      "grad_norm": 0.4653687936969232,
+      "learning_rate": 1.4766001663056422e-05,
+      "loss": 0.3155,
+      "step": 6949
+    },
+    {
+      "epoch": 1.370070977917981,
+      "grad_norm": 0.5286992819313316,
+      "learning_rate": 1.4764638946699275e-05,
+      "loss": 0.3449,
+      "step": 6950
+    },
+    {
+      "epoch": 1.3702681388012619,
+      "grad_norm": 0.5307937528972434,
+      "learning_rate": 1.476327611586784e-05,
+      "loss": 0.3482,
+      "step": 6951
+    },
+    {
+      "epoch": 1.3704652996845426,
+      "grad_norm": 0.47920732199670424,
+      "learning_rate": 1.4761913170594859e-05,
+      "loss": 0.3334,
+      "step": 6952
+    },
+    {
+      "epoch": 1.3706624605678233,
+      "grad_norm": 0.4986071058412949,
+      "learning_rate": 1.4760550110913081e-05,
+      "loss": 0.3575,
+      "step": 6953
+    },
+    {
+      "epoch": 1.3708596214511042,
+      "grad_norm": 0.48580007339999265,
+      "learning_rate": 1.4759186936855253e-05,
+      "loss": 0.3166,
+      "step": 6954
+    },
+    {
+      "epoch": 1.3710567823343849,
+      "grad_norm": 0.4847134680546892,
+      "learning_rate": 1.4757823648454124e-05,
+      "loss": 0.3355,
+      "step": 6955
+    },
+    {
+      "epoch": 1.3712539432176656,
+      "grad_norm": 0.49892520436181714,
+      "learning_rate": 1.475646024574245e-05,
+      "loss": 0.3437,
+      "step": 6956
+    },
+    {
+      "epoch": 1.3714511041009465,
+      "grad_norm": 0.46579178489145134,
+      "learning_rate": 1.4755096728752992e-05,
+      "loss": 0.3251,
+      "step": 6957
+    },
+    {
+      "epoch": 1.3716482649842272,
+      "grad_norm": 0.4854313491900763,
+      "learning_rate": 1.4753733097518503e-05,
+      "loss": 0.3197,
+      "step": 6958
+    },
+    {
+      "epoch": 1.371845425867508,
+      "grad_norm": 0.5084781508505444,
+      "learning_rate": 1.475236935207175e-05,
+      "loss": 0.3447,
+      "step": 6959
+    },
+    {
+      "epoch": 1.3720425867507886,
+      "grad_norm": 0.5154369206311532,
+      "learning_rate": 1.4751005492445496e-05,
+      "loss": 0.3022,
+      "step": 6960
+    },
+    {
+      "epoch": 1.3722397476340693,
+      "grad_norm": 0.5072733874919334,
+      "learning_rate": 1.4749641518672508e-05,
+      "loss": 0.3478,
+      "step": 6961
+    },
+    {
+      "epoch": 1.3724369085173502,
+      "grad_norm": 0.5276500649999918,
+      "learning_rate": 1.4748277430785557e-05,
+      "loss": 0.3433,
+      "step": 6962
+    },
+    {
+      "epoch": 1.372634069400631,
+      "grad_norm": 0.5068275820907489,
+      "learning_rate": 1.4746913228817416e-05,
+      "loss": 0.3324,
+      "step": 6963
+    },
+    {
+      "epoch": 1.3728312302839116,
+      "grad_norm": 0.5588764584629632,
+      "learning_rate": 1.4745548912800867e-05,
+      "loss": 0.381,
+      "step": 6964
+    },
+    {
+      "epoch": 1.3730283911671926,
+      "grad_norm": 0.48397923052902175,
+      "learning_rate": 1.4744184482768678e-05,
+      "loss": 0.3241,
+      "step": 6965
+    },
+    {
+      "epoch": 1.3732255520504733,
+      "grad_norm": 0.5009095018317993,
+      "learning_rate": 1.4742819938753641e-05,
+      "loss": 0.3374,
+      "step": 6966
+    },
+    {
+      "epoch": 1.373422712933754,
+      "grad_norm": 0.5214836401414918,
+      "learning_rate": 1.4741455280788533e-05,
+      "loss": 0.3338,
+      "step": 6967
+    },
+    {
+      "epoch": 1.3736198738170347,
+      "grad_norm": 0.4969774753340322,
+      "learning_rate": 1.4740090508906147e-05,
+      "loss": 0.3243,
+      "step": 6968
+    },
+    {
+      "epoch": 1.3738170347003154,
+      "grad_norm": 0.5559922169108119,
+      "learning_rate": 1.4738725623139263e-05,
+      "loss": 0.3567,
+      "step": 6969
+    },
+    {
+      "epoch": 1.3740141955835963,
+      "grad_norm": 0.5165083267867453,
+      "learning_rate": 1.4737360623520684e-05,
+      "loss": 0.3324,
+      "step": 6970
+    },
+    {
+      "epoch": 1.374211356466877,
+      "grad_norm": 0.5186856386291233,
+      "learning_rate": 1.47359955100832e-05,
+      "loss": 0.3333,
+      "step": 6971
+    },
+    {
+      "epoch": 1.3744085173501577,
+      "grad_norm": 0.5238837322487031,
+      "learning_rate": 1.473463028285961e-05,
+      "loss": 0.3484,
+      "step": 6972
+    },
+    {
+      "epoch": 1.3746056782334386,
+      "grad_norm": 2.292582007349146,
+      "learning_rate": 1.4733264941882714e-05,
+      "loss": 0.3815,
+      "step": 6973
+    },
+    {
+      "epoch": 1.3748028391167193,
+      "grad_norm": 0.5643806541463167,
+      "learning_rate": 1.4731899487185319e-05,
+      "loss": 0.3442,
+      "step": 6974
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 0.47507958786208415,
+      "learning_rate": 1.4730533918800227e-05,
+      "loss": 0.3075,
+      "step": 6975
+    },
+    {
+      "epoch": 1.3751971608832807,
+      "grad_norm": 0.4912700593505576,
+      "learning_rate": 1.4729168236760248e-05,
+      "loss": 0.3229,
+      "step": 6976
+    },
+    {
+      "epoch": 1.3753943217665614,
+      "grad_norm": 0.5116118374339484,
+      "learning_rate": 1.4727802441098193e-05,
+      "loss": 0.3345,
+      "step": 6977
+    },
+    {
+      "epoch": 1.3755914826498423,
+      "grad_norm": 0.5865259693172717,
+      "learning_rate": 1.4726436531846877e-05,
+      "loss": 0.3333,
+      "step": 6978
+    },
+    {
+      "epoch": 1.375788643533123,
+      "grad_norm": 0.49366649497140386,
+      "learning_rate": 1.4725070509039117e-05,
+      "loss": 0.3337,
+      "step": 6979
+    },
+    {
+      "epoch": 1.3759858044164037,
+      "grad_norm": 0.5372885703885392,
+      "learning_rate": 1.4723704372707734e-05,
+      "loss": 0.3287,
+      "step": 6980
+    },
+    {
+      "epoch": 1.3761829652996846,
+      "grad_norm": 0.5043319157124423,
+      "learning_rate": 1.4722338122885548e-05,
+      "loss": 0.3396,
+      "step": 6981
+    },
+    {
+      "epoch": 1.3763801261829653,
+      "grad_norm": 0.6345501977793165,
+      "learning_rate": 1.4720971759605387e-05,
+      "loss": 0.3289,
+      "step": 6982
+    },
+    {
+      "epoch": 1.376577287066246,
+      "grad_norm": 0.5375836546232604,
+      "learning_rate": 1.4719605282900077e-05,
+      "loss": 0.3757,
+      "step": 6983
+    },
+    {
+      "epoch": 1.3767744479495267,
+      "grad_norm": 0.5178077248278057,
+      "learning_rate": 1.4718238692802449e-05,
+      "loss": 0.3594,
+      "step": 6984
+    },
+    {
+      "epoch": 1.3769716088328074,
+      "grad_norm": 0.5147912712127123,
+      "learning_rate": 1.4716871989345338e-05,
+      "loss": 0.3474,
+      "step": 6985
+    },
+    {
+      "epoch": 1.3771687697160884,
+      "grad_norm": 0.5189874060272969,
+      "learning_rate": 1.4715505172561577e-05,
+      "loss": 0.3247,
+      "step": 6986
+    },
+    {
+      "epoch": 1.377365930599369,
+      "grad_norm": 0.5281430394013332,
+      "learning_rate": 1.471413824248401e-05,
+      "loss": 0.3568,
+      "step": 6987
+    },
+    {
+      "epoch": 1.3775630914826498,
+      "grad_norm": 0.5147078126304419,
+      "learning_rate": 1.4712771199145472e-05,
+      "loss": 0.3559,
+      "step": 6988
+    },
+    {
+      "epoch": 1.3777602523659307,
+      "grad_norm": 0.49074990134353796,
+      "learning_rate": 1.4711404042578814e-05,
+      "loss": 0.3416,
+      "step": 6989
+    },
+    {
+      "epoch": 1.3779574132492114,
+      "grad_norm": 0.4841551234965781,
+      "learning_rate": 1.4710036772816877e-05,
+      "loss": 0.3476,
+      "step": 6990
+    },
+    {
+      "epoch": 1.378154574132492,
+      "grad_norm": 0.5076124986937911,
+      "learning_rate": 1.4708669389892514e-05,
+      "loss": 0.3451,
+      "step": 6991
+    },
+    {
+      "epoch": 1.3783517350157728,
+      "grad_norm": 0.5101395115062465,
+      "learning_rate": 1.4707301893838578e-05,
+      "loss": 0.337,
+      "step": 6992
+    },
+    {
+      "epoch": 1.3785488958990535,
+      "grad_norm": 0.47965592126549544,
+      "learning_rate": 1.4705934284687923e-05,
+      "loss": 0.3242,
+      "step": 6993
+    },
+    {
+      "epoch": 1.3787460567823344,
+      "grad_norm": 0.5219389338261724,
+      "learning_rate": 1.4704566562473408e-05,
+      "loss": 0.3683,
+      "step": 6994
+    },
+    {
+      "epoch": 1.3789432176656151,
+      "grad_norm": 0.5205227382874077,
+      "learning_rate": 1.4703198727227892e-05,
+      "loss": 0.3318,
+      "step": 6995
+    },
+    {
+      "epoch": 1.3791403785488958,
+      "grad_norm": 0.48222107531876607,
+      "learning_rate": 1.4701830778984239e-05,
+      "loss": 0.3253,
+      "step": 6996
+    },
+    {
+      "epoch": 1.3793375394321767,
+      "grad_norm": 0.49353311997235216,
+      "learning_rate": 1.4700462717775317e-05,
+      "loss": 0.3219,
+      "step": 6997
+    },
+    {
+      "epoch": 1.3795347003154574,
+      "grad_norm": 0.5345425712355901,
+      "learning_rate": 1.4699094543633989e-05,
+      "loss": 0.3624,
+      "step": 6998
+    },
+    {
+      "epoch": 1.3797318611987381,
+      "grad_norm": 0.5050606461841181,
+      "learning_rate": 1.4697726256593132e-05,
+      "loss": 0.3441,
+      "step": 6999
+    },
+    {
+      "epoch": 1.379929022082019,
+      "grad_norm": 0.47478110398234236,
+      "learning_rate": 1.469635785668562e-05,
+      "loss": 0.3204,
+      "step": 7000
+    },
+    {
+      "epoch": 1.3801261829652998,
+      "grad_norm": 0.503768054032645,
+      "learning_rate": 1.4694989343944327e-05,
+      "loss": 0.3508,
+      "step": 7001
+    },
+    {
+      "epoch": 1.3803233438485805,
+      "grad_norm": 0.6492703807376875,
+      "learning_rate": 1.4693620718402137e-05,
+      "loss": 0.3212,
+      "step": 7002
+    },
+    {
+      "epoch": 1.3805205047318612,
+      "grad_norm": 0.4805328225116543,
+      "learning_rate": 1.4692251980091927e-05,
+      "loss": 0.3254,
+      "step": 7003
+    },
+    {
+      "epoch": 1.3807176656151419,
+      "grad_norm": 0.4842656353572685,
+      "learning_rate": 1.4690883129046585e-05,
+      "loss": 0.3372,
+      "step": 7004
+    },
+    {
+      "epoch": 1.3809148264984228,
+      "grad_norm": 0.5211693690974576,
+      "learning_rate": 1.4689514165298995e-05,
+      "loss": 0.3491,
+      "step": 7005
+    },
+    {
+      "epoch": 1.3811119873817035,
+      "grad_norm": 0.5885371549295725,
+      "learning_rate": 1.4688145088882056e-05,
+      "loss": 0.3444,
+      "step": 7006
+    },
+    {
+      "epoch": 1.3813091482649842,
+      "grad_norm": 0.4754832288654815,
+      "learning_rate": 1.4686775899828651e-05,
+      "loss": 0.3224,
+      "step": 7007
+    },
+    {
+      "epoch": 1.381506309148265,
+      "grad_norm": 0.4780150559237149,
+      "learning_rate": 1.4685406598171686e-05,
+      "loss": 0.3052,
+      "step": 7008
+    },
+    {
+      "epoch": 1.3817034700315458,
+      "grad_norm": 0.49765867952072046,
+      "learning_rate": 1.4684037183944051e-05,
+      "loss": 0.3434,
+      "step": 7009
+    },
+    {
+      "epoch": 1.3819006309148265,
+      "grad_norm": 0.47835652909606524,
+      "learning_rate": 1.4682667657178653e-05,
+      "loss": 0.3396,
+      "step": 7010
+    },
+    {
+      "epoch": 1.3820977917981072,
+      "grad_norm": 0.46329597154012175,
+      "learning_rate": 1.4681298017908391e-05,
+      "loss": 0.3226,
+      "step": 7011
+    },
+    {
+      "epoch": 1.382294952681388,
+      "grad_norm": 0.49817719336350075,
+      "learning_rate": 1.4679928266166175e-05,
+      "loss": 0.3357,
+      "step": 7012
+    },
+    {
+      "epoch": 1.3824921135646688,
+      "grad_norm": 0.8663152836331668,
+      "learning_rate": 1.4678558401984915e-05,
+      "loss": 0.3431,
+      "step": 7013
+    },
+    {
+      "epoch": 1.3826892744479495,
+      "grad_norm": 0.5197376334684173,
+      "learning_rate": 1.467718842539752e-05,
+      "loss": 0.3367,
+      "step": 7014
+    },
+    {
+      "epoch": 1.3828864353312302,
+      "grad_norm": 0.5621592048068038,
+      "learning_rate": 1.467581833643691e-05,
+      "loss": 0.3152,
+      "step": 7015
+    },
+    {
+      "epoch": 1.3830835962145112,
+      "grad_norm": 0.5392174642427607,
+      "learning_rate": 1.4674448135135993e-05,
+      "loss": 0.348,
+      "step": 7016
+    },
+    {
+      "epoch": 1.3832807570977919,
+      "grad_norm": 0.4767809373705618,
+      "learning_rate": 1.46730778215277e-05,
+      "loss": 0.3133,
+      "step": 7017
+    },
+    {
+      "epoch": 1.3834779179810726,
+      "grad_norm": 1.1165470822976469,
+      "learning_rate": 1.4671707395644946e-05,
+      "loss": 0.3479,
+      "step": 7018
+    },
+    {
+      "epoch": 1.3836750788643533,
+      "grad_norm": 0.5264241691242134,
+      "learning_rate": 1.4670336857520661e-05,
+      "loss": 0.3528,
+      "step": 7019
+    },
+    {
+      "epoch": 1.383872239747634,
+      "grad_norm": 0.4677220107524557,
+      "learning_rate": 1.4668966207187774e-05,
+      "loss": 0.3322,
+      "step": 7020
+    },
+    {
+      "epoch": 1.3840694006309149,
+      "grad_norm": 0.4729142635567184,
+      "learning_rate": 1.4667595444679212e-05,
+      "loss": 0.3265,
+      "step": 7021
+    },
+    {
+      "epoch": 1.3842665615141956,
+      "grad_norm": 0.5851446109658627,
+      "learning_rate": 1.466622457002791e-05,
+      "loss": 0.3491,
+      "step": 7022
+    },
+    {
+      "epoch": 1.3844637223974763,
+      "grad_norm": 0.5195972584107764,
+      "learning_rate": 1.4664853583266807e-05,
+      "loss": 0.3408,
+      "step": 7023
+    },
+    {
+      "epoch": 1.3846608832807572,
+      "grad_norm": 0.48331307744775936,
+      "learning_rate": 1.4663482484428839e-05,
+      "loss": 0.3438,
+      "step": 7024
+    },
+    {
+      "epoch": 1.384858044164038,
+      "grad_norm": 0.4872542154414702,
+      "learning_rate": 1.4662111273546949e-05,
+      "loss": 0.3116,
+      "step": 7025
+    },
+    {
+      "epoch": 1.3850552050473186,
+      "grad_norm": 0.5209465468531163,
+      "learning_rate": 1.4660739950654081e-05,
+      "loss": 0.3713,
+      "step": 7026
+    },
+    {
+      "epoch": 1.3852523659305993,
+      "grad_norm": 0.4946017796024822,
+      "learning_rate": 1.4659368515783183e-05,
+      "loss": 0.3238,
+      "step": 7027
+    },
+    {
+      "epoch": 1.38544952681388,
+      "grad_norm": 0.5092592005569798,
+      "learning_rate": 1.4657996968967202e-05,
+      "loss": 0.3456,
+      "step": 7028
+    },
+    {
+      "epoch": 1.385646687697161,
+      "grad_norm": 0.47093221898089077,
+      "learning_rate": 1.4656625310239095e-05,
+      "loss": 0.3068,
+      "step": 7029
+    },
+    {
+      "epoch": 1.3858438485804416,
+      "grad_norm": 0.5092749341911277,
+      "learning_rate": 1.4655253539631816e-05,
+      "loss": 0.3501,
+      "step": 7030
+    },
+    {
+      "epoch": 1.3860410094637223,
+      "grad_norm": 0.49688273138146666,
+      "learning_rate": 1.4653881657178317e-05,
+      "loss": 0.329,
+      "step": 7031
+    },
+    {
+      "epoch": 1.3862381703470033,
+      "grad_norm": 0.5210667444751383,
+      "learning_rate": 1.465250966291157e-05,
+      "loss": 0.3308,
+      "step": 7032
+    },
+    {
+      "epoch": 1.386435331230284,
+      "grad_norm": 0.5051826303648079,
+      "learning_rate": 1.4651137556864526e-05,
+      "loss": 0.3246,
+      "step": 7033
+    },
+    {
+      "epoch": 1.3866324921135647,
+      "grad_norm": 0.5131056295583875,
+      "learning_rate": 1.4649765339070161e-05,
+      "loss": 0.3472,
+      "step": 7034
+    },
+    {
+      "epoch": 1.3868296529968454,
+      "grad_norm": 0.5045828989240048,
+      "learning_rate": 1.4648393009561434e-05,
+      "loss": 0.35,
+      "step": 7035
+    },
+    {
+      "epoch": 1.387026813880126,
+      "grad_norm": 2.3607682081415566,
+      "learning_rate": 1.4647020568371329e-05,
+      "loss": 0.3819,
+      "step": 7036
+    },
+    {
+      "epoch": 1.387223974763407,
+      "grad_norm": 0.7063638677947639,
+      "learning_rate": 1.4645648015532806e-05,
+      "loss": 0.3035,
+      "step": 7037
+    },
+    {
+      "epoch": 1.3874211356466877,
+      "grad_norm": 0.5347651370289318,
+      "learning_rate": 1.464427535107885e-05,
+      "loss": 0.3415,
+      "step": 7038
+    },
+    {
+      "epoch": 1.3876182965299684,
+      "grad_norm": 0.5063799018311032,
+      "learning_rate": 1.4642902575042439e-05,
+      "loss": 0.3682,
+      "step": 7039
+    },
+    {
+      "epoch": 1.3878154574132493,
+      "grad_norm": 0.5418753232993281,
+      "learning_rate": 1.4641529687456558e-05,
+      "loss": 0.3801,
+      "step": 7040
+    },
+    {
+      "epoch": 1.38801261829653,
+      "grad_norm": 0.510285067684664,
+      "learning_rate": 1.4640156688354183e-05,
+      "loss": 0.3417,
+      "step": 7041
+    },
+    {
+      "epoch": 1.3882097791798107,
+      "grad_norm": 0.5074567362081557,
+      "learning_rate": 1.4638783577768312e-05,
+      "loss": 0.3426,
+      "step": 7042
+    },
+    {
+      "epoch": 1.3884069400630916,
+      "grad_norm": 0.4986711709526979,
+      "learning_rate": 1.4637410355731927e-05,
+      "loss": 0.3517,
+      "step": 7043
+    },
+    {
+      "epoch": 1.3886041009463723,
+      "grad_norm": 0.4834879827214969,
+      "learning_rate": 1.4636037022278022e-05,
+      "loss": 0.325,
+      "step": 7044
+    },
+    {
+      "epoch": 1.388801261829653,
+      "grad_norm": 0.4725810540995025,
+      "learning_rate": 1.4634663577439598e-05,
+      "loss": 0.3329,
+      "step": 7045
+    },
+    {
+      "epoch": 1.3889984227129337,
+      "grad_norm": 0.499687944712635,
+      "learning_rate": 1.4633290021249646e-05,
+      "loss": 0.3261,
+      "step": 7046
+    },
+    {
+      "epoch": 1.3891955835962144,
+      "grad_norm": 0.49600002260141485,
+      "learning_rate": 1.4631916353741174e-05,
+      "loss": 0.3625,
+      "step": 7047
+    },
+    {
+      "epoch": 1.3893927444794953,
+      "grad_norm": 0.5362771118840135,
+      "learning_rate": 1.4630542574947177e-05,
+      "loss": 0.3511,
+      "step": 7048
+    },
+    {
+      "epoch": 1.389589905362776,
+      "grad_norm": 0.49021230005905664,
+      "learning_rate": 1.462916868490067e-05,
+      "loss": 0.3459,
+      "step": 7049
+    },
+    {
+      "epoch": 1.3897870662460567,
+      "grad_norm": 0.5226203596978446,
+      "learning_rate": 1.4627794683634655e-05,
+      "loss": 0.3359,
+      "step": 7050
+    },
+    {
+      "epoch": 1.3899842271293377,
+      "grad_norm": 0.4892949036036693,
+      "learning_rate": 1.4626420571182146e-05,
+      "loss": 0.3354,
+      "step": 7051
+    },
+    {
+      "epoch": 1.3901813880126184,
+      "grad_norm": 0.49319072511760836,
+      "learning_rate": 1.4625046347576155e-05,
+      "loss": 0.3213,
+      "step": 7052
+    },
+    {
+      "epoch": 1.390378548895899,
+      "grad_norm": 0.48026160014473557,
+      "learning_rate": 1.4623672012849705e-05,
+      "loss": 0.3328,
+      "step": 7053
+    },
+    {
+      "epoch": 1.3905757097791798,
+      "grad_norm": 0.519599979648436,
+      "learning_rate": 1.462229756703581e-05,
+      "loss": 0.3768,
+      "step": 7054
+    },
+    {
+      "epoch": 1.3907728706624605,
+      "grad_norm": 0.5158052836985166,
+      "learning_rate": 1.4620923010167496e-05,
+      "loss": 0.3472,
+      "step": 7055
+    },
+    {
+      "epoch": 1.3909700315457414,
+      "grad_norm": 0.49003976941752064,
+      "learning_rate": 1.461954834227778e-05,
+      "loss": 0.3197,
+      "step": 7056
+    },
+    {
+      "epoch": 1.391167192429022,
+      "grad_norm": 0.47468028680465263,
+      "learning_rate": 1.46181735633997e-05,
+      "loss": 0.3163,
+      "step": 7057
+    },
+    {
+      "epoch": 1.3913643533123028,
+      "grad_norm": 0.4951911330562906,
+      "learning_rate": 1.4616798673566276e-05,
+      "loss": 0.3479,
+      "step": 7058
+    },
+    {
+      "epoch": 1.3915615141955837,
+      "grad_norm": 0.4785035909764805,
+      "learning_rate": 1.4615423672810549e-05,
+      "loss": 0.3244,
+      "step": 7059
+    },
+    {
+      "epoch": 1.3917586750788644,
+      "grad_norm": 0.46972733617436485,
+      "learning_rate": 1.4614048561165552e-05,
+      "loss": 0.3264,
+      "step": 7060
+    },
+    {
+      "epoch": 1.3919558359621451,
+      "grad_norm": 0.5045776545128406,
+      "learning_rate": 1.4612673338664322e-05,
+      "loss": 0.3555,
+      "step": 7061
+    },
+    {
+      "epoch": 1.3921529968454258,
+      "grad_norm": 0.4873523714181044,
+      "learning_rate": 1.46112980053399e-05,
+      "loss": 0.3357,
+      "step": 7062
+    },
+    {
+      "epoch": 1.3923501577287065,
+      "grad_norm": 0.5568338246839404,
+      "learning_rate": 1.460992256122533e-05,
+      "loss": 0.3606,
+      "step": 7063
+    },
+    {
+      "epoch": 1.3925473186119874,
+      "grad_norm": 0.5056504428860361,
+      "learning_rate": 1.4608547006353661e-05,
+      "loss": 0.3515,
+      "step": 7064
+    },
+    {
+      "epoch": 1.3927444794952681,
+      "grad_norm": 0.5637743705182227,
+      "learning_rate": 1.4607171340757935e-05,
+      "loss": 0.387,
+      "step": 7065
+    },
+    {
+      "epoch": 1.3929416403785488,
+      "grad_norm": 0.5012198726529813,
+      "learning_rate": 1.460579556447121e-05,
+      "loss": 0.3339,
+      "step": 7066
+    },
+    {
+      "epoch": 1.3931388012618298,
+      "grad_norm": 0.5255783486737358,
+      "learning_rate": 1.4604419677526536e-05,
+      "loss": 0.34,
+      "step": 7067
+    },
+    {
+      "epoch": 1.3933359621451105,
+      "grad_norm": 0.49793007788388866,
+      "learning_rate": 1.4603043679956972e-05,
+      "loss": 0.3364,
+      "step": 7068
+    },
+    {
+      "epoch": 1.3935331230283912,
+      "grad_norm": 0.5065921955646553,
+      "learning_rate": 1.4601667571795577e-05,
+      "loss": 0.3439,
+      "step": 7069
+    },
+    {
+      "epoch": 1.3937302839116719,
+      "grad_norm": 0.46887161071080763,
+      "learning_rate": 1.4600291353075413e-05,
+      "loss": 0.3174,
+      "step": 7070
+    },
+    {
+      "epoch": 1.3939274447949526,
+      "grad_norm": 0.5473282371408335,
+      "learning_rate": 1.4598915023829543e-05,
+      "loss": 0.3446,
+      "step": 7071
+    },
+    {
+      "epoch": 1.3941246056782335,
+      "grad_norm": 0.5483532718507378,
+      "learning_rate": 1.4597538584091038e-05,
+      "loss": 0.356,
+      "step": 7072
+    },
+    {
+      "epoch": 1.3943217665615142,
+      "grad_norm": 0.5316150064778802,
+      "learning_rate": 1.4596162033892962e-05,
+      "loss": 0.3371,
+      "step": 7073
+    },
+    {
+      "epoch": 1.3945189274447949,
+      "grad_norm": 0.5159110622307619,
+      "learning_rate": 1.4594785373268399e-05,
+      "loss": 0.3488,
+      "step": 7074
+    },
+    {
+      "epoch": 1.3947160883280758,
+      "grad_norm": 0.48598790594632707,
+      "learning_rate": 1.4593408602250412e-05,
+      "loss": 0.3235,
+      "step": 7075
+    },
+    {
+      "epoch": 1.3949132492113565,
+      "grad_norm": 0.4724557712538457,
+      "learning_rate": 1.4592031720872086e-05,
+      "loss": 0.3286,
+      "step": 7076
+    },
+    {
+      "epoch": 1.3951104100946372,
+      "grad_norm": 0.48240302693285425,
+      "learning_rate": 1.45906547291665e-05,
+      "loss": 0.3269,
+      "step": 7077
+    },
+    {
+      "epoch": 1.395307570977918,
+      "grad_norm": 0.5090930573384124,
+      "learning_rate": 1.4589277627166738e-05,
+      "loss": 0.3376,
+      "step": 7078
+    },
+    {
+      "epoch": 1.3955047318611986,
+      "grad_norm": 0.49297802997817497,
+      "learning_rate": 1.4587900414905884e-05,
+      "loss": 0.341,
+      "step": 7079
+    },
+    {
+      "epoch": 1.3957018927444795,
+      "grad_norm": 0.4997166046095957,
+      "learning_rate": 1.4586523092417023e-05,
+      "loss": 0.3496,
+      "step": 7080
+    },
+    {
+      "epoch": 1.3958990536277602,
+      "grad_norm": 0.5090637102230009,
+      "learning_rate": 1.4585145659733261e-05,
+      "loss": 0.3348,
+      "step": 7081
+    },
+    {
+      "epoch": 1.396096214511041,
+      "grad_norm": 0.5273133534088523,
+      "learning_rate": 1.4583768116887675e-05,
+      "loss": 0.3552,
+      "step": 7082
+    },
+    {
+      "epoch": 1.3962933753943219,
+      "grad_norm": 0.49036343465671867,
+      "learning_rate": 1.4582390463913374e-05,
+      "loss": 0.3291,
+      "step": 7083
+    },
+    {
+      "epoch": 1.3964905362776026,
+      "grad_norm": 0.5038344983388285,
+      "learning_rate": 1.4581012700843447e-05,
+      "loss": 0.333,
+      "step": 7084
+    },
+    {
+      "epoch": 1.3966876971608833,
+      "grad_norm": 0.4900487095066304,
+      "learning_rate": 1.4579634827711004e-05,
+      "loss": 0.3483,
+      "step": 7085
+    },
+    {
+      "epoch": 1.396884858044164,
+      "grad_norm": 0.5031078803878702,
+      "learning_rate": 1.4578256844549144e-05,
+      "loss": 0.3157,
+      "step": 7086
+    },
+    {
+      "epoch": 1.3970820189274447,
+      "grad_norm": 0.6475198113556349,
+      "learning_rate": 1.4576878751390977e-05,
+      "loss": 0.3807,
+      "step": 7087
+    },
+    {
+      "epoch": 1.3972791798107256,
+      "grad_norm": 0.5187308668664504,
+      "learning_rate": 1.4575500548269612e-05,
+      "loss": 0.3417,
+      "step": 7088
+    },
+    {
+      "epoch": 1.3974763406940063,
+      "grad_norm": 0.4702071372526483,
+      "learning_rate": 1.4574122235218165e-05,
+      "loss": 0.3175,
+      "step": 7089
+    },
+    {
+      "epoch": 1.397673501577287,
+      "grad_norm": 0.5261987476073774,
+      "learning_rate": 1.4572743812269742e-05,
+      "loss": 0.3641,
+      "step": 7090
+    },
+    {
+      "epoch": 1.397870662460568,
+      "grad_norm": 0.5232677422093927,
+      "learning_rate": 1.457136527945747e-05,
+      "loss": 0.3366,
+      "step": 7091
+    },
+    {
+      "epoch": 1.3980678233438486,
+      "grad_norm": 0.5413611690474402,
+      "learning_rate": 1.4569986636814467e-05,
+      "loss": 0.3387,
+      "step": 7092
+    },
+    {
+      "epoch": 1.3982649842271293,
+      "grad_norm": 0.504814762397322,
+      "learning_rate": 1.4568607884373853e-05,
+      "loss": 0.3559,
+      "step": 7093
+    },
+    {
+      "epoch": 1.3984621451104102,
+      "grad_norm": 0.5127986043906394,
+      "learning_rate": 1.4567229022168756e-05,
+      "loss": 0.3346,
+      "step": 7094
+    },
+    {
+      "epoch": 1.398659305993691,
+      "grad_norm": 0.5194678556274392,
+      "learning_rate": 1.4565850050232303e-05,
+      "loss": 0.3264,
+      "step": 7095
+    },
+    {
+      "epoch": 1.3988564668769716,
+      "grad_norm": 1.36315390724496,
+      "learning_rate": 1.4564470968597629e-05,
+      "loss": 0.3374,
+      "step": 7096
+    },
+    {
+      "epoch": 1.3990536277602523,
+      "grad_norm": 0.5377737743898511,
+      "learning_rate": 1.456309177729786e-05,
+      "loss": 0.3548,
+      "step": 7097
+    },
+    {
+      "epoch": 1.399250788643533,
+      "grad_norm": 0.5141750880497071,
+      "learning_rate": 1.4561712476366138e-05,
+      "loss": 0.3328,
+      "step": 7098
+    },
+    {
+      "epoch": 1.399447949526814,
+      "grad_norm": 0.48903622056427165,
+      "learning_rate": 1.4560333065835597e-05,
+      "loss": 0.3118,
+      "step": 7099
+    },
+    {
+      "epoch": 1.3996451104100947,
+      "grad_norm": 0.5462273751987472,
+      "learning_rate": 1.4558953545739386e-05,
+      "loss": 0.3234,
+      "step": 7100
+    },
+    {
+      "epoch": 1.3998422712933754,
+      "grad_norm": 0.4955309077143716,
+      "learning_rate": 1.4557573916110643e-05,
+      "loss": 0.3108,
+      "step": 7101
+    },
+    {
+      "epoch": 1.4000394321766563,
+      "grad_norm": 0.5069308732041122,
+      "learning_rate": 1.4556194176982521e-05,
+      "loss": 0.3231,
+      "step": 7102
+    },
+    {
+      "epoch": 1.400236593059937,
+      "grad_norm": 0.48081025706983266,
+      "learning_rate": 1.4554814328388158e-05,
+      "loss": 0.311,
+      "step": 7103
+    },
+    {
+      "epoch": 1.4004337539432177,
+      "grad_norm": 0.5215098010308659,
+      "learning_rate": 1.4553434370360718e-05,
+      "loss": 0.3516,
+      "step": 7104
+    },
+    {
+      "epoch": 1.4006309148264984,
+      "grad_norm": 0.4849700152957339,
+      "learning_rate": 1.4552054302933344e-05,
+      "loss": 0.3482,
+      "step": 7105
+    },
+    {
+      "epoch": 1.400828075709779,
+      "grad_norm": 0.5043874074673499,
+      "learning_rate": 1.4550674126139206e-05,
+      "loss": 0.3562,
+      "step": 7106
+    },
+    {
+      "epoch": 1.40102523659306,
+      "grad_norm": 0.4972388280378365,
+      "learning_rate": 1.4549293840011453e-05,
+      "loss": 0.3371,
+      "step": 7107
+    },
+    {
+      "epoch": 1.4012223974763407,
+      "grad_norm": 0.501745399798661,
+      "learning_rate": 1.4547913444583254e-05,
+      "loss": 0.3435,
+      "step": 7108
+    },
+    {
+      "epoch": 1.4014195583596214,
+      "grad_norm": 0.5121863899401545,
+      "learning_rate": 1.4546532939887775e-05,
+      "loss": 0.3126,
+      "step": 7109
+    },
+    {
+      "epoch": 1.4016167192429023,
+      "grad_norm": 0.5122597557427135,
+      "learning_rate": 1.4545152325958176e-05,
+      "loss": 0.3241,
+      "step": 7110
+    },
+    {
+      "epoch": 1.401813880126183,
+      "grad_norm": 0.5356092124198039,
+      "learning_rate": 1.4543771602827635e-05,
+      "loss": 0.3352,
+      "step": 7111
+    },
+    {
+      "epoch": 1.4020110410094637,
+      "grad_norm": 0.5046731418348064,
+      "learning_rate": 1.454239077052932e-05,
+      "loss": 0.3255,
+      "step": 7112
+    },
+    {
+      "epoch": 1.4022082018927444,
+      "grad_norm": 0.5329137134815505,
+      "learning_rate": 1.4541009829096411e-05,
+      "loss": 0.3668,
+      "step": 7113
+    },
+    {
+      "epoch": 1.4024053627760251,
+      "grad_norm": 0.5089935811054527,
+      "learning_rate": 1.4539628778562082e-05,
+      "loss": 0.3246,
+      "step": 7114
+    },
+    {
+      "epoch": 1.402602523659306,
+      "grad_norm": 0.5703923513720676,
+      "learning_rate": 1.4538247618959519e-05,
+      "loss": 0.3691,
+      "step": 7115
+    },
+    {
+      "epoch": 1.4027996845425867,
+      "grad_norm": 0.48847066057622457,
+      "learning_rate": 1.4536866350321899e-05,
+      "loss": 0.3268,
+      "step": 7116
+    },
+    {
+      "epoch": 1.4029968454258674,
+      "grad_norm": 0.5073745379809114,
+      "learning_rate": 1.4535484972682412e-05,
+      "loss": 0.3149,
+      "step": 7117
+    },
+    {
+      "epoch": 1.4031940063091484,
+      "grad_norm": 0.6462283290424533,
+      "learning_rate": 1.4534103486074246e-05,
+      "loss": 0.3336,
+      "step": 7118
+    },
+    {
+      "epoch": 1.403391167192429,
+      "grad_norm": 0.515565073309167,
+      "learning_rate": 1.4532721890530594e-05,
+      "loss": 0.3456,
+      "step": 7119
+    },
+    {
+      "epoch": 1.4035883280757098,
+      "grad_norm": 0.4990119046287184,
+      "learning_rate": 1.4531340186084647e-05,
+      "loss": 0.3122,
+      "step": 7120
+    },
+    {
+      "epoch": 1.4037854889589905,
+      "grad_norm": 0.510752738342485,
+      "learning_rate": 1.4529958372769603e-05,
+      "loss": 0.3478,
+      "step": 7121
+    },
+    {
+      "epoch": 1.4039826498422712,
+      "grad_norm": 0.533223751953392,
+      "learning_rate": 1.4528576450618661e-05,
+      "loss": 0.3166,
+      "step": 7122
+    },
+    {
+      "epoch": 1.404179810725552,
+      "grad_norm": 0.5250407115618655,
+      "learning_rate": 1.4527194419665027e-05,
+      "loss": 0.3405,
+      "step": 7123
+    },
+    {
+      "epoch": 1.4043769716088328,
+      "grad_norm": 0.5492621619572747,
+      "learning_rate": 1.4525812279941896e-05,
+      "loss": 0.3307,
+      "step": 7124
+    },
+    {
+      "epoch": 1.4045741324921135,
+      "grad_norm": 0.5808645619860187,
+      "learning_rate": 1.4524430031482483e-05,
+      "loss": 0.3236,
+      "step": 7125
+    },
+    {
+      "epoch": 1.4047712933753944,
+      "grad_norm": 0.516938354436358,
+      "learning_rate": 1.4523047674319992e-05,
+      "loss": 0.3397,
+      "step": 7126
+    },
+    {
+      "epoch": 1.4049684542586751,
+      "grad_norm": 0.5446571797503498,
+      "learning_rate": 1.452166520848764e-05,
+      "loss": 0.3739,
+      "step": 7127
+    },
+    {
+      "epoch": 1.4051656151419558,
+      "grad_norm": 0.5247398532256848,
+      "learning_rate": 1.4520282634018642e-05,
+      "loss": 0.33,
+      "step": 7128
+    },
+    {
+      "epoch": 1.4053627760252365,
+      "grad_norm": 0.5174783252477193,
+      "learning_rate": 1.451889995094621e-05,
+      "loss": 0.3291,
+      "step": 7129
+    },
+    {
+      "epoch": 1.4055599369085172,
+      "grad_norm": 0.45618206010019524,
+      "learning_rate": 1.4517517159303573e-05,
+      "loss": 0.3154,
+      "step": 7130
+    },
+    {
+      "epoch": 1.4057570977917981,
+      "grad_norm": 0.5302269877465452,
+      "learning_rate": 1.4516134259123944e-05,
+      "loss": 0.3305,
+      "step": 7131
+    },
+    {
+      "epoch": 1.4059542586750788,
+      "grad_norm": 0.49691950071685087,
+      "learning_rate": 1.4514751250440556e-05,
+      "loss": 0.3465,
+      "step": 7132
+    },
+    {
+      "epoch": 1.4061514195583595,
+      "grad_norm": 0.5204660479559051,
+      "learning_rate": 1.4513368133286628e-05,
+      "loss": 0.3484,
+      "step": 7133
+    },
+    {
+      "epoch": 1.4063485804416405,
+      "grad_norm": 0.5209911663466499,
+      "learning_rate": 1.45119849076954e-05,
+      "loss": 0.3324,
+      "step": 7134
+    },
+    {
+      "epoch": 1.4065457413249212,
+      "grad_norm": 0.52870920201794,
+      "learning_rate": 1.4510601573700098e-05,
+      "loss": 0.3464,
+      "step": 7135
+    },
+    {
+      "epoch": 1.4067429022082019,
+      "grad_norm": 0.5337821985731359,
+      "learning_rate": 1.4509218131333964e-05,
+      "loss": 0.3547,
+      "step": 7136
+    },
+    {
+      "epoch": 1.4069400630914828,
+      "grad_norm": 0.4726297001749275,
+      "learning_rate": 1.4507834580630231e-05,
+      "loss": 0.3287,
+      "step": 7137
+    },
+    {
+      "epoch": 1.4071372239747635,
+      "grad_norm": 0.5411170706465964,
+      "learning_rate": 1.4506450921622144e-05,
+      "loss": 0.3249,
+      "step": 7138
+    },
+    {
+      "epoch": 1.4073343848580442,
+      "grad_norm": 0.5280250037967184,
+      "learning_rate": 1.4505067154342944e-05,
+      "loss": 0.3481,
+      "step": 7139
+    },
+    {
+      "epoch": 1.4075315457413249,
+      "grad_norm": 0.5254200953659547,
+      "learning_rate": 1.4503683278825877e-05,
+      "loss": 0.3102,
+      "step": 7140
+    },
+    {
+      "epoch": 1.4077287066246056,
+      "grad_norm": 0.5167455740268537,
+      "learning_rate": 1.4502299295104194e-05,
+      "loss": 0.3626,
+      "step": 7141
+    },
+    {
+      "epoch": 1.4079258675078865,
+      "grad_norm": 0.5231802604371326,
+      "learning_rate": 1.4500915203211144e-05,
+      "loss": 0.3298,
+      "step": 7142
+    },
+    {
+      "epoch": 1.4081230283911672,
+      "grad_norm": 0.5273278352996336,
+      "learning_rate": 1.449953100317998e-05,
+      "loss": 0.3262,
+      "step": 7143
+    },
+    {
+      "epoch": 1.408320189274448,
+      "grad_norm": 0.467342150805326,
+      "learning_rate": 1.4498146695043963e-05,
+      "loss": 0.3136,
+      "step": 7144
+    },
+    {
+      "epoch": 1.4085173501577288,
+      "grad_norm": 0.6192865687389703,
+      "learning_rate": 1.4496762278836347e-05,
+      "loss": 0.3684,
+      "step": 7145
+    },
+    {
+      "epoch": 1.4087145110410095,
+      "grad_norm": 0.5257953374768632,
+      "learning_rate": 1.4495377754590396e-05,
+      "loss": 0.3528,
+      "step": 7146
+    },
+    {
+      "epoch": 1.4089116719242902,
+      "grad_norm": 0.5430432195240142,
+      "learning_rate": 1.4493993122339375e-05,
+      "loss": 0.3398,
+      "step": 7147
+    },
+    {
+      "epoch": 1.409108832807571,
+      "grad_norm": 0.5071243838122111,
+      "learning_rate": 1.4492608382116548e-05,
+      "loss": 0.3306,
+      "step": 7148
+    },
+    {
+      "epoch": 1.4093059936908516,
+      "grad_norm": 0.591861854929956,
+      "learning_rate": 1.4491223533955191e-05,
+      "loss": 0.3736,
+      "step": 7149
+    },
+    {
+      "epoch": 1.4095031545741326,
+      "grad_norm": 0.5388065248530421,
+      "learning_rate": 1.4489838577888569e-05,
+      "loss": 0.3513,
+      "step": 7150
+    },
+    {
+      "epoch": 1.4097003154574133,
+      "grad_norm": 0.517788510937729,
+      "learning_rate": 1.4488453513949963e-05,
+      "loss": 0.3479,
+      "step": 7151
+    },
+    {
+      "epoch": 1.409897476340694,
+      "grad_norm": 0.5610229696228063,
+      "learning_rate": 1.4487068342172642e-05,
+      "loss": 0.3469,
+      "step": 7152
+    },
+    {
+      "epoch": 1.4100946372239749,
+      "grad_norm": 0.4712793595470132,
+      "learning_rate": 1.4485683062589895e-05,
+      "loss": 0.3182,
+      "step": 7153
+    },
+    {
+      "epoch": 1.4102917981072556,
+      "grad_norm": 0.5668593758772468,
+      "learning_rate": 1.4484297675234995e-05,
+      "loss": 0.3809,
+      "step": 7154
+    },
+    {
+      "epoch": 1.4104889589905363,
+      "grad_norm": 0.4925454763252612,
+      "learning_rate": 1.4482912180141236e-05,
+      "loss": 0.3475,
+      "step": 7155
+    },
+    {
+      "epoch": 1.410686119873817,
+      "grad_norm": 0.5510156258206566,
+      "learning_rate": 1.44815265773419e-05,
+      "loss": 0.3579,
+      "step": 7156
+    },
+    {
+      "epoch": 1.4108832807570977,
+      "grad_norm": 0.5405436736712104,
+      "learning_rate": 1.4480140866870281e-05,
+      "loss": 0.3426,
+      "step": 7157
+    },
+    {
+      "epoch": 1.4110804416403786,
+      "grad_norm": 0.5237178960169243,
+      "learning_rate": 1.4478755048759668e-05,
+      "loss": 0.3632,
+      "step": 7158
+    },
+    {
+      "epoch": 1.4112776025236593,
+      "grad_norm": 0.5175312490148417,
+      "learning_rate": 1.4477369123043358e-05,
+      "loss": 0.3236,
+      "step": 7159
+    },
+    {
+      "epoch": 1.41147476340694,
+      "grad_norm": 0.4913933173819947,
+      "learning_rate": 1.447598308975465e-05,
+      "loss": 0.324,
+      "step": 7160
+    },
+    {
+      "epoch": 1.411671924290221,
+      "grad_norm": 0.4826321503807621,
+      "learning_rate": 1.4474596948926844e-05,
+      "loss": 0.326,
+      "step": 7161
+    },
+    {
+      "epoch": 1.4118690851735016,
+      "grad_norm": 0.4915727914760078,
+      "learning_rate": 1.4473210700593242e-05,
+      "loss": 0.33,
+      "step": 7162
+    },
+    {
+      "epoch": 1.4120662460567823,
+      "grad_norm": 0.5060788214654445,
+      "learning_rate": 1.4471824344787153e-05,
+      "loss": 0.314,
+      "step": 7163
+    },
+    {
+      "epoch": 1.412263406940063,
+      "grad_norm": 0.491241758641821,
+      "learning_rate": 1.4470437881541882e-05,
+      "loss": 0.3421,
+      "step": 7164
+    },
+    {
+      "epoch": 1.4124605678233437,
+      "grad_norm": 0.4942260983754986,
+      "learning_rate": 1.446905131089074e-05,
+      "loss": 0.3491,
+      "step": 7165
+    },
+    {
+      "epoch": 1.4126577287066246,
+      "grad_norm": 0.48529396730532887,
+      "learning_rate": 1.4467664632867042e-05,
+      "loss": 0.332,
+      "step": 7166
+    },
+    {
+      "epoch": 1.4128548895899053,
+      "grad_norm": 0.5474445100151097,
+      "learning_rate": 1.44662778475041e-05,
+      "loss": 0.3503,
+      "step": 7167
+    },
+    {
+      "epoch": 1.413052050473186,
+      "grad_norm": 0.49293254269828907,
+      "learning_rate": 1.4464890954835242e-05,
+      "loss": 0.326,
+      "step": 7168
+    },
+    {
+      "epoch": 1.413249211356467,
+      "grad_norm": 0.5017859566166725,
+      "learning_rate": 1.4463503954893778e-05,
+      "loss": 0.3244,
+      "step": 7169
+    },
+    {
+      "epoch": 1.4134463722397477,
+      "grad_norm": 0.5240485239593689,
+      "learning_rate": 1.446211684771304e-05,
+      "loss": 0.3708,
+      "step": 7170
+    },
+    {
+      "epoch": 1.4136435331230284,
+      "grad_norm": 0.4870947916252577,
+      "learning_rate": 1.4460729633326351e-05,
+      "loss": 0.3162,
+      "step": 7171
+    },
+    {
+      "epoch": 1.413840694006309,
+      "grad_norm": 0.4593378993420768,
+      "learning_rate": 1.4459342311767041e-05,
+      "loss": 0.3221,
+      "step": 7172
+    },
+    {
+      "epoch": 1.4140378548895898,
+      "grad_norm": 0.48340606797854385,
+      "learning_rate": 1.445795488306844e-05,
+      "loss": 0.3228,
+      "step": 7173
+    },
+    {
+      "epoch": 1.4142350157728707,
+      "grad_norm": 0.4713202093530444,
+      "learning_rate": 1.445656734726388e-05,
+      "loss": 0.3172,
+      "step": 7174
+    },
+    {
+      "epoch": 1.4144321766561514,
+      "grad_norm": 0.4780513871592405,
+      "learning_rate": 1.4455179704386706e-05,
+      "loss": 0.3551,
+      "step": 7175
+    },
+    {
+      "epoch": 1.414629337539432,
+      "grad_norm": 0.48058366057305824,
+      "learning_rate": 1.4453791954470248e-05,
+      "loss": 0.3257,
+      "step": 7176
+    },
+    {
+      "epoch": 1.414826498422713,
+      "grad_norm": 0.49793083984186,
+      "learning_rate": 1.4452404097547855e-05,
+      "loss": 0.3592,
+      "step": 7177
+    },
+    {
+      "epoch": 1.4150236593059937,
+      "grad_norm": 0.472865401917883,
+      "learning_rate": 1.4451016133652864e-05,
+      "loss": 0.3218,
+      "step": 7178
+    },
+    {
+      "epoch": 1.4152208201892744,
+      "grad_norm": 0.46575544454257445,
+      "learning_rate": 1.4449628062818628e-05,
+      "loss": 0.3077,
+      "step": 7179
+    },
+    {
+      "epoch": 1.4154179810725553,
+      "grad_norm": 0.48121486650008277,
+      "learning_rate": 1.4448239885078494e-05,
+      "loss": 0.3326,
+      "step": 7180
+    },
+    {
+      "epoch": 1.415615141955836,
+      "grad_norm": 0.473468751152917,
+      "learning_rate": 1.4446851600465817e-05,
+      "loss": 0.3376,
+      "step": 7181
+    },
+    {
+      "epoch": 1.4158123028391167,
+      "grad_norm": 0.5137837266489765,
+      "learning_rate": 1.4445463209013948e-05,
+      "loss": 0.3401,
+      "step": 7182
+    },
+    {
+      "epoch": 1.4160094637223974,
+      "grad_norm": 0.497315989831655,
+      "learning_rate": 1.4444074710756244e-05,
+      "loss": 0.3391,
+      "step": 7183
+    },
+    {
+      "epoch": 1.4162066246056781,
+      "grad_norm": 0.5170085322910686,
+      "learning_rate": 1.4442686105726066e-05,
+      "loss": 0.3612,
+      "step": 7184
+    },
+    {
+      "epoch": 1.416403785488959,
+      "grad_norm": 0.5055898381412588,
+      "learning_rate": 1.4441297393956779e-05,
+      "loss": 0.3368,
+      "step": 7185
+    },
+    {
+      "epoch": 1.4166009463722398,
+      "grad_norm": 0.4588445495464938,
+      "learning_rate": 1.4439908575481744e-05,
+      "loss": 0.3215,
+      "step": 7186
+    },
+    {
+      "epoch": 1.4167981072555205,
+      "grad_norm": 0.4940514211342652,
+      "learning_rate": 1.443851965033433e-05,
+      "loss": 0.3229,
+      "step": 7187
+    },
+    {
+      "epoch": 1.4169952681388014,
+      "grad_norm": 0.5237805317857336,
+      "learning_rate": 1.4437130618547905e-05,
+      "loss": 0.3746,
+      "step": 7188
+    },
+    {
+      "epoch": 1.417192429022082,
+      "grad_norm": 0.46802246011483617,
+      "learning_rate": 1.443574148015585e-05,
+      "loss": 0.3247,
+      "step": 7189
+    },
+    {
+      "epoch": 1.4173895899053628,
+      "grad_norm": 0.4760962059204285,
+      "learning_rate": 1.4434352235191526e-05,
+      "loss": 0.3145,
+      "step": 7190
+    },
+    {
+      "epoch": 1.4175867507886435,
+      "grad_norm": 0.46609718235500464,
+      "learning_rate": 1.4432962883688327e-05,
+      "loss": 0.3309,
+      "step": 7191
+    },
+    {
+      "epoch": 1.4177839116719242,
+      "grad_norm": 0.5063549571401547,
+      "learning_rate": 1.443157342567962e-05,
+      "loss": 0.3656,
+      "step": 7192
+    },
+    {
+      "epoch": 1.4179810725552051,
+      "grad_norm": 0.5161414428496777,
+      "learning_rate": 1.4430183861198792e-05,
+      "loss": 0.3811,
+      "step": 7193
+    },
+    {
+      "epoch": 1.4181782334384858,
+      "grad_norm": 0.46919572179206825,
+      "learning_rate": 1.4428794190279231e-05,
+      "loss": 0.3344,
+      "step": 7194
+    },
+    {
+      "epoch": 1.4183753943217665,
+      "grad_norm": 0.5020263012424411,
+      "learning_rate": 1.442740441295432e-05,
+      "loss": 0.3387,
+      "step": 7195
+    },
+    {
+      "epoch": 1.4185725552050474,
+      "grad_norm": 0.4593456839363708,
+      "learning_rate": 1.4426014529257457e-05,
+      "loss": 0.3161,
+      "step": 7196
+    },
+    {
+      "epoch": 1.4187697160883281,
+      "grad_norm": 0.50798569973899,
+      "learning_rate": 1.4424624539222028e-05,
+      "loss": 0.3444,
+      "step": 7197
+    },
+    {
+      "epoch": 1.4189668769716088,
+      "grad_norm": 0.4790042993417883,
+      "learning_rate": 1.4423234442881433e-05,
+      "loss": 0.3375,
+      "step": 7198
+    },
+    {
+      "epoch": 1.4191640378548895,
+      "grad_norm": 0.5051516627231682,
+      "learning_rate": 1.4421844240269064e-05,
+      "loss": 0.3502,
+      "step": 7199
+    },
+    {
+      "epoch": 1.4193611987381702,
+      "grad_norm": 0.49347178341054193,
+      "learning_rate": 1.4420453931418332e-05,
+      "loss": 0.3503,
+      "step": 7200
+    },
+    {
+      "epoch": 1.4195583596214512,
+      "grad_norm": 0.4764569323870533,
+      "learning_rate": 1.4419063516362633e-05,
+      "loss": 0.3142,
+      "step": 7201
+    },
+    {
+      "epoch": 1.4197555205047319,
+      "grad_norm": 0.48966644810736204,
+      "learning_rate": 1.4417672995135372e-05,
+      "loss": 0.3348,
+      "step": 7202
+    },
+    {
+      "epoch": 1.4199526813880126,
+      "grad_norm": 0.4897579050663058,
+      "learning_rate": 1.4416282367769961e-05,
+      "loss": 0.3284,
+      "step": 7203
+    },
+    {
+      "epoch": 1.4201498422712935,
+      "grad_norm": 0.4709690024262115,
+      "learning_rate": 1.441489163429981e-05,
+      "loss": 0.3158,
+      "step": 7204
+    },
+    {
+      "epoch": 1.4203470031545742,
+      "grad_norm": 0.5083654874998398,
+      "learning_rate": 1.4413500794758333e-05,
+      "loss": 0.3569,
+      "step": 7205
+    },
+    {
+      "epoch": 1.4205441640378549,
+      "grad_norm": 0.4977170017599856,
+      "learning_rate": 1.4412109849178944e-05,
+      "loss": 0.3207,
+      "step": 7206
+    },
+    {
+      "epoch": 1.4207413249211356,
+      "grad_norm": 0.49785411016808323,
+      "learning_rate": 1.4410718797595063e-05,
+      "loss": 0.3385,
+      "step": 7207
+    },
+    {
+      "epoch": 1.4209384858044163,
+      "grad_norm": 0.4985331494542473,
+      "learning_rate": 1.4409327640040111e-05,
+      "loss": 0.3256,
+      "step": 7208
+    },
+    {
+      "epoch": 1.4211356466876972,
+      "grad_norm": 12.637841434778313,
+      "learning_rate": 1.440793637654751e-05,
+      "loss": 0.4233,
+      "step": 7209
+    },
+    {
+      "epoch": 1.421332807570978,
+      "grad_norm": 0.5318624726951542,
+      "learning_rate": 1.4406545007150693e-05,
+      "loss": 0.3293,
+      "step": 7210
+    },
+    {
+      "epoch": 1.4215299684542586,
+      "grad_norm": 0.5031296805718974,
+      "learning_rate": 1.440515353188308e-05,
+      "loss": 0.3301,
+      "step": 7211
+    },
+    {
+      "epoch": 1.4217271293375395,
+      "grad_norm": 0.5445707070467908,
+      "learning_rate": 1.4403761950778106e-05,
+      "loss": 0.3248,
+      "step": 7212
+    },
+    {
+      "epoch": 1.4219242902208202,
+      "grad_norm": 0.4659455732252377,
+      "learning_rate": 1.4402370263869205e-05,
+      "loss": 0.3216,
+      "step": 7213
+    },
+    {
+      "epoch": 1.422121451104101,
+      "grad_norm": 0.5288243979500802,
+      "learning_rate": 1.4400978471189812e-05,
+      "loss": 0.3693,
+      "step": 7214
+    },
+    {
+      "epoch": 1.4223186119873816,
+      "grad_norm": 0.528449022273337,
+      "learning_rate": 1.439958657277337e-05,
+      "loss": 0.3402,
+      "step": 7215
+    },
+    {
+      "epoch": 1.4225157728706623,
+      "grad_norm": 0.5119944243955784,
+      "learning_rate": 1.4398194568653313e-05,
+      "loss": 0.3494,
+      "step": 7216
+    },
+    {
+      "epoch": 1.4227129337539433,
+      "grad_norm": 0.4916670845158821,
+      "learning_rate": 1.4396802458863095e-05,
+      "loss": 0.3217,
+      "step": 7217
+    },
+    {
+      "epoch": 1.422910094637224,
+      "grad_norm": 0.48113900294298484,
+      "learning_rate": 1.4395410243436153e-05,
+      "loss": 0.3314,
+      "step": 7218
+    },
+    {
+      "epoch": 1.4231072555205047,
+      "grad_norm": 0.47058896620118573,
+      "learning_rate": 1.4394017922405943e-05,
+      "loss": 0.3159,
+      "step": 7219
+    },
+    {
+      "epoch": 1.4233044164037856,
+      "grad_norm": 0.4849453680724847,
+      "learning_rate": 1.4392625495805913e-05,
+      "loss": 0.3395,
+      "step": 7220
+    },
+    {
+      "epoch": 1.4235015772870663,
+      "grad_norm": 0.5373072375771156,
+      "learning_rate": 1.439123296366952e-05,
+      "loss": 0.3288,
+      "step": 7221
+    },
+    {
+      "epoch": 1.423698738170347,
+      "grad_norm": 0.4693872279830995,
+      "learning_rate": 1.4389840326030213e-05,
+      "loss": 0.3302,
+      "step": 7222
+    },
+    {
+      "epoch": 1.4238958990536277,
+      "grad_norm": 0.5319368606240615,
+      "learning_rate": 1.438844758292146e-05,
+      "loss": 0.3312,
+      "step": 7223
+    },
+    {
+      "epoch": 1.4240930599369084,
+      "grad_norm": 0.5817496986395633,
+      "learning_rate": 1.4387054734376722e-05,
+      "loss": 0.3249,
+      "step": 7224
+    },
+    {
+      "epoch": 1.4242902208201893,
+      "grad_norm": 0.48934565333610924,
+      "learning_rate": 1.4385661780429461e-05,
+      "loss": 0.334,
+      "step": 7225
+    },
+    {
+      "epoch": 1.42448738170347,
+      "grad_norm": 0.49779432157118675,
+      "learning_rate": 1.438426872111314e-05,
+      "loss": 0.3428,
+      "step": 7226
+    },
+    {
+      "epoch": 1.4246845425867507,
+      "grad_norm": 0.4940338187493709,
+      "learning_rate": 1.4382875556461238e-05,
+      "loss": 0.3252,
+      "step": 7227
+    },
+    {
+      "epoch": 1.4248817034700316,
+      "grad_norm": 0.4815774455413699,
+      "learning_rate": 1.4381482286507216e-05,
+      "loss": 0.3158,
+      "step": 7228
+    },
+    {
+      "epoch": 1.4250788643533123,
+      "grad_norm": 0.4826791184071724,
+      "learning_rate": 1.4380088911284557e-05,
+      "loss": 0.3388,
+      "step": 7229
+    },
+    {
+      "epoch": 1.425276025236593,
+      "grad_norm": 0.5434358688212808,
+      "learning_rate": 1.4378695430826732e-05,
+      "loss": 0.3457,
+      "step": 7230
+    },
+    {
+      "epoch": 1.425473186119874,
+      "grad_norm": 0.496948379843039,
+      "learning_rate": 1.4377301845167227e-05,
+      "loss": 0.3431,
+      "step": 7231
+    },
+    {
+      "epoch": 1.4256703470031546,
+      "grad_norm": 0.4980057220909191,
+      "learning_rate": 1.4375908154339517e-05,
+      "loss": 0.3411,
+      "step": 7232
+    },
+    {
+      "epoch": 1.4258675078864353,
+      "grad_norm": 0.5381831341243695,
+      "learning_rate": 1.437451435837709e-05,
+      "loss": 0.3364,
+      "step": 7233
+    },
+    {
+      "epoch": 1.426064668769716,
+      "grad_norm": 0.5286642410721416,
+      "learning_rate": 1.4373120457313435e-05,
+      "loss": 0.3471,
+      "step": 7234
+    },
+    {
+      "epoch": 1.4262618296529967,
+      "grad_norm": 0.7724157082129258,
+      "learning_rate": 1.4371726451182038e-05,
+      "loss": 0.349,
+      "step": 7235
+    },
+    {
+      "epoch": 1.4264589905362777,
+      "grad_norm": 0.5265279436116093,
+      "learning_rate": 1.437033234001639e-05,
+      "loss": 0.3557,
+      "step": 7236
+    },
+    {
+      "epoch": 1.4266561514195584,
+      "grad_norm": 1.479067454447673,
+      "learning_rate": 1.436893812384999e-05,
+      "loss": 0.3668,
+      "step": 7237
+    },
+    {
+      "epoch": 1.426853312302839,
+      "grad_norm": 0.5212610233545351,
+      "learning_rate": 1.4367543802716334e-05,
+      "loss": 0.3513,
+      "step": 7238
+    },
+    {
+      "epoch": 1.42705047318612,
+      "grad_norm": 0.5107655933355794,
+      "learning_rate": 1.436614937664892e-05,
+      "loss": 0.3375,
+      "step": 7239
+    },
+    {
+      "epoch": 1.4272476340694007,
+      "grad_norm": 0.5255564965953446,
+      "learning_rate": 1.4364754845681253e-05,
+      "loss": 0.344,
+      "step": 7240
+    },
+    {
+      "epoch": 1.4274447949526814,
+      "grad_norm": 0.5736229739284229,
+      "learning_rate": 1.4363360209846833e-05,
+      "loss": 0.357,
+      "step": 7241
+    },
+    {
+      "epoch": 1.427641955835962,
+      "grad_norm": 0.500621320395277,
+      "learning_rate": 1.4361965469179173e-05,
+      "loss": 0.3204,
+      "step": 7242
+    },
+    {
+      "epoch": 1.4278391167192428,
+      "grad_norm": 0.49627772986305124,
+      "learning_rate": 1.4360570623711778e-05,
+      "loss": 0.3196,
+      "step": 7243
+    },
+    {
+      "epoch": 1.4280362776025237,
+      "grad_norm": 0.5137643214444078,
+      "learning_rate": 1.4359175673478163e-05,
+      "loss": 0.3455,
+      "step": 7244
+    },
+    {
+      "epoch": 1.4282334384858044,
+      "grad_norm": 0.5139351480323677,
+      "learning_rate": 1.435778061851184e-05,
+      "loss": 0.3494,
+      "step": 7245
+    },
+    {
+      "epoch": 1.4284305993690851,
+      "grad_norm": 0.5154850389726875,
+      "learning_rate": 1.435638545884633e-05,
+      "loss": 0.3538,
+      "step": 7246
+    },
+    {
+      "epoch": 1.428627760252366,
+      "grad_norm": 0.49555806148409276,
+      "learning_rate": 1.4354990194515155e-05,
+      "loss": 0.345,
+      "step": 7247
+    },
+    {
+      "epoch": 1.4288249211356467,
+      "grad_norm": 0.5168230858439605,
+      "learning_rate": 1.4353594825551827e-05,
+      "loss": 0.3459,
+      "step": 7248
+    },
+    {
+      "epoch": 1.4290220820189274,
+      "grad_norm": 0.5471309658606096,
+      "learning_rate": 1.4352199351989881e-05,
+      "loss": 0.3419,
+      "step": 7249
+    },
+    {
+      "epoch": 1.4292192429022081,
+      "grad_norm": 0.47810122148847417,
+      "learning_rate": 1.4350803773862841e-05,
+      "loss": 0.3093,
+      "step": 7250
+    },
+    {
+      "epoch": 1.4294164037854888,
+      "grad_norm": 0.4924167515842062,
+      "learning_rate": 1.4349408091204234e-05,
+      "loss": 0.3169,
+      "step": 7251
+    },
+    {
+      "epoch": 1.4296135646687698,
+      "grad_norm": 0.48173871551750974,
+      "learning_rate": 1.4348012304047596e-05,
+      "loss": 0.3336,
+      "step": 7252
+    },
+    {
+      "epoch": 1.4298107255520505,
+      "grad_norm": 0.49482226426234466,
+      "learning_rate": 1.4346616412426464e-05,
+      "loss": 0.3356,
+      "step": 7253
+    },
+    {
+      "epoch": 1.4300078864353312,
+      "grad_norm": 0.505868389409233,
+      "learning_rate": 1.434522041637437e-05,
+      "loss": 0.3555,
+      "step": 7254
+    },
+    {
+      "epoch": 1.430205047318612,
+      "grad_norm": 0.46830331430343086,
+      "learning_rate": 1.4343824315924855e-05,
+      "loss": 0.3163,
+      "step": 7255
+    },
+    {
+      "epoch": 1.4304022082018928,
+      "grad_norm": 0.4854035751744796,
+      "learning_rate": 1.4342428111111461e-05,
+      "loss": 0.3245,
+      "step": 7256
+    },
+    {
+      "epoch": 1.4305993690851735,
+      "grad_norm": 0.4948071078265386,
+      "learning_rate": 1.4341031801967742e-05,
+      "loss": 0.3319,
+      "step": 7257
+    },
+    {
+      "epoch": 1.4307965299684542,
+      "grad_norm": 0.5325267099042197,
+      "learning_rate": 1.4339635388527231e-05,
+      "loss": 0.3412,
+      "step": 7258
+    },
+    {
+      "epoch": 1.430993690851735,
+      "grad_norm": 0.6445122208735229,
+      "learning_rate": 1.433823887082349e-05,
+      "loss": 0.3409,
+      "step": 7259
+    },
+    {
+      "epoch": 1.4311908517350158,
+      "grad_norm": 0.5057312476219762,
+      "learning_rate": 1.4336842248890065e-05,
+      "loss": 0.3433,
+      "step": 7260
+    },
+    {
+      "epoch": 1.4313880126182965,
+      "grad_norm": 0.5013879369726504,
+      "learning_rate": 1.4335445522760512e-05,
+      "loss": 0.3373,
+      "step": 7261
+    },
+    {
+      "epoch": 1.4315851735015772,
+      "grad_norm": 0.5361669119631592,
+      "learning_rate": 1.433404869246839e-05,
+      "loss": 0.3599,
+      "step": 7262
+    },
+    {
+      "epoch": 1.4317823343848581,
+      "grad_norm": 0.5096029349380119,
+      "learning_rate": 1.4332651758047254e-05,
+      "loss": 0.3446,
+      "step": 7263
+    },
+    {
+      "epoch": 1.4319794952681388,
+      "grad_norm": 0.5265991854311413,
+      "learning_rate": 1.4331254719530676e-05,
+      "loss": 0.3717,
+      "step": 7264
+    },
+    {
+      "epoch": 1.4321766561514195,
+      "grad_norm": 0.4825505433718403,
+      "learning_rate": 1.4329857576952212e-05,
+      "loss": 0.3206,
+      "step": 7265
+    },
+    {
+      "epoch": 1.4323738170347002,
+      "grad_norm": 0.509655525792269,
+      "learning_rate": 1.4328460330345434e-05,
+      "loss": 0.3277,
+      "step": 7266
+    },
+    {
+      "epoch": 1.432570977917981,
+      "grad_norm": 0.4686753391448895,
+      "learning_rate": 1.432706297974391e-05,
+      "loss": 0.3028,
+      "step": 7267
+    },
+    {
+      "epoch": 1.4327681388012619,
+      "grad_norm": 0.5074945700987945,
+      "learning_rate": 1.4325665525181213e-05,
+      "loss": 0.3524,
+      "step": 7268
+    },
+    {
+      "epoch": 1.4329652996845426,
+      "grad_norm": 0.4994918673135275,
+      "learning_rate": 1.4324267966690919e-05,
+      "loss": 0.3373,
+      "step": 7269
+    },
+    {
+      "epoch": 1.4331624605678233,
+      "grad_norm": 0.4840679030796876,
+      "learning_rate": 1.4322870304306604e-05,
+      "loss": 0.3397,
+      "step": 7270
+    },
+    {
+      "epoch": 1.4333596214511042,
+      "grad_norm": 0.4620784380959485,
+      "learning_rate": 1.4321472538061852e-05,
+      "loss": 0.3314,
+      "step": 7271
+    },
+    {
+      "epoch": 1.4335567823343849,
+      "grad_norm": 0.4551298101484555,
+      "learning_rate": 1.4320074667990237e-05,
+      "loss": 0.3147,
+      "step": 7272
+    },
+    {
+      "epoch": 1.4337539432176656,
+      "grad_norm": 0.48354743821185125,
+      "learning_rate": 1.4318676694125353e-05,
+      "loss": 0.3441,
+      "step": 7273
+    },
+    {
+      "epoch": 1.4339511041009465,
+      "grad_norm": 0.44671416102776385,
+      "learning_rate": 1.4317278616500785e-05,
+      "loss": 0.3095,
+      "step": 7274
+    },
+    {
+      "epoch": 1.4341482649842272,
+      "grad_norm": 0.48747299202730604,
+      "learning_rate": 1.4315880435150119e-05,
+      "loss": 0.3416,
+      "step": 7275
+    },
+    {
+      "epoch": 1.434345425867508,
+      "grad_norm": 0.4931741101936371,
+      "learning_rate": 1.431448215010695e-05,
+      "loss": 0.3516,
+      "step": 7276
+    },
+    {
+      "epoch": 1.4345425867507886,
+      "grad_norm": 0.49240050671361324,
+      "learning_rate": 1.4313083761404874e-05,
+      "loss": 0.3476,
+      "step": 7277
+    },
+    {
+      "epoch": 1.4347397476340693,
+      "grad_norm": 0.4827511773398998,
+      "learning_rate": 1.4311685269077484e-05,
+      "loss": 0.3185,
+      "step": 7278
+    },
+    {
+      "epoch": 1.4349369085173502,
+      "grad_norm": 0.4971900913026981,
+      "learning_rate": 1.4310286673158387e-05,
+      "loss": 0.3614,
+      "step": 7279
+    },
+    {
+      "epoch": 1.435134069400631,
+      "grad_norm": 0.4857373079412297,
+      "learning_rate": 1.430888797368118e-05,
+      "loss": 0.3184,
+      "step": 7280
+    },
+    {
+      "epoch": 1.4353312302839116,
+      "grad_norm": 0.5759392585568717,
+      "learning_rate": 1.430748917067947e-05,
+      "loss": 0.3404,
+      "step": 7281
+    },
+    {
+      "epoch": 1.4355283911671926,
+      "grad_norm": 0.48435593637452573,
+      "learning_rate": 1.4306090264186863e-05,
+      "loss": 0.3346,
+      "step": 7282
+    },
+    {
+      "epoch": 1.4357255520504733,
+      "grad_norm": 0.5183943537528527,
+      "learning_rate": 1.430469125423697e-05,
+      "loss": 0.3635,
+      "step": 7283
+    },
+    {
+      "epoch": 1.435922712933754,
+      "grad_norm": 0.4874012749572438,
+      "learning_rate": 1.4303292140863402e-05,
+      "loss": 0.3417,
+      "step": 7284
+    },
+    {
+      "epoch": 1.4361198738170347,
+      "grad_norm": 0.47674893494615467,
+      "learning_rate": 1.4301892924099778e-05,
+      "loss": 0.326,
+      "step": 7285
+    },
+    {
+      "epoch": 1.4363170347003154,
+      "grad_norm": 0.4975791293304716,
+      "learning_rate": 1.430049360397971e-05,
+      "loss": 0.3426,
+      "step": 7286
+    },
+    {
+      "epoch": 1.4365141955835963,
+      "grad_norm": 0.47977676406998915,
+      "learning_rate": 1.4299094180536821e-05,
+      "loss": 0.3338,
+      "step": 7287
+    },
+    {
+      "epoch": 1.436711356466877,
+      "grad_norm": 0.4849615911471788,
+      "learning_rate": 1.429769465380473e-05,
+      "loss": 0.3365,
+      "step": 7288
+    },
+    {
+      "epoch": 1.4369085173501577,
+      "grad_norm": 0.50892477699892,
+      "learning_rate": 1.4296295023817068e-05,
+      "loss": 0.3194,
+      "step": 7289
+    },
+    {
+      "epoch": 1.4371056782334386,
+      "grad_norm": 0.5145007377570271,
+      "learning_rate": 1.4294895290607454e-05,
+      "loss": 0.3497,
+      "step": 7290
+    },
+    {
+      "epoch": 1.4373028391167193,
+      "grad_norm": 0.5513464913968162,
+      "learning_rate": 1.4293495454209525e-05,
+      "loss": 0.3693,
+      "step": 7291
+    },
+    {
+      "epoch": 1.4375,
+      "grad_norm": 0.5079800533958371,
+      "learning_rate": 1.4292095514656907e-05,
+      "loss": 0.3527,
+      "step": 7292
+    },
+    {
+      "epoch": 1.4376971608832807,
+      "grad_norm": 0.48977789204818567,
+      "learning_rate": 1.4290695471983243e-05,
+      "loss": 0.3207,
+      "step": 7293
+    },
+    {
+      "epoch": 1.4378943217665614,
+      "grad_norm": 0.6229406915573323,
+      "learning_rate": 1.4289295326222161e-05,
+      "loss": 0.3519,
+      "step": 7294
+    },
+    {
+      "epoch": 1.4380914826498423,
+      "grad_norm": 0.5705871966971622,
+      "learning_rate": 1.4287895077407306e-05,
+      "loss": 0.3602,
+      "step": 7295
+    },
+    {
+      "epoch": 1.438288643533123,
+      "grad_norm": 0.47501369073246813,
+      "learning_rate": 1.4286494725572317e-05,
+      "loss": 0.3289,
+      "step": 7296
+    },
+    {
+      "epoch": 1.4384858044164037,
+      "grad_norm": 0.4981376094438177,
+      "learning_rate": 1.4285094270750843e-05,
+      "loss": 0.3381,
+      "step": 7297
+    },
+    {
+      "epoch": 1.4386829652996846,
+      "grad_norm": 0.5229391942882657,
+      "learning_rate": 1.4283693712976527e-05,
+      "loss": 0.3332,
+      "step": 7298
+    },
+    {
+      "epoch": 1.4388801261829653,
+      "grad_norm": 0.5054756208664721,
+      "learning_rate": 1.4282293052283019e-05,
+      "loss": 0.352,
+      "step": 7299
+    },
+    {
+      "epoch": 1.439077287066246,
+      "grad_norm": 0.4951109600486945,
+      "learning_rate": 1.4280892288703974e-05,
+      "loss": 0.331,
+      "step": 7300
+    },
+    {
+      "epoch": 1.4392744479495267,
+      "grad_norm": 0.46364365237898975,
+      "learning_rate": 1.4279491422273043e-05,
+      "loss": 0.318,
+      "step": 7301
+    },
+    {
+      "epoch": 1.4394716088328074,
+      "grad_norm": 0.4947030225121084,
+      "learning_rate": 1.4278090453023885e-05,
+      "loss": 0.3423,
+      "step": 7302
+    },
+    {
+      "epoch": 1.4396687697160884,
+      "grad_norm": 0.4751068461401763,
+      "learning_rate": 1.4276689380990156e-05,
+      "loss": 0.3291,
+      "step": 7303
+    },
+    {
+      "epoch": 1.439865930599369,
+      "grad_norm": 0.5071969159877269,
+      "learning_rate": 1.4275288206205525e-05,
+      "loss": 0.3428,
+      "step": 7304
+    },
+    {
+      "epoch": 1.4400630914826498,
+      "grad_norm": 0.4964103377634554,
+      "learning_rate": 1.4273886928703648e-05,
+      "loss": 0.3302,
+      "step": 7305
+    },
+    {
+      "epoch": 1.4402602523659307,
+      "grad_norm": 0.4786434613936568,
+      "learning_rate": 1.4272485548518198e-05,
+      "loss": 0.3171,
+      "step": 7306
+    },
+    {
+      "epoch": 1.4404574132492114,
+      "grad_norm": 0.4821496603078477,
+      "learning_rate": 1.427108406568284e-05,
+      "loss": 0.3286,
+      "step": 7307
+    },
+    {
+      "epoch": 1.440654574132492,
+      "grad_norm": 0.4899274116908069,
+      "learning_rate": 1.4269682480231253e-05,
+      "loss": 0.3384,
+      "step": 7308
+    },
+    {
+      "epoch": 1.4408517350157728,
+      "grad_norm": 0.496373158989974,
+      "learning_rate": 1.42682807921971e-05,
+      "loss": 0.3355,
+      "step": 7309
+    },
+    {
+      "epoch": 1.4410488958990535,
+      "grad_norm": 0.509985991828358,
+      "learning_rate": 1.4266879001614067e-05,
+      "loss": 0.3311,
+      "step": 7310
+    },
+    {
+      "epoch": 1.4412460567823344,
+      "grad_norm": 0.4956371703782259,
+      "learning_rate": 1.4265477108515828e-05,
+      "loss": 0.3319,
+      "step": 7311
+    },
+    {
+      "epoch": 1.4414432176656151,
+      "grad_norm": 0.490193768475847,
+      "learning_rate": 1.426407511293607e-05,
+      "loss": 0.3395,
+      "step": 7312
+    },
+    {
+      "epoch": 1.4416403785488958,
+      "grad_norm": 0.4955718459701483,
+      "learning_rate": 1.4262673014908472e-05,
+      "loss": 0.3417,
+      "step": 7313
+    },
+    {
+      "epoch": 1.4418375394321767,
+      "grad_norm": 0.5089810792820509,
+      "learning_rate": 1.4261270814466719e-05,
+      "loss": 0.3719,
+      "step": 7314
+    },
+    {
+      "epoch": 1.4420347003154574,
+      "grad_norm": 0.5373269993400187,
+      "learning_rate": 1.4259868511644508e-05,
+      "loss": 0.3366,
+      "step": 7315
+    },
+    {
+      "epoch": 1.4422318611987381,
+      "grad_norm": 0.5191745083925651,
+      "learning_rate": 1.4258466106475522e-05,
+      "loss": 0.3412,
+      "step": 7316
+    },
+    {
+      "epoch": 1.442429022082019,
+      "grad_norm": 0.49462707282721535,
+      "learning_rate": 1.4257063598993458e-05,
+      "loss": 0.3463,
+      "step": 7317
+    },
+    {
+      "epoch": 1.4426261829652998,
+      "grad_norm": 0.4818656214076428,
+      "learning_rate": 1.4255660989232014e-05,
+      "loss": 0.321,
+      "step": 7318
+    },
+    {
+      "epoch": 1.4428233438485805,
+      "grad_norm": 0.5247053748007937,
+      "learning_rate": 1.4254258277224888e-05,
+      "loss": 0.3496,
+      "step": 7319
+    },
+    {
+      "epoch": 1.4430205047318612,
+      "grad_norm": 0.4934106851982142,
+      "learning_rate": 1.4252855463005782e-05,
+      "loss": 0.3463,
+      "step": 7320
+    },
+    {
+      "epoch": 1.4432176656151419,
+      "grad_norm": 0.4941318734425298,
+      "learning_rate": 1.4251452546608397e-05,
+      "loss": 0.3258,
+      "step": 7321
+    },
+    {
+      "epoch": 1.4434148264984228,
+      "grad_norm": 0.4740782133591454,
+      "learning_rate": 1.4250049528066441e-05,
+      "loss": 0.3002,
+      "step": 7322
+    },
+    {
+      "epoch": 1.4436119873817035,
+      "grad_norm": 0.528973383583317,
+      "learning_rate": 1.4248646407413622e-05,
+      "loss": 0.3539,
+      "step": 7323
+    },
+    {
+      "epoch": 1.4438091482649842,
+      "grad_norm": 0.49936440814980976,
+      "learning_rate": 1.424724318468365e-05,
+      "loss": 0.3563,
+      "step": 7324
+    },
+    {
+      "epoch": 1.444006309148265,
+      "grad_norm": 0.5037879267205015,
+      "learning_rate": 1.4245839859910247e-05,
+      "loss": 0.3528,
+      "step": 7325
+    },
+    {
+      "epoch": 1.4442034700315458,
+      "grad_norm": 0.5022670283442432,
+      "learning_rate": 1.4244436433127118e-05,
+      "loss": 0.3321,
+      "step": 7326
+    },
+    {
+      "epoch": 1.4444006309148265,
+      "grad_norm": 0.49067613703910734,
+      "learning_rate": 1.4243032904367984e-05,
+      "loss": 0.3502,
+      "step": 7327
+    },
+    {
+      "epoch": 1.4445977917981072,
+      "grad_norm": 0.49245160744368666,
+      "learning_rate": 1.424162927366657e-05,
+      "loss": 0.3459,
+      "step": 7328
+    },
+    {
+      "epoch": 1.444794952681388,
+      "grad_norm": 0.4882492569918211,
+      "learning_rate": 1.4240225541056596e-05,
+      "loss": 0.3348,
+      "step": 7329
+    },
+    {
+      "epoch": 1.4449921135646688,
+      "grad_norm": 0.5009716845668267,
+      "learning_rate": 1.423882170657179e-05,
+      "loss": 0.3276,
+      "step": 7330
+    },
+    {
+      "epoch": 1.4451892744479495,
+      "grad_norm": 0.46911913556435136,
+      "learning_rate": 1.4237417770245877e-05,
+      "loss": 0.3144,
+      "step": 7331
+    },
+    {
+      "epoch": 1.4453864353312302,
+      "grad_norm": 0.5206696861698887,
+      "learning_rate": 1.423601373211259e-05,
+      "loss": 0.3623,
+      "step": 7332
+    },
+    {
+      "epoch": 1.4455835962145112,
+      "grad_norm": 0.4906485716791266,
+      "learning_rate": 1.4234609592205662e-05,
+      "loss": 0.3442,
+      "step": 7333
+    },
+    {
+      "epoch": 1.4457807570977919,
+      "grad_norm": 0.5090573234966967,
+      "learning_rate": 1.423320535055883e-05,
+      "loss": 0.3448,
+      "step": 7334
+    },
+    {
+      "epoch": 1.4459779179810726,
+      "grad_norm": 0.4712940869922216,
+      "learning_rate": 1.4231801007205827e-05,
+      "loss": 0.3334,
+      "step": 7335
+    },
+    {
+      "epoch": 1.4461750788643533,
+      "grad_norm": 0.5092882547025841,
+      "learning_rate": 1.4230396562180401e-05,
+      "loss": 0.3217,
+      "step": 7336
+    },
+    {
+      "epoch": 1.446372239747634,
+      "grad_norm": 0.498188825420168,
+      "learning_rate": 1.4228992015516287e-05,
+      "loss": 0.3587,
+      "step": 7337
+    },
+    {
+      "epoch": 1.4465694006309149,
+      "grad_norm": 0.49312249691916427,
+      "learning_rate": 1.4227587367247238e-05,
+      "loss": 0.3465,
+      "step": 7338
+    },
+    {
+      "epoch": 1.4467665615141956,
+      "grad_norm": 0.5293802799153761,
+      "learning_rate": 1.4226182617406996e-05,
+      "loss": 0.3549,
+      "step": 7339
+    },
+    {
+      "epoch": 1.4469637223974763,
+      "grad_norm": 0.45293571133841104,
+      "learning_rate": 1.4224777766029311e-05,
+      "loss": 0.3161,
+      "step": 7340
+    },
+    {
+      "epoch": 1.4471608832807572,
+      "grad_norm": 0.46222311219300893,
+      "learning_rate": 1.4223372813147942e-05,
+      "loss": 0.3056,
+      "step": 7341
+    },
+    {
+      "epoch": 1.447358044164038,
+      "grad_norm": 0.5057284253651947,
+      "learning_rate": 1.422196775879664e-05,
+      "loss": 0.3432,
+      "step": 7342
+    },
+    {
+      "epoch": 1.4475552050473186,
+      "grad_norm": 0.48672729425679273,
+      "learning_rate": 1.422056260300916e-05,
+      "loss": 0.3328,
+      "step": 7343
+    },
+    {
+      "epoch": 1.4477523659305993,
+      "grad_norm": 0.4625625209082356,
+      "learning_rate": 1.4219157345819268e-05,
+      "loss": 0.3003,
+      "step": 7344
+    },
+    {
+      "epoch": 1.44794952681388,
+      "grad_norm": 0.5101487287795781,
+      "learning_rate": 1.421775198726072e-05,
+      "loss": 0.3384,
+      "step": 7345
+    },
+    {
+      "epoch": 1.448146687697161,
+      "grad_norm": 0.4940574773463017,
+      "learning_rate": 1.4216346527367284e-05,
+      "loss": 0.3219,
+      "step": 7346
+    },
+    {
+      "epoch": 1.4483438485804416,
+      "grad_norm": 0.4876148749981708,
+      "learning_rate": 1.421494096617273e-05,
+      "loss": 0.3428,
+      "step": 7347
+    },
+    {
+      "epoch": 1.4485410094637223,
+      "grad_norm": 0.47062050672221883,
+      "learning_rate": 1.4213535303710822e-05,
+      "loss": 0.3042,
+      "step": 7348
+    },
+    {
+      "epoch": 1.4487381703470033,
+      "grad_norm": 0.48615991071089154,
+      "learning_rate": 1.4212129540015339e-05,
+      "loss": 0.3421,
+      "step": 7349
+    },
+    {
+      "epoch": 1.448935331230284,
+      "grad_norm": 0.45576427389164237,
+      "learning_rate": 1.4210723675120049e-05,
+      "loss": 0.3115,
+      "step": 7350
+    },
+    {
+      "epoch": 1.4491324921135647,
+      "grad_norm": 0.5118194739469882,
+      "learning_rate": 1.420931770905873e-05,
+      "loss": 0.3674,
+      "step": 7351
+    },
+    {
+      "epoch": 1.4493296529968454,
+      "grad_norm": 0.5135213363178154,
+      "learning_rate": 1.4207911641865164e-05,
+      "loss": 0.349,
+      "step": 7352
+    },
+    {
+      "epoch": 1.449526813880126,
+      "grad_norm": 0.5010570752976321,
+      "learning_rate": 1.4206505473573135e-05,
+      "loss": 0.3526,
+      "step": 7353
+    },
+    {
+      "epoch": 1.449723974763407,
+      "grad_norm": 0.48809906237733963,
+      "learning_rate": 1.4205099204216421e-05,
+      "loss": 0.3339,
+      "step": 7354
+    },
+    {
+      "epoch": 1.4499211356466877,
+      "grad_norm": 0.48260002387765977,
+      "learning_rate": 1.4203692833828817e-05,
+      "loss": 0.315,
+      "step": 7355
+    },
+    {
+      "epoch": 1.4501182965299684,
+      "grad_norm": 0.4752614271985504,
+      "learning_rate": 1.4202286362444105e-05,
+      "loss": 0.3108,
+      "step": 7356
+    },
+    {
+      "epoch": 1.4503154574132493,
+      "grad_norm": 0.48050012591845676,
+      "learning_rate": 1.4200879790096078e-05,
+      "loss": 0.3118,
+      "step": 7357
+    },
+    {
+      "epoch": 1.45051261829653,
+      "grad_norm": 0.49612980028187825,
+      "learning_rate": 1.419947311681853e-05,
+      "loss": 0.3452,
+      "step": 7358
+    },
+    {
+      "epoch": 1.4507097791798107,
+      "grad_norm": 6.6692877091280165,
+      "learning_rate": 1.4198066342645262e-05,
+      "loss": 0.3363,
+      "step": 7359
+    },
+    {
+      "epoch": 1.4509069400630916,
+      "grad_norm": 0.5452461743710111,
+      "learning_rate": 1.4196659467610068e-05,
+      "loss": 0.3699,
+      "step": 7360
+    },
+    {
+      "epoch": 1.4511041009463723,
+      "grad_norm": 0.5175861598786076,
+      "learning_rate": 1.419525249174675e-05,
+      "loss": 0.3358,
+      "step": 7361
+    },
+    {
+      "epoch": 1.451301261829653,
+      "grad_norm": 0.5004638206871369,
+      "learning_rate": 1.4193845415089113e-05,
+      "loss": 0.3465,
+      "step": 7362
+    },
+    {
+      "epoch": 1.4514984227129337,
+      "grad_norm": 0.5018482035564551,
+      "learning_rate": 1.4192438237670962e-05,
+      "loss": 0.3553,
+      "step": 7363
+    },
+    {
+      "epoch": 1.4516955835962144,
+      "grad_norm": 0.6472551081778795,
+      "learning_rate": 1.4191030959526106e-05,
+      "loss": 0.3311,
+      "step": 7364
+    },
+    {
+      "epoch": 1.4518927444794953,
+      "grad_norm": 0.48615491438153513,
+      "learning_rate": 1.4189623580688358e-05,
+      "loss": 0.3209,
+      "step": 7365
+    },
+    {
+      "epoch": 1.452089905362776,
+      "grad_norm": 0.4996419475157942,
+      "learning_rate": 1.418821610119153e-05,
+      "loss": 0.3374,
+      "step": 7366
+    },
+    {
+      "epoch": 1.4522870662460567,
+      "grad_norm": 0.503614653074225,
+      "learning_rate": 1.4186808521069436e-05,
+      "loss": 0.3366,
+      "step": 7367
+    },
+    {
+      "epoch": 1.4524842271293377,
+      "grad_norm": 0.6183156108882925,
+      "learning_rate": 1.4185400840355895e-05,
+      "loss": 0.3291,
+      "step": 7368
+    },
+    {
+      "epoch": 1.4526813880126184,
+      "grad_norm": 0.4913587909312157,
+      "learning_rate": 1.4183993059084728e-05,
+      "loss": 0.3561,
+      "step": 7369
+    },
+    {
+      "epoch": 1.452878548895899,
+      "grad_norm": 0.5001590541017276,
+      "learning_rate": 1.418258517728976e-05,
+      "loss": 0.3561,
+      "step": 7370
+    },
+    {
+      "epoch": 1.4530757097791798,
+      "grad_norm": 0.4642898097782909,
+      "learning_rate": 1.4181177195004814e-05,
+      "loss": 0.3192,
+      "step": 7371
+    },
+    {
+      "epoch": 1.4532728706624605,
+      "grad_norm": 0.5378165455257576,
+      "learning_rate": 1.4179769112263719e-05,
+      "loss": 0.3453,
+      "step": 7372
+    },
+    {
+      "epoch": 1.4534700315457414,
+      "grad_norm": 0.4847498894723453,
+      "learning_rate": 1.4178360929100303e-05,
+      "loss": 0.3151,
+      "step": 7373
+    },
+    {
+      "epoch": 1.453667192429022,
+      "grad_norm": 0.5067406612634322,
+      "learning_rate": 1.4176952645548406e-05,
+      "loss": 0.3362,
+      "step": 7374
+    },
+    {
+      "epoch": 1.4538643533123028,
+      "grad_norm": 0.4918446350110439,
+      "learning_rate": 1.4175544261641854e-05,
+      "loss": 0.3319,
+      "step": 7375
+    },
+    {
+      "epoch": 1.4540615141955837,
+      "grad_norm": 0.49347879868509636,
+      "learning_rate": 1.417413577741449e-05,
+      "loss": 0.342,
+      "step": 7376
+    },
+    {
+      "epoch": 1.4542586750788644,
+      "grad_norm": 0.4783914406259114,
+      "learning_rate": 1.417272719290015e-05,
+      "loss": 0.3255,
+      "step": 7377
+    },
+    {
+      "epoch": 1.4544558359621451,
+      "grad_norm": 0.4941216512113406,
+      "learning_rate": 1.4171318508132683e-05,
+      "loss": 0.3196,
+      "step": 7378
+    },
+    {
+      "epoch": 1.4546529968454258,
+      "grad_norm": 0.49485434742875883,
+      "learning_rate": 1.416990972314593e-05,
+      "loss": 0.3293,
+      "step": 7379
+    },
+    {
+      "epoch": 1.4548501577287065,
+      "grad_norm": 0.4788373125318391,
+      "learning_rate": 1.4168500837973733e-05,
+      "loss": 0.3059,
+      "step": 7380
+    },
+    {
+      "epoch": 1.4550473186119874,
+      "grad_norm": 0.4834085395567045,
+      "learning_rate": 1.416709185264995e-05,
+      "loss": 0.3363,
+      "step": 7381
+    },
+    {
+      "epoch": 1.4552444794952681,
+      "grad_norm": 0.5134730730417215,
+      "learning_rate": 1.4165682767208426e-05,
+      "loss": 0.3416,
+      "step": 7382
+    },
+    {
+      "epoch": 1.4554416403785488,
+      "grad_norm": 0.5044382136084319,
+      "learning_rate": 1.4164273581683023e-05,
+      "loss": 0.3297,
+      "step": 7383
+    },
+    {
+      "epoch": 1.4556388012618298,
+      "grad_norm": 0.5714557060481008,
+      "learning_rate": 1.4162864296107593e-05,
+      "loss": 0.3529,
+      "step": 7384
+    },
+    {
+      "epoch": 1.4558359621451105,
+      "grad_norm": 0.5054705611777254,
+      "learning_rate": 1.4161454910515997e-05,
+      "loss": 0.3306,
+      "step": 7385
+    },
+    {
+      "epoch": 1.4560331230283912,
+      "grad_norm": 0.47824473710019705,
+      "learning_rate": 1.416004542494209e-05,
+      "loss": 0.3287,
+      "step": 7386
+    },
+    {
+      "epoch": 1.4562302839116719,
+      "grad_norm": 0.4835230550928211,
+      "learning_rate": 1.4158635839419745e-05,
+      "loss": 0.3193,
+      "step": 7387
+    },
+    {
+      "epoch": 1.4564274447949526,
+      "grad_norm": 0.5966821280918011,
+      "learning_rate": 1.4157226153982826e-05,
+      "loss": 0.3512,
+      "step": 7388
+    },
+    {
+      "epoch": 1.4566246056782335,
+      "grad_norm": 0.487708074097899,
+      "learning_rate": 1.4155816368665201e-05,
+      "loss": 0.3356,
+      "step": 7389
+    },
+    {
+      "epoch": 1.4568217665615142,
+      "grad_norm": 0.47577454166544947,
+      "learning_rate": 1.415440648350074e-05,
+      "loss": 0.3373,
+      "step": 7390
+    },
+    {
+      "epoch": 1.4570189274447949,
+      "grad_norm": 0.4872308941864148,
+      "learning_rate": 1.4152996498523317e-05,
+      "loss": 0.3309,
+      "step": 7391
+    },
+    {
+      "epoch": 1.4572160883280758,
+      "grad_norm": 0.5225407802789235,
+      "learning_rate": 1.4151586413766811e-05,
+      "loss": 0.3419,
+      "step": 7392
+    },
+    {
+      "epoch": 1.4574132492113565,
+      "grad_norm": 0.4998933279734873,
+      "learning_rate": 1.4150176229265096e-05,
+      "loss": 0.3668,
+      "step": 7393
+    },
+    {
+      "epoch": 1.4576104100946372,
+      "grad_norm": 0.5050373414529986,
+      "learning_rate": 1.4148765945052056e-05,
+      "loss": 0.3389,
+      "step": 7394
+    },
+    {
+      "epoch": 1.457807570977918,
+      "grad_norm": 0.5064131922922323,
+      "learning_rate": 1.4147355561161574e-05,
+      "loss": 0.3561,
+      "step": 7395
+    },
+    {
+      "epoch": 1.4580047318611986,
+      "grad_norm": 0.4948700452474314,
+      "learning_rate": 1.4145945077627531e-05,
+      "loss": 0.3395,
+      "step": 7396
+    },
+    {
+      "epoch": 1.4582018927444795,
+      "grad_norm": 0.5239164208524786,
+      "learning_rate": 1.4144534494483824e-05,
+      "loss": 0.3643,
+      "step": 7397
+    },
+    {
+      "epoch": 1.4583990536277602,
+      "grad_norm": 0.5345022125344085,
+      "learning_rate": 1.4143123811764335e-05,
+      "loss": 0.3509,
+      "step": 7398
+    },
+    {
+      "epoch": 1.458596214511041,
+      "grad_norm": 0.49581376523712395,
+      "learning_rate": 1.414171302950296e-05,
+      "loss": 0.3298,
+      "step": 7399
+    },
+    {
+      "epoch": 1.4587933753943219,
+      "grad_norm": 0.4921239829922169,
+      "learning_rate": 1.4140302147733596e-05,
+      "loss": 0.3322,
+      "step": 7400
+    },
+    {
+      "epoch": 1.4589905362776026,
+      "grad_norm": 0.5592026245468484,
+      "learning_rate": 1.4138891166490135e-05,
+      "loss": 0.3293,
+      "step": 7401
+    },
+    {
+      "epoch": 1.4591876971608833,
+      "grad_norm": 0.5442208251383976,
+      "learning_rate": 1.4137480085806486e-05,
+      "loss": 0.3361,
+      "step": 7402
+    },
+    {
+      "epoch": 1.459384858044164,
+      "grad_norm": 0.5170050108654181,
+      "learning_rate": 1.413606890571654e-05,
+      "loss": 0.3538,
+      "step": 7403
+    },
+    {
+      "epoch": 1.4595820189274447,
+      "grad_norm": 0.510647100212481,
+      "learning_rate": 1.4134657626254214e-05,
+      "loss": 0.337,
+      "step": 7404
+    },
+    {
+      "epoch": 1.4597791798107256,
+      "grad_norm": 0.5306321326123213,
+      "learning_rate": 1.4133246247453403e-05,
+      "loss": 0.3282,
+      "step": 7405
+    },
+    {
+      "epoch": 1.4599763406940063,
+      "grad_norm": 0.5133341234040529,
+      "learning_rate": 1.4131834769348026e-05,
+      "loss": 0.3366,
+      "step": 7406
+    },
+    {
+      "epoch": 1.460173501577287,
+      "grad_norm": 0.5395277590006632,
+      "learning_rate": 1.4130423191971992e-05,
+      "loss": 0.3333,
+      "step": 7407
+    },
+    {
+      "epoch": 1.460370662460568,
+      "grad_norm": 0.47995198018193463,
+      "learning_rate": 1.4129011515359212e-05,
+      "loss": 0.3362,
+      "step": 7408
+    },
+    {
+      "epoch": 1.4605678233438486,
+      "grad_norm": 0.4987772919977686,
+      "learning_rate": 1.4127599739543606e-05,
+      "loss": 0.3347,
+      "step": 7409
+    },
+    {
+      "epoch": 1.4607649842271293,
+      "grad_norm": 0.5453925939490976,
+      "learning_rate": 1.4126187864559094e-05,
+      "loss": 0.3689,
+      "step": 7410
+    },
+    {
+      "epoch": 1.4609621451104102,
+      "grad_norm": 0.4959830516006325,
+      "learning_rate": 1.4124775890439595e-05,
+      "loss": 0.3154,
+      "step": 7411
+    },
+    {
+      "epoch": 1.461159305993691,
+      "grad_norm": 0.5124401018607515,
+      "learning_rate": 1.4123363817219034e-05,
+      "loss": 0.321,
+      "step": 7412
+    },
+    {
+      "epoch": 1.4613564668769716,
+      "grad_norm": 1.2270349005748566,
+      "learning_rate": 1.4121951644931336e-05,
+      "loss": 0.3614,
+      "step": 7413
+    },
+    {
+      "epoch": 1.4615536277602523,
+      "grad_norm": 0.5138791690617409,
+      "learning_rate": 1.4120539373610429e-05,
+      "loss": 0.3237,
+      "step": 7414
+    },
+    {
+      "epoch": 1.461750788643533,
+      "grad_norm": 0.4521273911127816,
+      "learning_rate": 1.4119127003290248e-05,
+      "loss": 0.3186,
+      "step": 7415
+    },
+    {
+      "epoch": 1.461947949526814,
+      "grad_norm": 0.49874325861714064,
+      "learning_rate": 1.411771453400472e-05,
+      "loss": 0.3281,
+      "step": 7416
+    },
+    {
+      "epoch": 1.4621451104100947,
+      "grad_norm": 0.5021452602032279,
+      "learning_rate": 1.4116301965787786e-05,
+      "loss": 0.3404,
+      "step": 7417
+    },
+    {
+      "epoch": 1.4623422712933754,
+      "grad_norm": 0.4973741088361569,
+      "learning_rate": 1.4114889298673383e-05,
+      "loss": 0.3309,
+      "step": 7418
+    },
+    {
+      "epoch": 1.4625394321766563,
+      "grad_norm": 0.516601612052012,
+      "learning_rate": 1.4113476532695452e-05,
+      "loss": 0.3076,
+      "step": 7419
+    },
+    {
+      "epoch": 1.462736593059937,
+      "grad_norm": 0.5526911269279787,
+      "learning_rate": 1.4112063667887932e-05,
+      "loss": 0.3539,
+      "step": 7420
+    },
+    {
+      "epoch": 1.4629337539432177,
+      "grad_norm": 0.5562535752274906,
+      "learning_rate": 1.4110650704284773e-05,
+      "loss": 0.3569,
+      "step": 7421
+    },
+    {
+      "epoch": 1.4631309148264984,
+      "grad_norm": 0.5488213791263362,
+      "learning_rate": 1.410923764191992e-05,
+      "loss": 0.357,
+      "step": 7422
+    },
+    {
+      "epoch": 1.463328075709779,
+      "grad_norm": 0.5107788895192845,
+      "learning_rate": 1.4107824480827324e-05,
+      "loss": 0.3328,
+      "step": 7423
+    },
+    {
+      "epoch": 1.46352523659306,
+      "grad_norm": 0.5171838627063462,
+      "learning_rate": 1.4106411221040935e-05,
+      "loss": 0.3217,
+      "step": 7424
+    },
+    {
+      "epoch": 1.4637223974763407,
+      "grad_norm": 0.4922378526729013,
+      "learning_rate": 1.4104997862594711e-05,
+      "loss": 0.3343,
+      "step": 7425
+    },
+    {
+      "epoch": 1.4639195583596214,
+      "grad_norm": 0.5037026945262372,
+      "learning_rate": 1.4103584405522605e-05,
+      "loss": 0.3284,
+      "step": 7426
+    },
+    {
+      "epoch": 1.4641167192429023,
+      "grad_norm": 0.49326551249285355,
+      "learning_rate": 1.4102170849858583e-05,
+      "loss": 0.3443,
+      "step": 7427
+    },
+    {
+      "epoch": 1.464313880126183,
+      "grad_norm": 0.546820130602552,
+      "learning_rate": 1.41007571956366e-05,
+      "loss": 0.3602,
+      "step": 7428
+    },
+    {
+      "epoch": 1.4645110410094637,
+      "grad_norm": 0.4837797980660489,
+      "learning_rate": 1.4099343442890624e-05,
+      "loss": 0.3333,
+      "step": 7429
+    },
+    {
+      "epoch": 1.4647082018927444,
+      "grad_norm": 0.5126857949066812,
+      "learning_rate": 1.4097929591654621e-05,
+      "loss": 0.3383,
+      "step": 7430
+    },
+    {
+      "epoch": 1.4649053627760251,
+      "grad_norm": 0.46279281738729217,
+      "learning_rate": 1.4096515641962563e-05,
+      "loss": 0.3435,
+      "step": 7431
+    },
+    {
+      "epoch": 1.465102523659306,
+      "grad_norm": 2.09692903793275,
+      "learning_rate": 1.4095101593848415e-05,
+      "loss": 0.3672,
+      "step": 7432
+    },
+    {
+      "epoch": 1.4652996845425867,
+      "grad_norm": 0.5501540341610509,
+      "learning_rate": 1.4093687447346151e-05,
+      "loss": 0.3876,
+      "step": 7433
+    },
+    {
+      "epoch": 1.4654968454258674,
+      "grad_norm": 0.47609912782149055,
+      "learning_rate": 1.409227320248975e-05,
+      "loss": 0.3153,
+      "step": 7434
+    },
+    {
+      "epoch": 1.4656940063091484,
+      "grad_norm": 0.525537823437008,
+      "learning_rate": 1.4090858859313193e-05,
+      "loss": 0.3046,
+      "step": 7435
+    },
+    {
+      "epoch": 1.465891167192429,
+      "grad_norm": 0.5028179572182535,
+      "learning_rate": 1.4089444417850455e-05,
+      "loss": 0.3421,
+      "step": 7436
+    },
+    {
+      "epoch": 1.4660883280757098,
+      "grad_norm": 0.5163272751032175,
+      "learning_rate": 1.408802987813552e-05,
+      "loss": 0.3294,
+      "step": 7437
+    },
+    {
+      "epoch": 1.4662854889589905,
+      "grad_norm": 0.5132796172325643,
+      "learning_rate": 1.408661524020238e-05,
+      "loss": 0.3264,
+      "step": 7438
+    },
+    {
+      "epoch": 1.4664826498422712,
+      "grad_norm": 0.5108505262340117,
+      "learning_rate": 1.4085200504085013e-05,
+      "loss": 0.3491,
+      "step": 7439
+    },
+    {
+      "epoch": 1.466679810725552,
+      "grad_norm": 0.48158384519210207,
+      "learning_rate": 1.4083785669817417e-05,
+      "loss": 0.3258,
+      "step": 7440
+    },
+    {
+      "epoch": 1.4668769716088328,
+      "grad_norm": 0.6121936078703637,
+      "learning_rate": 1.408237073743358e-05,
+      "loss": 0.3299,
+      "step": 7441
+    },
+    {
+      "epoch": 1.4670741324921135,
+      "grad_norm": 0.5452188941885944,
+      "learning_rate": 1.4080955706967501e-05,
+      "loss": 0.3629,
+      "step": 7442
+    },
+    {
+      "epoch": 1.4672712933753944,
+      "grad_norm": 0.467809566032474,
+      "learning_rate": 1.407954057845317e-05,
+      "loss": 0.3111,
+      "step": 7443
+    },
+    {
+      "epoch": 1.4674684542586751,
+      "grad_norm": 0.5136940118705474,
+      "learning_rate": 1.4078125351924597e-05,
+      "loss": 0.3638,
+      "step": 7444
+    },
+    {
+      "epoch": 1.4676656151419558,
+      "grad_norm": 0.511820180671828,
+      "learning_rate": 1.4076710027415776e-05,
+      "loss": 0.357,
+      "step": 7445
+    },
+    {
+      "epoch": 1.4678627760252365,
+      "grad_norm": 0.490168596701873,
+      "learning_rate": 1.4075294604960715e-05,
+      "loss": 0.3198,
+      "step": 7446
+    },
+    {
+      "epoch": 1.4680599369085172,
+      "grad_norm": 0.5058740263507188,
+      "learning_rate": 1.4073879084593416e-05,
+      "loss": 0.3195,
+      "step": 7447
+    },
+    {
+      "epoch": 1.4682570977917981,
+      "grad_norm": 0.4674312373171209,
+      "learning_rate": 1.4072463466347892e-05,
+      "loss": 0.3201,
+      "step": 7448
+    },
+    {
+      "epoch": 1.4684542586750788,
+      "grad_norm": 2.1568178975708383,
+      "learning_rate": 1.4071047750258156e-05,
+      "loss": 0.3337,
+      "step": 7449
+    },
+    {
+      "epoch": 1.4686514195583595,
+      "grad_norm": 0.5546196457316546,
+      "learning_rate": 1.4069631936358214e-05,
+      "loss": 0.3781,
+      "step": 7450
+    },
+    {
+      "epoch": 1.4688485804416405,
+      "grad_norm": 0.5357099924800648,
+      "learning_rate": 1.4068216024682095e-05,
+      "loss": 0.3631,
+      "step": 7451
+    },
+    {
+      "epoch": 1.4690457413249212,
+      "grad_norm": 0.507066065309645,
+      "learning_rate": 1.4066800015263807e-05,
+      "loss": 0.3288,
+      "step": 7452
+    },
+    {
+      "epoch": 1.4692429022082019,
+      "grad_norm": 0.4874279119555861,
+      "learning_rate": 1.4065383908137373e-05,
+      "loss": 0.3218,
+      "step": 7453
+    },
+    {
+      "epoch": 1.4694400630914828,
+      "grad_norm": 0.4762728451473061,
+      "learning_rate": 1.4063967703336814e-05,
+      "loss": 0.3198,
+      "step": 7454
+    },
+    {
+      "epoch": 1.4696372239747635,
+      "grad_norm": 0.4854069150886455,
+      "learning_rate": 1.4062551400896163e-05,
+      "loss": 0.3287,
+      "step": 7455
+    },
+    {
+      "epoch": 1.4698343848580442,
+      "grad_norm": 0.4715608643695716,
+      "learning_rate": 1.406113500084944e-05,
+      "loss": 0.2892,
+      "step": 7456
+    },
+    {
+      "epoch": 1.4700315457413249,
+      "grad_norm": 0.5002411384785533,
+      "learning_rate": 1.405971850323068e-05,
+      "loss": 0.3071,
+      "step": 7457
+    },
+    {
+      "epoch": 1.4702287066246056,
+      "grad_norm": 0.5388064136656453,
+      "learning_rate": 1.4058301908073912e-05,
+      "loss": 0.3851,
+      "step": 7458
+    },
+    {
+      "epoch": 1.4704258675078865,
+      "grad_norm": 0.49296401636550136,
+      "learning_rate": 1.4056885215413174e-05,
+      "loss": 0.3298,
+      "step": 7459
+    },
+    {
+      "epoch": 1.4706230283911672,
+      "grad_norm": 0.5355211557498206,
+      "learning_rate": 1.4055468425282502e-05,
+      "loss": 0.338,
+      "step": 7460
+    },
+    {
+      "epoch": 1.470820189274448,
+      "grad_norm": 0.48977540245567913,
+      "learning_rate": 1.4054051537715933e-05,
+      "loss": 0.317,
+      "step": 7461
+    },
+    {
+      "epoch": 1.4710173501577288,
+      "grad_norm": 0.5326327177590915,
+      "learning_rate": 1.4052634552747512e-05,
+      "loss": 0.346,
+      "step": 7462
+    },
+    {
+      "epoch": 1.4712145110410095,
+      "grad_norm": 0.526993033584732,
+      "learning_rate": 1.4051217470411284e-05,
+      "loss": 0.3519,
+      "step": 7463
+    },
+    {
+      "epoch": 1.4714116719242902,
+      "grad_norm": 0.5258328026865123,
+      "learning_rate": 1.4049800290741293e-05,
+      "loss": 0.3475,
+      "step": 7464
+    },
+    {
+      "epoch": 1.471608832807571,
+      "grad_norm": 0.5203038982752369,
+      "learning_rate": 1.4048383013771588e-05,
+      "loss": 0.3321,
+      "step": 7465
+    },
+    {
+      "epoch": 1.4718059936908516,
+      "grad_norm": 0.5113074927167442,
+      "learning_rate": 1.404696563953622e-05,
+      "loss": 0.3354,
+      "step": 7466
+    },
+    {
+      "epoch": 1.4720031545741326,
+      "grad_norm": 0.49919882528097426,
+      "learning_rate": 1.4045548168069246e-05,
+      "loss": 0.3539,
+      "step": 7467
+    },
+    {
+      "epoch": 1.4722003154574133,
+      "grad_norm": 0.5016200068739187,
+      "learning_rate": 1.4044130599404717e-05,
+      "loss": 0.325,
+      "step": 7468
+    },
+    {
+      "epoch": 1.472397476340694,
+      "grad_norm": 0.5564339720636347,
+      "learning_rate": 1.4042712933576694e-05,
+      "loss": 0.3908,
+      "step": 7469
+    },
+    {
+      "epoch": 1.4725946372239749,
+      "grad_norm": 0.47101819777928333,
+      "learning_rate": 1.4041295170619241e-05,
+      "loss": 0.3266,
+      "step": 7470
+    },
+    {
+      "epoch": 1.4727917981072556,
+      "grad_norm": 0.4968101062786291,
+      "learning_rate": 1.403987731056641e-05,
+      "loss": 0.3045,
+      "step": 7471
+    },
+    {
+      "epoch": 1.4729889589905363,
+      "grad_norm": 0.5063663744372873,
+      "learning_rate": 1.403845935345228e-05,
+      "loss": 0.323,
+      "step": 7472
+    },
+    {
+      "epoch": 1.473186119873817,
+      "grad_norm": 0.4892904205195065,
+      "learning_rate": 1.4037041299310908e-05,
+      "loss": 0.3245,
+      "step": 7473
+    },
+    {
+      "epoch": 1.4733832807570977,
+      "grad_norm": 0.5159500037402428,
+      "learning_rate": 1.4035623148176369e-05,
+      "loss": 0.3428,
+      "step": 7474
+    },
+    {
+      "epoch": 1.4735804416403786,
+      "grad_norm": 0.48891170305663467,
+      "learning_rate": 1.4034204900082734e-05,
+      "loss": 0.3305,
+      "step": 7475
+    },
+    {
+      "epoch": 1.4737776025236593,
+      "grad_norm": 0.557799558291397,
+      "learning_rate": 1.4032786555064077e-05,
+      "loss": 0.3915,
+      "step": 7476
+    },
+    {
+      "epoch": 1.47397476340694,
+      "grad_norm": 0.49527625272519055,
+      "learning_rate": 1.4031368113154478e-05,
+      "loss": 0.323,
+      "step": 7477
+    },
+    {
+      "epoch": 1.474171924290221,
+      "grad_norm": 0.5316789203889967,
+      "learning_rate": 1.4029949574388009e-05,
+      "loss": 0.3531,
+      "step": 7478
+    },
+    {
+      "epoch": 1.4743690851735016,
+      "grad_norm": 0.47098548601018697,
+      "learning_rate": 1.4028530938798759e-05,
+      "loss": 0.3212,
+      "step": 7479
+    },
+    {
+      "epoch": 1.4745662460567823,
+      "grad_norm": 0.490959684927149,
+      "learning_rate": 1.402711220642081e-05,
+      "loss": 0.328,
+      "step": 7480
+    },
+    {
+      "epoch": 1.474763406940063,
+      "grad_norm": 0.5168859380738673,
+      "learning_rate": 1.4025693377288246e-05,
+      "loss": 0.3489,
+      "step": 7481
+    },
+    {
+      "epoch": 1.4749605678233437,
+      "grad_norm": 0.49571446645437645,
+      "learning_rate": 1.4024274451435157e-05,
+      "loss": 0.3375,
+      "step": 7482
+    },
+    {
+      "epoch": 1.4751577287066246,
+      "grad_norm": 0.5197521991945561,
+      "learning_rate": 1.4022855428895632e-05,
+      "loss": 0.3343,
+      "step": 7483
+    },
+    {
+      "epoch": 1.4753548895899053,
+      "grad_norm": 0.48589615662818103,
+      "learning_rate": 1.4021436309703766e-05,
+      "loss": 0.3201,
+      "step": 7484
+    },
+    {
+      "epoch": 1.475552050473186,
+      "grad_norm": 0.4995993722216193,
+      "learning_rate": 1.4020017093893656e-05,
+      "loss": 0.3371,
+      "step": 7485
+    },
+    {
+      "epoch": 1.475749211356467,
+      "grad_norm": 0.5036745093017024,
+      "learning_rate": 1.4018597781499399e-05,
+      "loss": 0.3505,
+      "step": 7486
+    },
+    {
+      "epoch": 1.4759463722397477,
+      "grad_norm": 0.4878243600702588,
+      "learning_rate": 1.4017178372555092e-05,
+      "loss": 0.3085,
+      "step": 7487
+    },
+    {
+      "epoch": 1.4761435331230284,
+      "grad_norm": 0.508433396912967,
+      "learning_rate": 1.4015758867094837e-05,
+      "loss": 0.3382,
+      "step": 7488
+    },
+    {
+      "epoch": 1.476340694006309,
+      "grad_norm": 0.5127101875396283,
+      "learning_rate": 1.4014339265152748e-05,
+      "loss": 0.3524,
+      "step": 7489
+    },
+    {
+      "epoch": 1.4765378548895898,
+      "grad_norm": 0.4724562924935433,
+      "learning_rate": 1.401291956676292e-05,
+      "loss": 0.3308,
+      "step": 7490
+    },
+    {
+      "epoch": 1.4767350157728707,
+      "grad_norm": 0.49909821615758254,
+      "learning_rate": 1.4011499771959469e-05,
+      "loss": 0.3278,
+      "step": 7491
+    },
+    {
+      "epoch": 1.4769321766561514,
+      "grad_norm": 0.500929277106934,
+      "learning_rate": 1.4010079880776505e-05,
+      "loss": 0.3428,
+      "step": 7492
+    },
+    {
+      "epoch": 1.477129337539432,
+      "grad_norm": 0.4578591298461616,
+      "learning_rate": 1.4008659893248147e-05,
+      "loss": 0.3263,
+      "step": 7493
+    },
+    {
+      "epoch": 1.477326498422713,
+      "grad_norm": 0.4829431282504663,
+      "learning_rate": 1.40072398094085e-05,
+      "loss": 0.3338,
+      "step": 7494
+    },
+    {
+      "epoch": 1.4775236593059937,
+      "grad_norm": 0.48641913532636866,
+      "learning_rate": 1.4005819629291692e-05,
+      "loss": 0.3367,
+      "step": 7495
+    },
+    {
+      "epoch": 1.4777208201892744,
+      "grad_norm": 0.5087064304556296,
+      "learning_rate": 1.4004399352931846e-05,
+      "loss": 0.3482,
+      "step": 7496
+    },
+    {
+      "epoch": 1.4779179810725553,
+      "grad_norm": 0.4930581875396616,
+      "learning_rate": 1.4002978980363075e-05,
+      "loss": 0.3192,
+      "step": 7497
+    },
+    {
+      "epoch": 1.478115141955836,
+      "grad_norm": 0.4957220120205278,
+      "learning_rate": 1.4001558511619515e-05,
+      "loss": 0.3341,
+      "step": 7498
+    },
+    {
+      "epoch": 1.4783123028391167,
+      "grad_norm": 0.514560592615134,
+      "learning_rate": 1.4000137946735284e-05,
+      "loss": 0.3032,
+      "step": 7499
+    },
+    {
+      "epoch": 1.4785094637223974,
+      "grad_norm": 0.5043148636879016,
+      "learning_rate": 1.3998717285744524e-05,
+      "loss": 0.3387,
+      "step": 7500
+    },
+    {
+      "epoch": 1.4787066246056781,
+      "grad_norm": 0.5094476950875343,
+      "learning_rate": 1.3997296528681355e-05,
+      "loss": 0.3536,
+      "step": 7501
+    },
+    {
+      "epoch": 1.478903785488959,
+      "grad_norm": 0.509271590468565,
+      "learning_rate": 1.3995875675579922e-05,
+      "loss": 0.3335,
+      "step": 7502
+    },
+    {
+      "epoch": 1.4791009463722398,
+      "grad_norm": 0.475098165612926,
+      "learning_rate": 1.3994454726474355e-05,
+      "loss": 0.3232,
+      "step": 7503
+    },
+    {
+      "epoch": 1.4792981072555205,
+      "grad_norm": 0.49409393273104263,
+      "learning_rate": 1.3993033681398797e-05,
+      "loss": 0.336,
+      "step": 7504
+    },
+    {
+      "epoch": 1.4794952681388014,
+      "grad_norm": 0.5025889163503731,
+      "learning_rate": 1.399161254038739e-05,
+      "loss": 0.3479,
+      "step": 7505
+    },
+    {
+      "epoch": 1.479692429022082,
+      "grad_norm": 0.5213687127491038,
+      "learning_rate": 1.399019130347428e-05,
+      "loss": 0.3591,
+      "step": 7506
+    },
+    {
+      "epoch": 1.4798895899053628,
+      "grad_norm": 0.49494600597377825,
+      "learning_rate": 1.3988769970693607e-05,
+      "loss": 0.3501,
+      "step": 7507
+    },
+    {
+      "epoch": 1.4800867507886435,
+      "grad_norm": 0.5047526802460451,
+      "learning_rate": 1.3987348542079526e-05,
+      "loss": 0.355,
+      "step": 7508
+    },
+    {
+      "epoch": 1.4802839116719242,
+      "grad_norm": 0.48534628796917784,
+      "learning_rate": 1.3985927017666183e-05,
+      "loss": 0.3306,
+      "step": 7509
+    },
+    {
+      "epoch": 1.4804810725552051,
+      "grad_norm": 0.5163215036754177,
+      "learning_rate": 1.3984505397487736e-05,
+      "loss": 0.3277,
+      "step": 7510
+    },
+    {
+      "epoch": 1.4806782334384858,
+      "grad_norm": 0.48771109504055543,
+      "learning_rate": 1.3983083681578336e-05,
+      "loss": 0.3142,
+      "step": 7511
+    },
+    {
+      "epoch": 1.4808753943217665,
+      "grad_norm": 0.49100954381314643,
+      "learning_rate": 1.3981661869972143e-05,
+      "loss": 0.3508,
+      "step": 7512
+    },
+    {
+      "epoch": 1.4810725552050474,
+      "grad_norm": 0.4964769898882689,
+      "learning_rate": 1.3980239962703316e-05,
+      "loss": 0.3195,
+      "step": 7513
+    },
+    {
+      "epoch": 1.4812697160883281,
+      "grad_norm": 0.4933646357348641,
+      "learning_rate": 1.3978817959806022e-05,
+      "loss": 0.3346,
+      "step": 7514
+    },
+    {
+      "epoch": 1.4814668769716088,
+      "grad_norm": 0.5237222580406612,
+      "learning_rate": 1.397739586131442e-05,
+      "loss": 0.3534,
+      "step": 7515
+    },
+    {
+      "epoch": 1.4816640378548895,
+      "grad_norm": 0.5499841344032339,
+      "learning_rate": 1.3975973667262678e-05,
+      "loss": 0.3752,
+      "step": 7516
+    },
+    {
+      "epoch": 1.4818611987381702,
+      "grad_norm": 0.5281225252481041,
+      "learning_rate": 1.397455137768497e-05,
+      "loss": 0.3612,
+      "step": 7517
+    },
+    {
+      "epoch": 1.4820583596214512,
+      "grad_norm": 0.4754396639001578,
+      "learning_rate": 1.3973128992615461e-05,
+      "loss": 0.3294,
+      "step": 7518
+    },
+    {
+      "epoch": 1.4822555205047319,
+      "grad_norm": 0.5106352743028668,
+      "learning_rate": 1.3971706512088334e-05,
+      "loss": 0.3489,
+      "step": 7519
+    },
+    {
+      "epoch": 1.4824526813880126,
+      "grad_norm": 0.4899386431725224,
+      "learning_rate": 1.3970283936137755e-05,
+      "loss": 0.326,
+      "step": 7520
+    },
+    {
+      "epoch": 1.4826498422712935,
+      "grad_norm": 0.49818208559529764,
+      "learning_rate": 1.3968861264797911e-05,
+      "loss": 0.3261,
+      "step": 7521
+    },
+    {
+      "epoch": 1.4828470031545742,
+      "grad_norm": 0.49436959103979455,
+      "learning_rate": 1.3967438498102971e-05,
+      "loss": 0.3511,
+      "step": 7522
+    },
+    {
+      "epoch": 1.4830441640378549,
+      "grad_norm": 0.5033269127969884,
+      "learning_rate": 1.3966015636087133e-05,
+      "loss": 0.3574,
+      "step": 7523
+    },
+    {
+      "epoch": 1.4832413249211356,
+      "grad_norm": 0.4906719889384929,
+      "learning_rate": 1.3964592678784574e-05,
+      "loss": 0.3317,
+      "step": 7524
+    },
+    {
+      "epoch": 1.4834384858044163,
+      "grad_norm": 0.516805390516583,
+      "learning_rate": 1.3963169626229485e-05,
+      "loss": 0.3533,
+      "step": 7525
+    },
+    {
+      "epoch": 1.4836356466876972,
+      "grad_norm": 0.4704857874190068,
+      "learning_rate": 1.396174647845605e-05,
+      "loss": 0.3244,
+      "step": 7526
+    },
+    {
+      "epoch": 1.483832807570978,
+      "grad_norm": 0.8886639274014726,
+      "learning_rate": 1.396032323549847e-05,
+      "loss": 0.36,
+      "step": 7527
+    },
+    {
+      "epoch": 1.4840299684542586,
+      "grad_norm": 0.5070409296139642,
+      "learning_rate": 1.3958899897390935e-05,
+      "loss": 0.3479,
+      "step": 7528
+    },
+    {
+      "epoch": 1.4842271293375395,
+      "grad_norm": 0.46156188122605435,
+      "learning_rate": 1.3957476464167639e-05,
+      "loss": 0.317,
+      "step": 7529
+    },
+    {
+      "epoch": 1.4844242902208202,
+      "grad_norm": 0.4798500007246751,
+      "learning_rate": 1.3956052935862782e-05,
+      "loss": 0.3307,
+      "step": 7530
+    },
+    {
+      "epoch": 1.484621451104101,
+      "grad_norm": 0.4699403853866003,
+      "learning_rate": 1.3954629312510573e-05,
+      "loss": 0.3395,
+      "step": 7531
+    },
+    {
+      "epoch": 1.4848186119873816,
+      "grad_norm": 0.5019467243741862,
+      "learning_rate": 1.3953205594145207e-05,
+      "loss": 0.3489,
+      "step": 7532
+    },
+    {
+      "epoch": 1.4850157728706623,
+      "grad_norm": 0.49998290248687316,
+      "learning_rate": 1.3951781780800892e-05,
+      "loss": 0.3493,
+      "step": 7533
+    },
+    {
+      "epoch": 1.4852129337539433,
+      "grad_norm": 0.4952129809254002,
+      "learning_rate": 1.3950357872511839e-05,
+      "loss": 0.3442,
+      "step": 7534
+    },
+    {
+      "epoch": 1.485410094637224,
+      "grad_norm": 0.5444354708755205,
+      "learning_rate": 1.3948933869312258e-05,
+      "loss": 0.3442,
+      "step": 7535
+    },
+    {
+      "epoch": 1.4856072555205047,
+      "grad_norm": 0.4968530516895542,
+      "learning_rate": 1.3947509771236361e-05,
+      "loss": 0.3205,
+      "step": 7536
+    },
+    {
+      "epoch": 1.4858044164037856,
+      "grad_norm": 0.6219558232573202,
+      "learning_rate": 1.3946085578318358e-05,
+      "loss": 0.3735,
+      "step": 7537
+    },
+    {
+      "epoch": 1.4860015772870663,
+      "grad_norm": 0.48319973377602,
+      "learning_rate": 1.3944661290592476e-05,
+      "loss": 0.3647,
+      "step": 7538
+    },
+    {
+      "epoch": 1.486198738170347,
+      "grad_norm": 0.4717457826676787,
+      "learning_rate": 1.3943236908092926e-05,
+      "loss": 0.3193,
+      "step": 7539
+    },
+    {
+      "epoch": 1.4863958990536277,
+      "grad_norm": 0.544058207471016,
+      "learning_rate": 1.3941812430853938e-05,
+      "loss": 0.3496,
+      "step": 7540
+    },
+    {
+      "epoch": 1.4865930599369084,
+      "grad_norm": 0.5298496126711097,
+      "learning_rate": 1.394038785890973e-05,
+      "loss": 0.3244,
+      "step": 7541
+    },
+    {
+      "epoch": 1.4867902208201893,
+      "grad_norm": 0.4946581847477271,
+      "learning_rate": 1.3938963192294533e-05,
+      "loss": 0.355,
+      "step": 7542
+    },
+    {
+      "epoch": 1.48698738170347,
+      "grad_norm": 0.46001185450442234,
+      "learning_rate": 1.3937538431042567e-05,
+      "loss": 0.3039,
+      "step": 7543
+    },
+    {
+      "epoch": 1.4871845425867507,
+      "grad_norm": 0.6003857261401706,
+      "learning_rate": 1.3936113575188074e-05,
+      "loss": 0.3356,
+      "step": 7544
+    },
+    {
+      "epoch": 1.4873817034700316,
+      "grad_norm": 0.48039873611738615,
+      "learning_rate": 1.3934688624765282e-05,
+      "loss": 0.3188,
+      "step": 7545
+    },
+    {
+      "epoch": 1.4875788643533123,
+      "grad_norm": 0.458905469054827,
+      "learning_rate": 1.3933263579808426e-05,
+      "loss": 0.3186,
+      "step": 7546
+    },
+    {
+      "epoch": 1.487776025236593,
+      "grad_norm": 0.5281694152795835,
+      "learning_rate": 1.3931838440351748e-05,
+      "loss": 0.3689,
+      "step": 7547
+    },
+    {
+      "epoch": 1.487973186119874,
+      "grad_norm": 0.48164249983072344,
+      "learning_rate": 1.3930413206429483e-05,
+      "loss": 0.3264,
+      "step": 7548
+    },
+    {
+      "epoch": 1.4881703470031546,
+      "grad_norm": 0.4698429076871976,
+      "learning_rate": 1.3928987878075874e-05,
+      "loss": 0.3312,
+      "step": 7549
+    },
+    {
+      "epoch": 1.4883675078864353,
+      "grad_norm": 0.4930590803919952,
+      "learning_rate": 1.392756245532517e-05,
+      "loss": 0.3312,
+      "step": 7550
+    },
+    {
+      "epoch": 1.488564668769716,
+      "grad_norm": 0.4898819634829872,
+      "learning_rate": 1.3926136938211615e-05,
+      "loss": 0.3485,
+      "step": 7551
+    },
+    {
+      "epoch": 1.4887618296529967,
+      "grad_norm": 0.48401145600289563,
+      "learning_rate": 1.3924711326769457e-05,
+      "loss": 0.3347,
+      "step": 7552
+    },
+    {
+      "epoch": 1.4889589905362777,
+      "grad_norm": 0.47592844367724896,
+      "learning_rate": 1.392328562103295e-05,
+      "loss": 0.3189,
+      "step": 7553
+    },
+    {
+      "epoch": 1.4891561514195584,
+      "grad_norm": 0.47313537818185825,
+      "learning_rate": 1.3921859821036345e-05,
+      "loss": 0.3195,
+      "step": 7554
+    },
+    {
+      "epoch": 1.489353312302839,
+      "grad_norm": 0.4828381063664307,
+      "learning_rate": 1.3920433926813901e-05,
+      "loss": 0.348,
+      "step": 7555
+    },
+    {
+      "epoch": 1.48955047318612,
+      "grad_norm": 0.5075800480469936,
+      "learning_rate": 1.3919007938399873e-05,
+      "loss": 0.353,
+      "step": 7556
+    },
+    {
+      "epoch": 1.4897476340694007,
+      "grad_norm": 0.4605077583110866,
+      "learning_rate": 1.3917581855828526e-05,
+      "loss": 0.3158,
+      "step": 7557
+    },
+    {
+      "epoch": 1.4899447949526814,
+      "grad_norm": 0.523185615476558,
+      "learning_rate": 1.3916155679134118e-05,
+      "loss": 0.3629,
+      "step": 7558
+    },
+    {
+      "epoch": 1.490141955835962,
+      "grad_norm": 4.614531591833265,
+      "learning_rate": 1.3914729408350918e-05,
+      "loss": 0.3198,
+      "step": 7559
+    },
+    {
+      "epoch": 1.4903391167192428,
+      "grad_norm": 0.6690308387896126,
+      "learning_rate": 1.3913303043513188e-05,
+      "loss": 0.3414,
+      "step": 7560
+    },
+    {
+      "epoch": 1.4905362776025237,
+      "grad_norm": 0.497169553741172,
+      "learning_rate": 1.3911876584655206e-05,
+      "loss": 0.3556,
+      "step": 7561
+    },
+    {
+      "epoch": 1.4907334384858044,
+      "grad_norm": 0.47652650301196414,
+      "learning_rate": 1.3910450031811235e-05,
+      "loss": 0.3258,
+      "step": 7562
+    },
+    {
+      "epoch": 1.4909305993690851,
+      "grad_norm": 0.500693022752151,
+      "learning_rate": 1.3909023385015551e-05,
+      "loss": 0.3462,
+      "step": 7563
+    },
+    {
+      "epoch": 1.491127760252366,
+      "grad_norm": 0.5068560385443583,
+      "learning_rate": 1.390759664430244e-05,
+      "loss": 0.3506,
+      "step": 7564
+    },
+    {
+      "epoch": 1.4913249211356467,
+      "grad_norm": 0.5227101584388434,
+      "learning_rate": 1.3906169809706165e-05,
+      "loss": 0.3384,
+      "step": 7565
+    },
+    {
+      "epoch": 1.4915220820189274,
+      "grad_norm": 0.49830512467585986,
+      "learning_rate": 1.390474288126102e-05,
+      "loss": 0.3357,
+      "step": 7566
+    },
+    {
+      "epoch": 1.4917192429022081,
+      "grad_norm": 0.5135981865878078,
+      "learning_rate": 1.3903315859001278e-05,
+      "loss": 0.3515,
+      "step": 7567
+    },
+    {
+      "epoch": 1.4919164037854888,
+      "grad_norm": 0.5072032920741137,
+      "learning_rate": 1.3901888742961233e-05,
+      "loss": 0.3636,
+      "step": 7568
+    },
+    {
+      "epoch": 1.4921135646687698,
+      "grad_norm": 0.46652743790054163,
+      "learning_rate": 1.3900461533175167e-05,
+      "loss": 0.3255,
+      "step": 7569
+    },
+    {
+      "epoch": 1.4923107255520505,
+      "grad_norm": 0.4559665210107085,
+      "learning_rate": 1.3899034229677373e-05,
+      "loss": 0.3187,
+      "step": 7570
+    },
+    {
+      "epoch": 1.4925078864353312,
+      "grad_norm": 0.4955869955690288,
+      "learning_rate": 1.389760683250214e-05,
+      "loss": 0.3473,
+      "step": 7571
+    },
+    {
+      "epoch": 1.492705047318612,
+      "grad_norm": 0.532343767493093,
+      "learning_rate": 1.3896179341683763e-05,
+      "loss": 0.3666,
+      "step": 7572
+    },
+    {
+      "epoch": 1.4929022082018928,
+      "grad_norm": 0.49706921835508877,
+      "learning_rate": 1.3894751757256544e-05,
+      "loss": 0.3403,
+      "step": 7573
+    },
+    {
+      "epoch": 1.4930993690851735,
+      "grad_norm": 0.5079181565121896,
+      "learning_rate": 1.3893324079254776e-05,
+      "loss": 0.3687,
+      "step": 7574
+    },
+    {
+      "epoch": 1.4932965299684542,
+      "grad_norm": 0.4908701711814409,
+      "learning_rate": 1.389189630771276e-05,
+      "loss": 0.3468,
+      "step": 7575
+    },
+    {
+      "epoch": 1.493493690851735,
+      "grad_norm": 0.4992855170423717,
+      "learning_rate": 1.3890468442664801e-05,
+      "loss": 0.3304,
+      "step": 7576
+    },
+    {
+      "epoch": 1.4936908517350158,
+      "grad_norm": 0.5074458565945166,
+      "learning_rate": 1.3889040484145206e-05,
+      "loss": 0.349,
+      "step": 7577
+    },
+    {
+      "epoch": 1.4938880126182965,
+      "grad_norm": 0.4776291728277284,
+      "learning_rate": 1.3887612432188282e-05,
+      "loss": 0.3391,
+      "step": 7578
+    },
+    {
+      "epoch": 1.4940851735015772,
+      "grad_norm": 0.47517474086457523,
+      "learning_rate": 1.388618428682834e-05,
+      "loss": 0.3163,
+      "step": 7579
+    },
+    {
+      "epoch": 1.4942823343848581,
+      "grad_norm": 0.4757824327904514,
+      "learning_rate": 1.3884756048099688e-05,
+      "loss": 0.3221,
+      "step": 7580
+    },
+    {
+      "epoch": 1.4944794952681388,
+      "grad_norm": 0.5268640892228607,
+      "learning_rate": 1.3883327716036643e-05,
+      "loss": 0.3628,
+      "step": 7581
+    },
+    {
+      "epoch": 1.4946766561514195,
+      "grad_norm": 0.4902773690445549,
+      "learning_rate": 1.3881899290673526e-05,
+      "loss": 0.3641,
+      "step": 7582
+    },
+    {
+      "epoch": 1.4948738170347002,
+      "grad_norm": 0.4598374139054406,
+      "learning_rate": 1.388047077204465e-05,
+      "loss": 0.3223,
+      "step": 7583
+    },
+    {
+      "epoch": 1.495070977917981,
+      "grad_norm": 0.4847424035881873,
+      "learning_rate": 1.3879042160184337e-05,
+      "loss": 0.3355,
+      "step": 7584
+    },
+    {
+      "epoch": 1.4952681388012619,
+      "grad_norm": 0.5130377501847655,
+      "learning_rate": 1.3877613455126918e-05,
+      "loss": 0.3306,
+      "step": 7585
+    },
+    {
+      "epoch": 1.4954652996845426,
+      "grad_norm": 0.4886195265619356,
+      "learning_rate": 1.3876184656906706e-05,
+      "loss": 0.3358,
+      "step": 7586
+    },
+    {
+      "epoch": 1.4956624605678233,
+      "grad_norm": 0.48902748634085175,
+      "learning_rate": 1.387475576555804e-05,
+      "loss": 0.339,
+      "step": 7587
+    },
+    {
+      "epoch": 1.4958596214511042,
+      "grad_norm": 0.47601101485961045,
+      "learning_rate": 1.3873326781115247e-05,
+      "loss": 0.3206,
+      "step": 7588
+    },
+    {
+      "epoch": 1.4960567823343849,
+      "grad_norm": 0.4942695337736247,
+      "learning_rate": 1.3871897703612658e-05,
+      "loss": 0.3439,
+      "step": 7589
+    },
+    {
+      "epoch": 1.4962539432176656,
+      "grad_norm": 0.4695829997055428,
+      "learning_rate": 1.3870468533084606e-05,
+      "loss": 0.3288,
+      "step": 7590
+    },
+    {
+      "epoch": 1.4964511041009465,
+      "grad_norm": 0.5104474370022767,
+      "learning_rate": 1.3869039269565434e-05,
+      "loss": 0.3388,
+      "step": 7591
+    },
+    {
+      "epoch": 1.4966482649842272,
+      "grad_norm": 0.508559710861152,
+      "learning_rate": 1.3867609913089476e-05,
+      "loss": 0.3353,
+      "step": 7592
+    },
+    {
+      "epoch": 1.496845425867508,
+      "grad_norm": 0.5039315719219875,
+      "learning_rate": 1.3866180463691077e-05,
+      "loss": 0.3208,
+      "step": 7593
+    },
+    {
+      "epoch": 1.4970425867507886,
+      "grad_norm": 0.5191040517201374,
+      "learning_rate": 1.386475092140458e-05,
+      "loss": 0.3468,
+      "step": 7594
+    },
+    {
+      "epoch": 1.4972397476340693,
+      "grad_norm": 0.4731795958829586,
+      "learning_rate": 1.3863321286264326e-05,
+      "loss": 0.2784,
+      "step": 7595
+    },
+    {
+      "epoch": 1.4974369085173502,
+      "grad_norm": 0.5740007403730863,
+      "learning_rate": 1.386189155830467e-05,
+      "loss": 0.3461,
+      "step": 7596
+    },
+    {
+      "epoch": 1.497634069400631,
+      "grad_norm": 0.5117098865886438,
+      "learning_rate": 1.3860461737559958e-05,
+      "loss": 0.3335,
+      "step": 7597
+    },
+    {
+      "epoch": 1.4978312302839116,
+      "grad_norm": 0.4963354358560534,
+      "learning_rate": 1.3859031824064543e-05,
+      "loss": 0.3213,
+      "step": 7598
+    },
+    {
+      "epoch": 1.4980283911671926,
+      "grad_norm": 0.5362349288534428,
+      "learning_rate": 1.3857601817852785e-05,
+      "loss": 0.3503,
+      "step": 7599
+    },
+    {
+      "epoch": 1.4982255520504733,
+      "grad_norm": 0.46854667444421344,
+      "learning_rate": 1.3856171718959033e-05,
+      "loss": 0.3113,
+      "step": 7600
+    },
+    {
+      "epoch": 1.498422712933754,
+      "grad_norm": 0.4976815417371681,
+      "learning_rate": 1.385474152741765e-05,
+      "loss": 0.3528,
+      "step": 7601
+    },
+    {
+      "epoch": 1.4986198738170347,
+      "grad_norm": 0.5129716538025957,
+      "learning_rate": 1.3853311243262999e-05,
+      "loss": 0.3332,
+      "step": 7602
+    },
+    {
+      "epoch": 1.4988170347003154,
+      "grad_norm": 0.4604531892193082,
+      "learning_rate": 1.3851880866529444e-05,
+      "loss": 0.3217,
+      "step": 7603
+    },
+    {
+      "epoch": 1.4990141955835963,
+      "grad_norm": 0.6867111001620063,
+      "learning_rate": 1.3850450397251344e-05,
+      "loss": 0.3352,
+      "step": 7604
+    },
+    {
+      "epoch": 1.499211356466877,
+      "grad_norm": 0.49556889703029594,
+      "learning_rate": 1.3849019835463076e-05,
+      "loss": 0.3441,
+      "step": 7605
+    },
+    {
+      "epoch": 1.4994085173501577,
+      "grad_norm": 0.6455148770924305,
+      "learning_rate": 1.3847589181199009e-05,
+      "loss": 0.3688,
+      "step": 7606
+    },
+    {
+      "epoch": 1.4996056782334386,
+      "grad_norm": 0.49503800910765083,
+      "learning_rate": 1.3846158434493507e-05,
+      "loss": 0.3615,
+      "step": 7607
+    },
+    {
+      "epoch": 1.4998028391167193,
+      "grad_norm": 0.5347889365402132,
+      "learning_rate": 1.3844727595380958e-05,
+      "loss": 0.3554,
+      "step": 7608
+    },
+    {
+      "epoch": 1.4998028391167193,
+      "eval_loss": 0.43272438645362854,
+      "eval_runtime": 344.4401,
+      "eval_samples_per_second": 23.604,
+      "eval_steps_per_second": 1.478,
+      "step": 7608
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.409221739880106,
+      "learning_rate": 1.3843296663895726e-05,
+      "loss": 0.421,
+      "step": 7609
+    },
+    {
+      "epoch": 1.500197160883281,
+      "grad_norm": 0.5293429847721727,
+      "learning_rate": 1.3841865640072203e-05,
+      "loss": 0.3249,
+      "step": 7610
+    },
+    {
+      "epoch": 1.5003943217665614,
+      "grad_norm": 0.5879361449704711,
+      "learning_rate": 1.3840434523944759e-05,
+      "loss": 0.3574,
+      "step": 7611
+    },
+    {
+      "epoch": 1.5005914826498423,
+      "grad_norm": 0.507948237489469,
+      "learning_rate": 1.3839003315547785e-05,
+      "loss": 0.3477,
+      "step": 7612
+    },
+    {
+      "epoch": 1.500788643533123,
+      "grad_norm": 0.5174406330344764,
+      "learning_rate": 1.3837572014915669e-05,
+      "loss": 0.3464,
+      "step": 7613
+    },
+    {
+      "epoch": 1.5009858044164037,
+      "grad_norm": 0.507543096114278,
+      "learning_rate": 1.3836140622082788e-05,
+      "loss": 0.3351,
+      "step": 7614
+    },
+    {
+      "epoch": 1.5011829652996846,
+      "grad_norm": 0.5397406725670891,
+      "learning_rate": 1.3834709137083544e-05,
+      "loss": 0.3506,
+      "step": 7615
+    },
+    {
+      "epoch": 1.5013801261829653,
+      "grad_norm": 0.5232479139909186,
+      "learning_rate": 1.3833277559952323e-05,
+      "loss": 0.3521,
+      "step": 7616
+    },
+    {
+      "epoch": 1.501577287066246,
+      "grad_norm": 0.4913998494274104,
+      "learning_rate": 1.3831845890723523e-05,
+      "loss": 0.351,
+      "step": 7617
+    },
+    {
+      "epoch": 1.501774447949527,
+      "grad_norm": 0.5928180766843834,
+      "learning_rate": 1.3830414129431538e-05,
+      "loss": 0.3554,
+      "step": 7618
+    },
+    {
+      "epoch": 1.5019716088328074,
+      "grad_norm": 0.47768284134792766,
+      "learning_rate": 1.3828982276110767e-05,
+      "loss": 0.323,
+      "step": 7619
+    },
+    {
+      "epoch": 1.5021687697160884,
+      "grad_norm": 0.4784865791461987,
+      "learning_rate": 1.3827550330795618e-05,
+      "loss": 0.2985,
+      "step": 7620
+    },
+    {
+      "epoch": 1.502365930599369,
+      "grad_norm": 0.5109879323464163,
+      "learning_rate": 1.3826118293520488e-05,
+      "loss": 0.3387,
+      "step": 7621
+    },
+    {
+      "epoch": 1.5025630914826498,
+      "grad_norm": 0.4942414542354045,
+      "learning_rate": 1.3824686164319782e-05,
+      "loss": 0.3306,
+      "step": 7622
+    },
+    {
+      "epoch": 1.5027602523659307,
+      "grad_norm": 0.5035209979496461,
+      "learning_rate": 1.3823253943227916e-05,
+      "loss": 0.3313,
+      "step": 7623
+    },
+    {
+      "epoch": 1.5029574132492114,
+      "grad_norm": 0.5438482479157711,
+      "learning_rate": 1.382182163027929e-05,
+      "loss": 0.3251,
+      "step": 7624
+    },
+    {
+      "epoch": 1.503154574132492,
+      "grad_norm": 0.48477404769438337,
+      "learning_rate": 1.3820389225508327e-05,
+      "loss": 0.3358,
+      "step": 7625
+    },
+    {
+      "epoch": 1.503351735015773,
+      "grad_norm": 0.49598419627564255,
+      "learning_rate": 1.3818956728949432e-05,
+      "loss": 0.3176,
+      "step": 7626
+    },
+    {
+      "epoch": 1.5035488958990535,
+      "grad_norm": 0.6022310575721551,
+      "learning_rate": 1.3817524140637029e-05,
+      "loss": 0.3363,
+      "step": 7627
+    },
+    {
+      "epoch": 1.5037460567823344,
+      "grad_norm": 0.4892461050892043,
+      "learning_rate": 1.3816091460605534e-05,
+      "loss": 0.3255,
+      "step": 7628
+    },
+    {
+      "epoch": 1.5039432176656151,
+      "grad_norm": 0.552459885881935,
+      "learning_rate": 1.3814658688889369e-05,
+      "loss": 0.3272,
+      "step": 7629
+    },
+    {
+      "epoch": 1.5041403785488958,
+      "grad_norm": 0.5011577635235227,
+      "learning_rate": 1.3813225825522954e-05,
+      "loss": 0.3263,
+      "step": 7630
+    },
+    {
+      "epoch": 1.5043375394321767,
+      "grad_norm": 0.5071339854073511,
+      "learning_rate": 1.3811792870540717e-05,
+      "loss": 0.3306,
+      "step": 7631
+    },
+    {
+      "epoch": 1.5045347003154574,
+      "grad_norm": 0.4690112256108807,
+      "learning_rate": 1.3810359823977094e-05,
+      "loss": 0.3142,
+      "step": 7632
+    },
+    {
+      "epoch": 1.5047318611987381,
+      "grad_norm": 0.507389006042168,
+      "learning_rate": 1.38089266858665e-05,
+      "loss": 0.3595,
+      "step": 7633
+    },
+    {
+      "epoch": 1.504929022082019,
+      "grad_norm": 0.5136892463140303,
+      "learning_rate": 1.380749345624338e-05,
+      "loss": 0.3552,
+      "step": 7634
+    },
+    {
+      "epoch": 1.5051261829652995,
+      "grad_norm": 0.5230467446626003,
+      "learning_rate": 1.3806060135142159e-05,
+      "loss": 0.3361,
+      "step": 7635
+    },
+    {
+      "epoch": 1.5053233438485805,
+      "grad_norm": 0.47365615277402534,
+      "learning_rate": 1.3804626722597283e-05,
+      "loss": 0.3181,
+      "step": 7636
+    },
+    {
+      "epoch": 1.5055205047318612,
+      "grad_norm": 0.5263364289856198,
+      "learning_rate": 1.3803193218643181e-05,
+      "loss": 0.3324,
+      "step": 7637
+    },
+    {
+      "epoch": 1.5057176656151419,
+      "grad_norm": 0.5134844859050957,
+      "learning_rate": 1.3801759623314302e-05,
+      "loss": 0.3429,
+      "step": 7638
+    },
+    {
+      "epoch": 1.5059148264984228,
+      "grad_norm": 0.4801245661280436,
+      "learning_rate": 1.3800325936645087e-05,
+      "loss": 0.2968,
+      "step": 7639
+    },
+    {
+      "epoch": 1.5061119873817035,
+      "grad_norm": 0.4700963225645677,
+      "learning_rate": 1.379889215866998e-05,
+      "loss": 0.3203,
+      "step": 7640
+    },
+    {
+      "epoch": 1.5063091482649842,
+      "grad_norm": 0.4710374994497814,
+      "learning_rate": 1.3797458289423431e-05,
+      "loss": 0.3241,
+      "step": 7641
+    },
+    {
+      "epoch": 1.506506309148265,
+      "grad_norm": 0.4957756081340281,
+      "learning_rate": 1.3796024328939887e-05,
+      "loss": 0.3512,
+      "step": 7642
+    },
+    {
+      "epoch": 1.5067034700315456,
+      "grad_norm": 0.4761451630851939,
+      "learning_rate": 1.3794590277253803e-05,
+      "loss": 0.334,
+      "step": 7643
+    },
+    {
+      "epoch": 1.5069006309148265,
+      "grad_norm": 0.4814926896127318,
+      "learning_rate": 1.3793156134399633e-05,
+      "loss": 0.3246,
+      "step": 7644
+    },
+    {
+      "epoch": 1.5070977917981072,
+      "grad_norm": 0.4828137836425216,
+      "learning_rate": 1.379172190041183e-05,
+      "loss": 0.3414,
+      "step": 7645
+    },
+    {
+      "epoch": 1.507294952681388,
+      "grad_norm": 0.5194210689710018,
+      "learning_rate": 1.3790287575324854e-05,
+      "loss": 0.3533,
+      "step": 7646
+    },
+    {
+      "epoch": 1.5074921135646688,
+      "grad_norm": 0.4962183001455974,
+      "learning_rate": 1.3788853159173169e-05,
+      "loss": 0.3336,
+      "step": 7647
+    },
+    {
+      "epoch": 1.5076892744479495,
+      "grad_norm": 0.4950050985316144,
+      "learning_rate": 1.3787418651991233e-05,
+      "loss": 0.3565,
+      "step": 7648
+    },
+    {
+      "epoch": 1.5078864353312302,
+      "grad_norm": 0.49021770978943213,
+      "learning_rate": 1.3785984053813517e-05,
+      "loss": 0.3356,
+      "step": 7649
+    },
+    {
+      "epoch": 1.5080835962145112,
+      "grad_norm": 0.47294234808211555,
+      "learning_rate": 1.3784549364674485e-05,
+      "loss": 0.3198,
+      "step": 7650
+    },
+    {
+      "epoch": 1.5082807570977916,
+      "grad_norm": 0.5189806281206693,
+      "learning_rate": 1.3783114584608605e-05,
+      "loss": 0.3294,
+      "step": 7651
+    },
+    {
+      "epoch": 1.5084779179810726,
+      "grad_norm": 0.5185822245810467,
+      "learning_rate": 1.3781679713650349e-05,
+      "loss": 0.3446,
+      "step": 7652
+    },
+    {
+      "epoch": 1.5086750788643533,
+      "grad_norm": 0.4825839494350987,
+      "learning_rate": 1.3780244751834197e-05,
+      "loss": 0.3422,
+      "step": 7653
+    },
+    {
+      "epoch": 1.508872239747634,
+      "grad_norm": 0.4796252236159735,
+      "learning_rate": 1.3778809699194616e-05,
+      "loss": 0.3431,
+      "step": 7654
+    },
+    {
+      "epoch": 1.5090694006309149,
+      "grad_norm": 0.5021963528108637,
+      "learning_rate": 1.3777374555766093e-05,
+      "loss": 0.3401,
+      "step": 7655
+    },
+    {
+      "epoch": 1.5092665615141956,
+      "grad_norm": 0.5751823831914286,
+      "learning_rate": 1.37759393215831e-05,
+      "loss": 0.3532,
+      "step": 7656
+    },
+    {
+      "epoch": 1.5094637223974763,
+      "grad_norm": 0.4777026387116764,
+      "learning_rate": 1.3774503996680128e-05,
+      "loss": 0.3461,
+      "step": 7657
+    },
+    {
+      "epoch": 1.5096608832807572,
+      "grad_norm": 0.47046319672734455,
+      "learning_rate": 1.3773068581091655e-05,
+      "loss": 0.3252,
+      "step": 7658
+    },
+    {
+      "epoch": 1.509858044164038,
+      "grad_norm": 0.5047250594337946,
+      "learning_rate": 1.3771633074852173e-05,
+      "loss": 0.3539,
+      "step": 7659
+    },
+    {
+      "epoch": 1.5100552050473186,
+      "grad_norm": 0.5141111962164151,
+      "learning_rate": 1.3770197477996168e-05,
+      "loss": 0.3272,
+      "step": 7660
+    },
+    {
+      "epoch": 1.5102523659305995,
+      "grad_norm": 0.4811169395000833,
+      "learning_rate": 1.3768761790558134e-05,
+      "loss": 0.3245,
+      "step": 7661
+    },
+    {
+      "epoch": 1.51044952681388,
+      "grad_norm": 0.5244771681249779,
+      "learning_rate": 1.3767326012572561e-05,
+      "loss": 0.3566,
+      "step": 7662
+    },
+    {
+      "epoch": 1.510646687697161,
+      "grad_norm": 0.4704331253968268,
+      "learning_rate": 1.376589014407395e-05,
+      "loss": 0.3154,
+      "step": 7663
+    },
+    {
+      "epoch": 1.5108438485804416,
+      "grad_norm": 0.49390026411452725,
+      "learning_rate": 1.3764454185096792e-05,
+      "loss": 0.3377,
+      "step": 7664
+    },
+    {
+      "epoch": 1.5110410094637223,
+      "grad_norm": 0.45874452901953416,
+      "learning_rate": 1.3763018135675592e-05,
+      "loss": 0.3388,
+      "step": 7665
+    },
+    {
+      "epoch": 1.5112381703470033,
+      "grad_norm": 0.4768940389299864,
+      "learning_rate": 1.3761581995844852e-05,
+      "loss": 0.3355,
+      "step": 7666
+    },
+    {
+      "epoch": 1.511435331230284,
+      "grad_norm": 0.511936222268746,
+      "learning_rate": 1.3760145765639075e-05,
+      "loss": 0.3519,
+      "step": 7667
+    },
+    {
+      "epoch": 1.5116324921135647,
+      "grad_norm": 0.7261444629938607,
+      "learning_rate": 1.3758709445092767e-05,
+      "loss": 0.3503,
+      "step": 7668
+    },
+    {
+      "epoch": 1.5118296529968456,
+      "grad_norm": 0.47879946943440466,
+      "learning_rate": 1.3757273034240437e-05,
+      "loss": 0.3403,
+      "step": 7669
+    },
+    {
+      "epoch": 1.512026813880126,
+      "grad_norm": 0.5050574898314102,
+      "learning_rate": 1.3755836533116597e-05,
+      "loss": 0.3364,
+      "step": 7670
+    },
+    {
+      "epoch": 1.512223974763407,
+      "grad_norm": 0.49140249592667906,
+      "learning_rate": 1.3754399941755763e-05,
+      "loss": 0.3237,
+      "step": 7671
+    },
+    {
+      "epoch": 1.5124211356466877,
+      "grad_norm": 0.5364739642991352,
+      "learning_rate": 1.3752963260192442e-05,
+      "loss": 0.3594,
+      "step": 7672
+    },
+    {
+      "epoch": 1.5126182965299684,
+      "grad_norm": 0.5184206685342432,
+      "learning_rate": 1.3751526488461158e-05,
+      "loss": 0.3346,
+      "step": 7673
+    },
+    {
+      "epoch": 1.5128154574132493,
+      "grad_norm": 0.5054659832613505,
+      "learning_rate": 1.375008962659643e-05,
+      "loss": 0.3309,
+      "step": 7674
+    },
+    {
+      "epoch": 1.51301261829653,
+      "grad_norm": 0.5555312582768909,
+      "learning_rate": 1.3748652674632779e-05,
+      "loss": 0.359,
+      "step": 7675
+    },
+    {
+      "epoch": 1.5132097791798107,
+      "grad_norm": 0.5258696149826486,
+      "learning_rate": 1.374721563260473e-05,
+      "loss": 0.3485,
+      "step": 7676
+    },
+    {
+      "epoch": 1.5134069400630916,
+      "grad_norm": 7.007449812708166,
+      "learning_rate": 1.3745778500546805e-05,
+      "loss": 0.326,
+      "step": 7677
+    },
+    {
+      "epoch": 1.513604100946372,
+      "grad_norm": 0.5503065900084199,
+      "learning_rate": 1.3744341278493535e-05,
+      "loss": 0.333,
+      "step": 7678
+    },
+    {
+      "epoch": 1.513801261829653,
+      "grad_norm": 0.46540872566930663,
+      "learning_rate": 1.374290396647945e-05,
+      "loss": 0.308,
+      "step": 7679
+    },
+    {
+      "epoch": 1.5139984227129337,
+      "grad_norm": 0.48222335793406707,
+      "learning_rate": 1.3741466564539085e-05,
+      "loss": 0.3374,
+      "step": 7680
+    },
+    {
+      "epoch": 1.5141955835962144,
+      "grad_norm": 0.5340584301274428,
+      "learning_rate": 1.3740029072706975e-05,
+      "loss": 0.3339,
+      "step": 7681
+    },
+    {
+      "epoch": 1.5143927444794953,
+      "grad_norm": 0.4798870462171238,
+      "learning_rate": 1.373859149101765e-05,
+      "loss": 0.327,
+      "step": 7682
+    },
+    {
+      "epoch": 1.514589905362776,
+      "grad_norm": 0.5198175974996748,
+      "learning_rate": 1.3737153819505658e-05,
+      "loss": 0.3557,
+      "step": 7683
+    },
+    {
+      "epoch": 1.5147870662460567,
+      "grad_norm": 0.5400239741407015,
+      "learning_rate": 1.3735716058205533e-05,
+      "loss": 0.3415,
+      "step": 7684
+    },
+    {
+      "epoch": 1.5149842271293377,
+      "grad_norm": 0.740156822222281,
+      "learning_rate": 1.3734278207151824e-05,
+      "loss": 0.353,
+      "step": 7685
+    },
+    {
+      "epoch": 1.5151813880126181,
+      "grad_norm": 0.5022300864303025,
+      "learning_rate": 1.3732840266379071e-05,
+      "loss": 0.3161,
+      "step": 7686
+    },
+    {
+      "epoch": 1.515378548895899,
+      "grad_norm": 0.4932017068142421,
+      "learning_rate": 1.3731402235921824e-05,
+      "loss": 0.3324,
+      "step": 7687
+    },
+    {
+      "epoch": 1.5155757097791798,
+      "grad_norm": 0.5065348898290931,
+      "learning_rate": 1.3729964115814636e-05,
+      "loss": 0.3333,
+      "step": 7688
+    },
+    {
+      "epoch": 1.5157728706624605,
+      "grad_norm": 0.5417643838253485,
+      "learning_rate": 1.3728525906092056e-05,
+      "loss": 0.3521,
+      "step": 7689
+    },
+    {
+      "epoch": 1.5159700315457414,
+      "grad_norm": 0.4907116291133275,
+      "learning_rate": 1.3727087606788639e-05,
+      "loss": 0.3293,
+      "step": 7690
+    },
+    {
+      "epoch": 1.516167192429022,
+      "grad_norm": 0.47069234491091616,
+      "learning_rate": 1.3725649217938938e-05,
+      "loss": 0.328,
+      "step": 7691
+    },
+    {
+      "epoch": 1.5163643533123028,
+      "grad_norm": 0.6967452358004222,
+      "learning_rate": 1.3724210739577516e-05,
+      "loss": 0.3358,
+      "step": 7692
+    },
+    {
+      "epoch": 1.5165615141955837,
+      "grad_norm": 0.5071542817692918,
+      "learning_rate": 1.3722772171738932e-05,
+      "loss": 0.3593,
+      "step": 7693
+    },
+    {
+      "epoch": 1.5167586750788642,
+      "grad_norm": 0.48669388997244545,
+      "learning_rate": 1.3721333514457748e-05,
+      "loss": 0.3457,
+      "step": 7694
+    },
+    {
+      "epoch": 1.5169558359621451,
+      "grad_norm": 0.49829646210372097,
+      "learning_rate": 1.3719894767768532e-05,
+      "loss": 0.3418,
+      "step": 7695
+    },
+    {
+      "epoch": 1.5171529968454258,
+      "grad_norm": 0.5104164374213708,
+      "learning_rate": 1.3718455931705845e-05,
+      "loss": 0.3544,
+      "step": 7696
+    },
+    {
+      "epoch": 1.5173501577287065,
+      "grad_norm": 0.5017546986684652,
+      "learning_rate": 1.371701700630426e-05,
+      "loss": 0.355,
+      "step": 7697
+    },
+    {
+      "epoch": 1.5175473186119874,
+      "grad_norm": 0.5926037563656988,
+      "learning_rate": 1.3715577991598352e-05,
+      "loss": 0.3789,
+      "step": 7698
+    },
+    {
+      "epoch": 1.5177444794952681,
+      "grad_norm": 0.4963131869513678,
+      "learning_rate": 1.3714138887622685e-05,
+      "loss": 0.3448,
+      "step": 7699
+    },
+    {
+      "epoch": 1.5179416403785488,
+      "grad_norm": 0.5147685242406805,
+      "learning_rate": 1.3712699694411846e-05,
+      "loss": 0.3254,
+      "step": 7700
+    },
+    {
+      "epoch": 1.5181388012618298,
+      "grad_norm": 0.5075260322768886,
+      "learning_rate": 1.3711260412000403e-05,
+      "loss": 0.3396,
+      "step": 7701
+    },
+    {
+      "epoch": 1.5183359621451105,
+      "grad_norm": 0.4938057819218849,
+      "learning_rate": 1.3709821040422944e-05,
+      "loss": 0.3434,
+      "step": 7702
+    },
+    {
+      "epoch": 1.5185331230283912,
+      "grad_norm": 0.48016800863389886,
+      "learning_rate": 1.370838157971404e-05,
+      "loss": 0.3161,
+      "step": 7703
+    },
+    {
+      "epoch": 1.518730283911672,
+      "grad_norm": 0.4947174895321279,
+      "learning_rate": 1.370694202990829e-05,
+      "loss": 0.3636,
+      "step": 7704
+    },
+    {
+      "epoch": 1.5189274447949526,
+      "grad_norm": 0.4617872910169582,
+      "learning_rate": 1.3705502391040266e-05,
+      "loss": 0.3193,
+      "step": 7705
+    },
+    {
+      "epoch": 1.5191246056782335,
+      "grad_norm": 0.49049250891250884,
+      "learning_rate": 1.3704062663144569e-05,
+      "loss": 0.3336,
+      "step": 7706
+    },
+    {
+      "epoch": 1.5193217665615142,
+      "grad_norm": 0.4457461204930738,
+      "learning_rate": 1.370262284625578e-05,
+      "loss": 0.2909,
+      "step": 7707
+    },
+    {
+      "epoch": 1.5195189274447949,
+      "grad_norm": 0.5403694722432415,
+      "learning_rate": 1.3701182940408495e-05,
+      "loss": 0.3567,
+      "step": 7708
+    },
+    {
+      "epoch": 1.5197160883280758,
+      "grad_norm": 0.4816431559603944,
+      "learning_rate": 1.3699742945637312e-05,
+      "loss": 0.3163,
+      "step": 7709
+    },
+    {
+      "epoch": 1.5199132492113565,
+      "grad_norm": 0.5071851732106253,
+      "learning_rate": 1.3698302861976822e-05,
+      "loss": 0.3197,
+      "step": 7710
+    },
+    {
+      "epoch": 1.5201104100946372,
+      "grad_norm": 0.5036103402901476,
+      "learning_rate": 1.369686268946163e-05,
+      "loss": 0.3594,
+      "step": 7711
+    },
+    {
+      "epoch": 1.5203075709779181,
+      "grad_norm": 0.5082597299605005,
+      "learning_rate": 1.3695422428126335e-05,
+      "loss": 0.3452,
+      "step": 7712
+    },
+    {
+      "epoch": 1.5205047318611986,
+      "grad_norm": 0.5034774870521724,
+      "learning_rate": 1.3693982078005538e-05,
+      "loss": 0.3415,
+      "step": 7713
+    },
+    {
+      "epoch": 1.5207018927444795,
+      "grad_norm": 0.49017821736880307,
+      "learning_rate": 1.3692541639133849e-05,
+      "loss": 0.3503,
+      "step": 7714
+    },
+    {
+      "epoch": 1.5208990536277602,
+      "grad_norm": 0.5033328297669176,
+      "learning_rate": 1.3691101111545873e-05,
+      "loss": 0.3306,
+      "step": 7715
+    },
+    {
+      "epoch": 1.521096214511041,
+      "grad_norm": 0.5089744822677265,
+      "learning_rate": 1.368966049527622e-05,
+      "loss": 0.3579,
+      "step": 7716
+    },
+    {
+      "epoch": 1.5212933753943219,
+      "grad_norm": 0.5390487095974562,
+      "learning_rate": 1.3688219790359503e-05,
+      "loss": 0.369,
+      "step": 7717
+    },
+    {
+      "epoch": 1.5214905362776026,
+      "grad_norm": 0.4996281857585761,
+      "learning_rate": 1.3686778996830335e-05,
+      "loss": 0.317,
+      "step": 7718
+    },
+    {
+      "epoch": 1.5216876971608833,
+      "grad_norm": 0.5360343337780741,
+      "learning_rate": 1.3685338114723331e-05,
+      "loss": 0.3536,
+      "step": 7719
+    },
+    {
+      "epoch": 1.5218848580441642,
+      "grad_norm": 0.48695308863803854,
+      "learning_rate": 1.3683897144073111e-05,
+      "loss": 0.3319,
+      "step": 7720
+    },
+    {
+      "epoch": 1.5220820189274447,
+      "grad_norm": 0.5111947489161915,
+      "learning_rate": 1.36824560849143e-05,
+      "loss": 0.3506,
+      "step": 7721
+    },
+    {
+      "epoch": 1.5222791798107256,
+      "grad_norm": 0.515635778087092,
+      "learning_rate": 1.3681014937281509e-05,
+      "loss": 0.2918,
+      "step": 7722
+    },
+    {
+      "epoch": 1.5224763406940063,
+      "grad_norm": 0.7712050067029159,
+      "learning_rate": 1.3679573701209376e-05,
+      "loss": 0.3375,
+      "step": 7723
+    },
+    {
+      "epoch": 1.522673501577287,
+      "grad_norm": 0.5100802873571525,
+      "learning_rate": 1.3678132376732518e-05,
+      "loss": 0.3686,
+      "step": 7724
+    },
+    {
+      "epoch": 1.522870662460568,
+      "grad_norm": 0.4995731991498597,
+      "learning_rate": 1.367669096388557e-05,
+      "loss": 0.3459,
+      "step": 7725
+    },
+    {
+      "epoch": 1.5230678233438486,
+      "grad_norm": 0.49021205630646913,
+      "learning_rate": 1.367524946270316e-05,
+      "loss": 0.3568,
+      "step": 7726
+    },
+    {
+      "epoch": 1.5232649842271293,
+      "grad_norm": 0.5069952766234989,
+      "learning_rate": 1.3673807873219921e-05,
+      "loss": 0.3652,
+      "step": 7727
+    },
+    {
+      "epoch": 1.5234621451104102,
+      "grad_norm": 0.5044160352391255,
+      "learning_rate": 1.367236619547049e-05,
+      "loss": 0.3359,
+      "step": 7728
+    },
+    {
+      "epoch": 1.5236593059936907,
+      "grad_norm": 0.5003901713256061,
+      "learning_rate": 1.3670924429489505e-05,
+      "loss": 0.345,
+      "step": 7729
+    },
+    {
+      "epoch": 1.5238564668769716,
+      "grad_norm": 0.4700435713643861,
+      "learning_rate": 1.3669482575311604e-05,
+      "loss": 0.3138,
+      "step": 7730
+    },
+    {
+      "epoch": 1.5240536277602523,
+      "grad_norm": 0.489703666578115,
+      "learning_rate": 1.366804063297143e-05,
+      "loss": 0.3327,
+      "step": 7731
+    },
+    {
+      "epoch": 1.524250788643533,
+      "grad_norm": 0.5862569684725426,
+      "learning_rate": 1.3666598602503622e-05,
+      "loss": 0.3267,
+      "step": 7732
+    },
+    {
+      "epoch": 1.524447949526814,
+      "grad_norm": 0.4842290548690208,
+      "learning_rate": 1.3665156483942834e-05,
+      "loss": 0.3538,
+      "step": 7733
+    },
+    {
+      "epoch": 1.5246451104100947,
+      "grad_norm": 0.5540354619028578,
+      "learning_rate": 1.3663714277323707e-05,
+      "loss": 0.373,
+      "step": 7734
+    },
+    {
+      "epoch": 1.5248422712933754,
+      "grad_norm": 0.49143503772872,
+      "learning_rate": 1.3662271982680895e-05,
+      "loss": 0.3312,
+      "step": 7735
+    },
+    {
+      "epoch": 1.5250394321766563,
+      "grad_norm": 0.52098501675389,
+      "learning_rate": 1.366082960004905e-05,
+      "loss": 0.3509,
+      "step": 7736
+    },
+    {
+      "epoch": 1.5252365930599368,
+      "grad_norm": 0.4553592686652839,
+      "learning_rate": 1.3659387129462826e-05,
+      "loss": 0.3139,
+      "step": 7737
+    },
+    {
+      "epoch": 1.5254337539432177,
+      "grad_norm": 0.519004995842705,
+      "learning_rate": 1.365794457095688e-05,
+      "loss": 0.3671,
+      "step": 7738
+    },
+    {
+      "epoch": 1.5256309148264984,
+      "grad_norm": 0.5279828914728227,
+      "learning_rate": 1.3656501924565867e-05,
+      "loss": 0.3654,
+      "step": 7739
+    },
+    {
+      "epoch": 1.525828075709779,
+      "grad_norm": 0.47358808531757013,
+      "learning_rate": 1.3655059190324453e-05,
+      "loss": 0.2977,
+      "step": 7740
+    },
+    {
+      "epoch": 1.52602523659306,
+      "grad_norm": 0.5047664002702137,
+      "learning_rate": 1.3653616368267297e-05,
+      "loss": 0.3231,
+      "step": 7741
+    },
+    {
+      "epoch": 1.5262223974763407,
+      "grad_norm": 0.49780512154971746,
+      "learning_rate": 1.3652173458429068e-05,
+      "loss": 0.3339,
+      "step": 7742
+    },
+    {
+      "epoch": 1.5264195583596214,
+      "grad_norm": 0.49842257988288025,
+      "learning_rate": 1.3650730460844428e-05,
+      "loss": 0.3231,
+      "step": 7743
+    },
+    {
+      "epoch": 1.5266167192429023,
+      "grad_norm": 0.5112605303580525,
+      "learning_rate": 1.3649287375548052e-05,
+      "loss": 0.3416,
+      "step": 7744
+    },
+    {
+      "epoch": 1.526813880126183,
+      "grad_norm": 0.49859797442001647,
+      "learning_rate": 1.3647844202574603e-05,
+      "loss": 0.3594,
+      "step": 7745
+    },
+    {
+      "epoch": 1.5270110410094637,
+      "grad_norm": 0.5055274577950344,
+      "learning_rate": 1.3646400941958766e-05,
+      "loss": 0.3519,
+      "step": 7746
+    },
+    {
+      "epoch": 1.5272082018927446,
+      "grad_norm": 0.5512290570751095,
+      "learning_rate": 1.3644957593735206e-05,
+      "loss": 0.3529,
+      "step": 7747
+    },
+    {
+      "epoch": 1.5274053627760251,
+      "grad_norm": 0.47973030402608713,
+      "learning_rate": 1.3643514157938603e-05,
+      "loss": 0.3028,
+      "step": 7748
+    },
+    {
+      "epoch": 1.527602523659306,
+      "grad_norm": 0.5198514112156248,
+      "learning_rate": 1.364207063460364e-05,
+      "loss": 0.3197,
+      "step": 7749
+    },
+    {
+      "epoch": 1.5277996845425867,
+      "grad_norm": 0.720977186938447,
+      "learning_rate": 1.3640627023764998e-05,
+      "loss": 0.3224,
+      "step": 7750
+    },
+    {
+      "epoch": 1.5279968454258674,
+      "grad_norm": 0.4966494106627751,
+      "learning_rate": 1.363918332545736e-05,
+      "loss": 0.3673,
+      "step": 7751
+    },
+    {
+      "epoch": 1.5281940063091484,
+      "grad_norm": 0.4806489488052334,
+      "learning_rate": 1.363773953971541e-05,
+      "loss": 0.3246,
+      "step": 7752
+    },
+    {
+      "epoch": 1.528391167192429,
+      "grad_norm": 0.5090026005116808,
+      "learning_rate": 1.3636295666573841e-05,
+      "loss": 0.3489,
+      "step": 7753
+    },
+    {
+      "epoch": 1.5285883280757098,
+      "grad_norm": 0.5410349262919377,
+      "learning_rate": 1.3634851706067335e-05,
+      "loss": 0.3381,
+      "step": 7754
+    },
+    {
+      "epoch": 1.5287854889589907,
+      "grad_norm": 0.49058400059922475,
+      "learning_rate": 1.3633407658230596e-05,
+      "loss": 0.3363,
+      "step": 7755
+    },
+    {
+      "epoch": 1.5289826498422712,
+      "grad_norm": 0.45828128853910793,
+      "learning_rate": 1.3631963523098308e-05,
+      "loss": 0.334,
+      "step": 7756
+    },
+    {
+      "epoch": 1.529179810725552,
+      "grad_norm": 0.4813482445211936,
+      "learning_rate": 1.3630519300705171e-05,
+      "loss": 0.3347,
+      "step": 7757
+    },
+    {
+      "epoch": 1.5293769716088328,
+      "grad_norm": 0.484990899943496,
+      "learning_rate": 1.3629074991085886e-05,
+      "loss": 0.3298,
+      "step": 7758
+    },
+    {
+      "epoch": 1.5295741324921135,
+      "grad_norm": 0.49602894149932886,
+      "learning_rate": 1.3627630594275151e-05,
+      "loss": 0.3363,
+      "step": 7759
+    },
+    {
+      "epoch": 1.5297712933753944,
+      "grad_norm": 0.46335979722560355,
+      "learning_rate": 1.3626186110307673e-05,
+      "loss": 0.3197,
+      "step": 7760
+    },
+    {
+      "epoch": 1.5299684542586751,
+      "grad_norm": 0.4907689809200294,
+      "learning_rate": 1.3624741539218151e-05,
+      "loss": 0.3183,
+      "step": 7761
+    },
+    {
+      "epoch": 1.5301656151419558,
+      "grad_norm": 0.48534404933301745,
+      "learning_rate": 1.3623296881041294e-05,
+      "loss": 0.3298,
+      "step": 7762
+    },
+    {
+      "epoch": 1.5303627760252367,
+      "grad_norm": 0.44875294362215107,
+      "learning_rate": 1.3621852135811812e-05,
+      "loss": 0.2968,
+      "step": 7763
+    },
+    {
+      "epoch": 1.5305599369085172,
+      "grad_norm": 0.5797019681078132,
+      "learning_rate": 1.3620407303564416e-05,
+      "loss": 0.353,
+      "step": 7764
+    },
+    {
+      "epoch": 1.5307570977917981,
+      "grad_norm": 0.5028625687192622,
+      "learning_rate": 1.3618962384333818e-05,
+      "loss": 0.319,
+      "step": 7765
+    },
+    {
+      "epoch": 1.5309542586750788,
+      "grad_norm": 0.5099435059525019,
+      "learning_rate": 1.3617517378154737e-05,
+      "loss": 0.3495,
+      "step": 7766
+    },
+    {
+      "epoch": 1.5311514195583595,
+      "grad_norm": 0.502597011605929,
+      "learning_rate": 1.3616072285061886e-05,
+      "loss": 0.3425,
+      "step": 7767
+    },
+    {
+      "epoch": 1.5313485804416405,
+      "grad_norm": 0.48907353787315166,
+      "learning_rate": 1.3614627105089986e-05,
+      "loss": 0.3266,
+      "step": 7768
+    },
+    {
+      "epoch": 1.5315457413249212,
+      "grad_norm": 0.519646837650501,
+      "learning_rate": 1.3613181838273758e-05,
+      "loss": 0.3502,
+      "step": 7769
+    },
+    {
+      "epoch": 1.5317429022082019,
+      "grad_norm": 0.51805181382411,
+      "learning_rate": 1.3611736484647928e-05,
+      "loss": 0.3447,
+      "step": 7770
+    },
+    {
+      "epoch": 1.5319400630914828,
+      "grad_norm": 0.5093499251255805,
+      "learning_rate": 1.3610291044247218e-05,
+      "loss": 0.3474,
+      "step": 7771
+    },
+    {
+      "epoch": 1.5321372239747633,
+      "grad_norm": 0.5281367810739265,
+      "learning_rate": 1.3608845517106364e-05,
+      "loss": 0.3572,
+      "step": 7772
+    },
+    {
+      "epoch": 1.5323343848580442,
+      "grad_norm": 0.4742137497487593,
+      "learning_rate": 1.3607399903260085e-05,
+      "loss": 0.3142,
+      "step": 7773
+    },
+    {
+      "epoch": 1.5325315457413249,
+      "grad_norm": 0.506813178653113,
+      "learning_rate": 1.3605954202743118e-05,
+      "loss": 0.3509,
+      "step": 7774
+    },
+    {
+      "epoch": 1.5327287066246056,
+      "grad_norm": 0.4904880273353688,
+      "learning_rate": 1.36045084155902e-05,
+      "loss": 0.3103,
+      "step": 7775
+    },
+    {
+      "epoch": 1.5329258675078865,
+      "grad_norm": 0.49044062116786236,
+      "learning_rate": 1.3603062541836068e-05,
+      "loss": 0.3517,
+      "step": 7776
+    },
+    {
+      "epoch": 1.5331230283911672,
+      "grad_norm": 0.4634871291754897,
+      "learning_rate": 1.3601616581515451e-05,
+      "loss": 0.328,
+      "step": 7777
+    },
+    {
+      "epoch": 1.533320189274448,
+      "grad_norm": 0.47077009351413396,
+      "learning_rate": 1.3600170534663097e-05,
+      "loss": 0.3169,
+      "step": 7778
+    },
+    {
+      "epoch": 1.5335173501577288,
+      "grad_norm": 0.4899373883327769,
+      "learning_rate": 1.3598724401313748e-05,
+      "loss": 0.3151,
+      "step": 7779
+    },
+    {
+      "epoch": 1.5337145110410093,
+      "grad_norm": 0.519208902268232,
+      "learning_rate": 1.3597278181502146e-05,
+      "loss": 0.3416,
+      "step": 7780
+    },
+    {
+      "epoch": 1.5339116719242902,
+      "grad_norm": 0.504751981494511,
+      "learning_rate": 1.3595831875263038e-05,
+      "loss": 0.354,
+      "step": 7781
+    },
+    {
+      "epoch": 1.534108832807571,
+      "grad_norm": 0.5020543827284227,
+      "learning_rate": 1.3594385482631176e-05,
+      "loss": 0.3447,
+      "step": 7782
+    },
+    {
+      "epoch": 1.5343059936908516,
+      "grad_norm": 0.47429364811655356,
+      "learning_rate": 1.3592939003641308e-05,
+      "loss": 0.3392,
+      "step": 7783
+    },
+    {
+      "epoch": 1.5345031545741326,
+      "grad_norm": 6.77906927343203,
+      "learning_rate": 1.3591492438328185e-05,
+      "loss": 0.3508,
+      "step": 7784
+    },
+    {
+      "epoch": 1.5347003154574133,
+      "grad_norm": 0.5933469489722475,
+      "learning_rate": 1.3590045786726565e-05,
+      "loss": 0.3581,
+      "step": 7785
+    },
+    {
+      "epoch": 1.534897476340694,
+      "grad_norm": 0.4826236402550896,
+      "learning_rate": 1.3588599048871202e-05,
+      "loss": 0.3339,
+      "step": 7786
+    },
+    {
+      "epoch": 1.5350946372239749,
+      "grad_norm": 0.4830828768348468,
+      "learning_rate": 1.358715222479686e-05,
+      "loss": 0.3166,
+      "step": 7787
+    },
+    {
+      "epoch": 1.5352917981072554,
+      "grad_norm": 0.5336248468657963,
+      "learning_rate": 1.3585705314538293e-05,
+      "loss": 0.3423,
+      "step": 7788
+    },
+    {
+      "epoch": 1.5354889589905363,
+      "grad_norm": 0.5287306065268992,
+      "learning_rate": 1.3584258318130274e-05,
+      "loss": 0.3569,
+      "step": 7789
+    },
+    {
+      "epoch": 1.535686119873817,
+      "grad_norm": 0.5579283549576503,
+      "learning_rate": 1.3582811235607559e-05,
+      "loss": 0.3587,
+      "step": 7790
+    },
+    {
+      "epoch": 1.5358832807570977,
+      "grad_norm": 0.5054259962581725,
+      "learning_rate": 1.358136406700492e-05,
+      "loss": 0.3343,
+      "step": 7791
+    },
+    {
+      "epoch": 1.5360804416403786,
+      "grad_norm": 0.49852211546975295,
+      "learning_rate": 1.3579916812357123e-05,
+      "loss": 0.3143,
+      "step": 7792
+    },
+    {
+      "epoch": 1.5362776025236593,
+      "grad_norm": 0.4885451610000915,
+      "learning_rate": 1.3578469471698946e-05,
+      "loss": 0.3347,
+      "step": 7793
+    },
+    {
+      "epoch": 1.53647476340694,
+      "grad_norm": 1.097411098465949,
+      "learning_rate": 1.3577022045065154e-05,
+      "loss": 0.3622,
+      "step": 7794
+    },
+    {
+      "epoch": 1.536671924290221,
+      "grad_norm": 0.48941610871328045,
+      "learning_rate": 1.3575574532490528e-05,
+      "loss": 0.3244,
+      "step": 7795
+    },
+    {
+      "epoch": 1.5368690851735016,
+      "grad_norm": 0.5230416955206131,
+      "learning_rate": 1.3574126934009843e-05,
+      "loss": 0.3488,
+      "step": 7796
+    },
+    {
+      "epoch": 1.5370662460567823,
+      "grad_norm": 0.47356891013201374,
+      "learning_rate": 1.3572679249657883e-05,
+      "loss": 0.3262,
+      "step": 7797
+    },
+    {
+      "epoch": 1.5372634069400632,
+      "grad_norm": 0.5633856591758994,
+      "learning_rate": 1.3571231479469428e-05,
+      "loss": 0.3206,
+      "step": 7798
+    },
+    {
+      "epoch": 1.5374605678233437,
+      "grad_norm": 0.48005492038383385,
+      "learning_rate": 1.3569783623479259e-05,
+      "loss": 0.3251,
+      "step": 7799
+    },
+    {
+      "epoch": 1.5376577287066246,
+      "grad_norm": 0.44407837297296465,
+      "learning_rate": 1.3568335681722165e-05,
+      "loss": 0.3067,
+      "step": 7800
+    },
+    {
+      "epoch": 1.5378548895899053,
+      "grad_norm": 0.49462114423942827,
+      "learning_rate": 1.3566887654232927e-05,
+      "loss": 0.3368,
+      "step": 7801
+    },
+    {
+      "epoch": 1.538052050473186,
+      "grad_norm": 0.5323311830491543,
+      "learning_rate": 1.3565439541046346e-05,
+      "loss": 0.3671,
+      "step": 7802
+    },
+    {
+      "epoch": 1.538249211356467,
+      "grad_norm": 0.4877716449979154,
+      "learning_rate": 1.3563991342197207e-05,
+      "loss": 0.3255,
+      "step": 7803
+    },
+    {
+      "epoch": 1.5384463722397477,
+      "grad_norm": 0.5199206793938796,
+      "learning_rate": 1.3562543057720308e-05,
+      "loss": 0.3473,
+      "step": 7804
+    },
+    {
+      "epoch": 1.5386435331230284,
+      "grad_norm": 0.48858101161536593,
+      "learning_rate": 1.356109468765044e-05,
+      "loss": 0.3512,
+      "step": 7805
+    },
+    {
+      "epoch": 1.5388406940063093,
+      "grad_norm": 0.6302130832925654,
+      "learning_rate": 1.3559646232022408e-05,
+      "loss": 0.3233,
+      "step": 7806
+    },
+    {
+      "epoch": 1.5390378548895898,
+      "grad_norm": 0.5264654250829359,
+      "learning_rate": 1.3558197690871004e-05,
+      "loss": 0.3619,
+      "step": 7807
+    },
+    {
+      "epoch": 1.5392350157728707,
+      "grad_norm": 1.142333806645509,
+      "learning_rate": 1.3556749064231038e-05,
+      "loss": 0.3357,
+      "step": 7808
+    },
+    {
+      "epoch": 1.5394321766561514,
+      "grad_norm": 0.5172634182082058,
+      "learning_rate": 1.3555300352137311e-05,
+      "loss": 0.3492,
+      "step": 7809
+    },
+    {
+      "epoch": 1.539629337539432,
+      "grad_norm": 0.521433933106068,
+      "learning_rate": 1.3553851554624631e-05,
+      "loss": 0.374,
+      "step": 7810
+    },
+    {
+      "epoch": 1.539826498422713,
+      "grad_norm": 0.46871626887951895,
+      "learning_rate": 1.3552402671727805e-05,
+      "loss": 0.325,
+      "step": 7811
+    },
+    {
+      "epoch": 1.5400236593059937,
+      "grad_norm": 0.46267385384546883,
+      "learning_rate": 1.3550953703481645e-05,
+      "loss": 0.307,
+      "step": 7812
+    },
+    {
+      "epoch": 1.5402208201892744,
+      "grad_norm": 0.5082065142815442,
+      "learning_rate": 1.3549504649920961e-05,
+      "loss": 0.3179,
+      "step": 7813
+    },
+    {
+      "epoch": 1.5404179810725553,
+      "grad_norm": 0.4835751507338688,
+      "learning_rate": 1.3548055511080568e-05,
+      "loss": 0.3464,
+      "step": 7814
+    },
+    {
+      "epoch": 1.5406151419558358,
+      "grad_norm": 0.5054295099711004,
+      "learning_rate": 1.3546606286995288e-05,
+      "loss": 0.3411,
+      "step": 7815
+    },
+    {
+      "epoch": 1.5408123028391167,
+      "grad_norm": 0.5124576121953441,
+      "learning_rate": 1.3545156977699931e-05,
+      "loss": 0.3484,
+      "step": 7816
+    },
+    {
+      "epoch": 1.5410094637223974,
+      "grad_norm": 0.5937412421888819,
+      "learning_rate": 1.3543707583229328e-05,
+      "loss": 0.347,
+      "step": 7817
+    },
+    {
+      "epoch": 1.5412066246056781,
+      "grad_norm": 0.47414570200167167,
+      "learning_rate": 1.3542258103618293e-05,
+      "loss": 0.333,
+      "step": 7818
+    },
+    {
+      "epoch": 1.541403785488959,
+      "grad_norm": 0.4826001472020254,
+      "learning_rate": 1.3540808538901658e-05,
+      "loss": 0.3439,
+      "step": 7819
+    },
+    {
+      "epoch": 1.5416009463722398,
+      "grad_norm": 0.46449722652318304,
+      "learning_rate": 1.353935888911424e-05,
+      "loss": 0.3237,
+      "step": 7820
+    },
+    {
+      "epoch": 1.5417981072555205,
+      "grad_norm": 0.48229434160473766,
+      "learning_rate": 1.3537909154290883e-05,
+      "loss": 0.3343,
+      "step": 7821
+    },
+    {
+      "epoch": 1.5419952681388014,
+      "grad_norm": 0.48229421460382765,
+      "learning_rate": 1.3536459334466403e-05,
+      "loss": 0.3375,
+      "step": 7822
+    },
+    {
+      "epoch": 1.5421924290220819,
+      "grad_norm": 0.5116523420598664,
+      "learning_rate": 1.3535009429675641e-05,
+      "loss": 0.3473,
+      "step": 7823
+    },
+    {
+      "epoch": 1.5423895899053628,
+      "grad_norm": 0.5776036825506357,
+      "learning_rate": 1.3533559439953429e-05,
+      "loss": 0.3598,
+      "step": 7824
+    },
+    {
+      "epoch": 1.5425867507886435,
+      "grad_norm": 0.4835560536957904,
+      "learning_rate": 1.3532109365334609e-05,
+      "loss": 0.3359,
+      "step": 7825
+    },
+    {
+      "epoch": 1.5427839116719242,
+      "grad_norm": 0.4967123924164706,
+      "learning_rate": 1.3530659205854018e-05,
+      "loss": 0.3343,
+      "step": 7826
+    },
+    {
+      "epoch": 1.5429810725552051,
+      "grad_norm": 0.5275219762814375,
+      "learning_rate": 1.3529208961546494e-05,
+      "loss": 0.357,
+      "step": 7827
+    },
+    {
+      "epoch": 1.5431782334384858,
+      "grad_norm": 0.5090525455291536,
+      "learning_rate": 1.3527758632446884e-05,
+      "loss": 0.3525,
+      "step": 7828
+    },
+    {
+      "epoch": 1.5433753943217665,
+      "grad_norm": 0.4942938895394185,
+      "learning_rate": 1.3526308218590032e-05,
+      "loss": 0.3466,
+      "step": 7829
+    },
+    {
+      "epoch": 1.5435725552050474,
+      "grad_norm": 0.485652733290798,
+      "learning_rate": 1.3524857720010784e-05,
+      "loss": 0.3549,
+      "step": 7830
+    },
+    {
+      "epoch": 1.543769716088328,
+      "grad_norm": 0.49030437167938445,
+      "learning_rate": 1.3523407136743992e-05,
+      "loss": 0.3342,
+      "step": 7831
+    },
+    {
+      "epoch": 1.5439668769716088,
+      "grad_norm": 0.48804412538773867,
+      "learning_rate": 1.3521956468824505e-05,
+      "loss": 0.3553,
+      "step": 7832
+    },
+    {
+      "epoch": 1.5441640378548895,
+      "grad_norm": 0.48837222737277775,
+      "learning_rate": 1.3520505716287178e-05,
+      "loss": 0.3262,
+      "step": 7833
+    },
+    {
+      "epoch": 1.5443611987381702,
+      "grad_norm": 0.47270681354383004,
+      "learning_rate": 1.3519054879166867e-05,
+      "loss": 0.3444,
+      "step": 7834
+    },
+    {
+      "epoch": 1.5445583596214512,
+      "grad_norm": 0.4926982400357956,
+      "learning_rate": 1.3517603957498426e-05,
+      "loss": 0.3336,
+      "step": 7835
+    },
+    {
+      "epoch": 1.5447555205047319,
+      "grad_norm": 0.49014739600256885,
+      "learning_rate": 1.351615295131672e-05,
+      "loss": 0.3392,
+      "step": 7836
+    },
+    {
+      "epoch": 1.5449526813880126,
+      "grad_norm": 0.4966773713684002,
+      "learning_rate": 1.3514701860656605e-05,
+      "loss": 0.3234,
+      "step": 7837
+    },
+    {
+      "epoch": 1.5451498422712935,
+      "grad_norm": 0.49585084596816736,
+      "learning_rate": 1.351325068555295e-05,
+      "loss": 0.3287,
+      "step": 7838
+    },
+    {
+      "epoch": 1.5453470031545742,
+      "grad_norm": 0.4772798404368484,
+      "learning_rate": 1.3511799426040617e-05,
+      "loss": 0.3295,
+      "step": 7839
+    },
+    {
+      "epoch": 1.5455441640378549,
+      "grad_norm": 0.49778402900715435,
+      "learning_rate": 1.3510348082154476e-05,
+      "loss": 0.3277,
+      "step": 7840
+    },
+    {
+      "epoch": 1.5457413249211358,
+      "grad_norm": 0.4984788297997575,
+      "learning_rate": 1.3508896653929392e-05,
+      "loss": 0.3276,
+      "step": 7841
+    },
+    {
+      "epoch": 1.5459384858044163,
+      "grad_norm": 1.3775125906338332,
+      "learning_rate": 1.3507445141400247e-05,
+      "loss": 0.3507,
+      "step": 7842
+    },
+    {
+      "epoch": 1.5461356466876972,
+      "grad_norm": 0.4962221209215611,
+      "learning_rate": 1.35059935446019e-05,
+      "loss": 0.363,
+      "step": 7843
+    },
+    {
+      "epoch": 1.546332807570978,
+      "grad_norm": 0.5016263606305518,
+      "learning_rate": 1.3504541863569237e-05,
+      "loss": 0.3119,
+      "step": 7844
+    },
+    {
+      "epoch": 1.5465299684542586,
+      "grad_norm": 0.5307167375631622,
+      "learning_rate": 1.3503090098337138e-05,
+      "loss": 0.332,
+      "step": 7845
+    },
+    {
+      "epoch": 1.5467271293375395,
+      "grad_norm": 0.46696657884783277,
+      "learning_rate": 1.3501638248940475e-05,
+      "loss": 0.3343,
+      "step": 7846
+    },
+    {
+      "epoch": 1.5469242902208202,
+      "grad_norm": 0.5225512410564163,
+      "learning_rate": 1.3500186315414133e-05,
+      "loss": 0.3373,
+      "step": 7847
+    },
+    {
+      "epoch": 1.547121451104101,
+      "grad_norm": 0.487864098453998,
+      "learning_rate": 1.3498734297792994e-05,
+      "loss": 0.3357,
+      "step": 7848
+    },
+    {
+      "epoch": 1.5473186119873819,
+      "grad_norm": 0.4768617570887559,
+      "learning_rate": 1.3497282196111949e-05,
+      "loss": 0.352,
+      "step": 7849
+    },
+    {
+      "epoch": 1.5475157728706623,
+      "grad_norm": 0.5331400342710803,
+      "learning_rate": 1.3495830010405884e-05,
+      "loss": 0.3285,
+      "step": 7850
+    },
+    {
+      "epoch": 1.5477129337539433,
+      "grad_norm": 0.48686256381769405,
+      "learning_rate": 1.3494377740709685e-05,
+      "loss": 0.3352,
+      "step": 7851
+    },
+    {
+      "epoch": 1.547910094637224,
+      "grad_norm": 0.4919653626792826,
+      "learning_rate": 1.3492925387058249e-05,
+      "loss": 0.349,
+      "step": 7852
+    },
+    {
+      "epoch": 1.5481072555205047,
+      "grad_norm": 0.5069611129563922,
+      "learning_rate": 1.3491472949486466e-05,
+      "loss": 0.3184,
+      "step": 7853
+    },
+    {
+      "epoch": 1.5483044164037856,
+      "grad_norm": 0.4979053390062899,
+      "learning_rate": 1.3490020428029236e-05,
+      "loss": 0.324,
+      "step": 7854
+    },
+    {
+      "epoch": 1.5485015772870663,
+      "grad_norm": 0.5968503457298566,
+      "learning_rate": 1.3488567822721453e-05,
+      "loss": 0.3675,
+      "step": 7855
+    },
+    {
+      "epoch": 1.548698738170347,
+      "grad_norm": 0.4798040440294239,
+      "learning_rate": 1.3487115133598017e-05,
+      "loss": 0.332,
+      "step": 7856
+    },
+    {
+      "epoch": 1.548895899053628,
+      "grad_norm": 0.9676067949318208,
+      "learning_rate": 1.3485662360693834e-05,
+      "loss": 0.3271,
+      "step": 7857
+    },
+    {
+      "epoch": 1.5490930599369084,
+      "grad_norm": 0.5290026297802422,
+      "learning_rate": 1.3484209504043804e-05,
+      "loss": 0.3601,
+      "step": 7858
+    },
+    {
+      "epoch": 1.5492902208201893,
+      "grad_norm": 0.5572104887123693,
+      "learning_rate": 1.3482756563682837e-05,
+      "loss": 0.3474,
+      "step": 7859
+    },
+    {
+      "epoch": 1.54948738170347,
+      "grad_norm": 0.48952981915700133,
+      "learning_rate": 1.3481303539645838e-05,
+      "loss": 0.349,
+      "step": 7860
+    },
+    {
+      "epoch": 1.5496845425867507,
+      "grad_norm": 0.49539223483029654,
+      "learning_rate": 1.347985043196772e-05,
+      "loss": 0.3374,
+      "step": 7861
+    },
+    {
+      "epoch": 1.5498817034700316,
+      "grad_norm": 0.4935516593268971,
+      "learning_rate": 1.3478397240683387e-05,
+      "loss": 0.3269,
+      "step": 7862
+    },
+    {
+      "epoch": 1.5500788643533123,
+      "grad_norm": 0.5036803822601013,
+      "learning_rate": 1.3476943965827765e-05,
+      "loss": 0.3285,
+      "step": 7863
+    },
+    {
+      "epoch": 1.550276025236593,
+      "grad_norm": 0.5028733026075429,
+      "learning_rate": 1.3475490607435764e-05,
+      "loss": 0.3494,
+      "step": 7864
+    },
+    {
+      "epoch": 1.550473186119874,
+      "grad_norm": 0.4764330154599795,
+      "learning_rate": 1.34740371655423e-05,
+      "loss": 0.3291,
+      "step": 7865
+    },
+    {
+      "epoch": 1.5506703470031544,
+      "grad_norm": 2.4174378790452895,
+      "learning_rate": 1.3472583640182298e-05,
+      "loss": 0.3636,
+      "step": 7866
+    },
+    {
+      "epoch": 1.5508675078864353,
+      "grad_norm": 0.5015765985791284,
+      "learning_rate": 1.3471130031390673e-05,
+      "loss": 0.332,
+      "step": 7867
+    },
+    {
+      "epoch": 1.551064668769716,
+      "grad_norm": 0.5309675714334715,
+      "learning_rate": 1.346967633920236e-05,
+      "loss": 0.3314,
+      "step": 7868
+    },
+    {
+      "epoch": 1.5512618296529967,
+      "grad_norm": 0.4868835503536096,
+      "learning_rate": 1.3468222563652274e-05,
+      "loss": 0.3377,
+      "step": 7869
+    },
+    {
+      "epoch": 1.5514589905362777,
+      "grad_norm": 0.5127927157453637,
+      "learning_rate": 1.3466768704775348e-05,
+      "loss": 0.3672,
+      "step": 7870
+    },
+    {
+      "epoch": 1.5516561514195584,
+      "grad_norm": 0.4827580717241643,
+      "learning_rate": 1.3465314762606513e-05,
+      "loss": 0.3218,
+      "step": 7871
+    },
+    {
+      "epoch": 1.551853312302839,
+      "grad_norm": 0.5736851092892326,
+      "learning_rate": 1.3463860737180703e-05,
+      "loss": 0.366,
+      "step": 7872
+    },
+    {
+      "epoch": 1.55205047318612,
+      "grad_norm": 0.5318567397637377,
+      "learning_rate": 1.3462406628532846e-05,
+      "loss": 0.3378,
+      "step": 7873
+    },
+    {
+      "epoch": 1.5522476340694005,
+      "grad_norm": 0.5155275111603366,
+      "learning_rate": 1.3460952436697883e-05,
+      "loss": 0.3438,
+      "step": 7874
+    },
+    {
+      "epoch": 1.5524447949526814,
+      "grad_norm": 0.4775879907549808,
+      "learning_rate": 1.345949816171075e-05,
+      "loss": 0.3427,
+      "step": 7875
+    },
+    {
+      "epoch": 1.552641955835962,
+      "grad_norm": 0.5168763642660567,
+      "learning_rate": 1.3458043803606386e-05,
+      "loss": 0.3597,
+      "step": 7876
+    },
+    {
+      "epoch": 1.5528391167192428,
+      "grad_norm": 0.5349163438973762,
+      "learning_rate": 1.3456589362419739e-05,
+      "loss": 0.3805,
+      "step": 7877
+    },
+    {
+      "epoch": 1.5530362776025237,
+      "grad_norm": 0.5844522900838138,
+      "learning_rate": 1.3455134838185746e-05,
+      "loss": 0.3263,
+      "step": 7878
+    },
+    {
+      "epoch": 1.5532334384858044,
+      "grad_norm": 0.4829639630596844,
+      "learning_rate": 1.3453680230939357e-05,
+      "loss": 0.3256,
+      "step": 7879
+    },
+    {
+      "epoch": 1.5534305993690851,
+      "grad_norm": 0.5218800494399819,
+      "learning_rate": 1.345222554071552e-05,
+      "loss": 0.3303,
+      "step": 7880
+    },
+    {
+      "epoch": 1.553627760252366,
+      "grad_norm": 0.521156173705382,
+      "learning_rate": 1.3450770767549181e-05,
+      "loss": 0.3395,
+      "step": 7881
+    },
+    {
+      "epoch": 1.5538249211356467,
+      "grad_norm": 0.4981289635139336,
+      "learning_rate": 1.34493159114753e-05,
+      "loss": 0.3327,
+      "step": 7882
+    },
+    {
+      "epoch": 1.5540220820189274,
+      "grad_norm": 0.5448350587172559,
+      "learning_rate": 1.3447860972528823e-05,
+      "loss": 0.3581,
+      "step": 7883
+    },
+    {
+      "epoch": 1.5542192429022084,
+      "grad_norm": 0.5056691581563505,
+      "learning_rate": 1.3446405950744709e-05,
+      "loss": 0.3432,
+      "step": 7884
+    },
+    {
+      "epoch": 1.5544164037854888,
+      "grad_norm": 0.4953282778846118,
+      "learning_rate": 1.344495084615792e-05,
+      "loss": 0.3137,
+      "step": 7885
+    },
+    {
+      "epoch": 1.5546135646687698,
+      "grad_norm": 0.4751287087730977,
+      "learning_rate": 1.344349565880341e-05,
+      "loss": 0.3207,
+      "step": 7886
+    },
+    {
+      "epoch": 1.5548107255520505,
+      "grad_norm": 0.5085147826900555,
+      "learning_rate": 1.3442040388716146e-05,
+      "loss": 0.3499,
+      "step": 7887
+    },
+    {
+      "epoch": 1.5550078864353312,
+      "grad_norm": 0.47206999350998413,
+      "learning_rate": 1.3440585035931089e-05,
+      "loss": 0.327,
+      "step": 7888
+    },
+    {
+      "epoch": 1.555205047318612,
+      "grad_norm": 0.4488347810163907,
+      "learning_rate": 1.3439129600483207e-05,
+      "loss": 0.318,
+      "step": 7889
+    },
+    {
+      "epoch": 1.5554022082018928,
+      "grad_norm": 0.5051353448066457,
+      "learning_rate": 1.3437674082407463e-05,
+      "loss": 0.3564,
+      "step": 7890
+    },
+    {
+      "epoch": 1.5555993690851735,
+      "grad_norm": 0.49268581152861246,
+      "learning_rate": 1.3436218481738834e-05,
+      "loss": 0.3415,
+      "step": 7891
+    },
+    {
+      "epoch": 1.5557965299684544,
+      "grad_norm": 0.45508654653177255,
+      "learning_rate": 1.343476279851229e-05,
+      "loss": 0.3255,
+      "step": 7892
+    },
+    {
+      "epoch": 1.555993690851735,
+      "grad_norm": 0.5114403490145619,
+      "learning_rate": 1.3433307032762799e-05,
+      "loss": 0.3453,
+      "step": 7893
+    },
+    {
+      "epoch": 1.5561908517350158,
+      "grad_norm": 0.5594515489170677,
+      "learning_rate": 1.3431851184525343e-05,
+      "loss": 0.3382,
+      "step": 7894
+    },
+    {
+      "epoch": 1.5563880126182965,
+      "grad_norm": 0.5360994056754633,
+      "learning_rate": 1.3430395253834902e-05,
+      "loss": 0.359,
+      "step": 7895
+    },
+    {
+      "epoch": 1.5565851735015772,
+      "grad_norm": 0.5154345101589657,
+      "learning_rate": 1.3428939240726451e-05,
+      "loss": 0.3317,
+      "step": 7896
+    },
+    {
+      "epoch": 1.5567823343848581,
+      "grad_norm": 0.5196905638444523,
+      "learning_rate": 1.3427483145234974e-05,
+      "loss": 0.3562,
+      "step": 7897
+    },
+    {
+      "epoch": 1.5569794952681388,
+      "grad_norm": 0.6914979302075561,
+      "learning_rate": 1.342602696739545e-05,
+      "loss": 0.3364,
+      "step": 7898
+    },
+    {
+      "epoch": 1.5571766561514195,
+      "grad_norm": 0.4739027137576714,
+      "learning_rate": 1.3424570707242875e-05,
+      "loss": 0.3069,
+      "step": 7899
+    },
+    {
+      "epoch": 1.5573738170347005,
+      "grad_norm": 0.4986733883254391,
+      "learning_rate": 1.3423114364812229e-05,
+      "loss": 0.3468,
+      "step": 7900
+    },
+    {
+      "epoch": 1.557570977917981,
+      "grad_norm": 0.5096555609944065,
+      "learning_rate": 1.3421657940138504e-05,
+      "loss": 0.3436,
+      "step": 7901
+    },
+    {
+      "epoch": 1.5577681388012619,
+      "grad_norm": 0.5170708802611125,
+      "learning_rate": 1.342020143325669e-05,
+      "loss": 0.3461,
+      "step": 7902
+    },
+    {
+      "epoch": 1.5579652996845426,
+      "grad_norm": 0.5134132452327762,
+      "learning_rate": 1.3418744844201783e-05,
+      "loss": 0.3173,
+      "step": 7903
+    },
+    {
+      "epoch": 1.5581624605678233,
+      "grad_norm": 0.48373651440601273,
+      "learning_rate": 1.3417288173008778e-05,
+      "loss": 0.3306,
+      "step": 7904
+    },
+    {
+      "epoch": 1.5583596214511042,
+      "grad_norm": 0.5305843575820182,
+      "learning_rate": 1.341583141971267e-05,
+      "loss": 0.359,
+      "step": 7905
+    },
+    {
+      "epoch": 1.5585567823343849,
+      "grad_norm": 0.4992440462111545,
+      "learning_rate": 1.3414374584348466e-05,
+      "loss": 0.3436,
+      "step": 7906
+    },
+    {
+      "epoch": 1.5587539432176656,
+      "grad_norm": 0.4815491827718251,
+      "learning_rate": 1.3412917666951159e-05,
+      "loss": 0.3462,
+      "step": 7907
+    },
+    {
+      "epoch": 1.5589511041009465,
+      "grad_norm": 0.4686319393529876,
+      "learning_rate": 1.3411460667555762e-05,
+      "loss": 0.3345,
+      "step": 7908
+    },
+    {
+      "epoch": 1.559148264984227,
+      "grad_norm": 0.4699483011052357,
+      "learning_rate": 1.341000358619727e-05,
+      "loss": 0.3093,
+      "step": 7909
+    },
+    {
+      "epoch": 1.559345425867508,
+      "grad_norm": 0.464632109167414,
+      "learning_rate": 1.34085464229107e-05,
+      "loss": 0.3376,
+      "step": 7910
+    },
+    {
+      "epoch": 1.5595425867507886,
+      "grad_norm": 0.4823143160131754,
+      "learning_rate": 1.3407089177731052e-05,
+      "loss": 0.3385,
+      "step": 7911
+    },
+    {
+      "epoch": 1.5597397476340693,
+      "grad_norm": 0.5218932690257738,
+      "learning_rate": 1.3405631850693347e-05,
+      "loss": 0.3565,
+      "step": 7912
+    },
+    {
+      "epoch": 1.5599369085173502,
+      "grad_norm": 0.45189898716468596,
+      "learning_rate": 1.3404174441832592e-05,
+      "loss": 0.3183,
+      "step": 7913
+    },
+    {
+      "epoch": 1.560134069400631,
+      "grad_norm": 0.5167028440182481,
+      "learning_rate": 1.3402716951183807e-05,
+      "loss": 0.3643,
+      "step": 7914
+    },
+    {
+      "epoch": 1.5603312302839116,
+      "grad_norm": 0.48376625167026677,
+      "learning_rate": 1.3401259378782005e-05,
+      "loss": 0.3401,
+      "step": 7915
+    },
+    {
+      "epoch": 1.5605283911671926,
+      "grad_norm": 0.4773795454869428,
+      "learning_rate": 1.3399801724662209e-05,
+      "loss": 0.3224,
+      "step": 7916
+    },
+    {
+      "epoch": 1.560725552050473,
+      "grad_norm": 0.4945502868783542,
+      "learning_rate": 1.3398343988859439e-05,
+      "loss": 0.3307,
+      "step": 7917
+    },
+    {
+      "epoch": 1.560922712933754,
+      "grad_norm": 0.5019400523084708,
+      "learning_rate": 1.3396886171408717e-05,
+      "loss": 0.3409,
+      "step": 7918
+    },
+    {
+      "epoch": 1.5611198738170347,
+      "grad_norm": 0.4691324729142486,
+      "learning_rate": 1.3395428272345067e-05,
+      "loss": 0.325,
+      "step": 7919
+    },
+    {
+      "epoch": 1.5613170347003154,
+      "grad_norm": 0.4672956168760278,
+      "learning_rate": 1.3393970291703523e-05,
+      "loss": 0.3032,
+      "step": 7920
+    },
+    {
+      "epoch": 1.5615141955835963,
+      "grad_norm": 0.6166235750122606,
+      "learning_rate": 1.3392512229519105e-05,
+      "loss": 0.3498,
+      "step": 7921
+    },
+    {
+      "epoch": 1.561711356466877,
+      "grad_norm": 0.4806514171065023,
+      "learning_rate": 1.339105408582685e-05,
+      "loss": 0.3212,
+      "step": 7922
+    },
+    {
+      "epoch": 1.5619085173501577,
+      "grad_norm": 0.48674489059156,
+      "learning_rate": 1.3389595860661793e-05,
+      "loss": 0.3374,
+      "step": 7923
+    },
+    {
+      "epoch": 1.5621056782334386,
+      "grad_norm": 0.502066062621961,
+      "learning_rate": 1.3388137554058961e-05,
+      "loss": 0.3328,
+      "step": 7924
+    },
+    {
+      "epoch": 1.562302839116719,
+      "grad_norm": 0.479052777119119,
+      "learning_rate": 1.33866791660534e-05,
+      "loss": 0.3363,
+      "step": 7925
+    },
+    {
+      "epoch": 1.5625,
+      "grad_norm": 0.512932795649453,
+      "learning_rate": 1.3385220696680142e-05,
+      "loss": 0.3632,
+      "step": 7926
+    },
+    {
+      "epoch": 1.562697160883281,
+      "grad_norm": 0.5291035392129729,
+      "learning_rate": 1.3383762145974233e-05,
+      "loss": 0.3619,
+      "step": 7927
+    },
+    {
+      "epoch": 1.5628943217665614,
+      "grad_norm": 0.49465140055838475,
+      "learning_rate": 1.338230351397071e-05,
+      "loss": 0.3417,
+      "step": 7928
+    },
+    {
+      "epoch": 1.5630914826498423,
+      "grad_norm": 0.4967349452813738,
+      "learning_rate": 1.3380844800704624e-05,
+      "loss": 0.3243,
+      "step": 7929
+    },
+    {
+      "epoch": 1.563288643533123,
+      "grad_norm": 0.47414238203469705,
+      "learning_rate": 1.3379386006211021e-05,
+      "loss": 0.3145,
+      "step": 7930
+    },
+    {
+      "epoch": 1.5634858044164037,
+      "grad_norm": 0.5112298596770326,
+      "learning_rate": 1.3377927130524943e-05,
+      "loss": 0.3532,
+      "step": 7931
+    },
+    {
+      "epoch": 1.5636829652996846,
+      "grad_norm": 0.49210192456304214,
+      "learning_rate": 1.337646817368145e-05,
+      "loss": 0.3367,
+      "step": 7932
+    },
+    {
+      "epoch": 1.5638801261829653,
+      "grad_norm": 0.5028321558489999,
+      "learning_rate": 1.3375009135715584e-05,
+      "loss": 0.3532,
+      "step": 7933
+    },
+    {
+      "epoch": 1.564077287066246,
+      "grad_norm": 0.45880232523040504,
+      "learning_rate": 1.3373550016662414e-05,
+      "loss": 0.3176,
+      "step": 7934
+    },
+    {
+      "epoch": 1.564274447949527,
+      "grad_norm": 0.5003629680329704,
+      "learning_rate": 1.337209081655698e-05,
+      "loss": 0.3133,
+      "step": 7935
+    },
+    {
+      "epoch": 1.5644716088328074,
+      "grad_norm": 0.5046115931642703,
+      "learning_rate": 1.3370631535434356e-05,
+      "loss": 0.3378,
+      "step": 7936
+    },
+    {
+      "epoch": 1.5646687697160884,
+      "grad_norm": 0.47177267681957447,
+      "learning_rate": 1.3369172173329588e-05,
+      "loss": 0.322,
+      "step": 7937
+    },
+    {
+      "epoch": 1.564865930599369,
+      "grad_norm": 0.5082813052750841,
+      "learning_rate": 1.3367712730277748e-05,
+      "loss": 0.3357,
+      "step": 7938
+    },
+    {
+      "epoch": 1.5650630914826498,
+      "grad_norm": 0.509547876262465,
+      "learning_rate": 1.33662532063139e-05,
+      "loss": 0.3505,
+      "step": 7939
+    },
+    {
+      "epoch": 1.5652602523659307,
+      "grad_norm": 0.4783658190779783,
+      "learning_rate": 1.3364793601473105e-05,
+      "loss": 0.3284,
+      "step": 7940
+    },
+    {
+      "epoch": 1.5654574132492114,
+      "grad_norm": 0.5056684800626197,
+      "learning_rate": 1.3363333915790435e-05,
+      "loss": 0.3331,
+      "step": 7941
+    },
+    {
+      "epoch": 1.565654574132492,
+      "grad_norm": 0.5243617531078861,
+      "learning_rate": 1.336187414930096e-05,
+      "loss": 0.3303,
+      "step": 7942
+    },
+    {
+      "epoch": 1.565851735015773,
+      "grad_norm": 0.49836190761234994,
+      "learning_rate": 1.336041430203975e-05,
+      "loss": 0.3452,
+      "step": 7943
+    },
+    {
+      "epoch": 1.5660488958990535,
+      "grad_norm": 0.4968929524695866,
+      "learning_rate": 1.3358954374041882e-05,
+      "loss": 0.3172,
+      "step": 7944
+    },
+    {
+      "epoch": 1.5662460567823344,
+      "grad_norm": 0.4356287567333736,
+      "learning_rate": 1.335749436534243e-05,
+      "loss": 0.2964,
+      "step": 7945
+    },
+    {
+      "epoch": 1.5664432176656151,
+      "grad_norm": 0.4984101635486525,
+      "learning_rate": 1.335603427597647e-05,
+      "loss": 0.355,
+      "step": 7946
+    },
+    {
+      "epoch": 1.5666403785488958,
+      "grad_norm": 0.47682247093029434,
+      "learning_rate": 1.3354574105979085e-05,
+      "loss": 0.3365,
+      "step": 7947
+    },
+    {
+      "epoch": 1.5668375394321767,
+      "grad_norm": 0.4845515350210312,
+      "learning_rate": 1.3353113855385356e-05,
+      "loss": 0.3208,
+      "step": 7948
+    },
+    {
+      "epoch": 1.5670347003154574,
+      "grad_norm": 0.5275550760728646,
+      "learning_rate": 1.3351653524230366e-05,
+      "loss": 0.3523,
+      "step": 7949
+    },
+    {
+      "epoch": 1.5672318611987381,
+      "grad_norm": 0.4735343037228871,
+      "learning_rate": 1.3350193112549202e-05,
+      "loss": 0.3279,
+      "step": 7950
+    },
+    {
+      "epoch": 1.567429022082019,
+      "grad_norm": 0.46954231120457435,
+      "learning_rate": 1.334873262037695e-05,
+      "loss": 0.3417,
+      "step": 7951
+    },
+    {
+      "epoch": 1.5676261829652995,
+      "grad_norm": 0.9705396445603504,
+      "learning_rate": 1.3347272047748696e-05,
+      "loss": 0.3745,
+      "step": 7952
+    },
+    {
+      "epoch": 1.5678233438485805,
+      "grad_norm": 0.513132943966501,
+      "learning_rate": 1.3345811394699542e-05,
+      "loss": 0.3437,
+      "step": 7953
+    },
+    {
+      "epoch": 1.5680205047318612,
+      "grad_norm": 0.4859610432771261,
+      "learning_rate": 1.3344350661264568e-05,
+      "loss": 0.3525,
+      "step": 7954
+    },
+    {
+      "epoch": 1.5682176656151419,
+      "grad_norm": 0.5006180525289092,
+      "learning_rate": 1.3342889847478884e-05,
+      "loss": 0.3054,
+      "step": 7955
+    },
+    {
+      "epoch": 1.5684148264984228,
+      "grad_norm": 0.49462467277505867,
+      "learning_rate": 1.3341428953377574e-05,
+      "loss": 0.3457,
+      "step": 7956
+    },
+    {
+      "epoch": 1.5686119873817035,
+      "grad_norm": 0.4960059003422695,
+      "learning_rate": 1.3339967978995746e-05,
+      "loss": 0.3262,
+      "step": 7957
+    },
+    {
+      "epoch": 1.5688091482649842,
+      "grad_norm": 0.5065110926656363,
+      "learning_rate": 1.3338506924368494e-05,
+      "loss": 0.3368,
+      "step": 7958
+    },
+    {
+      "epoch": 1.569006309148265,
+      "grad_norm": 0.49924087045412563,
+      "learning_rate": 1.3337045789530927e-05,
+      "loss": 0.3544,
+      "step": 7959
+    },
+    {
+      "epoch": 1.5692034700315456,
+      "grad_norm": 0.5450061812449806,
+      "learning_rate": 1.3335584574518148e-05,
+      "loss": 0.3455,
+      "step": 7960
+    },
+    {
+      "epoch": 1.5694006309148265,
+      "grad_norm": 0.5370940243152774,
+      "learning_rate": 1.333412327936526e-05,
+      "loss": 0.3212,
+      "step": 7961
+    },
+    {
+      "epoch": 1.5695977917981072,
+      "grad_norm": 0.48962467971522494,
+      "learning_rate": 1.333266190410738e-05,
+      "loss": 0.327,
+      "step": 7962
+    },
+    {
+      "epoch": 1.569794952681388,
+      "grad_norm": 0.4967444956543661,
+      "learning_rate": 1.3331200448779611e-05,
+      "loss": 0.3412,
+      "step": 7963
+    },
+    {
+      "epoch": 1.5699921135646688,
+      "grad_norm": 1.5137118599387835,
+      "learning_rate": 1.332973891341707e-05,
+      "loss": 0.3315,
+      "step": 7964
+    },
+    {
+      "epoch": 1.5701892744479495,
+      "grad_norm": 0.5024737388313264,
+      "learning_rate": 1.332827729805487e-05,
+      "loss": 0.3528,
+      "step": 7965
+    },
+    {
+      "epoch": 1.5703864353312302,
+      "grad_norm": 0.5259290240369583,
+      "learning_rate": 1.3326815602728127e-05,
+      "loss": 0.3713,
+      "step": 7966
+    },
+    {
+      "epoch": 1.5705835962145112,
+      "grad_norm": 0.5279028181427355,
+      "learning_rate": 1.332535382747196e-05,
+      "loss": 0.3557,
+      "step": 7967
+    },
+    {
+      "epoch": 1.5707807570977916,
+      "grad_norm": 0.7224209367198574,
+      "learning_rate": 1.332389197232149e-05,
+      "loss": 0.3688,
+      "step": 7968
+    },
+    {
+      "epoch": 1.5709779179810726,
+      "grad_norm": 0.48099183427409264,
+      "learning_rate": 1.3322430037311837e-05,
+      "loss": 0.3116,
+      "step": 7969
+    },
+    {
+      "epoch": 1.5711750788643533,
+      "grad_norm": 0.508998063080973,
+      "learning_rate": 1.332096802247813e-05,
+      "loss": 0.3365,
+      "step": 7970
+    },
+    {
+      "epoch": 1.571372239747634,
+      "grad_norm": 0.49220793045424055,
+      "learning_rate": 1.331950592785549e-05,
+      "loss": 0.3522,
+      "step": 7971
+    },
+    {
+      "epoch": 1.5715694006309149,
+      "grad_norm": 0.5332391139433758,
+      "learning_rate": 1.3318043753479047e-05,
+      "loss": 0.3892,
+      "step": 7972
+    },
+    {
+      "epoch": 1.5717665615141956,
+      "grad_norm": 0.4951350369532224,
+      "learning_rate": 1.3316581499383929e-05,
+      "loss": 0.3295,
+      "step": 7973
+    },
+    {
+      "epoch": 1.5719637223974763,
+      "grad_norm": 0.5140523975794388,
+      "learning_rate": 1.3315119165605273e-05,
+      "loss": 0.3443,
+      "step": 7974
+    },
+    {
+      "epoch": 1.5721608832807572,
+      "grad_norm": 0.5191099345240461,
+      "learning_rate": 1.3313656752178205e-05,
+      "loss": 0.3456,
+      "step": 7975
+    },
+    {
+      "epoch": 1.572358044164038,
+      "grad_norm": 0.5150428182108653,
+      "learning_rate": 1.331219425913787e-05,
+      "loss": 0.3262,
+      "step": 7976
+    },
+    {
+      "epoch": 1.5725552050473186,
+      "grad_norm": 0.4909428641563583,
+      "learning_rate": 1.3310731686519397e-05,
+      "loss": 0.3172,
+      "step": 7977
+    },
+    {
+      "epoch": 1.5727523659305995,
+      "grad_norm": 0.4961393599175189,
+      "learning_rate": 1.3309269034357931e-05,
+      "loss": 0.3097,
+      "step": 7978
+    },
+    {
+      "epoch": 1.57294952681388,
+      "grad_norm": 0.5318028453318919,
+      "learning_rate": 1.330780630268861e-05,
+      "loss": 0.3635,
+      "step": 7979
+    },
+    {
+      "epoch": 1.573146687697161,
+      "grad_norm": 0.5996323717685671,
+      "learning_rate": 1.3306343491546581e-05,
+      "loss": 0.3703,
+      "step": 7980
+    },
+    {
+      "epoch": 1.5733438485804416,
+      "grad_norm": 1.5419005107960566,
+      "learning_rate": 1.3304880600966985e-05,
+      "loss": 0.3559,
+      "step": 7981
+    },
+    {
+      "epoch": 1.5735410094637223,
+      "grad_norm": 0.4648861254663534,
+      "learning_rate": 1.3303417630984972e-05,
+      "loss": 0.31,
+      "step": 7982
+    },
+    {
+      "epoch": 1.5737381703470033,
+      "grad_norm": 0.5014405931273911,
+      "learning_rate": 1.3301954581635692e-05,
+      "loss": 0.3542,
+      "step": 7983
+    },
+    {
+      "epoch": 1.573935331230284,
+      "grad_norm": 0.47324297334032567,
+      "learning_rate": 1.3300491452954292e-05,
+      "loss": 0.3439,
+      "step": 7984
+    },
+    {
+      "epoch": 1.5741324921135647,
+      "grad_norm": 0.5098324583372497,
+      "learning_rate": 1.3299028244975929e-05,
+      "loss": 0.3528,
+      "step": 7985
+    },
+    {
+      "epoch": 1.5743296529968456,
+      "grad_norm": 0.48462830498298404,
+      "learning_rate": 1.3297564957735752e-05,
+      "loss": 0.3318,
+      "step": 7986
+    },
+    {
+      "epoch": 1.574526813880126,
+      "grad_norm": 0.5391632212698709,
+      "learning_rate": 1.3296101591268924e-05,
+      "loss": 0.3632,
+      "step": 7987
+    },
+    {
+      "epoch": 1.574723974763407,
+      "grad_norm": 0.48926713072275946,
+      "learning_rate": 1.3294638145610598e-05,
+      "loss": 0.314,
+      "step": 7988
+    },
+    {
+      "epoch": 1.5749211356466877,
+      "grad_norm": 0.5220045491815719,
+      "learning_rate": 1.3293174620795942e-05,
+      "loss": 0.3469,
+      "step": 7989
+    },
+    {
+      "epoch": 1.5751182965299684,
+      "grad_norm": 0.5103365539849146,
+      "learning_rate": 1.329171101686011e-05,
+      "loss": 0.3431,
+      "step": 7990
+    },
+    {
+      "epoch": 1.5753154574132493,
+      "grad_norm": 0.49515948764836015,
+      "learning_rate": 1.329024733383827e-05,
+      "loss": 0.312,
+      "step": 7991
+    },
+    {
+      "epoch": 1.57551261829653,
+      "grad_norm": 0.47764308725452986,
+      "learning_rate": 1.328878357176559e-05,
+      "loss": 0.3156,
+      "step": 7992
+    },
+    {
+      "epoch": 1.5757097791798107,
+      "grad_norm": 0.4913518129002428,
+      "learning_rate": 1.3287319730677237e-05,
+      "loss": 0.3449,
+      "step": 7993
+    },
+    {
+      "epoch": 1.5759069400630916,
+      "grad_norm": 0.5199989452514197,
+      "learning_rate": 1.3285855810608377e-05,
+      "loss": 0.3572,
+      "step": 7994
+    },
+    {
+      "epoch": 1.576104100946372,
+      "grad_norm": 0.46619302997703116,
+      "learning_rate": 1.3284391811594191e-05,
+      "loss": 0.3115,
+      "step": 7995
+    },
+    {
+      "epoch": 1.576301261829653,
+      "grad_norm": 0.532011015620942,
+      "learning_rate": 1.3282927733669842e-05,
+      "loss": 0.3377,
+      "step": 7996
+    },
+    {
+      "epoch": 1.5764984227129337,
+      "grad_norm": 1.440253937845809,
+      "learning_rate": 1.328146357687051e-05,
+      "loss": 0.3688,
+      "step": 7997
+    },
+    {
+      "epoch": 1.5766955835962144,
+      "grad_norm": 1.0850935873776184,
+      "learning_rate": 1.3279999341231375e-05,
+      "loss": 0.338,
+      "step": 7998
+    },
+    {
+      "epoch": 1.5768927444794953,
+      "grad_norm": 0.5168104366063799,
+      "learning_rate": 1.3278535026787614e-05,
+      "loss": 0.3501,
+      "step": 7999
+    },
+    {
+      "epoch": 1.577089905362776,
+      "grad_norm": 0.5139066874038835,
+      "learning_rate": 1.3277070633574409e-05,
+      "loss": 0.3414,
+      "step": 8000
+    },
+    {
+      "epoch": 1.5772870662460567,
+      "grad_norm": 0.48443540557771675,
+      "learning_rate": 1.3275606161626941e-05,
+      "loss": 0.3271,
+      "step": 8001
+    },
+    {
+      "epoch": 1.5774842271293377,
+      "grad_norm": 0.5239644881242183,
+      "learning_rate": 1.32741416109804e-05,
+      "loss": 0.3445,
+      "step": 8002
+    },
+    {
+      "epoch": 1.5776813880126181,
+      "grad_norm": 0.46561515482433635,
+      "learning_rate": 1.3272676981669968e-05,
+      "loss": 0.3232,
+      "step": 8003
+    },
+    {
+      "epoch": 1.577878548895899,
+      "grad_norm": 0.5355952021642951,
+      "learning_rate": 1.327121227373084e-05,
+      "loss": 0.3552,
+      "step": 8004
+    },
+    {
+      "epoch": 1.5780757097791798,
+      "grad_norm": 0.511194087803704,
+      "learning_rate": 1.3269747487198197e-05,
+      "loss": 0.3651,
+      "step": 8005
+    },
+    {
+      "epoch": 1.5782728706624605,
+      "grad_norm": 0.463393152472201,
+      "learning_rate": 1.326828262210724e-05,
+      "loss": 0.3282,
+      "step": 8006
+    },
+    {
+      "epoch": 1.5784700315457414,
+      "grad_norm": 0.49704658525209167,
+      "learning_rate": 1.326681767849316e-05,
+      "loss": 0.3174,
+      "step": 8007
+    },
+    {
+      "epoch": 1.578667192429022,
+      "grad_norm": 0.5412028232341971,
+      "learning_rate": 1.3265352656391158e-05,
+      "loss": 0.349,
+      "step": 8008
+    },
+    {
+      "epoch": 1.5788643533123028,
+      "grad_norm": 0.4698860322570366,
+      "learning_rate": 1.3263887555836425e-05,
+      "loss": 0.3106,
+      "step": 8009
+    },
+    {
+      "epoch": 1.5790615141955837,
+      "grad_norm": 0.4982618521435573,
+      "learning_rate": 1.3262422376864168e-05,
+      "loss": 0.3262,
+      "step": 8010
+    },
+    {
+      "epoch": 1.5792586750788642,
+      "grad_norm": 0.539325905160115,
+      "learning_rate": 1.3260957119509586e-05,
+      "loss": 0.3502,
+      "step": 8011
+    },
+    {
+      "epoch": 1.5794558359621451,
+      "grad_norm": 0.5288545896046527,
+      "learning_rate": 1.325949178380788e-05,
+      "loss": 0.3692,
+      "step": 8012
+    },
+    {
+      "epoch": 1.5796529968454258,
+      "grad_norm": 0.4758545656821426,
+      "learning_rate": 1.3258026369794261e-05,
+      "loss": 0.3283,
+      "step": 8013
+    },
+    {
+      "epoch": 1.5798501577287065,
+      "grad_norm": 0.5218405213561202,
+      "learning_rate": 1.3256560877503936e-05,
+      "loss": 0.3415,
+      "step": 8014
+    },
+    {
+      "epoch": 1.5800473186119874,
+      "grad_norm": 0.4894784063780892,
+      "learning_rate": 1.3255095306972112e-05,
+      "loss": 0.3269,
+      "step": 8015
+    },
+    {
+      "epoch": 1.5802444794952681,
+      "grad_norm": 0.5110496055324127,
+      "learning_rate": 1.3253629658234002e-05,
+      "loss": 0.3542,
+      "step": 8016
+    },
+    {
+      "epoch": 1.5804416403785488,
+      "grad_norm": 0.49174003178088704,
+      "learning_rate": 1.325216393132482e-05,
+      "loss": 0.3394,
+      "step": 8017
+    },
+    {
+      "epoch": 1.5806388012618298,
+      "grad_norm": 0.47299800973515416,
+      "learning_rate": 1.3250698126279781e-05,
+      "loss": 0.3419,
+      "step": 8018
+    },
+    {
+      "epoch": 1.5808359621451105,
+      "grad_norm": 0.4822158106803239,
+      "learning_rate": 1.32492322431341e-05,
+      "loss": 0.3539,
+      "step": 8019
+    },
+    {
+      "epoch": 1.5810331230283912,
+      "grad_norm": 0.47952679544937005,
+      "learning_rate": 1.3247766281922998e-05,
+      "loss": 0.3468,
+      "step": 8020
+    },
+    {
+      "epoch": 1.581230283911672,
+      "grad_norm": 0.46714001850575293,
+      "learning_rate": 1.3246300242681698e-05,
+      "loss": 0.3468,
+      "step": 8021
+    },
+    {
+      "epoch": 1.5814274447949526,
+      "grad_norm": 0.5770765745101369,
+      "learning_rate": 1.3244834125445415e-05,
+      "loss": 0.3658,
+      "step": 8022
+    },
+    {
+      "epoch": 1.5816246056782335,
+      "grad_norm": 0.466026202641457,
+      "learning_rate": 1.3243367930249386e-05,
+      "loss": 0.3342,
+      "step": 8023
+    },
+    {
+      "epoch": 1.5818217665615142,
+      "grad_norm": 0.5082981481974369,
+      "learning_rate": 1.3241901657128827e-05,
+      "loss": 0.3635,
+      "step": 8024
+    },
+    {
+      "epoch": 1.5820189274447949,
+      "grad_norm": 0.48542475906619664,
+      "learning_rate": 1.3240435306118973e-05,
+      "loss": 0.3553,
+      "step": 8025
+    },
+    {
+      "epoch": 1.5822160883280758,
+      "grad_norm": 0.4787474049672265,
+      "learning_rate": 1.3238968877255044e-05,
+      "loss": 0.3357,
+      "step": 8026
+    },
+    {
+      "epoch": 1.5824132492113565,
+      "grad_norm": 0.5075996681809781,
+      "learning_rate": 1.3237502370572287e-05,
+      "loss": 0.3479,
+      "step": 8027
+    },
+    {
+      "epoch": 1.5826104100946372,
+      "grad_norm": 0.5129452355663832,
+      "learning_rate": 1.3236035786105922e-05,
+      "loss": 0.3598,
+      "step": 8028
+    },
+    {
+      "epoch": 1.5828075709779181,
+      "grad_norm": 0.45167882333931725,
+      "learning_rate": 1.3234569123891197e-05,
+      "loss": 0.3199,
+      "step": 8029
+    },
+    {
+      "epoch": 1.5830047318611986,
+      "grad_norm": 0.4891940893562189,
+      "learning_rate": 1.3233102383963341e-05,
+      "loss": 0.3588,
+      "step": 8030
+    },
+    {
+      "epoch": 1.5832018927444795,
+      "grad_norm": 0.4924263041994256,
+      "learning_rate": 1.3231635566357599e-05,
+      "loss": 0.3558,
+      "step": 8031
+    },
+    {
+      "epoch": 1.5833990536277602,
+      "grad_norm": 0.5691351727091879,
+      "learning_rate": 1.3230168671109207e-05,
+      "loss": 0.3215,
+      "step": 8032
+    },
+    {
+      "epoch": 1.583596214511041,
+      "grad_norm": 0.5454873000853145,
+      "learning_rate": 1.3228701698253415e-05,
+      "loss": 0.3561,
+      "step": 8033
+    },
+    {
+      "epoch": 1.5837933753943219,
+      "grad_norm": 4.541252989056811,
+      "learning_rate": 1.3227234647825463e-05,
+      "loss": 0.329,
+      "step": 8034
+    },
+    {
+      "epoch": 1.5839905362776026,
+      "grad_norm": 0.5120743849675218,
+      "learning_rate": 1.3225767519860597e-05,
+      "loss": 0.342,
+      "step": 8035
+    },
+    {
+      "epoch": 1.5841876971608833,
+      "grad_norm": 0.5241749900315733,
+      "learning_rate": 1.3224300314394073e-05,
+      "loss": 0.3847,
+      "step": 8036
+    },
+    {
+      "epoch": 1.5843848580441642,
+      "grad_norm": 0.5165462915069103,
+      "learning_rate": 1.3222833031461133e-05,
+      "loss": 0.3414,
+      "step": 8037
+    },
+    {
+      "epoch": 1.5845820189274447,
+      "grad_norm": 0.5189206607460046,
+      "learning_rate": 1.3221365671097038e-05,
+      "loss": 0.3291,
+      "step": 8038
+    },
+    {
+      "epoch": 1.5847791798107256,
+      "grad_norm": 0.5059501509473748,
+      "learning_rate": 1.3219898233337036e-05,
+      "loss": 0.3249,
+      "step": 8039
+    },
+    {
+      "epoch": 1.5849763406940063,
+      "grad_norm": 0.4904431912501636,
+      "learning_rate": 1.321843071821639e-05,
+      "loss": 0.3435,
+      "step": 8040
+    },
+    {
+      "epoch": 1.585173501577287,
+      "grad_norm": 0.45852925927612104,
+      "learning_rate": 1.3216963125770345e-05,
+      "loss": 0.3055,
+      "step": 8041
+    },
+    {
+      "epoch": 1.585370662460568,
+      "grad_norm": 0.5589366409498242,
+      "learning_rate": 1.3215495456034179e-05,
+      "loss": 0.3253,
+      "step": 8042
+    },
+    {
+      "epoch": 1.5855678233438486,
+      "grad_norm": 0.49532146559960627,
+      "learning_rate": 1.3214027709043142e-05,
+      "loss": 0.3442,
+      "step": 8043
+    },
+    {
+      "epoch": 1.5857649842271293,
+      "grad_norm": 0.49902148390151746,
+      "learning_rate": 1.3212559884832503e-05,
+      "loss": 0.3499,
+      "step": 8044
+    },
+    {
+      "epoch": 1.5859621451104102,
+      "grad_norm": 0.5117292560455523,
+      "learning_rate": 1.3211091983437524e-05,
+      "loss": 0.3502,
+      "step": 8045
+    },
+    {
+      "epoch": 1.5861593059936907,
+      "grad_norm": 0.4829964496583425,
+      "learning_rate": 1.3209624004893476e-05,
+      "loss": 0.321,
+      "step": 8046
+    },
+    {
+      "epoch": 1.5863564668769716,
+      "grad_norm": 0.5776878181785877,
+      "learning_rate": 1.3208155949235621e-05,
+      "loss": 0.3194,
+      "step": 8047
+    },
+    {
+      "epoch": 1.5865536277602523,
+      "grad_norm": 0.44848690865415186,
+      "learning_rate": 1.3206687816499242e-05,
+      "loss": 0.3077,
+      "step": 8048
+    },
+    {
+      "epoch": 1.586750788643533,
+      "grad_norm": 0.4881456482953011,
+      "learning_rate": 1.3205219606719606e-05,
+      "loss": 0.3246,
+      "step": 8049
+    },
+    {
+      "epoch": 1.586947949526814,
+      "grad_norm": 0.5011767838202971,
+      "learning_rate": 1.3203751319931983e-05,
+      "loss": 0.355,
+      "step": 8050
+    },
+    {
+      "epoch": 1.5871451104100947,
+      "grad_norm": 0.5664107192411584,
+      "learning_rate": 1.320228295617166e-05,
+      "loss": 0.3579,
+      "step": 8051
+    },
+    {
+      "epoch": 1.5873422712933754,
+      "grad_norm": 3.4551884902309777,
+      "learning_rate": 1.3200814515473905e-05,
+      "loss": 0.3417,
+      "step": 8052
+    },
+    {
+      "epoch": 1.5875394321766563,
+      "grad_norm": 0.5286194871009565,
+      "learning_rate": 1.3199345997874007e-05,
+      "loss": 0.3351,
+      "step": 8053
+    },
+    {
+      "epoch": 1.5877365930599368,
+      "grad_norm": 0.4625651878543429,
+      "learning_rate": 1.3197877403407242e-05,
+      "loss": 0.3198,
+      "step": 8054
+    },
+    {
+      "epoch": 1.5879337539432177,
+      "grad_norm": 0.6352734118743283,
+      "learning_rate": 1.31964087321089e-05,
+      "loss": 0.3448,
+      "step": 8055
+    },
+    {
+      "epoch": 1.5881309148264984,
+      "grad_norm": 0.503782340045354,
+      "learning_rate": 1.3194939984014263e-05,
+      "loss": 0.354,
+      "step": 8056
+    },
+    {
+      "epoch": 1.588328075709779,
+      "grad_norm": 0.5794061421651925,
+      "learning_rate": 1.3193471159158621e-05,
+      "loss": 0.3621,
+      "step": 8057
+    },
+    {
+      "epoch": 1.58852523659306,
+      "grad_norm": 0.49275969055339675,
+      "learning_rate": 1.3192002257577263e-05,
+      "loss": 0.3333,
+      "step": 8058
+    },
+    {
+      "epoch": 1.5887223974763407,
+      "grad_norm": 0.4815640944601471,
+      "learning_rate": 1.319053327930548e-05,
+      "loss": 0.341,
+      "step": 8059
+    },
+    {
+      "epoch": 1.5889195583596214,
+      "grad_norm": 0.4613014999066501,
+      "learning_rate": 1.3189064224378562e-05,
+      "loss": 0.3101,
+      "step": 8060
+    },
+    {
+      "epoch": 1.5891167192429023,
+      "grad_norm": 0.48626447035569825,
+      "learning_rate": 1.3187595092831813e-05,
+      "loss": 0.3146,
+      "step": 8061
+    },
+    {
+      "epoch": 1.589313880126183,
+      "grad_norm": 0.4577848208645168,
+      "learning_rate": 1.3186125884700522e-05,
+      "loss": 0.3175,
+      "step": 8062
+    },
+    {
+      "epoch": 1.5895110410094637,
+      "grad_norm": 0.49457076312235915,
+      "learning_rate": 1.3184656600019992e-05,
+      "loss": 0.353,
+      "step": 8063
+    },
+    {
+      "epoch": 1.5897082018927446,
+      "grad_norm": 0.47999535138775207,
+      "learning_rate": 1.318318723882552e-05,
+      "loss": 0.3392,
+      "step": 8064
+    },
+    {
+      "epoch": 1.5899053627760251,
+      "grad_norm": 0.5418783986590079,
+      "learning_rate": 1.3181717801152414e-05,
+      "loss": 0.3415,
+      "step": 8065
+    },
+    {
+      "epoch": 1.590102523659306,
+      "grad_norm": 0.5096260948506632,
+      "learning_rate": 1.3180248287035977e-05,
+      "loss": 0.3262,
+      "step": 8066
+    },
+    {
+      "epoch": 1.5902996845425867,
+      "grad_norm": 0.4906247366527042,
+      "learning_rate": 1.3178778696511511e-05,
+      "loss": 0.3384,
+      "step": 8067
+    },
+    {
+      "epoch": 1.5904968454258674,
+      "grad_norm": 0.470211162945986,
+      "learning_rate": 1.317730902961433e-05,
+      "loss": 0.3244,
+      "step": 8068
+    },
+    {
+      "epoch": 1.5906940063091484,
+      "grad_norm": 0.49115378811432053,
+      "learning_rate": 1.3175839286379734e-05,
+      "loss": 0.3422,
+      "step": 8069
+    },
+    {
+      "epoch": 1.590891167192429,
+      "grad_norm": 0.5013805949559059,
+      "learning_rate": 1.3174369466843048e-05,
+      "loss": 0.3441,
+      "step": 8070
+    },
+    {
+      "epoch": 1.5910883280757098,
+      "grad_norm": 0.4876708210398381,
+      "learning_rate": 1.3172899571039577e-05,
+      "loss": 0.3444,
+      "step": 8071
+    },
+    {
+      "epoch": 1.5912854889589907,
+      "grad_norm": 0.45242714845948756,
+      "learning_rate": 1.3171429599004641e-05,
+      "loss": 0.3162,
+      "step": 8072
+    },
+    {
+      "epoch": 1.5914826498422712,
+      "grad_norm": 0.47227360673852276,
+      "learning_rate": 1.316995955077355e-05,
+      "loss": 0.3263,
+      "step": 8073
+    },
+    {
+      "epoch": 1.591679810725552,
+      "grad_norm": 0.4875936546980924,
+      "learning_rate": 1.3168489426381635e-05,
+      "loss": 0.3398,
+      "step": 8074
+    },
+    {
+      "epoch": 1.5918769716088328,
+      "grad_norm": 0.483433090727928,
+      "learning_rate": 1.3167019225864203e-05,
+      "loss": 0.3262,
+      "step": 8075
+    },
+    {
+      "epoch": 1.5920741324921135,
+      "grad_norm": 0.4983273920225542,
+      "learning_rate": 1.3165548949256586e-05,
+      "loss": 0.3528,
+      "step": 8076
+    },
+    {
+      "epoch": 1.5922712933753944,
+      "grad_norm": 0.48450953137363295,
+      "learning_rate": 1.3164078596594107e-05,
+      "loss": 0.3452,
+      "step": 8077
+    },
+    {
+      "epoch": 1.5924684542586751,
+      "grad_norm": 0.4623023529103843,
+      "learning_rate": 1.3162608167912091e-05,
+      "loss": 0.3421,
+      "step": 8078
+    },
+    {
+      "epoch": 1.5926656151419558,
+      "grad_norm": 1.1891945757617244,
+      "learning_rate": 1.3161137663245869e-05,
+      "loss": 0.335,
+      "step": 8079
+    },
+    {
+      "epoch": 1.5928627760252367,
+      "grad_norm": 0.46625342878642473,
+      "learning_rate": 1.3159667082630768e-05,
+      "loss": 0.3017,
+      "step": 8080
+    },
+    {
+      "epoch": 1.5930599369085172,
+      "grad_norm": 0.4885959593564677,
+      "learning_rate": 1.3158196426102121e-05,
+      "loss": 0.3207,
+      "step": 8081
+    },
+    {
+      "epoch": 1.5932570977917981,
+      "grad_norm": 0.5374079371161415,
+      "learning_rate": 1.315672569369526e-05,
+      "loss": 0.3292,
+      "step": 8082
+    },
+    {
+      "epoch": 1.5934542586750788,
+      "grad_norm": 0.47767332144300273,
+      "learning_rate": 1.3155254885445526e-05,
+      "loss": 0.3134,
+      "step": 8083
+    },
+    {
+      "epoch": 1.5936514195583595,
+      "grad_norm": 0.4563667095812108,
+      "learning_rate": 1.3153784001388249e-05,
+      "loss": 0.3181,
+      "step": 8084
+    },
+    {
+      "epoch": 1.5938485804416405,
+      "grad_norm": 0.4891329550595928,
+      "learning_rate": 1.315231304155877e-05,
+      "loss": 0.352,
+      "step": 8085
+    },
+    {
+      "epoch": 1.5940457413249212,
+      "grad_norm": 0.4746370559092902,
+      "learning_rate": 1.3150842005992434e-05,
+      "loss": 0.3107,
+      "step": 8086
+    },
+    {
+      "epoch": 1.5942429022082019,
+      "grad_norm": 0.5821961651457958,
+      "learning_rate": 1.3149370894724583e-05,
+      "loss": 0.3215,
+      "step": 8087
+    },
+    {
+      "epoch": 1.5944400630914828,
+      "grad_norm": 0.4692821907783642,
+      "learning_rate": 1.3147899707790557e-05,
+      "loss": 0.3383,
+      "step": 8088
+    },
+    {
+      "epoch": 1.5946372239747633,
+      "grad_norm": 0.6369717589355536,
+      "learning_rate": 1.3146428445225708e-05,
+      "loss": 0.3516,
+      "step": 8089
+    },
+    {
+      "epoch": 1.5948343848580442,
+      "grad_norm": 0.5091082227275152,
+      "learning_rate": 1.3144957107065379e-05,
+      "loss": 0.3598,
+      "step": 8090
+    },
+    {
+      "epoch": 1.5950315457413249,
+      "grad_norm": 0.4761782875190844,
+      "learning_rate": 1.3143485693344925e-05,
+      "loss": 0.3409,
+      "step": 8091
+    },
+    {
+      "epoch": 1.5952287066246056,
+      "grad_norm": 0.46602014777166734,
+      "learning_rate": 1.3142014204099696e-05,
+      "loss": 0.3135,
+      "step": 8092
+    },
+    {
+      "epoch": 1.5954258675078865,
+      "grad_norm": 0.4514528986309173,
+      "learning_rate": 1.3140542639365047e-05,
+      "loss": 0.3134,
+      "step": 8093
+    },
+    {
+      "epoch": 1.5956230283911672,
+      "grad_norm": 0.5297263361088499,
+      "learning_rate": 1.3139070999176326e-05,
+      "loss": 0.3419,
+      "step": 8094
+    },
+    {
+      "epoch": 1.595820189274448,
+      "grad_norm": 0.5469443697151118,
+      "learning_rate": 1.3137599283568902e-05,
+      "loss": 0.3563,
+      "step": 8095
+    },
+    {
+      "epoch": 1.5960173501577288,
+      "grad_norm": 0.4957278314999418,
+      "learning_rate": 1.3136127492578126e-05,
+      "loss": 0.355,
+      "step": 8096
+    },
+    {
+      "epoch": 1.5962145110410093,
+      "grad_norm": 0.5166791104046938,
+      "learning_rate": 1.3134655626239363e-05,
+      "loss": 0.3296,
+      "step": 8097
+    },
+    {
+      "epoch": 1.5964116719242902,
+      "grad_norm": 0.5016830354531797,
+      "learning_rate": 1.3133183684587974e-05,
+      "loss": 0.3412,
+      "step": 8098
+    },
+    {
+      "epoch": 1.596608832807571,
+      "grad_norm": 0.4893337636428677,
+      "learning_rate": 1.3131711667659323e-05,
+      "loss": 0.339,
+      "step": 8099
+    },
+    {
+      "epoch": 1.5968059936908516,
+      "grad_norm": 0.467330301371642,
+      "learning_rate": 1.3130239575488777e-05,
+      "loss": 0.2984,
+      "step": 8100
+    },
+    {
+      "epoch": 1.5970031545741326,
+      "grad_norm": 3.5270122225585343,
+      "learning_rate": 1.3128767408111704e-05,
+      "loss": 0.3816,
+      "step": 8101
+    },
+    {
+      "epoch": 1.5972003154574133,
+      "grad_norm": 0.46859234609245476,
+      "learning_rate": 1.3127295165563476e-05,
+      "loss": 0.3137,
+      "step": 8102
+    },
+    {
+      "epoch": 1.597397476340694,
+      "grad_norm": 0.5086912187536752,
+      "learning_rate": 1.3125822847879464e-05,
+      "loss": 0.364,
+      "step": 8103
+    },
+    {
+      "epoch": 1.5975946372239749,
+      "grad_norm": 0.46400065943464436,
+      "learning_rate": 1.312435045509504e-05,
+      "loss": 0.3094,
+      "step": 8104
+    },
+    {
+      "epoch": 1.5977917981072554,
+      "grad_norm": 0.5054767816862595,
+      "learning_rate": 1.3122877987245579e-05,
+      "loss": 0.3287,
+      "step": 8105
+    },
+    {
+      "epoch": 1.5979889589905363,
+      "grad_norm": 0.49607387723822316,
+      "learning_rate": 1.3121405444366459e-05,
+      "loss": 0.3377,
+      "step": 8106
+    },
+    {
+      "epoch": 1.598186119873817,
+      "grad_norm": 0.4921049056516874,
+      "learning_rate": 1.3119932826493063e-05,
+      "loss": 0.3322,
+      "step": 8107
+    },
+    {
+      "epoch": 1.5983832807570977,
+      "grad_norm": 0.4850703791480183,
+      "learning_rate": 1.3118460133660766e-05,
+      "loss": 0.317,
+      "step": 8108
+    },
+    {
+      "epoch": 1.5985804416403786,
+      "grad_norm": 0.4917675892451221,
+      "learning_rate": 1.3116987365904951e-05,
+      "loss": 0.3443,
+      "step": 8109
+    },
+    {
+      "epoch": 1.5987776025236593,
+      "grad_norm": 0.5527143499695473,
+      "learning_rate": 1.3115514523261008e-05,
+      "loss": 0.317,
+      "step": 8110
+    },
+    {
+      "epoch": 1.59897476340694,
+      "grad_norm": 0.4889167905883699,
+      "learning_rate": 1.3114041605764319e-05,
+      "loss": 0.3288,
+      "step": 8111
+    },
+    {
+      "epoch": 1.599171924290221,
+      "grad_norm": 0.4604739649545089,
+      "learning_rate": 1.3112568613450271e-05,
+      "loss": 0.3237,
+      "step": 8112
+    },
+    {
+      "epoch": 1.5993690851735016,
+      "grad_norm": 0.4826891223593189,
+      "learning_rate": 1.3111095546354257e-05,
+      "loss": 0.3372,
+      "step": 8113
+    },
+    {
+      "epoch": 1.5995662460567823,
+      "grad_norm": 0.4627352710214227,
+      "learning_rate": 1.3109622404511669e-05,
+      "loss": 0.3274,
+      "step": 8114
+    },
+    {
+      "epoch": 1.5997634069400632,
+      "grad_norm": 0.5183424926340384,
+      "learning_rate": 1.3108149187957895e-05,
+      "loss": 0.3668,
+      "step": 8115
+    },
+    {
+      "epoch": 1.5999605678233437,
+      "grad_norm": 0.4890735359860389,
+      "learning_rate": 1.3106675896728334e-05,
+      "loss": 0.3281,
+      "step": 8116
+    },
+    {
+      "epoch": 1.6001577287066246,
+      "grad_norm": 0.4663459951593309,
+      "learning_rate": 1.3105202530858386e-05,
+      "loss": 0.3157,
+      "step": 8117
+    },
+    {
+      "epoch": 1.6003548895899053,
+      "grad_norm": 0.4642219805501338,
+      "learning_rate": 1.310372909038344e-05,
+      "loss": 0.322,
+      "step": 8118
+    },
+    {
+      "epoch": 1.600552050473186,
+      "grad_norm": 0.4883535083604666,
+      "learning_rate": 1.3102255575338912e-05,
+      "loss": 0.3544,
+      "step": 8119
+    },
+    {
+      "epoch": 1.600749211356467,
+      "grad_norm": 0.4938965248583283,
+      "learning_rate": 1.3100781985760188e-05,
+      "loss": 0.3415,
+      "step": 8120
+    },
+    {
+      "epoch": 1.6009463722397477,
+      "grad_norm": 0.508009918172584,
+      "learning_rate": 1.3099308321682685e-05,
+      "loss": 0.3645,
+      "step": 8121
+    },
+    {
+      "epoch": 1.6011435331230284,
+      "grad_norm": 0.4977394454270282,
+      "learning_rate": 1.30978345831418e-05,
+      "loss": 0.37,
+      "step": 8122
+    },
+    {
+      "epoch": 1.6013406940063093,
+      "grad_norm": 0.4827810334398335,
+      "learning_rate": 1.3096360770172947e-05,
+      "loss": 0.3373,
+      "step": 8123
+    },
+    {
+      "epoch": 1.6015378548895898,
+      "grad_norm": 0.5238029508010671,
+      "learning_rate": 1.309488688281153e-05,
+      "loss": 0.3907,
+      "step": 8124
+    },
+    {
+      "epoch": 1.6017350157728707,
+      "grad_norm": 0.4897188747786206,
+      "learning_rate": 1.3093412921092967e-05,
+      "loss": 0.3295,
+      "step": 8125
+    },
+    {
+      "epoch": 1.6019321766561514,
+      "grad_norm": 0.4957516780778494,
+      "learning_rate": 1.3091938885052665e-05,
+      "loss": 0.35,
+      "step": 8126
+    },
+    {
+      "epoch": 1.602129337539432,
+      "grad_norm": 0.5216427832225966,
+      "learning_rate": 1.3090464774726042e-05,
+      "loss": 0.3486,
+      "step": 8127
+    },
+    {
+      "epoch": 1.602326498422713,
+      "grad_norm": 0.5159704540850552,
+      "learning_rate": 1.3088990590148516e-05,
+      "loss": 0.355,
+      "step": 8128
+    },
+    {
+      "epoch": 1.6025236593059937,
+      "grad_norm": 0.49842376144042355,
+      "learning_rate": 1.3087516331355501e-05,
+      "loss": 0.3439,
+      "step": 8129
+    },
+    {
+      "epoch": 1.6027208201892744,
+      "grad_norm": 0.4666686177899867,
+      "learning_rate": 1.3086041998382419e-05,
+      "loss": 0.3322,
+      "step": 8130
+    },
+    {
+      "epoch": 1.6029179810725553,
+      "grad_norm": 0.6474512146454534,
+      "learning_rate": 1.3084567591264694e-05,
+      "loss": 0.3618,
+      "step": 8131
+    },
+    {
+      "epoch": 1.6031151419558358,
+      "grad_norm": 0.5026757411329433,
+      "learning_rate": 1.308309311003775e-05,
+      "loss": 0.3374,
+      "step": 8132
+    },
+    {
+      "epoch": 1.6033123028391167,
+      "grad_norm": 0.4966462917789033,
+      "learning_rate": 1.308161855473701e-05,
+      "loss": 0.326,
+      "step": 8133
+    },
+    {
+      "epoch": 1.6035094637223974,
+      "grad_norm": 0.48096431260184735,
+      "learning_rate": 1.3080143925397904e-05,
+      "loss": 0.3368,
+      "step": 8134
+    },
+    {
+      "epoch": 1.6037066246056781,
+      "grad_norm": 0.48619845262827444,
+      "learning_rate": 1.3078669222055858e-05,
+      "loss": 0.3617,
+      "step": 8135
+    },
+    {
+      "epoch": 1.603903785488959,
+      "grad_norm": 1.8367614071058735,
+      "learning_rate": 1.3077194444746307e-05,
+      "loss": 0.3497,
+      "step": 8136
+    },
+    {
+      "epoch": 1.6041009463722398,
+      "grad_norm": 0.512344448724785,
+      "learning_rate": 1.3075719593504674e-05,
+      "loss": 0.3413,
+      "step": 8137
+    },
+    {
+      "epoch": 1.6042981072555205,
+      "grad_norm": 0.5211149611982421,
+      "learning_rate": 1.3074244668366412e-05,
+      "loss": 0.3447,
+      "step": 8138
+    },
+    {
+      "epoch": 1.6044952681388014,
+      "grad_norm": 0.48739207546058483,
+      "learning_rate": 1.3072769669366938e-05,
+      "loss": 0.3083,
+      "step": 8139
+    },
+    {
+      "epoch": 1.6046924290220819,
+      "grad_norm": 0.49212067031603235,
+      "learning_rate": 1.3071294596541701e-05,
+      "loss": 0.3196,
+      "step": 8140
+    },
+    {
+      "epoch": 1.6048895899053628,
+      "grad_norm": 3.7626457727782956,
+      "learning_rate": 1.3069819449926136e-05,
+      "loss": 0.4303,
+      "step": 8141
+    },
+    {
+      "epoch": 1.6050867507886435,
+      "grad_norm": 0.5830689224709581,
+      "learning_rate": 1.3068344229555692e-05,
+      "loss": 0.3308,
+      "step": 8142
+    },
+    {
+      "epoch": 1.6052839116719242,
+      "grad_norm": 0.49521045756115595,
+      "learning_rate": 1.30668689354658e-05,
+      "loss": 0.3391,
+      "step": 8143
+    },
+    {
+      "epoch": 1.6054810725552051,
+      "grad_norm": 0.5145784616836976,
+      "learning_rate": 1.3065393567691914e-05,
+      "loss": 0.3544,
+      "step": 8144
+    },
+    {
+      "epoch": 1.6056782334384858,
+      "grad_norm": 0.5028725718766839,
+      "learning_rate": 1.3063918126269483e-05,
+      "loss": 0.3444,
+      "step": 8145
+    },
+    {
+      "epoch": 1.6058753943217665,
+      "grad_norm": 0.5769436395830195,
+      "learning_rate": 1.3062442611233949e-05,
+      "loss": 0.3543,
+      "step": 8146
+    },
+    {
+      "epoch": 1.6060725552050474,
+      "grad_norm": 0.5022646683039049,
+      "learning_rate": 1.3060967022620766e-05,
+      "loss": 0.3165,
+      "step": 8147
+    },
+    {
+      "epoch": 1.606269716088328,
+      "grad_norm": 0.4827891396259431,
+      "learning_rate": 1.3059491360465384e-05,
+      "loss": 0.3357,
+      "step": 8148
+    },
+    {
+      "epoch": 1.6064668769716088,
+      "grad_norm": 0.49635418716653323,
+      "learning_rate": 1.305801562480326e-05,
+      "loss": 0.3121,
+      "step": 8149
+    },
+    {
+      "epoch": 1.6066640378548895,
+      "grad_norm": 0.4857542068874944,
+      "learning_rate": 1.3056539815669846e-05,
+      "loss": 0.3357,
+      "step": 8150
+    },
+    {
+      "epoch": 1.6068611987381702,
+      "grad_norm": 0.45147908050559366,
+      "learning_rate": 1.3055063933100602e-05,
+      "loss": 0.2883,
+      "step": 8151
+    },
+    {
+      "epoch": 1.6070583596214512,
+      "grad_norm": 0.46546726459791526,
+      "learning_rate": 1.3053587977130988e-05,
+      "loss": 0.3118,
+      "step": 8152
+    },
+    {
+      "epoch": 1.6072555205047319,
+      "grad_norm": 0.4993862309452187,
+      "learning_rate": 1.3052111947796463e-05,
+      "loss": 0.3594,
+      "step": 8153
+    },
+    {
+      "epoch": 1.6074526813880126,
+      "grad_norm": 0.4879439935041709,
+      "learning_rate": 1.305063584513249e-05,
+      "loss": 0.3151,
+      "step": 8154
+    },
+    {
+      "epoch": 1.6076498422712935,
+      "grad_norm": 0.48554820642690266,
+      "learning_rate": 1.3049159669174534e-05,
+      "loss": 0.3031,
+      "step": 8155
+    },
+    {
+      "epoch": 1.6078470031545742,
+      "grad_norm": 0.500184176222356,
+      "learning_rate": 1.3047683419958062e-05,
+      "loss": 0.3406,
+      "step": 8156
+    },
+    {
+      "epoch": 1.6080441640378549,
+      "grad_norm": 0.5063923589890583,
+      "learning_rate": 1.3046207097518542e-05,
+      "loss": 0.3499,
+      "step": 8157
+    },
+    {
+      "epoch": 1.6082413249211358,
+      "grad_norm": 0.49816615378885987,
+      "learning_rate": 1.3044730701891442e-05,
+      "loss": 0.3237,
+      "step": 8158
+    },
+    {
+      "epoch": 1.6084384858044163,
+      "grad_norm": 0.4925753536085062,
+      "learning_rate": 1.3043254233112237e-05,
+      "loss": 0.3368,
+      "step": 8159
+    },
+    {
+      "epoch": 1.6086356466876972,
+      "grad_norm": 0.5310015500392122,
+      "learning_rate": 1.3041777691216395e-05,
+      "loss": 0.3725,
+      "step": 8160
+    },
+    {
+      "epoch": 1.608832807570978,
+      "grad_norm": 0.49537678955838665,
+      "learning_rate": 1.3040301076239398e-05,
+      "loss": 0.3287,
+      "step": 8161
+    },
+    {
+      "epoch": 1.6090299684542586,
+      "grad_norm": 0.5527761110782208,
+      "learning_rate": 1.3038824388216718e-05,
+      "loss": 0.3726,
+      "step": 8162
+    },
+    {
+      "epoch": 1.6092271293375395,
+      "grad_norm": 0.447484220574138,
+      "learning_rate": 1.3037347627183835e-05,
+      "loss": 0.3021,
+      "step": 8163
+    },
+    {
+      "epoch": 1.6094242902208202,
+      "grad_norm": 0.5179693625577074,
+      "learning_rate": 1.3035870793176229e-05,
+      "loss": 0.3551,
+      "step": 8164
+    },
+    {
+      "epoch": 1.609621451104101,
+      "grad_norm": 0.46927906569820993,
+      "learning_rate": 1.3034393886229381e-05,
+      "loss": 0.3183,
+      "step": 8165
+    },
+    {
+      "epoch": 1.6098186119873819,
+      "grad_norm": 0.498692120573005,
+      "learning_rate": 1.3032916906378782e-05,
+      "loss": 0.3039,
+      "step": 8166
+    },
+    {
+      "epoch": 1.6100157728706623,
+      "grad_norm": 0.501698354263306,
+      "learning_rate": 1.3031439853659906e-05,
+      "loss": 0.3314,
+      "step": 8167
+    },
+    {
+      "epoch": 1.6102129337539433,
+      "grad_norm": 0.8075569329846617,
+      "learning_rate": 1.302996272810825e-05,
+      "loss": 0.3544,
+      "step": 8168
+    },
+    {
+      "epoch": 1.610410094637224,
+      "grad_norm": 0.4790951853547528,
+      "learning_rate": 1.3028485529759296e-05,
+      "loss": 0.3349,
+      "step": 8169
+    },
+    {
+      "epoch": 1.6106072555205047,
+      "grad_norm": 0.48026262439420764,
+      "learning_rate": 1.3027008258648538e-05,
+      "loss": 0.3649,
+      "step": 8170
+    },
+    {
+      "epoch": 1.6108044164037856,
+      "grad_norm": 0.4922180547792553,
+      "learning_rate": 1.3025530914811473e-05,
+      "loss": 0.3143,
+      "step": 8171
+    },
+    {
+      "epoch": 1.6110015772870663,
+      "grad_norm": 0.5198306988127153,
+      "learning_rate": 1.3024053498283588e-05,
+      "loss": 0.3275,
+      "step": 8172
+    },
+    {
+      "epoch": 1.611198738170347,
+      "grad_norm": 0.4742201615022806,
+      "learning_rate": 1.3022576009100382e-05,
+      "loss": 0.3558,
+      "step": 8173
+    },
+    {
+      "epoch": 1.611395899053628,
+      "grad_norm": 0.5014496532783586,
+      "learning_rate": 1.3021098447297358e-05,
+      "loss": 0.3412,
+      "step": 8174
+    },
+    {
+      "epoch": 1.6115930599369084,
+      "grad_norm": 0.49858321941500644,
+      "learning_rate": 1.3019620812910008e-05,
+      "loss": 0.3349,
+      "step": 8175
+    },
+    {
+      "epoch": 1.6117902208201893,
+      "grad_norm": 0.4876528140518932,
+      "learning_rate": 1.3018143105973835e-05,
+      "loss": 0.3295,
+      "step": 8176
+    },
+    {
+      "epoch": 1.61198738170347,
+      "grad_norm": 0.5014129411656993,
+      "learning_rate": 1.3016665326524343e-05,
+      "loss": 0.3169,
+      "step": 8177
+    },
+    {
+      "epoch": 1.6121845425867507,
+      "grad_norm": 0.4709155858213729,
+      "learning_rate": 1.301518747459704e-05,
+      "loss": 0.3033,
+      "step": 8178
+    },
+    {
+      "epoch": 1.6123817034700316,
+      "grad_norm": 0.569369280596523,
+      "learning_rate": 1.3013709550227429e-05,
+      "loss": 0.3567,
+      "step": 8179
+    },
+    {
+      "epoch": 1.6125788643533123,
+      "grad_norm": 0.49624380854953204,
+      "learning_rate": 1.3012231553451018e-05,
+      "loss": 0.3198,
+      "step": 8180
+    },
+    {
+      "epoch": 1.612776025236593,
+      "grad_norm": 0.4572480950811159,
+      "learning_rate": 1.301075348430332e-05,
+      "loss": 0.3102,
+      "step": 8181
+    },
+    {
+      "epoch": 1.612973186119874,
+      "grad_norm": 0.5168349496318023,
+      "learning_rate": 1.3009275342819842e-05,
+      "loss": 0.3265,
+      "step": 8182
+    },
+    {
+      "epoch": 1.6131703470031544,
+      "grad_norm": 0.6022754654599433,
+      "learning_rate": 1.3007797129036104e-05,
+      "loss": 0.3463,
+      "step": 8183
+    },
+    {
+      "epoch": 1.6133675078864353,
+      "grad_norm": 0.49921945706957105,
+      "learning_rate": 1.3006318842987615e-05,
+      "loss": 0.3402,
+      "step": 8184
+    },
+    {
+      "epoch": 1.613564668769716,
+      "grad_norm": 0.4786378011895483,
+      "learning_rate": 1.3004840484709897e-05,
+      "loss": 0.3151,
+      "step": 8185
+    },
+    {
+      "epoch": 1.6137618296529967,
+      "grad_norm": 0.5091965495386273,
+      "learning_rate": 1.3003362054238465e-05,
+      "loss": 0.3439,
+      "step": 8186
+    },
+    {
+      "epoch": 1.6139589905362777,
+      "grad_norm": 0.5075959256110124,
+      "learning_rate": 1.3001883551608843e-05,
+      "loss": 0.3539,
+      "step": 8187
+    },
+    {
+      "epoch": 1.6141561514195584,
+      "grad_norm": 0.51624022913439,
+      "learning_rate": 1.3000404976856546e-05,
+      "loss": 0.362,
+      "step": 8188
+    },
+    {
+      "epoch": 1.614353312302839,
+      "grad_norm": 0.5353531514337778,
+      "learning_rate": 1.2998926330017109e-05,
+      "loss": 0.3716,
+      "step": 8189
+    },
+    {
+      "epoch": 1.61455047318612,
+      "grad_norm": 0.48775573515644133,
+      "learning_rate": 1.2997447611126049e-05,
+      "loss": 0.3223,
+      "step": 8190
+    },
+    {
+      "epoch": 1.6147476340694005,
+      "grad_norm": 0.5058365439971066,
+      "learning_rate": 1.2995968820218896e-05,
+      "loss": 0.3351,
+      "step": 8191
+    },
+    {
+      "epoch": 1.6149447949526814,
+      "grad_norm": 0.4873527967441035,
+      "learning_rate": 1.2994489957331183e-05,
+      "loss": 0.3134,
+      "step": 8192
+    },
+    {
+      "epoch": 1.615141955835962,
+      "grad_norm": 0.5142383938117662,
+      "learning_rate": 1.2993011022498434e-05,
+      "loss": 0.365,
+      "step": 8193
+    },
+    {
+      "epoch": 1.6153391167192428,
+      "grad_norm": 0.5405740913312057,
+      "learning_rate": 1.2991532015756185e-05,
+      "loss": 0.3483,
+      "step": 8194
+    },
+    {
+      "epoch": 1.6155362776025237,
+      "grad_norm": 0.5168750664395806,
+      "learning_rate": 1.2990052937139972e-05,
+      "loss": 0.343,
+      "step": 8195
+    },
+    {
+      "epoch": 1.6157334384858044,
+      "grad_norm": 0.49456698674100874,
+      "learning_rate": 1.298857378668533e-05,
+      "loss": 0.3316,
+      "step": 8196
+    },
+    {
+      "epoch": 1.6159305993690851,
+      "grad_norm": 0.5197649456360487,
+      "learning_rate": 1.2987094564427794e-05,
+      "loss": 0.3375,
+      "step": 8197
+    },
+    {
+      "epoch": 1.616127760252366,
+      "grad_norm": 0.49653557262255904,
+      "learning_rate": 1.2985615270402904e-05,
+      "loss": 0.3166,
+      "step": 8198
+    },
+    {
+      "epoch": 1.6163249211356467,
+      "grad_norm": 0.5287246948423061,
+      "learning_rate": 1.2984135904646206e-05,
+      "loss": 0.3684,
+      "step": 8199
+    },
+    {
+      "epoch": 1.6165220820189274,
+      "grad_norm": 0.5612379255016995,
+      "learning_rate": 1.298265646719324e-05,
+      "loss": 0.3378,
+      "step": 8200
+    },
+    {
+      "epoch": 1.6167192429022084,
+      "grad_norm": 0.5045525899494582,
+      "learning_rate": 1.2981176958079549e-05,
+      "loss": 0.3534,
+      "step": 8201
+    },
+    {
+      "epoch": 1.6169164037854888,
+      "grad_norm": 0.5021046772968188,
+      "learning_rate": 1.2979697377340681e-05,
+      "loss": 0.3276,
+      "step": 8202
+    },
+    {
+      "epoch": 1.6171135646687698,
+      "grad_norm": 0.5041966851488991,
+      "learning_rate": 1.2978217725012183e-05,
+      "loss": 0.3378,
+      "step": 8203
+    },
+    {
+      "epoch": 1.6173107255520505,
+      "grad_norm": 0.4830420145017819,
+      "learning_rate": 1.2976738001129608e-05,
+      "loss": 0.3352,
+      "step": 8204
+    },
+    {
+      "epoch": 1.6175078864353312,
+      "grad_norm": 0.511796475168007,
+      "learning_rate": 1.2975258205728503e-05,
+      "loss": 0.3401,
+      "step": 8205
+    },
+    {
+      "epoch": 1.617705047318612,
+      "grad_norm": 0.4940584230053775,
+      "learning_rate": 1.2973778338844425e-05,
+      "loss": 0.3433,
+      "step": 8206
+    },
+    {
+      "epoch": 1.6179022082018928,
+      "grad_norm": 0.5293789678998656,
+      "learning_rate": 1.2972298400512926e-05,
+      "loss": 0.3809,
+      "step": 8207
+    },
+    {
+      "epoch": 1.6180993690851735,
+      "grad_norm": 0.49843756683541707,
+      "learning_rate": 1.2970818390769569e-05,
+      "loss": 0.3397,
+      "step": 8208
+    },
+    {
+      "epoch": 1.6182965299684544,
+      "grad_norm": 0.45831650251490347,
+      "learning_rate": 1.2969338309649901e-05,
+      "loss": 0.302,
+      "step": 8209
+    },
+    {
+      "epoch": 1.618493690851735,
+      "grad_norm": 0.4793435986752458,
+      "learning_rate": 1.2967858157189495e-05,
+      "loss": 0.3479,
+      "step": 8210
+    },
+    {
+      "epoch": 1.6186908517350158,
+      "grad_norm": 0.48039522018416736,
+      "learning_rate": 1.2966377933423901e-05,
+      "loss": 0.3395,
+      "step": 8211
+    },
+    {
+      "epoch": 1.6188880126182965,
+      "grad_norm": 0.49848149873721165,
+      "learning_rate": 1.2964897638388694e-05,
+      "loss": 0.3391,
+      "step": 8212
+    },
+    {
+      "epoch": 1.6190851735015772,
+      "grad_norm": 0.4735212283074649,
+      "learning_rate": 1.296341727211943e-05,
+      "loss": 0.3421,
+      "step": 8213
+    },
+    {
+      "epoch": 1.6192823343848581,
+      "grad_norm": 0.5087062921882801,
+      "learning_rate": 1.296193683465168e-05,
+      "loss": 0.342,
+      "step": 8214
+    },
+    {
+      "epoch": 1.6194794952681388,
+      "grad_norm": 0.5258996248721838,
+      "learning_rate": 1.2960456326021013e-05,
+      "loss": 0.354,
+      "step": 8215
+    },
+    {
+      "epoch": 1.6196766561514195,
+      "grad_norm": 0.47955351066783924,
+      "learning_rate": 1.2958975746263e-05,
+      "loss": 0.3241,
+      "step": 8216
+    },
+    {
+      "epoch": 1.6198738170347005,
+      "grad_norm": 0.49147931279397045,
+      "learning_rate": 1.295749509541321e-05,
+      "loss": 0.3606,
+      "step": 8217
+    },
+    {
+      "epoch": 1.620070977917981,
+      "grad_norm": 0.5002125077620726,
+      "learning_rate": 1.2956014373507219e-05,
+      "loss": 0.3514,
+      "step": 8218
+    },
+    {
+      "epoch": 1.6202681388012619,
+      "grad_norm": 0.5136920075059789,
+      "learning_rate": 1.2954533580580603e-05,
+      "loss": 0.3622,
+      "step": 8219
+    },
+    {
+      "epoch": 1.6204652996845426,
+      "grad_norm": 0.5015775386249709,
+      "learning_rate": 1.2953052716668939e-05,
+      "loss": 0.3395,
+      "step": 8220
+    },
+    {
+      "epoch": 1.6206624605678233,
+      "grad_norm": 0.48326937347111654,
+      "learning_rate": 1.2951571781807804e-05,
+      "loss": 0.3227,
+      "step": 8221
+    },
+    {
+      "epoch": 1.6208596214511042,
+      "grad_norm": 0.5751545472906557,
+      "learning_rate": 1.295009077603278e-05,
+      "loss": 0.3488,
+      "step": 8222
+    },
+    {
+      "epoch": 1.6210567823343849,
+      "grad_norm": 0.5320211920407559,
+      "learning_rate": 1.2948609699379451e-05,
+      "loss": 0.3413,
+      "step": 8223
+    },
+    {
+      "epoch": 1.6212539432176656,
+      "grad_norm": 0.4687064635801877,
+      "learning_rate": 1.2947128551883399e-05,
+      "loss": 0.3454,
+      "step": 8224
+    },
+    {
+      "epoch": 1.6214511041009465,
+      "grad_norm": 0.49744433210961403,
+      "learning_rate": 1.294564733358021e-05,
+      "loss": 0.3515,
+      "step": 8225
+    },
+    {
+      "epoch": 1.621648264984227,
+      "grad_norm": 0.5069074005833163,
+      "learning_rate": 1.2944166044505467e-05,
+      "loss": 0.3457,
+      "step": 8226
+    },
+    {
+      "epoch": 1.621845425867508,
+      "grad_norm": 0.47976757490607047,
+      "learning_rate": 1.294268468469477e-05,
+      "loss": 0.3453,
+      "step": 8227
+    },
+    {
+      "epoch": 1.6220425867507886,
+      "grad_norm": 0.4952595442891473,
+      "learning_rate": 1.29412032541837e-05,
+      "loss": 0.32,
+      "step": 8228
+    },
+    {
+      "epoch": 1.6222397476340693,
+      "grad_norm": 0.46297510160068384,
+      "learning_rate": 1.2939721753007857e-05,
+      "loss": 0.3111,
+      "step": 8229
+    },
+    {
+      "epoch": 1.6224369085173502,
+      "grad_norm": 0.5043310082743702,
+      "learning_rate": 1.2938240181202828e-05,
+      "loss": 0.3748,
+      "step": 8230
+    },
+    {
+      "epoch": 1.622634069400631,
+      "grad_norm": 0.5095253493670123,
+      "learning_rate": 1.2936758538804215e-05,
+      "loss": 0.3488,
+      "step": 8231
+    },
+    {
+      "epoch": 1.6228312302839116,
+      "grad_norm": 0.4503216039107211,
+      "learning_rate": 1.2935276825847614e-05,
+      "loss": 0.3134,
+      "step": 8232
+    },
+    {
+      "epoch": 1.6230283911671926,
+      "grad_norm": 0.5050115110615685,
+      "learning_rate": 1.293379504236862e-05,
+      "loss": 0.3343,
+      "step": 8233
+    },
+    {
+      "epoch": 1.623225552050473,
+      "grad_norm": 0.5141044664399913,
+      "learning_rate": 1.293231318840284e-05,
+      "loss": 0.3302,
+      "step": 8234
+    },
+    {
+      "epoch": 1.623422712933754,
+      "grad_norm": 0.4852074392315996,
+      "learning_rate": 1.2930831263985873e-05,
+      "loss": 0.3247,
+      "step": 8235
+    },
+    {
+      "epoch": 1.6236198738170347,
+      "grad_norm": 0.5009665060672993,
+      "learning_rate": 1.2929349269153326e-05,
+      "loss": 0.3452,
+      "step": 8236
+    },
+    {
+      "epoch": 1.6238170347003154,
+      "grad_norm": 0.48558116806406154,
+      "learning_rate": 1.29278672039408e-05,
+      "loss": 0.3345,
+      "step": 8237
+    },
+    {
+      "epoch": 1.6240141955835963,
+      "grad_norm": 0.45949344463355146,
+      "learning_rate": 1.292638506838391e-05,
+      "loss": 0.3232,
+      "step": 8238
+    },
+    {
+      "epoch": 1.624211356466877,
+      "grad_norm": 0.4606410064917042,
+      "learning_rate": 1.2924902862518262e-05,
+      "loss": 0.3166,
+      "step": 8239
+    },
+    {
+      "epoch": 1.6244085173501577,
+      "grad_norm": 0.47453261506907685,
+      "learning_rate": 1.2923420586379466e-05,
+      "loss": 0.3172,
+      "step": 8240
+    },
+    {
+      "epoch": 1.6246056782334386,
+      "grad_norm": 0.49860886417423067,
+      "learning_rate": 1.2921938240003138e-05,
+      "loss": 0.3468,
+      "step": 8241
+    },
+    {
+      "epoch": 1.624802839116719,
+      "grad_norm": 0.4842849858710501,
+      "learning_rate": 1.2920455823424892e-05,
+      "loss": 0.3301,
+      "step": 8242
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 0.4715791270451386,
+      "learning_rate": 1.2918973336680339e-05,
+      "loss": 0.3352,
+      "step": 8243
+    },
+    {
+      "epoch": 1.625197160883281,
+      "grad_norm": 0.4857708781749364,
+      "learning_rate": 1.2917490779805105e-05,
+      "loss": 0.3422,
+      "step": 8244
+    },
+    {
+      "epoch": 1.6253943217665614,
+      "grad_norm": 1.2910001595707934,
+      "learning_rate": 1.2916008152834803e-05,
+      "loss": 0.3522,
+      "step": 8245
+    },
+    {
+      "epoch": 1.6255914826498423,
+      "grad_norm": 0.46922968126191145,
+      "learning_rate": 1.2914525455805056e-05,
+      "loss": 0.3209,
+      "step": 8246
+    },
+    {
+      "epoch": 1.625788643533123,
+      "grad_norm": 0.5124348833505142,
+      "learning_rate": 1.291304268875149e-05,
+      "loss": 0.3444,
+      "step": 8247
+    },
+    {
+      "epoch": 1.6259858044164037,
+      "grad_norm": 0.4617262156966585,
+      "learning_rate": 1.2911559851709728e-05,
+      "loss": 0.3357,
+      "step": 8248
+    },
+    {
+      "epoch": 1.6261829652996846,
+      "grad_norm": 0.4682239861424123,
+      "learning_rate": 1.2910076944715394e-05,
+      "loss": 0.3258,
+      "step": 8249
+    },
+    {
+      "epoch": 1.6263801261829653,
+      "grad_norm": 0.4926977054765841,
+      "learning_rate": 1.2908593967804117e-05,
+      "loss": 0.3357,
+      "step": 8250
+    },
+    {
+      "epoch": 1.626577287066246,
+      "grad_norm": 0.47277698075135877,
+      "learning_rate": 1.290711092101153e-05,
+      "loss": 0.3236,
+      "step": 8251
+    },
+    {
+      "epoch": 1.626774447949527,
+      "grad_norm": 0.46612901619189784,
+      "learning_rate": 1.2905627804373259e-05,
+      "loss": 0.3218,
+      "step": 8252
+    },
+    {
+      "epoch": 1.6269716088328074,
+      "grad_norm": 0.46710838609859795,
+      "learning_rate": 1.2904144617924946e-05,
+      "loss": 0.32,
+      "step": 8253
+    },
+    {
+      "epoch": 1.6271687697160884,
+      "grad_norm": 0.46480402221238837,
+      "learning_rate": 1.2902661361702214e-05,
+      "loss": 0.322,
+      "step": 8254
+    },
+    {
+      "epoch": 1.627365930599369,
+      "grad_norm": 0.4992203321218696,
+      "learning_rate": 1.2901178035740709e-05,
+      "loss": 0.317,
+      "step": 8255
+    },
+    {
+      "epoch": 1.6275630914826498,
+      "grad_norm": 0.5101087513586251,
+      "learning_rate": 1.2899694640076062e-05,
+      "loss": 0.3448,
+      "step": 8256
+    },
+    {
+      "epoch": 1.6277602523659307,
+      "grad_norm": 0.48008287555221985,
+      "learning_rate": 1.289821117474392e-05,
+      "loss": 0.3251,
+      "step": 8257
+    },
+    {
+      "epoch": 1.6279574132492114,
+      "grad_norm": 0.4797668818388636,
+      "learning_rate": 1.2896727639779916e-05,
+      "loss": 0.3361,
+      "step": 8258
+    },
+    {
+      "epoch": 1.628154574132492,
+      "grad_norm": 0.4713917003117169,
+      "learning_rate": 1.2895244035219701e-05,
+      "loss": 0.3121,
+      "step": 8259
+    },
+    {
+      "epoch": 1.628351735015773,
+      "grad_norm": 0.8057901594512715,
+      "learning_rate": 1.2893760361098915e-05,
+      "loss": 0.3256,
+      "step": 8260
+    },
+    {
+      "epoch": 1.6285488958990535,
+      "grad_norm": 0.4876468535149187,
+      "learning_rate": 1.2892276617453208e-05,
+      "loss": 0.3393,
+      "step": 8261
+    },
+    {
+      "epoch": 1.6287460567823344,
+      "grad_norm": 0.5166514990575025,
+      "learning_rate": 1.2890792804318224e-05,
+      "loss": 0.3577,
+      "step": 8262
+    },
+    {
+      "epoch": 1.6289432176656151,
+      "grad_norm": 0.4791140344047714,
+      "learning_rate": 1.2889308921729616e-05,
+      "loss": 0.3388,
+      "step": 8263
+    },
+    {
+      "epoch": 1.6291403785488958,
+      "grad_norm": 0.4812157889038476,
+      "learning_rate": 1.2887824969723035e-05,
+      "loss": 0.3236,
+      "step": 8264
+    },
+    {
+      "epoch": 1.6293375394321767,
+      "grad_norm": 0.45798962884288835,
+      "learning_rate": 1.2886340948334132e-05,
+      "loss": 0.328,
+      "step": 8265
+    },
+    {
+      "epoch": 1.6295347003154574,
+      "grad_norm": 0.8222202615392995,
+      "learning_rate": 1.2884856857598564e-05,
+      "loss": 0.3577,
+      "step": 8266
+    },
+    {
+      "epoch": 1.6297318611987381,
+      "grad_norm": 0.5528989353996063,
+      "learning_rate": 1.2883372697551987e-05,
+      "loss": 0.3358,
+      "step": 8267
+    },
+    {
+      "epoch": 1.629929022082019,
+      "grad_norm": 0.49179851604347374,
+      "learning_rate": 1.2881888468230059e-05,
+      "loss": 0.3182,
+      "step": 8268
+    },
+    {
+      "epoch": 1.6301261829652995,
+      "grad_norm": 0.6056266880805475,
+      "learning_rate": 1.2880404169668438e-05,
+      "loss": 0.3249,
+      "step": 8269
+    },
+    {
+      "epoch": 1.6303233438485805,
+      "grad_norm": 0.7467253138778867,
+      "learning_rate": 1.2878919801902791e-05,
+      "loss": 0.3381,
+      "step": 8270
+    },
+    {
+      "epoch": 1.6305205047318612,
+      "grad_norm": 0.48977301540090146,
+      "learning_rate": 1.2877435364968776e-05,
+      "loss": 0.3175,
+      "step": 8271
+    },
+    {
+      "epoch": 1.6307176656151419,
+      "grad_norm": 0.507009351089707,
+      "learning_rate": 1.2875950858902057e-05,
+      "loss": 0.3186,
+      "step": 8272
+    },
+    {
+      "epoch": 1.6309148264984228,
+      "grad_norm": 0.48004983350279057,
+      "learning_rate": 1.2874466283738303e-05,
+      "loss": 0.3183,
+      "step": 8273
+    },
+    {
+      "epoch": 1.6311119873817035,
+      "grad_norm": 0.6427227001497235,
+      "learning_rate": 1.2872981639513187e-05,
+      "loss": 0.36,
+      "step": 8274
+    },
+    {
+      "epoch": 1.6313091482649842,
+      "grad_norm": 0.5043727739166707,
+      "learning_rate": 1.2871496926262365e-05,
+      "loss": 0.3388,
+      "step": 8275
+    },
+    {
+      "epoch": 1.631506309148265,
+      "grad_norm": 0.537296723687631,
+      "learning_rate": 1.2870012144021524e-05,
+      "loss": 0.3812,
+      "step": 8276
+    },
+    {
+      "epoch": 1.6317034700315456,
+      "grad_norm": 0.4600951770649924,
+      "learning_rate": 1.2868527292826325e-05,
+      "loss": 0.3042,
+      "step": 8277
+    },
+    {
+      "epoch": 1.6319006309148265,
+      "grad_norm": 0.4961373527723629,
+      "learning_rate": 1.2867042372712453e-05,
+      "loss": 0.3288,
+      "step": 8278
+    },
+    {
+      "epoch": 1.6320977917981072,
+      "grad_norm": 0.49399953361553006,
+      "learning_rate": 1.2865557383715574e-05,
+      "loss": 0.3309,
+      "step": 8279
+    },
+    {
+      "epoch": 1.632294952681388,
+      "grad_norm": 0.4951800992829749,
+      "learning_rate": 1.2864072325871372e-05,
+      "loss": 0.316,
+      "step": 8280
+    },
+    {
+      "epoch": 1.6324921135646688,
+      "grad_norm": 0.5172023636303266,
+      "learning_rate": 1.2862587199215528e-05,
+      "loss": 0.3361,
+      "step": 8281
+    },
+    {
+      "epoch": 1.6326892744479495,
+      "grad_norm": 0.44561120909851326,
+      "learning_rate": 1.2861102003783722e-05,
+      "loss": 0.3092,
+      "step": 8282
+    },
+    {
+      "epoch": 1.6328864353312302,
+      "grad_norm": 0.4804875200058334,
+      "learning_rate": 1.2859616739611636e-05,
+      "loss": 0.3166,
+      "step": 8283
+    },
+    {
+      "epoch": 1.6330835962145112,
+      "grad_norm": 1.2364757919941514,
+      "learning_rate": 1.2858131406734953e-05,
+      "loss": 0.3109,
+      "step": 8284
+    },
+    {
+      "epoch": 1.6332807570977916,
+      "grad_norm": 0.4655120000897447,
+      "learning_rate": 1.2856646005189367e-05,
+      "loss": 0.3197,
+      "step": 8285
+    },
+    {
+      "epoch": 1.6334779179810726,
+      "grad_norm": 0.49106292754923186,
+      "learning_rate": 1.285516053501055e-05,
+      "loss": 0.3198,
+      "step": 8286
+    },
+    {
+      "epoch": 1.6336750788643533,
+      "grad_norm": 0.48718403958723305,
+      "learning_rate": 1.2853674996234209e-05,
+      "loss": 0.3309,
+      "step": 8287
+    },
+    {
+      "epoch": 1.633872239747634,
+      "grad_norm": 0.5300102271215745,
+      "learning_rate": 1.2852189388896027e-05,
+      "loss": 0.3611,
+      "step": 8288
+    },
+    {
+      "epoch": 1.6340694006309149,
+      "grad_norm": 0.5758537724813633,
+      "learning_rate": 1.2850703713031698e-05,
+      "loss": 0.321,
+      "step": 8289
+    },
+    {
+      "epoch": 1.6342665615141956,
+      "grad_norm": 0.4930932386464715,
+      "learning_rate": 1.2849217968676916e-05,
+      "loss": 0.3411,
+      "step": 8290
+    },
+    {
+      "epoch": 1.6344637223974763,
+      "grad_norm": 0.47643339025266535,
+      "learning_rate": 1.284773215586738e-05,
+      "loss": 0.3325,
+      "step": 8291
+    },
+    {
+      "epoch": 1.6346608832807572,
+      "grad_norm": 0.5102517200602528,
+      "learning_rate": 1.2846246274638783e-05,
+      "loss": 0.3085,
+      "step": 8292
+    },
+    {
+      "epoch": 1.634858044164038,
+      "grad_norm": 0.4964470884863604,
+      "learning_rate": 1.2844760325026827e-05,
+      "loss": 0.3336,
+      "step": 8293
+    },
+    {
+      "epoch": 1.6350552050473186,
+      "grad_norm": 0.4830571944820674,
+      "learning_rate": 1.2843274307067212e-05,
+      "loss": 0.2993,
+      "step": 8294
+    },
+    {
+      "epoch": 1.6352523659305995,
+      "grad_norm": 0.604904792281781,
+      "learning_rate": 1.2841788220795648e-05,
+      "loss": 0.382,
+      "step": 8295
+    },
+    {
+      "epoch": 1.63544952681388,
+      "grad_norm": 0.5476960221303971,
+      "learning_rate": 1.2840302066247828e-05,
+      "loss": 0.3219,
+      "step": 8296
+    },
+    {
+      "epoch": 1.635646687697161,
+      "grad_norm": 0.5578041142423232,
+      "learning_rate": 1.2838815843459467e-05,
+      "loss": 0.3686,
+      "step": 8297
+    },
+    {
+      "epoch": 1.6358438485804416,
+      "grad_norm": 0.4786603253362986,
+      "learning_rate": 1.2837329552466268e-05,
+      "loss": 0.3073,
+      "step": 8298
+    },
+    {
+      "epoch": 1.6360410094637223,
+      "grad_norm": 0.5103008039139185,
+      "learning_rate": 1.2835843193303941e-05,
+      "loss": 0.3315,
+      "step": 8299
+    },
+    {
+      "epoch": 1.6362381703470033,
+      "grad_norm": 0.5547846433262801,
+      "learning_rate": 1.2834356766008198e-05,
+      "loss": 0.3745,
+      "step": 8300
+    },
+    {
+      "epoch": 1.636435331230284,
+      "grad_norm": 0.47914878602175826,
+      "learning_rate": 1.283287027061475e-05,
+      "loss": 0.3504,
+      "step": 8301
+    },
+    {
+      "epoch": 1.6366324921135647,
+      "grad_norm": 0.48804434383979656,
+      "learning_rate": 1.2831383707159316e-05,
+      "loss": 0.2993,
+      "step": 8302
+    },
+    {
+      "epoch": 1.6368296529968456,
+      "grad_norm": 0.4802703000186615,
+      "learning_rate": 1.2829897075677602e-05,
+      "loss": 0.3225,
+      "step": 8303
+    },
+    {
+      "epoch": 1.637026813880126,
+      "grad_norm": 0.4776966853109427,
+      "learning_rate": 1.2828410376205338e-05,
+      "loss": 0.3372,
+      "step": 8304
+    },
+    {
+      "epoch": 1.637223974763407,
+      "grad_norm": 0.4575440033807344,
+      "learning_rate": 1.2826923608778234e-05,
+      "loss": 0.3184,
+      "step": 8305
+    },
+    {
+      "epoch": 1.6374211356466877,
+      "grad_norm": 0.5436332250882948,
+      "learning_rate": 1.2825436773432014e-05,
+      "loss": 0.3861,
+      "step": 8306
+    },
+    {
+      "epoch": 1.6376182965299684,
+      "grad_norm": 0.48668387601687046,
+      "learning_rate": 1.2823949870202402e-05,
+      "loss": 0.3331,
+      "step": 8307
+    },
+    {
+      "epoch": 1.6378154574132493,
+      "grad_norm": 0.5285033833556001,
+      "learning_rate": 1.2822462899125118e-05,
+      "loss": 0.3535,
+      "step": 8308
+    },
+    {
+      "epoch": 1.63801261829653,
+      "grad_norm": 0.47806217212282204,
+      "learning_rate": 1.2820975860235892e-05,
+      "loss": 0.3247,
+      "step": 8309
+    },
+    {
+      "epoch": 1.6382097791798107,
+      "grad_norm": 0.47295818635975434,
+      "learning_rate": 1.2819488753570448e-05,
+      "loss": 0.3155,
+      "step": 8310
+    },
+    {
+      "epoch": 1.6384069400630916,
+      "grad_norm": 0.49044071641561593,
+      "learning_rate": 1.2818001579164516e-05,
+      "loss": 0.3129,
+      "step": 8311
+    },
+    {
+      "epoch": 1.638604100946372,
+      "grad_norm": 0.49641282989980273,
+      "learning_rate": 1.2816514337053829e-05,
+      "loss": 0.339,
+      "step": 8312
+    },
+    {
+      "epoch": 1.638801261829653,
+      "grad_norm": 0.49283456075636484,
+      "learning_rate": 1.2815027027274114e-05,
+      "loss": 0.3328,
+      "step": 8313
+    },
+    {
+      "epoch": 1.6389984227129337,
+      "grad_norm": 0.509363681934031,
+      "learning_rate": 1.281353964986111e-05,
+      "loss": 0.338,
+      "step": 8314
+    },
+    {
+      "epoch": 1.6391955835962144,
+      "grad_norm": 0.4973258630677114,
+      "learning_rate": 1.2812052204850547e-05,
+      "loss": 0.3491,
+      "step": 8315
+    },
+    {
+      "epoch": 1.6393927444794953,
+      "grad_norm": 0.5065183080300542,
+      "learning_rate": 1.2810564692278167e-05,
+      "loss": 0.3394,
+      "step": 8316
+    },
+    {
+      "epoch": 1.639589905362776,
+      "grad_norm": 0.4506338288617181,
+      "learning_rate": 1.2809077112179708e-05,
+      "loss": 0.3088,
+      "step": 8317
+    },
+    {
+      "epoch": 1.6397870662460567,
+      "grad_norm": 0.5475036275756043,
+      "learning_rate": 1.2807589464590908e-05,
+      "loss": 0.3665,
+      "step": 8318
+    },
+    {
+      "epoch": 1.6399842271293377,
+      "grad_norm": 0.5165355996297737,
+      "learning_rate": 1.280610174954751e-05,
+      "loss": 0.3362,
+      "step": 8319
+    },
+    {
+      "epoch": 1.6401813880126181,
+      "grad_norm": 0.4791278974491389,
+      "learning_rate": 1.2804613967085258e-05,
+      "loss": 0.3059,
+      "step": 8320
+    },
+    {
+      "epoch": 1.640378548895899,
+      "grad_norm": 0.4683684775788071,
+      "learning_rate": 1.28031261172399e-05,
+      "loss": 0.3264,
+      "step": 8321
+    },
+    {
+      "epoch": 1.6405757097791798,
+      "grad_norm": 0.4783064134174778,
+      "learning_rate": 1.2801638200047173e-05,
+      "loss": 0.3351,
+      "step": 8322
+    },
+    {
+      "epoch": 1.6407728706624605,
+      "grad_norm": 0.4818861779472011,
+      "learning_rate": 1.2800150215542839e-05,
+      "loss": 0.3351,
+      "step": 8323
+    },
+    {
+      "epoch": 1.6409700315457414,
+      "grad_norm": 0.46380844410707833,
+      "learning_rate": 1.2798662163762635e-05,
+      "loss": 0.3361,
+      "step": 8324
+    },
+    {
+      "epoch": 1.641167192429022,
+      "grad_norm": 1.7966899299353296,
+      "learning_rate": 1.2797174044742324e-05,
+      "loss": 0.3874,
+      "step": 8325
+    },
+    {
+      "epoch": 1.6413643533123028,
+      "grad_norm": 0.4735261266189423,
+      "learning_rate": 1.2795685858517651e-05,
+      "loss": 0.3191,
+      "step": 8326
+    },
+    {
+      "epoch": 1.6415615141955837,
+      "grad_norm": 0.4845463952330695,
+      "learning_rate": 1.2794197605124375e-05,
+      "loss": 0.356,
+      "step": 8327
+    },
+    {
+      "epoch": 1.6417586750788642,
+      "grad_norm": 0.503438921518367,
+      "learning_rate": 1.279270928459825e-05,
+      "loss": 0.3515,
+      "step": 8328
+    },
+    {
+      "epoch": 1.6419558359621451,
+      "grad_norm": 0.4899624719397338,
+      "learning_rate": 1.2791220896975037e-05,
+      "loss": 0.3158,
+      "step": 8329
+    },
+    {
+      "epoch": 1.6421529968454258,
+      "grad_norm": 0.4669534661764057,
+      "learning_rate": 1.2789732442290493e-05,
+      "loss": 0.3215,
+      "step": 8330
+    },
+    {
+      "epoch": 1.6423501577287065,
+      "grad_norm": 0.5173134441620917,
+      "learning_rate": 1.2788243920580381e-05,
+      "loss": 0.3774,
+      "step": 8331
+    },
+    {
+      "epoch": 1.6425473186119874,
+      "grad_norm": 0.4980287720201502,
+      "learning_rate": 1.2786755331880464e-05,
+      "loss": 0.3449,
+      "step": 8332
+    },
+    {
+      "epoch": 1.6427444794952681,
+      "grad_norm": 0.5047789983811025,
+      "learning_rate": 1.2785266676226507e-05,
+      "loss": 0.3501,
+      "step": 8333
+    },
+    {
+      "epoch": 1.6429416403785488,
+      "grad_norm": 0.468218461046473,
+      "learning_rate": 1.2783777953654273e-05,
+      "loss": 0.3191,
+      "step": 8334
+    },
+    {
+      "epoch": 1.6431388012618298,
+      "grad_norm": 0.48844626736338487,
+      "learning_rate": 1.2782289164199534e-05,
+      "loss": 0.3438,
+      "step": 8335
+    },
+    {
+      "epoch": 1.6433359621451105,
+      "grad_norm": 0.5111075206278436,
+      "learning_rate": 1.2780800307898057e-05,
+      "loss": 0.3595,
+      "step": 8336
+    },
+    {
+      "epoch": 1.6435331230283912,
+      "grad_norm": 0.5097957258324088,
+      "learning_rate": 1.2779311384785609e-05,
+      "loss": 0.3501,
+      "step": 8337
+    },
+    {
+      "epoch": 1.643730283911672,
+      "grad_norm": 0.4715524753760921,
+      "learning_rate": 1.2777822394897971e-05,
+      "loss": 0.3133,
+      "step": 8338
+    },
+    {
+      "epoch": 1.6439274447949526,
+      "grad_norm": 0.5158213364397016,
+      "learning_rate": 1.2776333338270912e-05,
+      "loss": 0.3584,
+      "step": 8339
+    },
+    {
+      "epoch": 1.6441246056782335,
+      "grad_norm": 0.5119422614957586,
+      "learning_rate": 1.277484421494021e-05,
+      "loss": 0.3578,
+      "step": 8340
+    },
+    {
+      "epoch": 1.6443217665615142,
+      "grad_norm": 0.4651826749966539,
+      "learning_rate": 1.2773355024941636e-05,
+      "loss": 0.3213,
+      "step": 8341
+    },
+    {
+      "epoch": 1.6445189274447949,
+      "grad_norm": 0.5550710370506761,
+      "learning_rate": 1.277186576831098e-05,
+      "loss": 0.3653,
+      "step": 8342
+    },
+    {
+      "epoch": 1.6447160883280758,
+      "grad_norm": 0.4788442285113236,
+      "learning_rate": 1.2770376445084014e-05,
+      "loss": 0.3207,
+      "step": 8343
+    },
+    {
+      "epoch": 1.6449132492113565,
+      "grad_norm": 0.5097909767918353,
+      "learning_rate": 1.2768887055296527e-05,
+      "loss": 0.3457,
+      "step": 8344
+    },
+    {
+      "epoch": 1.6451104100946372,
+      "grad_norm": 0.5052842170050793,
+      "learning_rate": 1.2767397598984293e-05,
+      "loss": 0.3592,
+      "step": 8345
+    },
+    {
+      "epoch": 1.6453075709779181,
+      "grad_norm": 0.525965947087887,
+      "learning_rate": 1.2765908076183107e-05,
+      "loss": 0.3717,
+      "step": 8346
+    },
+    {
+      "epoch": 1.6455047318611986,
+      "grad_norm": 0.4520388814224274,
+      "learning_rate": 1.2764418486928748e-05,
+      "loss": 0.3079,
+      "step": 8347
+    },
+    {
+      "epoch": 1.6457018927444795,
+      "grad_norm": 0.4948960396210387,
+      "learning_rate": 1.276292883125701e-05,
+      "loss": 0.3625,
+      "step": 8348
+    },
+    {
+      "epoch": 1.6458990536277602,
+      "grad_norm": 0.48398295575215156,
+      "learning_rate": 1.2761439109203683e-05,
+      "loss": 0.3362,
+      "step": 8349
+    },
+    {
+      "epoch": 1.646096214511041,
+      "grad_norm": 0.45308149007619875,
+      "learning_rate": 1.2759949320804559e-05,
+      "loss": 0.3072,
+      "step": 8350
+    },
+    {
+      "epoch": 1.6462933753943219,
+      "grad_norm": 0.47227300867152255,
+      "learning_rate": 1.2758459466095432e-05,
+      "loss": 0.3099,
+      "step": 8351
+    },
+    {
+      "epoch": 1.6464905362776026,
+      "grad_norm": 0.5261569725940197,
+      "learning_rate": 1.275696954511209e-05,
+      "loss": 0.3195,
+      "step": 8352
+    },
+    {
+      "epoch": 1.6466876971608833,
+      "grad_norm": 0.5172694259251647,
+      "learning_rate": 1.2755479557890337e-05,
+      "loss": 0.3382,
+      "step": 8353
+    },
+    {
+      "epoch": 1.6468848580441642,
+      "grad_norm": 0.4596032582706732,
+      "learning_rate": 1.2753989504465967e-05,
+      "loss": 0.3264,
+      "step": 8354
+    },
+    {
+      "epoch": 1.6470820189274447,
+      "grad_norm": 0.4700775556886168,
+      "learning_rate": 1.275249938487478e-05,
+      "loss": 0.3145,
+      "step": 8355
+    },
+    {
+      "epoch": 1.6472791798107256,
+      "grad_norm": 0.5182439165674259,
+      "learning_rate": 1.2751009199152584e-05,
+      "loss": 0.3204,
+      "step": 8356
+    },
+    {
+      "epoch": 1.6474763406940063,
+      "grad_norm": 0.4892848082259555,
+      "learning_rate": 1.2749518947335173e-05,
+      "loss": 0.3589,
+      "step": 8357
+    },
+    {
+      "epoch": 1.647673501577287,
+      "grad_norm": 0.5181167588333921,
+      "learning_rate": 1.2748028629458356e-05,
+      "loss": 0.3517,
+      "step": 8358
+    },
+    {
+      "epoch": 1.647870662460568,
+      "grad_norm": 0.5273555196791303,
+      "learning_rate": 1.2746538245557938e-05,
+      "loss": 0.3611,
+      "step": 8359
+    },
+    {
+      "epoch": 1.6480678233438486,
+      "grad_norm": 0.4769301710367526,
+      "learning_rate": 1.2745047795669728e-05,
+      "loss": 0.3385,
+      "step": 8360
+    },
+    {
+      "epoch": 1.6482649842271293,
+      "grad_norm": 0.5168568047514019,
+      "learning_rate": 1.274355727982953e-05,
+      "loss": 0.3357,
+      "step": 8361
+    },
+    {
+      "epoch": 1.6484621451104102,
+      "grad_norm": 0.47721045303944054,
+      "learning_rate": 1.2742066698073164e-05,
+      "loss": 0.3318,
+      "step": 8362
+    },
+    {
+      "epoch": 1.6486593059936907,
+      "grad_norm": 0.5244966742590467,
+      "learning_rate": 1.2740576050436433e-05,
+      "loss": 0.3629,
+      "step": 8363
+    },
+    {
+      "epoch": 1.6488564668769716,
+      "grad_norm": 0.4763494332427964,
+      "learning_rate": 1.273908533695516e-05,
+      "loss": 0.355,
+      "step": 8364
+    },
+    {
+      "epoch": 1.6490536277602523,
+      "grad_norm": 0.4791256346523753,
+      "learning_rate": 1.2737594557665152e-05,
+      "loss": 0.3291,
+      "step": 8365
+    },
+    {
+      "epoch": 1.649250788643533,
+      "grad_norm": 0.49789637045484475,
+      "learning_rate": 1.2736103712602232e-05,
+      "loss": 0.3177,
+      "step": 8366
+    },
+    {
+      "epoch": 1.649447949526814,
+      "grad_norm": 0.4568801611649524,
+      "learning_rate": 1.2734612801802217e-05,
+      "loss": 0.3141,
+      "step": 8367
+    },
+    {
+      "epoch": 1.6496451104100947,
+      "grad_norm": 0.49441685311553696,
+      "learning_rate": 1.2733121825300927e-05,
+      "loss": 0.3359,
+      "step": 8368
+    },
+    {
+      "epoch": 1.6498422712933754,
+      "grad_norm": 0.5050247460468098,
+      "learning_rate": 1.2731630783134182e-05,
+      "loss": 0.3715,
+      "step": 8369
+    },
+    {
+      "epoch": 1.6500394321766563,
+      "grad_norm": 0.5015943491120733,
+      "learning_rate": 1.273013967533781e-05,
+      "loss": 0.3383,
+      "step": 8370
+    },
+    {
+      "epoch": 1.6502365930599368,
+      "grad_norm": 0.506827984233287,
+      "learning_rate": 1.2728648501947633e-05,
+      "loss": 0.3357,
+      "step": 8371
+    },
+    {
+      "epoch": 1.6504337539432177,
+      "grad_norm": 0.44237770072174254,
+      "learning_rate": 1.2727157262999481e-05,
+      "loss": 0.2997,
+      "step": 8372
+    },
+    {
+      "epoch": 1.6506309148264984,
+      "grad_norm": 0.48053734591353375,
+      "learning_rate": 1.2725665958529177e-05,
+      "loss": 0.3193,
+      "step": 8373
+    },
+    {
+      "epoch": 1.650828075709779,
+      "grad_norm": 0.5146248090989906,
+      "learning_rate": 1.2724174588572556e-05,
+      "loss": 0.3481,
+      "step": 8374
+    },
+    {
+      "epoch": 1.65102523659306,
+      "grad_norm": 0.47551521631075744,
+      "learning_rate": 1.272268315316544e-05,
+      "loss": 0.327,
+      "step": 8375
+    },
+    {
+      "epoch": 1.6512223974763407,
+      "grad_norm": 0.4877727357397369,
+      "learning_rate": 1.2721191652343674e-05,
+      "loss": 0.3283,
+      "step": 8376
+    },
+    {
+      "epoch": 1.6514195583596214,
+      "grad_norm": 1.4180824108129741,
+      "learning_rate": 1.2719700086143088e-05,
+      "loss": 0.3418,
+      "step": 8377
+    },
+    {
+      "epoch": 1.6516167192429023,
+      "grad_norm": 0.5155193108625191,
+      "learning_rate": 1.2718208454599515e-05,
+      "loss": 0.3532,
+      "step": 8378
+    },
+    {
+      "epoch": 1.651813880126183,
+      "grad_norm": 0.47241976046112605,
+      "learning_rate": 1.2716716757748795e-05,
+      "loss": 0.3526,
+      "step": 8379
+    },
+    {
+      "epoch": 1.6520110410094637,
+      "grad_norm": 0.4883813868684643,
+      "learning_rate": 1.2715224995626769e-05,
+      "loss": 0.321,
+      "step": 8380
+    },
+    {
+      "epoch": 1.6522082018927446,
+      "grad_norm": 0.468988201610082,
+      "learning_rate": 1.2713733168269275e-05,
+      "loss": 0.3357,
+      "step": 8381
+    },
+    {
+      "epoch": 1.6524053627760251,
+      "grad_norm": 0.4474868601000789,
+      "learning_rate": 1.2712241275712156e-05,
+      "loss": 0.3386,
+      "step": 8382
+    },
+    {
+      "epoch": 1.652602523659306,
+      "grad_norm": 1.0602180659250304,
+      "learning_rate": 1.2710749317991255e-05,
+      "loss": 0.3677,
+      "step": 8383
+    },
+    {
+      "epoch": 1.6527996845425867,
+      "grad_norm": 0.4626736577510362,
+      "learning_rate": 1.2709257295142421e-05,
+      "loss": 0.3171,
+      "step": 8384
+    },
+    {
+      "epoch": 1.6529968454258674,
+      "grad_norm": 0.7287485310553602,
+      "learning_rate": 1.2707765207201497e-05,
+      "loss": 0.3376,
+      "step": 8385
+    },
+    {
+      "epoch": 1.6531940063091484,
+      "grad_norm": 0.49508196335788296,
+      "learning_rate": 1.2706273054204334e-05,
+      "loss": 0.3154,
+      "step": 8386
+    },
+    {
+      "epoch": 1.653391167192429,
+      "grad_norm": 0.5335563258052243,
+      "learning_rate": 1.2704780836186781e-05,
+      "loss": 0.335,
+      "step": 8387
+    },
+    {
+      "epoch": 1.6535883280757098,
+      "grad_norm": 0.49204084068864695,
+      "learning_rate": 1.270328855318469e-05,
+      "loss": 0.3457,
+      "step": 8388
+    },
+    {
+      "epoch": 1.6537854889589907,
+      "grad_norm": 0.7179327613194991,
+      "learning_rate": 1.2701796205233916e-05,
+      "loss": 0.3262,
+      "step": 8389
+    },
+    {
+      "epoch": 1.6539826498422712,
+      "grad_norm": 0.5875255445299163,
+      "learning_rate": 1.270030379237031e-05,
+      "loss": 0.3445,
+      "step": 8390
+    },
+    {
+      "epoch": 1.654179810725552,
+      "grad_norm": 0.48571857002344315,
+      "learning_rate": 1.2698811314629734e-05,
+      "loss": 0.3384,
+      "step": 8391
+    },
+    {
+      "epoch": 1.6543769716088328,
+      "grad_norm": 0.5420200873446145,
+      "learning_rate": 1.269731877204804e-05,
+      "loss": 0.3471,
+      "step": 8392
+    },
+    {
+      "epoch": 1.6545741324921135,
+      "grad_norm": 0.4978969822403536,
+      "learning_rate": 1.2695826164661093e-05,
+      "loss": 0.3361,
+      "step": 8393
+    },
+    {
+      "epoch": 1.6547712933753944,
+      "grad_norm": 0.5619686050490151,
+      "learning_rate": 1.269433349250475e-05,
+      "loss": 0.3448,
+      "step": 8394
+    },
+    {
+      "epoch": 1.6549684542586751,
+      "grad_norm": 0.5254844107289572,
+      "learning_rate": 1.2692840755614873e-05,
+      "loss": 0.3367,
+      "step": 8395
+    },
+    {
+      "epoch": 1.6551656151419558,
+      "grad_norm": 0.5011006680947409,
+      "learning_rate": 1.269134795402733e-05,
+      "loss": 0.3418,
+      "step": 8396
+    },
+    {
+      "epoch": 1.6553627760252367,
+      "grad_norm": 0.569081232736909,
+      "learning_rate": 1.2689855087777988e-05,
+      "loss": 0.3265,
+      "step": 8397
+    },
+    {
+      "epoch": 1.6555599369085172,
+      "grad_norm": 0.49741199855508267,
+      "learning_rate": 1.2688362156902707e-05,
+      "loss": 0.3367,
+      "step": 8398
+    },
+    {
+      "epoch": 1.6557570977917981,
+      "grad_norm": 0.5473920091128143,
+      "learning_rate": 1.2686869161437364e-05,
+      "loss": 0.3747,
+      "step": 8399
+    },
+    {
+      "epoch": 1.6559542586750788,
+      "grad_norm": 0.5076794965437509,
+      "learning_rate": 1.2685376101417823e-05,
+      "loss": 0.3506,
+      "step": 8400
+    },
+    {
+      "epoch": 1.6561514195583595,
+      "grad_norm": 0.5162556295971144,
+      "learning_rate": 1.268388297687996e-05,
+      "loss": 0.3288,
+      "step": 8401
+    },
+    {
+      "epoch": 1.6563485804416405,
+      "grad_norm": 0.5519110540607234,
+      "learning_rate": 1.2682389787859646e-05,
+      "loss": 0.3312,
+      "step": 8402
+    },
+    {
+      "epoch": 1.6565457413249212,
+      "grad_norm": 0.5118402740090534,
+      "learning_rate": 1.268089653439276e-05,
+      "loss": 0.2988,
+      "step": 8403
+    },
+    {
+      "epoch": 1.6567429022082019,
+      "grad_norm": 0.5055886550149764,
+      "learning_rate": 1.2679403216515171e-05,
+      "loss": 0.3293,
+      "step": 8404
+    },
+    {
+      "epoch": 1.6569400630914828,
+      "grad_norm": 0.5299784103251353,
+      "learning_rate": 1.2677909834262764e-05,
+      "loss": 0.3818,
+      "step": 8405
+    },
+    {
+      "epoch": 1.6571372239747633,
+      "grad_norm": 0.4721883334342438,
+      "learning_rate": 1.267641638767142e-05,
+      "loss": 0.3164,
+      "step": 8406
+    },
+    {
+      "epoch": 1.6573343848580442,
+      "grad_norm": 0.46598843868067574,
+      "learning_rate": 1.2674922876777014e-05,
+      "loss": 0.3242,
+      "step": 8407
+    },
+    {
+      "epoch": 1.6575315457413249,
+      "grad_norm": 0.4908947433651475,
+      "learning_rate": 1.2673429301615431e-05,
+      "loss": 0.361,
+      "step": 8408
+    },
+    {
+      "epoch": 1.6577287066246056,
+      "grad_norm": 0.5044190970850814,
+      "learning_rate": 1.2671935662222556e-05,
+      "loss": 0.3492,
+      "step": 8409
+    },
+    {
+      "epoch": 1.6579258675078865,
+      "grad_norm": 0.4885883648458966,
+      "learning_rate": 1.2670441958634278e-05,
+      "loss": 0.3405,
+      "step": 8410
+    },
+    {
+      "epoch": 1.6581230283911672,
+      "grad_norm": 0.4992792901348933,
+      "learning_rate": 1.2668948190886479e-05,
+      "loss": 0.3495,
+      "step": 8411
+    },
+    {
+      "epoch": 1.658320189274448,
+      "grad_norm": 0.4766119795887696,
+      "learning_rate": 1.2667454359015053e-05,
+      "loss": 0.324,
+      "step": 8412
+    },
+    {
+      "epoch": 1.6585173501577288,
+      "grad_norm": 0.49210628144064317,
+      "learning_rate": 1.2665960463055884e-05,
+      "loss": 0.3262,
+      "step": 8413
+    },
+    {
+      "epoch": 1.6587145110410093,
+      "grad_norm": 0.5123201240998976,
+      "learning_rate": 1.2664466503044872e-05,
+      "loss": 0.3375,
+      "step": 8414
+    },
+    {
+      "epoch": 1.6589116719242902,
+      "grad_norm": 0.48206936397556427,
+      "learning_rate": 1.2662972479017905e-05,
+      "loss": 0.3495,
+      "step": 8415
+    },
+    {
+      "epoch": 1.659108832807571,
+      "grad_norm": 0.46491900044693,
+      "learning_rate": 1.2661478391010877e-05,
+      "loss": 0.3371,
+      "step": 8416
+    },
+    {
+      "epoch": 1.6593059936908516,
+      "grad_norm": 0.4754604595322926,
+      "learning_rate": 1.2659984239059693e-05,
+      "loss": 0.345,
+      "step": 8417
+    },
+    {
+      "epoch": 1.6595031545741326,
+      "grad_norm": 0.485110553505231,
+      "learning_rate": 1.2658490023200238e-05,
+      "loss": 0.3457,
+      "step": 8418
+    },
+    {
+      "epoch": 1.6597003154574133,
+      "grad_norm": 0.5020242501635195,
+      "learning_rate": 1.2656995743468428e-05,
+      "loss": 0.3449,
+      "step": 8419
+    },
+    {
+      "epoch": 1.659897476340694,
+      "grad_norm": 0.47335087548835225,
+      "learning_rate": 1.2655501399900147e-05,
+      "loss": 0.3344,
+      "step": 8420
+    },
+    {
+      "epoch": 1.6600946372239749,
+      "grad_norm": 0.7292549190195249,
+      "learning_rate": 1.2654006992531314e-05,
+      "loss": 0.3618,
+      "step": 8421
+    },
+    {
+      "epoch": 1.6602917981072554,
+      "grad_norm": 0.46704737896112347,
+      "learning_rate": 1.265251252139782e-05,
+      "loss": 0.3379,
+      "step": 8422
+    },
+    {
+      "epoch": 1.6604889589905363,
+      "grad_norm": 0.47828307265868714,
+      "learning_rate": 1.2651017986535578e-05,
+      "loss": 0.3546,
+      "step": 8423
+    },
+    {
+      "epoch": 1.660686119873817,
+      "grad_norm": 0.5046362588351052,
+      "learning_rate": 1.2649523387980497e-05,
+      "loss": 0.3535,
+      "step": 8424
+    },
+    {
+      "epoch": 1.6608832807570977,
+      "grad_norm": 0.529973462052246,
+      "learning_rate": 1.264802872576848e-05,
+      "loss": 0.3614,
+      "step": 8425
+    },
+    {
+      "epoch": 1.6610804416403786,
+      "grad_norm": 0.47162369718261343,
+      "learning_rate": 1.2646533999935442e-05,
+      "loss": 0.3486,
+      "step": 8426
+    },
+    {
+      "epoch": 1.6612776025236593,
+      "grad_norm": 0.5125003435688313,
+      "learning_rate": 1.2645039210517291e-05,
+      "loss": 0.35,
+      "step": 8427
+    },
+    {
+      "epoch": 1.66147476340694,
+      "grad_norm": 0.49775247171182835,
+      "learning_rate": 1.2643544357549946e-05,
+      "loss": 0.3468,
+      "step": 8428
+    },
+    {
+      "epoch": 1.661671924290221,
+      "grad_norm": 0.4979574735742323,
+      "learning_rate": 1.2642049441069318e-05,
+      "loss": 0.3479,
+      "step": 8429
+    },
+    {
+      "epoch": 1.6618690851735016,
+      "grad_norm": 0.49155198733720284,
+      "learning_rate": 1.2640554461111324e-05,
+      "loss": 0.3534,
+      "step": 8430
+    },
+    {
+      "epoch": 1.6620662460567823,
+      "grad_norm": 0.49261265160510403,
+      "learning_rate": 1.2639059417711882e-05,
+      "loss": 0.3377,
+      "step": 8431
+    },
+    {
+      "epoch": 1.6622634069400632,
+      "grad_norm": 0.47430655424794055,
+      "learning_rate": 1.2637564310906913e-05,
+      "loss": 0.3409,
+      "step": 8432
+    },
+    {
+      "epoch": 1.6624605678233437,
+      "grad_norm": 0.4847102680872918,
+      "learning_rate": 1.2636069140732338e-05,
+      "loss": 0.3385,
+      "step": 8433
+    },
+    {
+      "epoch": 1.6626577287066246,
+      "grad_norm": 0.47723013743075726,
+      "learning_rate": 1.263457390722408e-05,
+      "loss": 0.3318,
+      "step": 8434
+    },
+    {
+      "epoch": 1.6628548895899053,
+      "grad_norm": 0.4724871482409827,
+      "learning_rate": 1.2633078610418062e-05,
+      "loss": 0.3122,
+      "step": 8435
+    },
+    {
+      "epoch": 1.663052050473186,
+      "grad_norm": 0.4828083436400746,
+      "learning_rate": 1.2631583250350208e-05,
+      "loss": 0.344,
+      "step": 8436
+    },
+    {
+      "epoch": 1.663249211356467,
+      "grad_norm": 0.4461653149295655,
+      "learning_rate": 1.2630087827056445e-05,
+      "loss": 0.3247,
+      "step": 8437
+    },
+    {
+      "epoch": 1.6634463722397477,
+      "grad_norm": 0.5069753338240224,
+      "learning_rate": 1.262859234057271e-05,
+      "loss": 0.3152,
+      "step": 8438
+    },
+    {
+      "epoch": 1.6636435331230284,
+      "grad_norm": 0.460521307249503,
+      "learning_rate": 1.2627096790934921e-05,
+      "loss": 0.3223,
+      "step": 8439
+    },
+    {
+      "epoch": 1.6638406940063093,
+      "grad_norm": 0.4604716994762646,
+      "learning_rate": 1.2625601178179021e-05,
+      "loss": 0.3283,
+      "step": 8440
+    },
+    {
+      "epoch": 1.6640378548895898,
+      "grad_norm": 0.5234481706217016,
+      "learning_rate": 1.2624105502340935e-05,
+      "loss": 0.3534,
+      "step": 8441
+    },
+    {
+      "epoch": 1.6642350157728707,
+      "grad_norm": 0.5252701025916288,
+      "learning_rate": 1.2622609763456604e-05,
+      "loss": 0.3829,
+      "step": 8442
+    },
+    {
+      "epoch": 1.6644321766561514,
+      "grad_norm": 0.48070551451863064,
+      "learning_rate": 1.262111396156196e-05,
+      "loss": 0.3495,
+      "step": 8443
+    },
+    {
+      "epoch": 1.664629337539432,
+      "grad_norm": 0.48384766288562747,
+      "learning_rate": 1.2619618096692942e-05,
+      "loss": 0.3316,
+      "step": 8444
+    },
+    {
+      "epoch": 1.664826498422713,
+      "grad_norm": 0.5062594443871286,
+      "learning_rate": 1.2618122168885489e-05,
+      "loss": 0.3348,
+      "step": 8445
+    },
+    {
+      "epoch": 1.6650236593059937,
+      "grad_norm": 0.4907130059730268,
+      "learning_rate": 1.2616626178175544e-05,
+      "loss": 0.3559,
+      "step": 8446
+    },
+    {
+      "epoch": 1.6652208201892744,
+      "grad_norm": 0.5082080738735469,
+      "learning_rate": 1.2615130124599047e-05,
+      "loss": 0.3292,
+      "step": 8447
+    },
+    {
+      "epoch": 1.6654179810725553,
+      "grad_norm": 0.5032607745506708,
+      "learning_rate": 1.2613634008191944e-05,
+      "loss": 0.3291,
+      "step": 8448
+    },
+    {
+      "epoch": 1.6656151419558358,
+      "grad_norm": 0.47422112873571903,
+      "learning_rate": 1.2612137828990178e-05,
+      "loss": 0.341,
+      "step": 8449
+    },
+    {
+      "epoch": 1.6658123028391167,
+      "grad_norm": 0.48408723981425067,
+      "learning_rate": 1.2610641587029697e-05,
+      "loss": 0.342,
+      "step": 8450
+    },
+    {
+      "epoch": 1.6660094637223974,
+      "grad_norm": 0.47815519859652317,
+      "learning_rate": 1.2609145282346452e-05,
+      "loss": 0.3042,
+      "step": 8451
+    },
+    {
+      "epoch": 1.6662066246056781,
+      "grad_norm": 0.45905147453962514,
+      "learning_rate": 1.2607648914976386e-05,
+      "loss": 0.3413,
+      "step": 8452
+    },
+    {
+      "epoch": 1.666403785488959,
+      "grad_norm": 0.4686501349889585,
+      "learning_rate": 1.2606152484955458e-05,
+      "loss": 0.32,
+      "step": 8453
+    },
+    {
+      "epoch": 1.6666009463722398,
+      "grad_norm": 0.5027813358909118,
+      "learning_rate": 1.2604655992319618e-05,
+      "loss": 0.3577,
+      "step": 8454
+    },
+    {
+      "epoch": 1.6667981072555205,
+      "grad_norm": 0.48965015962544906,
+      "learning_rate": 1.260315943710482e-05,
+      "loss": 0.3383,
+      "step": 8455
+    },
+    {
+      "epoch": 1.6669952681388014,
+      "grad_norm": 0.48574589911473337,
+      "learning_rate": 1.2601662819347017e-05,
+      "loss": 0.3396,
+      "step": 8456
+    },
+    {
+      "epoch": 1.6671924290220819,
+      "grad_norm": 0.44679877307708993,
+      "learning_rate": 1.2600166139082175e-05,
+      "loss": 0.3054,
+      "step": 8457
+    },
+    {
+      "epoch": 1.6673895899053628,
+      "grad_norm": 0.5131395778523016,
+      "learning_rate": 1.2598669396346244e-05,
+      "loss": 0.3606,
+      "step": 8458
+    },
+    {
+      "epoch": 1.6675867507886435,
+      "grad_norm": 0.48971170022752913,
+      "learning_rate": 1.259717259117519e-05,
+      "loss": 0.3588,
+      "step": 8459
+    },
+    {
+      "epoch": 1.6677839116719242,
+      "grad_norm": 0.49066716607132943,
+      "learning_rate": 1.259567572360497e-05,
+      "loss": 0.3364,
+      "step": 8460
+    },
+    {
+      "epoch": 1.6679810725552051,
+      "grad_norm": 0.4509623703464088,
+      "learning_rate": 1.2594178793671554e-05,
+      "loss": 0.3219,
+      "step": 8461
+    },
+    {
+      "epoch": 1.6681782334384858,
+      "grad_norm": 0.46371182929662663,
+      "learning_rate": 1.25926818014109e-05,
+      "loss": 0.3491,
+      "step": 8462
+    },
+    {
+      "epoch": 1.6683753943217665,
+      "grad_norm": 0.49080219225409744,
+      "learning_rate": 1.259118474685898e-05,
+      "loss": 0.3511,
+      "step": 8463
+    },
+    {
+      "epoch": 1.6685725552050474,
+      "grad_norm": 0.4985969802063767,
+      "learning_rate": 1.258968763005176e-05,
+      "loss": 0.3487,
+      "step": 8464
+    },
+    {
+      "epoch": 1.668769716088328,
+      "grad_norm": 0.49622170543145033,
+      "learning_rate": 1.2588190451025209e-05,
+      "loss": 0.333,
+      "step": 8465
+    },
+    {
+      "epoch": 1.6689668769716088,
+      "grad_norm": 0.4511106753090801,
+      "learning_rate": 1.2586693209815298e-05,
+      "loss": 0.3192,
+      "step": 8466
+    },
+    {
+      "epoch": 1.6691640378548895,
+      "grad_norm": 0.479506900351286,
+      "learning_rate": 1.2585195906457998e-05,
+      "loss": 0.3485,
+      "step": 8467
+    },
+    {
+      "epoch": 1.6693611987381702,
+      "grad_norm": 0.4715197277159651,
+      "learning_rate": 1.2583698540989288e-05,
+      "loss": 0.3276,
+      "step": 8468
+    },
+    {
+      "epoch": 1.6695583596214512,
+      "grad_norm": 0.4798793765470226,
+      "learning_rate": 1.2582201113445136e-05,
+      "loss": 0.3414,
+      "step": 8469
+    },
+    {
+      "epoch": 1.6697555205047319,
+      "grad_norm": 0.4505832073301154,
+      "learning_rate": 1.2580703623861525e-05,
+      "loss": 0.3292,
+      "step": 8470
+    },
+    {
+      "epoch": 1.6699526813880126,
+      "grad_norm": 0.4695096328518373,
+      "learning_rate": 1.257920607227443e-05,
+      "loss": 0.3511,
+      "step": 8471
+    },
+    {
+      "epoch": 1.6701498422712935,
+      "grad_norm": 0.475920897049246,
+      "learning_rate": 1.2577708458719836e-05,
+      "loss": 0.3333,
+      "step": 8472
+    },
+    {
+      "epoch": 1.6703470031545742,
+      "grad_norm": 0.4629881113952917,
+      "learning_rate": 1.2576210783233715e-05,
+      "loss": 0.3427,
+      "step": 8473
+    },
+    {
+      "epoch": 1.6705441640378549,
+      "grad_norm": 0.4579101805999366,
+      "learning_rate": 1.2574713045852059e-05,
+      "loss": 0.3249,
+      "step": 8474
+    },
+    {
+      "epoch": 1.6707413249211358,
+      "grad_norm": 0.5062337208811096,
+      "learning_rate": 1.2573215246610845e-05,
+      "loss": 0.3161,
+      "step": 8475
+    },
+    {
+      "epoch": 1.6709384858044163,
+      "grad_norm": 0.5716309498702405,
+      "learning_rate": 1.2571717385546067e-05,
+      "loss": 0.3362,
+      "step": 8476
+    },
+    {
+      "epoch": 1.6711356466876972,
+      "grad_norm": 0.45248380776560626,
+      "learning_rate": 1.2570219462693703e-05,
+      "loss": 0.3345,
+      "step": 8477
+    },
+    {
+      "epoch": 1.671332807570978,
+      "grad_norm": 0.4930136541420434,
+      "learning_rate": 1.2568721478089752e-05,
+      "loss": 0.3379,
+      "step": 8478
+    },
+    {
+      "epoch": 1.6715299684542586,
+      "grad_norm": 0.5282094905423682,
+      "learning_rate": 1.2567223431770193e-05,
+      "loss": 0.3808,
+      "step": 8479
+    },
+    {
+      "epoch": 1.6717271293375395,
+      "grad_norm": 0.49446422793627104,
+      "learning_rate": 1.256572532377103e-05,
+      "loss": 0.3307,
+      "step": 8480
+    },
+    {
+      "epoch": 1.6719242902208202,
+      "grad_norm": 0.4665091200670554,
+      "learning_rate": 1.2564227154128248e-05,
+      "loss": 0.3129,
+      "step": 8481
+    },
+    {
+      "epoch": 1.672121451104101,
+      "grad_norm": 0.5106508095556581,
+      "learning_rate": 1.256272892287784e-05,
+      "loss": 0.3647,
+      "step": 8482
+    },
+    {
+      "epoch": 1.6723186119873819,
+      "grad_norm": 0.48462370149465533,
+      "learning_rate": 1.256123063005581e-05,
+      "loss": 0.345,
+      "step": 8483
+    },
+    {
+      "epoch": 1.6725157728706623,
+      "grad_norm": 0.46959052479288327,
+      "learning_rate": 1.2559732275698147e-05,
+      "loss": 0.3305,
+      "step": 8484
+    },
+    {
+      "epoch": 1.6727129337539433,
+      "grad_norm": 0.4863277709967929,
+      "learning_rate": 1.2558233859840861e-05,
+      "loss": 0.3402,
+      "step": 8485
+    },
+    {
+      "epoch": 1.672910094637224,
+      "grad_norm": 0.48685079228972755,
+      "learning_rate": 1.255673538251994e-05,
+      "loss": 0.3153,
+      "step": 8486
+    },
+    {
+      "epoch": 1.6731072555205047,
+      "grad_norm": 0.4893988869401554,
+      "learning_rate": 1.2555236843771398e-05,
+      "loss": 0.3546,
+      "step": 8487
+    },
+    {
+      "epoch": 1.6733044164037856,
+      "grad_norm": 0.4806560567289344,
+      "learning_rate": 1.2553738243631228e-05,
+      "loss": 0.3762,
+      "step": 8488
+    },
+    {
+      "epoch": 1.6735015772870663,
+      "grad_norm": 0.4744744228607181,
+      "learning_rate": 1.2552239582135446e-05,
+      "loss": 0.3557,
+      "step": 8489
+    },
+    {
+      "epoch": 1.673698738170347,
+      "grad_norm": 0.5065072339151813,
+      "learning_rate": 1.2550740859320047e-05,
+      "loss": 0.3514,
+      "step": 8490
+    },
+    {
+      "epoch": 1.673895899053628,
+      "grad_norm": 0.49720037597529226,
+      "learning_rate": 1.2549242075221047e-05,
+      "loss": 0.3621,
+      "step": 8491
+    },
+    {
+      "epoch": 1.6740930599369084,
+      "grad_norm": 0.4862962373562821,
+      "learning_rate": 1.2547743229874452e-05,
+      "loss": 0.3278,
+      "step": 8492
+    },
+    {
+      "epoch": 1.6742902208201893,
+      "grad_norm": 0.4862623198322913,
+      "learning_rate": 1.2546244323316276e-05,
+      "loss": 0.3649,
+      "step": 8493
+    },
+    {
+      "epoch": 1.67448738170347,
+      "grad_norm": 0.46675920503742757,
+      "learning_rate": 1.254474535558253e-05,
+      "loss": 0.3142,
+      "step": 8494
+    },
+    {
+      "epoch": 1.6746845425867507,
+      "grad_norm": 6.314767157106249,
+      "learning_rate": 1.2543246326709227e-05,
+      "loss": 0.3485,
+      "step": 8495
+    },
+    {
+      "epoch": 1.6748817034700316,
+      "grad_norm": 0.5032826105490751,
+      "learning_rate": 1.2541747236732382e-05,
+      "loss": 0.334,
+      "step": 8496
+    },
+    {
+      "epoch": 1.6750788643533123,
+      "grad_norm": 0.4565883254189662,
+      "learning_rate": 1.2540248085688013e-05,
+      "loss": 0.3298,
+      "step": 8497
+    },
+    {
+      "epoch": 1.675276025236593,
+      "grad_norm": 0.47701955741607316,
+      "learning_rate": 1.253874887361214e-05,
+      "loss": 0.3419,
+      "step": 8498
+    },
+    {
+      "epoch": 1.675473186119874,
+      "grad_norm": 0.5192006825879065,
+      "learning_rate": 1.253724960054078e-05,
+      "loss": 0.349,
+      "step": 8499
+    },
+    {
+      "epoch": 1.6756703470031544,
+      "grad_norm": 0.498046580157121,
+      "learning_rate": 1.2535750266509955e-05,
+      "loss": 0.359,
+      "step": 8500
+    },
+    {
+      "epoch": 1.6758675078864353,
+      "grad_norm": 0.45577657594308446,
+      "learning_rate": 1.2534250871555687e-05,
+      "loss": 0.3262,
+      "step": 8501
+    },
+    {
+      "epoch": 1.676064668769716,
+      "grad_norm": 0.4735721377691079,
+      "learning_rate": 1.2532751415714001e-05,
+      "loss": 0.2878,
+      "step": 8502
+    },
+    {
+      "epoch": 1.6762618296529967,
+      "grad_norm": 0.4771932029118233,
+      "learning_rate": 1.2531251899020925e-05,
+      "loss": 0.3162,
+      "step": 8503
+    },
+    {
+      "epoch": 1.6764589905362777,
+      "grad_norm": 0.4757910741943044,
+      "learning_rate": 1.252975232151248e-05,
+      "loss": 0.3357,
+      "step": 8504
+    },
+    {
+      "epoch": 1.6766561514195584,
+      "grad_norm": 0.5121567601038779,
+      "learning_rate": 1.2528252683224697e-05,
+      "loss": 0.3436,
+      "step": 8505
+    },
+    {
+      "epoch": 1.676853312302839,
+      "grad_norm": 0.5520693211608968,
+      "learning_rate": 1.2526752984193613e-05,
+      "loss": 0.3188,
+      "step": 8506
+    },
+    {
+      "epoch": 1.67705047318612,
+      "grad_norm": 0.47244929893727605,
+      "learning_rate": 1.2525253224455249e-05,
+      "loss": 0.3074,
+      "step": 8507
+    },
+    {
+      "epoch": 1.6772476340694005,
+      "grad_norm": 0.4911902577750642,
+      "learning_rate": 1.252375340404565e-05,
+      "loss": 0.3437,
+      "step": 8508
+    },
+    {
+      "epoch": 1.6774447949526814,
+      "grad_norm": 0.4702219725532421,
+      "learning_rate": 1.2522253523000834e-05,
+      "loss": 0.2994,
+      "step": 8509
+    },
+    {
+      "epoch": 1.677641955835962,
+      "grad_norm": 0.4719943895332728,
+      "learning_rate": 1.2520753581356852e-05,
+      "loss": 0.3319,
+      "step": 8510
+    },
+    {
+      "epoch": 1.6778391167192428,
+      "grad_norm": 0.46273321896963937,
+      "learning_rate": 1.251925357914973e-05,
+      "loss": 0.3156,
+      "step": 8511
+    },
+    {
+      "epoch": 1.6780362776025237,
+      "grad_norm": 0.45943840807810565,
+      "learning_rate": 1.2517753516415516e-05,
+      "loss": 0.3108,
+      "step": 8512
+    },
+    {
+      "epoch": 1.6782334384858044,
+      "grad_norm": 0.4958574879886677,
+      "learning_rate": 1.2516253393190245e-05,
+      "loss": 0.3415,
+      "step": 8513
+    },
+    {
+      "epoch": 1.6784305993690851,
+      "grad_norm": 0.471413234518349,
+      "learning_rate": 1.251475320950996e-05,
+      "loss": 0.3336,
+      "step": 8514
+    },
+    {
+      "epoch": 1.678627760252366,
+      "grad_norm": 3.3034750186313615,
+      "learning_rate": 1.2513252965410706e-05,
+      "loss": 0.3529,
+      "step": 8515
+    },
+    {
+      "epoch": 1.6788249211356467,
+      "grad_norm": 0.5058575196517642,
+      "learning_rate": 1.2511752660928523e-05,
+      "loss": 0.3583,
+      "step": 8516
+    },
+    {
+      "epoch": 1.6790220820189274,
+      "grad_norm": 0.5210370780798075,
+      "learning_rate": 1.251025229609946e-05,
+      "loss": 0.3627,
+      "step": 8517
+    },
+    {
+      "epoch": 1.6792192429022084,
+      "grad_norm": 0.9679223222367713,
+      "learning_rate": 1.2508751870959563e-05,
+      "loss": 0.3305,
+      "step": 8518
+    },
+    {
+      "epoch": 1.6794164037854888,
+      "grad_norm": 0.955205787142468,
+      "learning_rate": 1.2507251385544885e-05,
+      "loss": 0.3384,
+      "step": 8519
+    },
+    {
+      "epoch": 1.6796135646687698,
+      "grad_norm": 0.5760713418529518,
+      "learning_rate": 1.2505750839891473e-05,
+      "loss": 0.3412,
+      "step": 8520
+    },
+    {
+      "epoch": 1.6798107255520505,
+      "grad_norm": 0.4672872118729301,
+      "learning_rate": 1.2504250234035378e-05,
+      "loss": 0.3238,
+      "step": 8521
+    },
+    {
+      "epoch": 1.6800078864353312,
+      "grad_norm": 0.46960680724931947,
+      "learning_rate": 1.2502749568012655e-05,
+      "loss": 0.3368,
+      "step": 8522
+    },
+    {
+      "epoch": 1.680205047318612,
+      "grad_norm": 0.4892990248131517,
+      "learning_rate": 1.2501248841859358e-05,
+      "loss": 0.3222,
+      "step": 8523
+    },
+    {
+      "epoch": 1.6804022082018928,
+      "grad_norm": 0.9521603633960574,
+      "learning_rate": 1.2499748055611543e-05,
+      "loss": 0.3473,
+      "step": 8524
+    },
+    {
+      "epoch": 1.6805993690851735,
+      "grad_norm": 0.4961229036962787,
+      "learning_rate": 1.2498247209305267e-05,
+      "loss": 0.34,
+      "step": 8525
+    },
+    {
+      "epoch": 1.6807965299684544,
+      "grad_norm": 0.4641387089623987,
+      "learning_rate": 1.2496746302976588e-05,
+      "loss": 0.3262,
+      "step": 8526
+    },
+    {
+      "epoch": 1.680993690851735,
+      "grad_norm": 0.5208046380242989,
+      "learning_rate": 1.2495245336661575e-05,
+      "loss": 0.357,
+      "step": 8527
+    },
+    {
+      "epoch": 1.6811908517350158,
+      "grad_norm": 0.6377815976714571,
+      "learning_rate": 1.2493744310396276e-05,
+      "loss": 0.3446,
+      "step": 8528
+    },
+    {
+      "epoch": 1.6813880126182965,
+      "grad_norm": 0.48508566333163605,
+      "learning_rate": 1.249224322421677e-05,
+      "loss": 0.3265,
+      "step": 8529
+    },
+    {
+      "epoch": 1.6815851735015772,
+      "grad_norm": 0.4640062435177377,
+      "learning_rate": 1.2490742078159107e-05,
+      "loss": 0.3453,
+      "step": 8530
+    },
+    {
+      "epoch": 1.6817823343848581,
+      "grad_norm": 0.48082524719773534,
+      "learning_rate": 1.248924087225936e-05,
+      "loss": 0.3428,
+      "step": 8531
+    },
+    {
+      "epoch": 1.6819794952681388,
+      "grad_norm": 0.46881859443066326,
+      "learning_rate": 1.24877396065536e-05,
+      "loss": 0.3073,
+      "step": 8532
+    },
+    {
+      "epoch": 1.6821766561514195,
+      "grad_norm": 0.4883737738622232,
+      "learning_rate": 1.248623828107789e-05,
+      "loss": 0.3472,
+      "step": 8533
+    },
+    {
+      "epoch": 1.6823738170347005,
+      "grad_norm": 0.45689282889724375,
+      "learning_rate": 1.2484736895868306e-05,
+      "loss": 0.3097,
+      "step": 8534
+    },
+    {
+      "epoch": 1.682570977917981,
+      "grad_norm": 0.47112167138042593,
+      "learning_rate": 1.2483235450960914e-05,
+      "loss": 0.3444,
+      "step": 8535
+    },
+    {
+      "epoch": 1.6827681388012619,
+      "grad_norm": 0.46214308971573514,
+      "learning_rate": 1.2481733946391792e-05,
+      "loss": 0.3189,
+      "step": 8536
+    },
+    {
+      "epoch": 1.6829652996845426,
+      "grad_norm": 0.46005114791319535,
+      "learning_rate": 1.2480232382197013e-05,
+      "loss": 0.3178,
+      "step": 8537
+    },
+    {
+      "epoch": 1.6831624605678233,
+      "grad_norm": 0.4767717118901013,
+      "learning_rate": 1.2478730758412652e-05,
+      "loss": 0.3362,
+      "step": 8538
+    },
+    {
+      "epoch": 1.6833596214511042,
+      "grad_norm": 0.4599960998922992,
+      "learning_rate": 1.247722907507479e-05,
+      "loss": 0.3195,
+      "step": 8539
+    },
+    {
+      "epoch": 1.6835567823343849,
+      "grad_norm": 0.4951705881711061,
+      "learning_rate": 1.2475727332219505e-05,
+      "loss": 0.346,
+      "step": 8540
+    },
+    {
+      "epoch": 1.6837539432176656,
+      "grad_norm": 0.5046721506851792,
+      "learning_rate": 1.2474225529882878e-05,
+      "loss": 0.3255,
+      "step": 8541
+    },
+    {
+      "epoch": 1.6839511041009465,
+      "grad_norm": 0.4829302359712838,
+      "learning_rate": 1.247272366810099e-05,
+      "loss": 0.3429,
+      "step": 8542
+    },
+    {
+      "epoch": 1.684148264984227,
+      "grad_norm": 0.46391312466042783,
+      "learning_rate": 1.2471221746909923e-05,
+      "loss": 0.3325,
+      "step": 8543
+    },
+    {
+      "epoch": 1.684345425867508,
+      "grad_norm": 0.47200457284310665,
+      "learning_rate": 1.2469719766345763e-05,
+      "loss": 0.3388,
+      "step": 8544
+    },
+    {
+      "epoch": 1.6845425867507886,
+      "grad_norm": 0.4746525342051994,
+      "learning_rate": 1.2468217726444595e-05,
+      "loss": 0.3293,
+      "step": 8545
+    },
+    {
+      "epoch": 1.6847397476340693,
+      "grad_norm": 0.523726717599289,
+      "learning_rate": 1.2466715627242514e-05,
+      "loss": 0.3564,
+      "step": 8546
+    },
+    {
+      "epoch": 1.6849369085173502,
+      "grad_norm": 0.47113498305611295,
+      "learning_rate": 1.2465213468775602e-05,
+      "loss": 0.3459,
+      "step": 8547
+    },
+    {
+      "epoch": 1.685134069400631,
+      "grad_norm": 0.4768683631487932,
+      "learning_rate": 1.2463711251079951e-05,
+      "loss": 0.3419,
+      "step": 8548
+    },
+    {
+      "epoch": 1.6853312302839116,
+      "grad_norm": 0.4796081831360892,
+      "learning_rate": 1.2462208974191652e-05,
+      "loss": 0.337,
+      "step": 8549
+    },
+    {
+      "epoch": 1.6855283911671926,
+      "grad_norm": 0.47022166990324243,
+      "learning_rate": 1.24607066381468e-05,
+      "loss": 0.3343,
+      "step": 8550
+    },
+    {
+      "epoch": 1.685725552050473,
+      "grad_norm": 0.47554608946434046,
+      "learning_rate": 1.245920424298149e-05,
+      "loss": 0.3205,
+      "step": 8551
+    },
+    {
+      "epoch": 1.685922712933754,
+      "grad_norm": 0.616505992571845,
+      "learning_rate": 1.2457701788731812e-05,
+      "loss": 0.3296,
+      "step": 8552
+    },
+    {
+      "epoch": 1.6861198738170347,
+      "grad_norm": 0.4878539182332263,
+      "learning_rate": 1.2456199275433878e-05,
+      "loss": 0.3525,
+      "step": 8553
+    },
+    {
+      "epoch": 1.6863170347003154,
+      "grad_norm": 0.4414899321505671,
+      "learning_rate": 1.2454696703123773e-05,
+      "loss": 0.3206,
+      "step": 8554
+    },
+    {
+      "epoch": 1.6865141955835963,
+      "grad_norm": 0.48286881495321865,
+      "learning_rate": 1.2453194071837606e-05,
+      "loss": 0.3238,
+      "step": 8555
+    },
+    {
+      "epoch": 1.686711356466877,
+      "grad_norm": 0.4742179590973439,
+      "learning_rate": 1.2451691381611472e-05,
+      "loss": 0.3197,
+      "step": 8556
+    },
+    {
+      "epoch": 1.6869085173501577,
+      "grad_norm": 0.47834655573536045,
+      "learning_rate": 1.2450188632481484e-05,
+      "loss": 0.3265,
+      "step": 8557
+    },
+    {
+      "epoch": 1.6871056782334386,
+      "grad_norm": 0.4343626858698186,
+      "learning_rate": 1.2448685824483735e-05,
+      "loss": 0.3035,
+      "step": 8558
+    },
+    {
+      "epoch": 1.687302839116719,
+      "grad_norm": 0.48021663657417696,
+      "learning_rate": 1.244718295765434e-05,
+      "loss": 0.3275,
+      "step": 8559
+    },
+    {
+      "epoch": 1.6875,
+      "grad_norm": 0.4680145546040287,
+      "learning_rate": 1.2445680032029403e-05,
+      "loss": 0.3164,
+      "step": 8560
+    },
+    {
+      "epoch": 1.687697160883281,
+      "grad_norm": 0.46324619631494696,
+      "learning_rate": 1.2444177047645036e-05,
+      "loss": 0.3164,
+      "step": 8561
+    },
+    {
+      "epoch": 1.6878943217665614,
+      "grad_norm": 0.4788496971926555,
+      "learning_rate": 1.2442674004537345e-05,
+      "loss": 0.3412,
+      "step": 8562
+    },
+    {
+      "epoch": 1.6880914826498423,
+      "grad_norm": 0.47439352264543966,
+      "learning_rate": 1.2441170902742445e-05,
+      "loss": 0.3272,
+      "step": 8563
+    },
+    {
+      "epoch": 1.688288643533123,
+      "grad_norm": 0.49561553090817284,
+      "learning_rate": 1.2439667742296448e-05,
+      "loss": 0.3302,
+      "step": 8564
+    },
+    {
+      "epoch": 1.6884858044164037,
+      "grad_norm": 0.4800560302995904,
+      "learning_rate": 1.2438164523235467e-05,
+      "loss": 0.3406,
+      "step": 8565
+    },
+    {
+      "epoch": 1.6886829652996846,
+      "grad_norm": 0.49067895256143157,
+      "learning_rate": 1.2436661245595623e-05,
+      "loss": 0.325,
+      "step": 8566
+    },
+    {
+      "epoch": 1.6888801261829653,
+      "grad_norm": 0.47817903605159684,
+      "learning_rate": 1.2435157909413029e-05,
+      "loss": 0.3247,
+      "step": 8567
+    },
+    {
+      "epoch": 1.689077287066246,
+      "grad_norm": 0.48136123183849766,
+      "learning_rate": 1.2433654514723806e-05,
+      "loss": 0.3454,
+      "step": 8568
+    },
+    {
+      "epoch": 1.689274447949527,
+      "grad_norm": 0.5051580981062126,
+      "learning_rate": 1.2432151061564071e-05,
+      "loss": 0.3601,
+      "step": 8569
+    },
+    {
+      "epoch": 1.6894716088328074,
+      "grad_norm": 0.461511250819562,
+      "learning_rate": 1.2430647549969949e-05,
+      "loss": 0.3233,
+      "step": 8570
+    },
+    {
+      "epoch": 1.6896687697160884,
+      "grad_norm": 0.47678716555909567,
+      "learning_rate": 1.2429143979977562e-05,
+      "loss": 0.3427,
+      "step": 8571
+    },
+    {
+      "epoch": 1.689865930599369,
+      "grad_norm": 0.4486388389800133,
+      "learning_rate": 1.2427640351623037e-05,
+      "loss": 0.3183,
+      "step": 8572
+    },
+    {
+      "epoch": 1.6900630914826498,
+      "grad_norm": 0.5299380107151339,
+      "learning_rate": 1.2426136664942495e-05,
+      "loss": 0.3656,
+      "step": 8573
+    },
+    {
+      "epoch": 1.6902602523659307,
+      "grad_norm": 0.46937533374559753,
+      "learning_rate": 1.2424632919972068e-05,
+      "loss": 0.33,
+      "step": 8574
+    },
+    {
+      "epoch": 1.6904574132492114,
+      "grad_norm": 0.4979192584567516,
+      "learning_rate": 1.2423129116747878e-05,
+      "loss": 0.3224,
+      "step": 8575
+    },
+    {
+      "epoch": 1.690654574132492,
+      "grad_norm": 1.0231877748006113,
+      "learning_rate": 1.2421625255306067e-05,
+      "loss": 0.3506,
+      "step": 8576
+    },
+    {
+      "epoch": 1.690851735015773,
+      "grad_norm": 0.47898806414813694,
+      "learning_rate": 1.242012133568275e-05,
+      "loss": 0.3408,
+      "step": 8577
+    },
+    {
+      "epoch": 1.6910488958990535,
+      "grad_norm": 0.47674507248699505,
+      "learning_rate": 1.2418617357914078e-05,
+      "loss": 0.3402,
+      "step": 8578
+    },
+    {
+      "epoch": 1.6912460567823344,
+      "grad_norm": 0.45179630277653227,
+      "learning_rate": 1.2417113322036172e-05,
+      "loss": 0.3154,
+      "step": 8579
+    },
+    {
+      "epoch": 1.6914432176656151,
+      "grad_norm": 0.46052841582017556,
+      "learning_rate": 1.2415609228085171e-05,
+      "loss": 0.3126,
+      "step": 8580
+    },
+    {
+      "epoch": 1.6916403785488958,
+      "grad_norm": 0.44611469096894074,
+      "learning_rate": 1.2414105076097214e-05,
+      "loss": 0.3166,
+      "step": 8581
+    },
+    {
+      "epoch": 1.6918375394321767,
+      "grad_norm": 0.4576223803492312,
+      "learning_rate": 1.241260086610844e-05,
+      "loss": 0.2939,
+      "step": 8582
+    },
+    {
+      "epoch": 1.6920347003154574,
+      "grad_norm": 0.4557877387492244,
+      "learning_rate": 1.2411096598154985e-05,
+      "loss": 0.3268,
+      "step": 8583
+    },
+    {
+      "epoch": 1.6922318611987381,
+      "grad_norm": 0.46572552067045964,
+      "learning_rate": 1.2409592272272995e-05,
+      "loss": 0.351,
+      "step": 8584
+    },
+    {
+      "epoch": 1.692429022082019,
+      "grad_norm": 0.5122610064442623,
+      "learning_rate": 1.2408087888498608e-05,
+      "loss": 0.3605,
+      "step": 8585
+    },
+    {
+      "epoch": 1.6926261829652995,
+      "grad_norm": 0.49403230477515936,
+      "learning_rate": 1.2406583446867972e-05,
+      "loss": 0.3342,
+      "step": 8586
+    },
+    {
+      "epoch": 1.6928233438485805,
+      "grad_norm": 0.5017202216659983,
+      "learning_rate": 1.240507894741723e-05,
+      "loss": 0.3491,
+      "step": 8587
+    },
+    {
+      "epoch": 1.6930205047318612,
+      "grad_norm": 0.4899331141459395,
+      "learning_rate": 1.2403574390182529e-05,
+      "loss": 0.3607,
+      "step": 8588
+    },
+    {
+      "epoch": 1.6932176656151419,
+      "grad_norm": 0.4930643507100292,
+      "learning_rate": 1.2402069775200018e-05,
+      "loss": 0.3346,
+      "step": 8589
+    },
+    {
+      "epoch": 1.6934148264984228,
+      "grad_norm": 0.47713702195641194,
+      "learning_rate": 1.2400565102505846e-05,
+      "loss": 0.3293,
+      "step": 8590
+    },
+    {
+      "epoch": 1.6936119873817035,
+      "grad_norm": 0.4783126509399956,
+      "learning_rate": 1.2399060372136165e-05,
+      "loss": 0.343,
+      "step": 8591
+    },
+    {
+      "epoch": 1.6938091482649842,
+      "grad_norm": 0.48348496039775457,
+      "learning_rate": 1.2397555584127127e-05,
+      "loss": 0.3281,
+      "step": 8592
+    },
+    {
+      "epoch": 1.694006309148265,
+      "grad_norm": 0.5160059605204272,
+      "learning_rate": 1.2396050738514884e-05,
+      "loss": 0.3423,
+      "step": 8593
+    },
+    {
+      "epoch": 1.6942034700315456,
+      "grad_norm": 0.45441215261046347,
+      "learning_rate": 1.2394545835335591e-05,
+      "loss": 0.314,
+      "step": 8594
+    },
+    {
+      "epoch": 1.6944006309148265,
+      "grad_norm": 0.4836399537284616,
+      "learning_rate": 1.239304087462541e-05,
+      "loss": 0.327,
+      "step": 8595
+    },
+    {
+      "epoch": 1.6945977917981072,
+      "grad_norm": 0.46510484102779165,
+      "learning_rate": 1.2391535856420492e-05,
+      "loss": 0.3528,
+      "step": 8596
+    },
+    {
+      "epoch": 1.694794952681388,
+      "grad_norm": 0.4430475484441221,
+      "learning_rate": 1.2390030780757e-05,
+      "loss": 0.31,
+      "step": 8597
+    },
+    {
+      "epoch": 1.6949921135646688,
+      "grad_norm": 0.4674652671591372,
+      "learning_rate": 1.2388525647671092e-05,
+      "loss": 0.3148,
+      "step": 8598
+    },
+    {
+      "epoch": 1.6951892744479495,
+      "grad_norm": 0.5054421188784337,
+      "learning_rate": 1.2387020457198937e-05,
+      "loss": 0.3795,
+      "step": 8599
+    },
+    {
+      "epoch": 1.6953864353312302,
+      "grad_norm": 0.4930491695671332,
+      "learning_rate": 1.238551520937669e-05,
+      "loss": 0.3515,
+      "step": 8600
+    },
+    {
+      "epoch": 1.6955835962145112,
+      "grad_norm": 0.464472471942234,
+      "learning_rate": 1.2384009904240517e-05,
+      "loss": 0.3178,
+      "step": 8601
+    },
+    {
+      "epoch": 1.6957807570977916,
+      "grad_norm": 0.46922888632843546,
+      "learning_rate": 1.238250454182659e-05,
+      "loss": 0.3421,
+      "step": 8602
+    },
+    {
+      "epoch": 1.6959779179810726,
+      "grad_norm": 0.5117715606317115,
+      "learning_rate": 1.238099912217107e-05,
+      "loss": 0.3657,
+      "step": 8603
+    },
+    {
+      "epoch": 1.6961750788643533,
+      "grad_norm": 0.45140287026735587,
+      "learning_rate": 1.237949364531013e-05,
+      "loss": 0.3224,
+      "step": 8604
+    },
+    {
+      "epoch": 1.696372239747634,
+      "grad_norm": 0.49574769053532836,
+      "learning_rate": 1.2377988111279937e-05,
+      "loss": 0.3462,
+      "step": 8605
+    },
+    {
+      "epoch": 1.6965694006309149,
+      "grad_norm": 0.46230478380221307,
+      "learning_rate": 1.2376482520116666e-05,
+      "loss": 0.3373,
+      "step": 8606
+    },
+    {
+      "epoch": 1.6967665615141956,
+      "grad_norm": 0.5059151436193712,
+      "learning_rate": 1.237497687185649e-05,
+      "loss": 0.3359,
+      "step": 8607
+    },
+    {
+      "epoch": 1.6969637223974763,
+      "grad_norm": 0.48234255694436523,
+      "learning_rate": 1.237347116653558e-05,
+      "loss": 0.3479,
+      "step": 8608
+    },
+    {
+      "epoch": 1.6971608832807572,
+      "grad_norm": 0.4396592143423356,
+      "learning_rate": 1.2371965404190116e-05,
+      "loss": 0.3067,
+      "step": 8609
+    },
+    {
+      "epoch": 1.697358044164038,
+      "grad_norm": 0.49065585389386196,
+      "learning_rate": 1.2370459584856271e-05,
+      "loss": 0.3445,
+      "step": 8610
+    },
+    {
+      "epoch": 1.6975552050473186,
+      "grad_norm": 0.49051628604620695,
+      "learning_rate": 1.2368953708570226e-05,
+      "loss": 0.3402,
+      "step": 8611
+    },
+    {
+      "epoch": 1.6977523659305995,
+      "grad_norm": 0.5254870988260015,
+      "learning_rate": 1.2367447775368163e-05,
+      "loss": 0.3454,
+      "step": 8612
+    },
+    {
+      "epoch": 1.69794952681388,
+      "grad_norm": 0.4562980109810731,
+      "learning_rate": 1.2365941785286258e-05,
+      "loss": 0.3074,
+      "step": 8613
+    },
+    {
+      "epoch": 1.698146687697161,
+      "grad_norm": 0.5678148667222066,
+      "learning_rate": 1.2364435738360696e-05,
+      "loss": 0.3396,
+      "step": 8614
+    },
+    {
+      "epoch": 1.6983438485804416,
+      "grad_norm": 0.4911585049022961,
+      "learning_rate": 1.2362929634627663e-05,
+      "loss": 0.3346,
+      "step": 8615
+    },
+    {
+      "epoch": 1.6985410094637223,
+      "grad_norm": 0.4681297072028885,
+      "learning_rate": 1.2361423474123343e-05,
+      "loss": 0.3082,
+      "step": 8616
+    },
+    {
+      "epoch": 1.6987381703470033,
+      "grad_norm": 0.5174736460144449,
+      "learning_rate": 1.235991725688392e-05,
+      "loss": 0.3814,
+      "step": 8617
+    },
+    {
+      "epoch": 1.698935331230284,
+      "grad_norm": 0.46929685353298345,
+      "learning_rate": 1.2358410982945586e-05,
+      "loss": 0.3355,
+      "step": 8618
+    },
+    {
+      "epoch": 1.6991324921135647,
+      "grad_norm": 0.4922370451892625,
+      "learning_rate": 1.2356904652344528e-05,
+      "loss": 0.3268,
+      "step": 8619
+    },
+    {
+      "epoch": 1.6993296529968456,
+      "grad_norm": 0.4596598424424867,
+      "learning_rate": 1.2355398265116937e-05,
+      "loss": 0.3293,
+      "step": 8620
+    },
+    {
+      "epoch": 1.699526813880126,
+      "grad_norm": 0.5001218713531277,
+      "learning_rate": 1.235389182129901e-05,
+      "loss": 0.3669,
+      "step": 8621
+    },
+    {
+      "epoch": 1.699723974763407,
+      "grad_norm": 0.4896954521807538,
+      "learning_rate": 1.2352385320926929e-05,
+      "loss": 0.3477,
+      "step": 8622
+    },
+    {
+      "epoch": 1.6999211356466877,
+      "grad_norm": 0.49063376422553284,
+      "learning_rate": 1.2350878764036904e-05,
+      "loss": 0.372,
+      "step": 8623
+    },
+    {
+      "epoch": 1.7001182965299684,
+      "grad_norm": 0.46066188918467005,
+      "learning_rate": 1.2349372150665117e-05,
+      "loss": 0.3251,
+      "step": 8624
+    },
+    {
+      "epoch": 1.7003154574132493,
+      "grad_norm": 0.48615185578096526,
+      "learning_rate": 1.2347865480847778e-05,
+      "loss": 0.3356,
+      "step": 8625
+    },
+    {
+      "epoch": 1.70051261829653,
+      "grad_norm": 0.46350483669286957,
+      "learning_rate": 1.2346358754621078e-05,
+      "loss": 0.3367,
+      "step": 8626
+    },
+    {
+      "epoch": 1.7007097791798107,
+      "grad_norm": 0.4534163321107968,
+      "learning_rate": 1.2344851972021219e-05,
+      "loss": 0.3297,
+      "step": 8627
+    },
+    {
+      "epoch": 1.7009069400630916,
+      "grad_norm": 0.46024896781400154,
+      "learning_rate": 1.2343345133084403e-05,
+      "loss": 0.3161,
+      "step": 8628
+    },
+    {
+      "epoch": 1.701104100946372,
+      "grad_norm": 0.4597911141388873,
+      "learning_rate": 1.2341838237846833e-05,
+      "loss": 0.3251,
+      "step": 8629
+    },
+    {
+      "epoch": 1.701301261829653,
+      "grad_norm": 0.47755657347184777,
+      "learning_rate": 1.2340331286344713e-05,
+      "loss": 0.3439,
+      "step": 8630
+    },
+    {
+      "epoch": 1.7014984227129337,
+      "grad_norm": 0.4595385546420116,
+      "learning_rate": 1.233882427861425e-05,
+      "loss": 0.3236,
+      "step": 8631
+    },
+    {
+      "epoch": 1.7016955835962144,
+      "grad_norm": 0.47878136809716293,
+      "learning_rate": 1.233731721469165e-05,
+      "loss": 0.3182,
+      "step": 8632
+    },
+    {
+      "epoch": 1.7018927444794953,
+      "grad_norm": 0.46749188234424455,
+      "learning_rate": 1.2335810094613123e-05,
+      "loss": 0.3346,
+      "step": 8633
+    },
+    {
+      "epoch": 1.702089905362776,
+      "grad_norm": 0.4620951279690006,
+      "learning_rate": 1.2334302918414875e-05,
+      "loss": 0.324,
+      "step": 8634
+    },
+    {
+      "epoch": 1.7022870662460567,
+      "grad_norm": 0.489807733516085,
+      "learning_rate": 1.2332795686133121e-05,
+      "loss": 0.3252,
+      "step": 8635
+    },
+    {
+      "epoch": 1.7024842271293377,
+      "grad_norm": 0.4804183436816162,
+      "learning_rate": 1.2331288397804072e-05,
+      "loss": 0.3598,
+      "step": 8636
+    },
+    {
+      "epoch": 1.7026813880126181,
+      "grad_norm": 0.5026867518086552,
+      "learning_rate": 1.2329781053463944e-05,
+      "loss": 0.36,
+      "step": 8637
+    },
+    {
+      "epoch": 1.702878548895899,
+      "grad_norm": 0.4759113828318395,
+      "learning_rate": 1.2328273653148945e-05,
+      "loss": 0.3411,
+      "step": 8638
+    },
+    {
+      "epoch": 1.7030757097791798,
+      "grad_norm": 0.48031817244407715,
+      "learning_rate": 1.2326766196895301e-05,
+      "loss": 0.3496,
+      "step": 8639
+    },
+    {
+      "epoch": 1.7032728706624605,
+      "grad_norm": 0.47282002859928735,
+      "learning_rate": 1.2325258684739223e-05,
+      "loss": 0.3215,
+      "step": 8640
+    },
+    {
+      "epoch": 1.7034700315457414,
+      "grad_norm": 3.191557994037556,
+      "learning_rate": 1.2323751116716932e-05,
+      "loss": 0.3359,
+      "step": 8641
+    },
+    {
+      "epoch": 1.703667192429022,
+      "grad_norm": 0.5967525800282695,
+      "learning_rate": 1.2322243492864651e-05,
+      "loss": 0.3469,
+      "step": 8642
+    },
+    {
+      "epoch": 1.7038643533123028,
+      "grad_norm": 0.4551107597567062,
+      "learning_rate": 1.2320735813218599e-05,
+      "loss": 0.3291,
+      "step": 8643
+    },
+    {
+      "epoch": 1.7040615141955837,
+      "grad_norm": 0.4620889896670438,
+      "learning_rate": 1.2319228077815001e-05,
+      "loss": 0.33,
+      "step": 8644
+    },
+    {
+      "epoch": 1.7042586750788642,
+      "grad_norm": 0.4853258209863878,
+      "learning_rate": 1.231772028669008e-05,
+      "loss": 0.3261,
+      "step": 8645
+    },
+    {
+      "epoch": 1.7044558359621451,
+      "grad_norm": 0.48368889919551544,
+      "learning_rate": 1.2316212439880065e-05,
+      "loss": 0.354,
+      "step": 8646
+    },
+    {
+      "epoch": 1.7046529968454258,
+      "grad_norm": 0.5542920344424767,
+      "learning_rate": 1.2314704537421177e-05,
+      "loss": 0.3408,
+      "step": 8647
+    },
+    {
+      "epoch": 1.7048501577287065,
+      "grad_norm": 0.47647351780117475,
+      "learning_rate": 1.2313196579349648e-05,
+      "loss": 0.3274,
+      "step": 8648
+    },
+    {
+      "epoch": 1.7050473186119874,
+      "grad_norm": 0.46934633413559235,
+      "learning_rate": 1.2311688565701711e-05,
+      "loss": 0.3418,
+      "step": 8649
+    },
+    {
+      "epoch": 1.7052444794952681,
+      "grad_norm": 0.48432919465079566,
+      "learning_rate": 1.2310180496513595e-05,
+      "loss": 0.3193,
+      "step": 8650
+    },
+    {
+      "epoch": 1.7054416403785488,
+      "grad_norm": 0.4661153258537868,
+      "learning_rate": 1.2308672371821532e-05,
+      "loss": 0.3321,
+      "step": 8651
+    },
+    {
+      "epoch": 1.7056388012618298,
+      "grad_norm": 0.4826086437714458,
+      "learning_rate": 1.2307164191661756e-05,
+      "loss": 0.338,
+      "step": 8652
+    },
+    {
+      "epoch": 1.7058359621451105,
+      "grad_norm": 0.4637877696238651,
+      "learning_rate": 1.2305655956070504e-05,
+      "loss": 0.3245,
+      "step": 8653
+    },
+    {
+      "epoch": 1.7060331230283912,
+      "grad_norm": 0.4640340301452536,
+      "learning_rate": 1.2304147665084007e-05,
+      "loss": 0.3176,
+      "step": 8654
+    },
+    {
+      "epoch": 1.706230283911672,
+      "grad_norm": 0.49786805036849485,
+      "learning_rate": 1.230263931873851e-05,
+      "loss": 0.3619,
+      "step": 8655
+    },
+    {
+      "epoch": 1.7064274447949526,
+      "grad_norm": 0.5747013253276394,
+      "learning_rate": 1.2301130917070245e-05,
+      "loss": 0.3151,
+      "step": 8656
+    },
+    {
+      "epoch": 1.7066246056782335,
+      "grad_norm": 0.5511325174270215,
+      "learning_rate": 1.2299622460115461e-05,
+      "loss": 0.36,
+      "step": 8657
+    },
+    {
+      "epoch": 1.7068217665615142,
+      "grad_norm": 0.5072499961226788,
+      "learning_rate": 1.2298113947910393e-05,
+      "loss": 0.3349,
+      "step": 8658
+    },
+    {
+      "epoch": 1.7070189274447949,
+      "grad_norm": 0.48091952766514195,
+      "learning_rate": 1.2296605380491288e-05,
+      "loss": 0.3491,
+      "step": 8659
+    },
+    {
+      "epoch": 1.7072160883280758,
+      "grad_norm": 0.4798754544496718,
+      "learning_rate": 1.2295096757894389e-05,
+      "loss": 0.3471,
+      "step": 8660
+    },
+    {
+      "epoch": 1.7074132492113565,
+      "grad_norm": 0.6307603113857772,
+      "learning_rate": 1.2293588080155943e-05,
+      "loss": 0.3413,
+      "step": 8661
+    },
+    {
+      "epoch": 1.7076104100946372,
+      "grad_norm": 0.4779048637501002,
+      "learning_rate": 1.2292079347312194e-05,
+      "loss": 0.3337,
+      "step": 8662
+    },
+    {
+      "epoch": 1.7078075709779181,
+      "grad_norm": 0.5247659710582877,
+      "learning_rate": 1.2290570559399395e-05,
+      "loss": 0.37,
+      "step": 8663
+    },
+    {
+      "epoch": 1.7080047318611986,
+      "grad_norm": 0.4617016132383226,
+      "learning_rate": 1.2289061716453795e-05,
+      "loss": 0.3223,
+      "step": 8664
+    },
+    {
+      "epoch": 1.7082018927444795,
+      "grad_norm": 0.47698713563356776,
+      "learning_rate": 1.2287552818511641e-05,
+      "loss": 0.3247,
+      "step": 8665
+    },
+    {
+      "epoch": 1.7083990536277602,
+      "grad_norm": 0.46484751275332564,
+      "learning_rate": 1.2286043865609188e-05,
+      "loss": 0.3319,
+      "step": 8666
+    },
+    {
+      "epoch": 1.708596214511041,
+      "grad_norm": 0.473092049126572,
+      "learning_rate": 1.2284534857782694e-05,
+      "loss": 0.3515,
+      "step": 8667
+    },
+    {
+      "epoch": 1.7087933753943219,
+      "grad_norm": 0.49275200785968454,
+      "learning_rate": 1.2283025795068407e-05,
+      "loss": 0.3342,
+      "step": 8668
+    },
+    {
+      "epoch": 1.7089905362776026,
+      "grad_norm": 0.5026513536426332,
+      "learning_rate": 1.2281516677502586e-05,
+      "loss": 0.3411,
+      "step": 8669
+    },
+    {
+      "epoch": 1.7091876971608833,
+      "grad_norm": 0.46869592654591175,
+      "learning_rate": 1.2280007505121491e-05,
+      "loss": 0.332,
+      "step": 8670
+    },
+    {
+      "epoch": 1.7093848580441642,
+      "grad_norm": 0.5099270759799205,
+      "learning_rate": 1.2278498277961377e-05,
+      "loss": 0.3446,
+      "step": 8671
+    },
+    {
+      "epoch": 1.7095820189274447,
+      "grad_norm": 2.050432417214253,
+      "learning_rate": 1.2276988996058511e-05,
+      "loss": 0.4219,
+      "step": 8672
+    },
+    {
+      "epoch": 1.7097791798107256,
+      "grad_norm": 0.4725441540525176,
+      "learning_rate": 1.227547965944915e-05,
+      "loss": 0.3432,
+      "step": 8673
+    },
+    {
+      "epoch": 1.7099763406940063,
+      "grad_norm": 0.4993708255622457,
+      "learning_rate": 1.227397026816956e-05,
+      "loss": 0.3513,
+      "step": 8674
+    },
+    {
+      "epoch": 1.710173501577287,
+      "grad_norm": 0.4887912737158361,
+      "learning_rate": 1.2272460822255996e-05,
+      "loss": 0.322,
+      "step": 8675
+    },
+    {
+      "epoch": 1.710370662460568,
+      "grad_norm": 0.46993623199235296,
+      "learning_rate": 1.2270951321744736e-05,
+      "loss": 0.3599,
+      "step": 8676
+    },
+    {
+      "epoch": 1.7105678233438486,
+      "grad_norm": 0.4643011227098681,
+      "learning_rate": 1.2269441766672042e-05,
+      "loss": 0.3012,
+      "step": 8677
+    },
+    {
+      "epoch": 1.7107649842271293,
+      "grad_norm": 0.5506263319435446,
+      "learning_rate": 1.2267932157074178e-05,
+      "loss": 0.3621,
+      "step": 8678
+    },
+    {
+      "epoch": 1.7109621451104102,
+      "grad_norm": 0.4919225236599655,
+      "learning_rate": 1.2266422492987423e-05,
+      "loss": 0.338,
+      "step": 8679
+    },
+    {
+      "epoch": 1.7111593059936907,
+      "grad_norm": 0.4896517148433233,
+      "learning_rate": 1.2264912774448037e-05,
+      "loss": 0.3527,
+      "step": 8680
+    },
+    {
+      "epoch": 1.7113564668769716,
+      "grad_norm": 0.48839509275100296,
+      "learning_rate": 1.22634030014923e-05,
+      "loss": 0.3464,
+      "step": 8681
+    },
+    {
+      "epoch": 1.7115536277602523,
+      "grad_norm": 0.4721548932185601,
+      "learning_rate": 1.2261893174156485e-05,
+      "loss": 0.3297,
+      "step": 8682
+    },
+    {
+      "epoch": 1.711750788643533,
+      "grad_norm": 0.4911637333121339,
+      "learning_rate": 1.2260383292476862e-05,
+      "loss": 0.3285,
+      "step": 8683
+    },
+    {
+      "epoch": 1.711947949526814,
+      "grad_norm": 0.459378104968867,
+      "learning_rate": 1.2258873356489713e-05,
+      "loss": 0.3292,
+      "step": 8684
+    },
+    {
+      "epoch": 1.7121451104100947,
+      "grad_norm": 0.4643594445198744,
+      "learning_rate": 1.2257363366231311e-05,
+      "loss": 0.3354,
+      "step": 8685
+    },
+    {
+      "epoch": 1.7123422712933754,
+      "grad_norm": 0.4875758149219096,
+      "learning_rate": 1.2255853321737935e-05,
+      "loss": 0.3461,
+      "step": 8686
+    },
+    {
+      "epoch": 1.7125394321766563,
+      "grad_norm": 0.48234734526715584,
+      "learning_rate": 1.225434322304587e-05,
+      "loss": 0.3432,
+      "step": 8687
+    },
+    {
+      "epoch": 1.7127365930599368,
+      "grad_norm": 0.4765321853764992,
+      "learning_rate": 1.2252833070191388e-05,
+      "loss": 0.335,
+      "step": 8688
+    },
+    {
+      "epoch": 1.7129337539432177,
+      "grad_norm": 0.4922172475680278,
+      "learning_rate": 1.2251322863210785e-05,
+      "loss": 0.3532,
+      "step": 8689
+    },
+    {
+      "epoch": 1.7131309148264984,
+      "grad_norm": 0.49132588894060586,
+      "learning_rate": 1.224981260214033e-05,
+      "loss": 0.3471,
+      "step": 8690
+    },
+    {
+      "epoch": 1.713328075709779,
+      "grad_norm": 0.4726091865577661,
+      "learning_rate": 1.2248302287016321e-05,
+      "loss": 0.3148,
+      "step": 8691
+    },
+    {
+      "epoch": 1.71352523659306,
+      "grad_norm": 0.44365990430314184,
+      "learning_rate": 1.2246791917875034e-05,
+      "loss": 0.2972,
+      "step": 8692
+    },
+    {
+      "epoch": 1.7137223974763407,
+      "grad_norm": 0.5487336291279901,
+      "learning_rate": 1.2245281494752765e-05,
+      "loss": 0.3856,
+      "step": 8693
+    },
+    {
+      "epoch": 1.7139195583596214,
+      "grad_norm": 0.4920272311780737,
+      "learning_rate": 1.2243771017685797e-05,
+      "loss": 0.3366,
+      "step": 8694
+    },
+    {
+      "epoch": 1.7141167192429023,
+      "grad_norm": 0.5015165415124818,
+      "learning_rate": 1.2242260486710427e-05,
+      "loss": 0.3512,
+      "step": 8695
+    },
+    {
+      "epoch": 1.714313880126183,
+      "grad_norm": 0.5239711415916627,
+      "learning_rate": 1.224074990186294e-05,
+      "loss": 0.381,
+      "step": 8696
+    },
+    {
+      "epoch": 1.7145110410094637,
+      "grad_norm": 0.4568955886255038,
+      "learning_rate": 1.2239239263179635e-05,
+      "loss": 0.3209,
+      "step": 8697
+    },
+    {
+      "epoch": 1.7147082018927446,
+      "grad_norm": 0.49060093715398756,
+      "learning_rate": 1.2237728570696801e-05,
+      "loss": 0.3439,
+      "step": 8698
+    },
+    {
+      "epoch": 1.7149053627760251,
+      "grad_norm": 0.47975117071356893,
+      "learning_rate": 1.2236217824450739e-05,
+      "loss": 0.344,
+      "step": 8699
+    },
+    {
+      "epoch": 1.715102523659306,
+      "grad_norm": 0.5258786604080068,
+      "learning_rate": 1.2234707024477742e-05,
+      "loss": 0.3348,
+      "step": 8700
+    },
+    {
+      "epoch": 1.7152996845425867,
+      "grad_norm": 0.48137432039450917,
+      "learning_rate": 1.2233196170814105e-05,
+      "loss": 0.3525,
+      "step": 8701
+    },
+    {
+      "epoch": 1.7154968454258674,
+      "grad_norm": 0.5014397647686436,
+      "learning_rate": 1.2231685263496137e-05,
+      "loss": 0.3438,
+      "step": 8702
+    },
+    {
+      "epoch": 1.7156940063091484,
+      "grad_norm": 0.4708182769624843,
+      "learning_rate": 1.2230174302560132e-05,
+      "loss": 0.3342,
+      "step": 8703
+    },
+    {
+      "epoch": 1.715891167192429,
+      "grad_norm": 0.512195415565872,
+      "learning_rate": 1.2228663288042392e-05,
+      "loss": 0.3553,
+      "step": 8704
+    },
+    {
+      "epoch": 1.7160883280757098,
+      "grad_norm": 0.4788767330139541,
+      "learning_rate": 1.2227152219979224e-05,
+      "loss": 0.3536,
+      "step": 8705
+    },
+    {
+      "epoch": 1.7162854889589907,
+      "grad_norm": 0.4625920435496106,
+      "learning_rate": 1.2225641098406928e-05,
+      "loss": 0.2878,
+      "step": 8706
+    },
+    {
+      "epoch": 1.7164826498422712,
+      "grad_norm": 0.5070101532510752,
+      "learning_rate": 1.2224129923361813e-05,
+      "loss": 0.3453,
+      "step": 8707
+    },
+    {
+      "epoch": 1.716679810725552,
+      "grad_norm": 0.4988378522423907,
+      "learning_rate": 1.2222618694880187e-05,
+      "loss": 0.3697,
+      "step": 8708
+    },
+    {
+      "epoch": 1.7168769716088328,
+      "grad_norm": 0.5033574544758712,
+      "learning_rate": 1.2221107412998352e-05,
+      "loss": 0.3118,
+      "step": 8709
+    },
+    {
+      "epoch": 1.7170741324921135,
+      "grad_norm": 0.4890282116096399,
+      "learning_rate": 1.2219596077752629e-05,
+      "loss": 0.325,
+      "step": 8710
+    },
+    {
+      "epoch": 1.7172712933753944,
+      "grad_norm": 0.4760094568747338,
+      "learning_rate": 1.221808468917932e-05,
+      "loss": 0.3281,
+      "step": 8711
+    },
+    {
+      "epoch": 1.7174684542586751,
+      "grad_norm": 0.48764166466975545,
+      "learning_rate": 1.221657324731474e-05,
+      "loss": 0.3331,
+      "step": 8712
+    },
+    {
+      "epoch": 1.7176656151419558,
+      "grad_norm": 0.4517243985708021,
+      "learning_rate": 1.22150617521952e-05,
+      "loss": 0.34,
+      "step": 8713
+    },
+    {
+      "epoch": 1.7178627760252367,
+      "grad_norm": 0.5139727951740557,
+      "learning_rate": 1.2213550203857025e-05,
+      "loss": 0.3511,
+      "step": 8714
+    },
+    {
+      "epoch": 1.7180599369085172,
+      "grad_norm": 0.45072979492101384,
+      "learning_rate": 1.2212038602336518e-05,
+      "loss": 0.3228,
+      "step": 8715
+    },
+    {
+      "epoch": 1.7182570977917981,
+      "grad_norm": 0.4844004632125416,
+      "learning_rate": 1.2210526947670003e-05,
+      "loss": 0.3361,
+      "step": 8716
+    },
+    {
+      "epoch": 1.7184542586750788,
+      "grad_norm": 0.48156386430034853,
+      "learning_rate": 1.22090152398938e-05,
+      "loss": 0.3579,
+      "step": 8717
+    },
+    {
+      "epoch": 1.7186514195583595,
+      "grad_norm": 0.4801222110814163,
+      "learning_rate": 1.2207503479044224e-05,
+      "loss": 0.3327,
+      "step": 8718
+    },
+    {
+      "epoch": 1.7188485804416405,
+      "grad_norm": 0.4633028361510336,
+      "learning_rate": 1.2205991665157604e-05,
+      "loss": 0.3361,
+      "step": 8719
+    },
+    {
+      "epoch": 1.7190457413249212,
+      "grad_norm": 0.4954639943287471,
+      "learning_rate": 1.2204479798270252e-05,
+      "loss": 0.3555,
+      "step": 8720
+    },
+    {
+      "epoch": 1.7192429022082019,
+      "grad_norm": 0.5071392340252393,
+      "learning_rate": 1.2202967878418504e-05,
+      "loss": 0.3311,
+      "step": 8721
+    },
+    {
+      "epoch": 1.7194400630914828,
+      "grad_norm": 0.4924112369717431,
+      "learning_rate": 1.2201455905638673e-05,
+      "loss": 0.3343,
+      "step": 8722
+    },
+    {
+      "epoch": 1.7196372239747633,
+      "grad_norm": 0.5281013648224745,
+      "learning_rate": 1.2199943879967092e-05,
+      "loss": 0.3661,
+      "step": 8723
+    },
+    {
+      "epoch": 1.7198343848580442,
+      "grad_norm": 0.4188521660470496,
+      "learning_rate": 1.2198431801440087e-05,
+      "loss": 0.2792,
+      "step": 8724
+    },
+    {
+      "epoch": 1.7200315457413249,
+      "grad_norm": 0.514554223490542,
+      "learning_rate": 1.2196919670093989e-05,
+      "loss": 0.351,
+      "step": 8725
+    },
+    {
+      "epoch": 1.7202287066246056,
+      "grad_norm": 0.4773669903370585,
+      "learning_rate": 1.2195407485965129e-05,
+      "loss": 0.3371,
+      "step": 8726
+    },
+    {
+      "epoch": 1.7204258675078865,
+      "grad_norm": 0.4486097522414558,
+      "learning_rate": 1.2193895249089833e-05,
+      "loss": 0.3221,
+      "step": 8727
+    },
+    {
+      "epoch": 1.7206230283911672,
+      "grad_norm": 0.4514574538208784,
+      "learning_rate": 1.2192382959504438e-05,
+      "loss": 0.3146,
+      "step": 8728
+    },
+    {
+      "epoch": 1.720820189274448,
+      "grad_norm": 0.4906075691840977,
+      "learning_rate": 1.2190870617245279e-05,
+      "loss": 0.3452,
+      "step": 8729
+    },
+    {
+      "epoch": 1.7210173501577288,
+      "grad_norm": 0.4728367195594932,
+      "learning_rate": 1.2189358222348685e-05,
+      "loss": 0.34,
+      "step": 8730
+    },
+    {
+      "epoch": 1.7212145110410093,
+      "grad_norm": 0.47426809412562415,
+      "learning_rate": 1.2187845774850999e-05,
+      "loss": 0.2974,
+      "step": 8731
+    },
+    {
+      "epoch": 1.7214116719242902,
+      "grad_norm": 0.4676329793425546,
+      "learning_rate": 1.2186333274788558e-05,
+      "loss": 0.3375,
+      "step": 8732
+    },
+    {
+      "epoch": 1.721608832807571,
+      "grad_norm": 0.467140693302685,
+      "learning_rate": 1.2184820722197696e-05,
+      "loss": 0.3224,
+      "step": 8733
+    },
+    {
+      "epoch": 1.7218059936908516,
+      "grad_norm": 0.492299563007178,
+      "learning_rate": 1.2183308117114759e-05,
+      "loss": 0.3401,
+      "step": 8734
+    },
+    {
+      "epoch": 1.7220031545741326,
+      "grad_norm": 0.47469376128492763,
+      "learning_rate": 1.2181795459576085e-05,
+      "loss": 0.3676,
+      "step": 8735
+    },
+    {
+      "epoch": 1.7222003154574133,
+      "grad_norm": 0.482130929800137,
+      "learning_rate": 1.2180282749618017e-05,
+      "loss": 0.332,
+      "step": 8736
+    },
+    {
+      "epoch": 1.722397476340694,
+      "grad_norm": 0.5031498208900129,
+      "learning_rate": 1.2178769987276902e-05,
+      "loss": 0.3525,
+      "step": 8737
+    },
+    {
+      "epoch": 1.7225946372239749,
+      "grad_norm": 0.48826854933895736,
+      "learning_rate": 1.2177257172589086e-05,
+      "loss": 0.3464,
+      "step": 8738
+    },
+    {
+      "epoch": 1.7227917981072554,
+      "grad_norm": 0.475972420949733,
+      "learning_rate": 1.2175744305590907e-05,
+      "loss": 0.3193,
+      "step": 8739
+    },
+    {
+      "epoch": 1.7229889589905363,
+      "grad_norm": 0.49949539660408465,
+      "learning_rate": 1.2174231386318724e-05,
+      "loss": 0.3567,
+      "step": 8740
+    },
+    {
+      "epoch": 1.723186119873817,
+      "grad_norm": 0.49325281815052363,
+      "learning_rate": 1.2172718414808877e-05,
+      "loss": 0.3581,
+      "step": 8741
+    },
+    {
+      "epoch": 1.7233832807570977,
+      "grad_norm": 0.4777530691912601,
+      "learning_rate": 1.2171205391097724e-05,
+      "loss": 0.3277,
+      "step": 8742
+    },
+    {
+      "epoch": 1.7235804416403786,
+      "grad_norm": 0.47301218354355706,
+      "learning_rate": 1.216969231522161e-05,
+      "loss": 0.3346,
+      "step": 8743
+    },
+    {
+      "epoch": 1.7237776025236593,
+      "grad_norm": 0.6201310559734784,
+      "learning_rate": 1.2168179187216893e-05,
+      "loss": 0.3247,
+      "step": 8744
+    },
+    {
+      "epoch": 1.72397476340694,
+      "grad_norm": 0.4669955469618345,
+      "learning_rate": 1.2166666007119925e-05,
+      "loss": 0.3382,
+      "step": 8745
+    },
+    {
+      "epoch": 1.724171924290221,
+      "grad_norm": 0.47156930793491,
+      "learning_rate": 1.2165152774967061e-05,
+      "loss": 0.348,
+      "step": 8746
+    },
+    {
+      "epoch": 1.7243690851735016,
+      "grad_norm": 0.5146898131222495,
+      "learning_rate": 1.2163639490794659e-05,
+      "loss": 0.3423,
+      "step": 8747
+    },
+    {
+      "epoch": 1.7245662460567823,
+      "grad_norm": 0.4748624471052609,
+      "learning_rate": 1.2162126154639073e-05,
+      "loss": 0.3329,
+      "step": 8748
+    },
+    {
+      "epoch": 1.7247634069400632,
+      "grad_norm": 0.5041545803945943,
+      "learning_rate": 1.2160612766536668e-05,
+      "loss": 0.3429,
+      "step": 8749
+    },
+    {
+      "epoch": 1.7249605678233437,
+      "grad_norm": 0.4560606048941914,
+      "learning_rate": 1.21590993265238e-05,
+      "loss": 0.3329,
+      "step": 8750
+    },
+    {
+      "epoch": 1.7251577287066246,
+      "grad_norm": 0.4920455686195489,
+      "learning_rate": 1.2157585834636834e-05,
+      "loss": 0.3369,
+      "step": 8751
+    },
+    {
+      "epoch": 1.7253548895899053,
+      "grad_norm": 0.4817930597858489,
+      "learning_rate": 1.2156072290912126e-05,
+      "loss": 0.3123,
+      "step": 8752
+    },
+    {
+      "epoch": 1.725552050473186,
+      "grad_norm": 0.4933196910433103,
+      "learning_rate": 1.2154558695386049e-05,
+      "loss": 0.3625,
+      "step": 8753
+    },
+    {
+      "epoch": 1.725749211356467,
+      "grad_norm": 0.43787576840008613,
+      "learning_rate": 1.2153045048094963e-05,
+      "loss": 0.3305,
+      "step": 8754
+    },
+    {
+      "epoch": 1.7259463722397477,
+      "grad_norm": 0.4646297207054528,
+      "learning_rate": 1.2151531349075236e-05,
+      "loss": 0.3276,
+      "step": 8755
+    },
+    {
+      "epoch": 1.7261435331230284,
+      "grad_norm": 0.459618597411763,
+      "learning_rate": 1.2150017598363236e-05,
+      "loss": 0.3039,
+      "step": 8756
+    },
+    {
+      "epoch": 1.7263406940063093,
+      "grad_norm": 0.48044932107882354,
+      "learning_rate": 1.2148503795995332e-05,
+      "loss": 0.3209,
+      "step": 8757
+    },
+    {
+      "epoch": 1.7265378548895898,
+      "grad_norm": 0.48647578029445776,
+      "learning_rate": 1.2146989942007891e-05,
+      "loss": 0.3308,
+      "step": 8758
+    },
+    {
+      "epoch": 1.7267350157728707,
+      "grad_norm": 0.4518527350354567,
+      "learning_rate": 1.2145476036437294e-05,
+      "loss": 0.3233,
+      "step": 8759
+    },
+    {
+      "epoch": 1.7269321766561514,
+      "grad_norm": 0.48074502143691444,
+      "learning_rate": 1.21439620793199e-05,
+      "loss": 0.3538,
+      "step": 8760
+    },
+    {
+      "epoch": 1.727129337539432,
+      "grad_norm": 0.4662467836506099,
+      "learning_rate": 1.2142448070692096e-05,
+      "loss": 0.3411,
+      "step": 8761
+    },
+    {
+      "epoch": 1.727326498422713,
+      "grad_norm": 0.5184102450636691,
+      "learning_rate": 1.2140934010590249e-05,
+      "loss": 0.3508,
+      "step": 8762
+    },
+    {
+      "epoch": 1.7275236593059937,
+      "grad_norm": 0.54097531724488,
+      "learning_rate": 1.213941989905074e-05,
+      "loss": 0.3291,
+      "step": 8763
+    },
+    {
+      "epoch": 1.7277208201892744,
+      "grad_norm": 0.4815838736915963,
+      "learning_rate": 1.2137905736109946e-05,
+      "loss": 0.3279,
+      "step": 8764
+    },
+    {
+      "epoch": 1.7279179810725553,
+      "grad_norm": 0.47938770142115866,
+      "learning_rate": 1.213639152180424e-05,
+      "loss": 0.3222,
+      "step": 8765
+    },
+    {
+      "epoch": 1.7281151419558358,
+      "grad_norm": 0.44462625843755743,
+      "learning_rate": 1.2134877256170012e-05,
+      "loss": 0.3056,
+      "step": 8766
+    },
+    {
+      "epoch": 1.7283123028391167,
+      "grad_norm": 0.48782541427748793,
+      "learning_rate": 1.2133362939243638e-05,
+      "loss": 0.3382,
+      "step": 8767
+    },
+    {
+      "epoch": 1.7285094637223974,
+      "grad_norm": 0.47686312944629256,
+      "learning_rate": 1.2131848571061501e-05,
+      "loss": 0.3336,
+      "step": 8768
+    },
+    {
+      "epoch": 1.7287066246056781,
+      "grad_norm": 0.4972392941654052,
+      "learning_rate": 1.2130334151659987e-05,
+      "loss": 0.3457,
+      "step": 8769
+    },
+    {
+      "epoch": 1.728903785488959,
+      "grad_norm": 0.44658596263964356,
+      "learning_rate": 1.2128819681075476e-05,
+      "loss": 0.3113,
+      "step": 8770
+    },
+    {
+      "epoch": 1.7291009463722398,
+      "grad_norm": 0.6132762769001334,
+      "learning_rate": 1.2127305159344358e-05,
+      "loss": 0.2762,
+      "step": 8771
+    },
+    {
+      "epoch": 1.7292981072555205,
+      "grad_norm": 0.5184653807356759,
+      "learning_rate": 1.2125790586503024e-05,
+      "loss": 0.3589,
+      "step": 8772
+    },
+    {
+      "epoch": 1.7294952681388014,
+      "grad_norm": 0.48887264222267973,
+      "learning_rate": 1.2124275962587857e-05,
+      "loss": 0.3458,
+      "step": 8773
+    },
+    {
+      "epoch": 1.7296924290220819,
+      "grad_norm": 0.48006157666774124,
+      "learning_rate": 1.212276128763525e-05,
+      "loss": 0.3261,
+      "step": 8774
+    },
+    {
+      "epoch": 1.7298895899053628,
+      "grad_norm": 0.4705936894192923,
+      "learning_rate": 1.2121246561681592e-05,
+      "loss": 0.3318,
+      "step": 8775
+    },
+    {
+      "epoch": 1.7300867507886435,
+      "grad_norm": 0.4680969494433612,
+      "learning_rate": 1.2119731784763278e-05,
+      "loss": 0.3278,
+      "step": 8776
+    },
+    {
+      "epoch": 1.7302839116719242,
+      "grad_norm": 0.5204916184909986,
+      "learning_rate": 1.21182169569167e-05,
+      "loss": 0.3461,
+      "step": 8777
+    },
+    {
+      "epoch": 1.7304810725552051,
+      "grad_norm": 0.4682144834045916,
+      "learning_rate": 1.2116702078178255e-05,
+      "loss": 0.3115,
+      "step": 8778
+    },
+    {
+      "epoch": 1.7306782334384858,
+      "grad_norm": 0.4684797179301652,
+      "learning_rate": 1.2115187148584338e-05,
+      "loss": 0.3297,
+      "step": 8779
+    },
+    {
+      "epoch": 1.7308753943217665,
+      "grad_norm": 0.5149570001266986,
+      "learning_rate": 1.2113672168171347e-05,
+      "loss": 0.3411,
+      "step": 8780
+    },
+    {
+      "epoch": 1.7310725552050474,
+      "grad_norm": 0.4918293028719042,
+      "learning_rate": 1.2112157136975678e-05,
+      "loss": 0.3302,
+      "step": 8781
+    },
+    {
+      "epoch": 1.731269716088328,
+      "grad_norm": 0.4680392185456968,
+      "learning_rate": 1.2110642055033737e-05,
+      "loss": 0.313,
+      "step": 8782
+    },
+    {
+      "epoch": 1.7314668769716088,
+      "grad_norm": 0.4675694840904993,
+      "learning_rate": 1.2109126922381917e-05,
+      "loss": 0.3383,
+      "step": 8783
+    },
+    {
+      "epoch": 1.7316640378548895,
+      "grad_norm": 0.5051346681197235,
+      "learning_rate": 1.2107611739056624e-05,
+      "loss": 0.3679,
+      "step": 8784
+    },
+    {
+      "epoch": 1.7318611987381702,
+      "grad_norm": 0.48965432777518686,
+      "learning_rate": 1.2106096505094264e-05,
+      "loss": 0.3412,
+      "step": 8785
+    },
+    {
+      "epoch": 1.7320583596214512,
+      "grad_norm": 0.5179854164790161,
+      "learning_rate": 1.2104581220531237e-05,
+      "loss": 0.3659,
+      "step": 8786
+    },
+    {
+      "epoch": 1.7322555205047319,
+      "grad_norm": 0.4814569942652689,
+      "learning_rate": 1.2103065885403955e-05,
+      "loss": 0.3369,
+      "step": 8787
+    },
+    {
+      "epoch": 1.7324526813880126,
+      "grad_norm": 0.4885561961779997,
+      "learning_rate": 1.2101550499748818e-05,
+      "loss": 0.3446,
+      "step": 8788
+    },
+    {
+      "epoch": 1.7326498422712935,
+      "grad_norm": 0.49161198978038273,
+      "learning_rate": 1.210003506360224e-05,
+      "loss": 0.3177,
+      "step": 8789
+    },
+    {
+      "epoch": 1.7328470031545742,
+      "grad_norm": 0.48634200842172337,
+      "learning_rate": 1.2098519577000627e-05,
+      "loss": 0.3553,
+      "step": 8790
+    },
+    {
+      "epoch": 1.7330441640378549,
+      "grad_norm": 0.48325106568491305,
+      "learning_rate": 1.2097004039980391e-05,
+      "loss": 0.3519,
+      "step": 8791
+    },
+    {
+      "epoch": 1.7332413249211358,
+      "grad_norm": 0.4641044710550502,
+      "learning_rate": 1.2095488452577946e-05,
+      "loss": 0.3123,
+      "step": 8792
+    },
+    {
+      "epoch": 1.7334384858044163,
+      "grad_norm": 0.47358093899817655,
+      "learning_rate": 1.2093972814829701e-05,
+      "loss": 0.3162,
+      "step": 8793
+    },
+    {
+      "epoch": 1.7336356466876972,
+      "grad_norm": 0.4913256828106219,
+      "learning_rate": 1.2092457126772074e-05,
+      "loss": 0.3531,
+      "step": 8794
+    },
+    {
+      "epoch": 1.733832807570978,
+      "grad_norm": 0.4912311679309798,
+      "learning_rate": 1.2090941388441482e-05,
+      "loss": 0.3359,
+      "step": 8795
+    },
+    {
+      "epoch": 1.7340299684542586,
+      "grad_norm": 0.47304417855315983,
+      "learning_rate": 1.2089425599874335e-05,
+      "loss": 0.3022,
+      "step": 8796
+    },
+    {
+      "epoch": 1.7342271293375395,
+      "grad_norm": 0.4672250909616135,
+      "learning_rate": 1.208790976110706e-05,
+      "loss": 0.3027,
+      "step": 8797
+    },
+    {
+      "epoch": 1.7344242902208202,
+      "grad_norm": 0.4647836250132962,
+      "learning_rate": 1.2086393872176067e-05,
+      "loss": 0.3144,
+      "step": 8798
+    },
+    {
+      "epoch": 1.734621451104101,
+      "grad_norm": 0.5010956167290279,
+      "learning_rate": 1.2084877933117784e-05,
+      "loss": 0.3641,
+      "step": 8799
+    },
+    {
+      "epoch": 1.7348186119873819,
+      "grad_norm": 0.46610623517280136,
+      "learning_rate": 1.2083361943968628e-05,
+      "loss": 0.3291,
+      "step": 8800
+    },
+    {
+      "epoch": 1.7350157728706623,
+      "grad_norm": 0.5069304515479853,
+      "learning_rate": 1.2081845904765026e-05,
+      "loss": 0.3766,
+      "step": 8801
+    },
+    {
+      "epoch": 1.7352129337539433,
+      "grad_norm": 0.4418811498806394,
+      "learning_rate": 1.2080329815543398e-05,
+      "loss": 0.3065,
+      "step": 8802
+    },
+    {
+      "epoch": 1.735410094637224,
+      "grad_norm": 0.5417087924566732,
+      "learning_rate": 1.2078813676340171e-05,
+      "loss": 0.322,
+      "step": 8803
+    },
+    {
+      "epoch": 1.7356072555205047,
+      "grad_norm": 0.4885878208354627,
+      "learning_rate": 1.2077297487191771e-05,
+      "loss": 0.3166,
+      "step": 8804
+    },
+    {
+      "epoch": 1.7358044164037856,
+      "grad_norm": 0.46819488040628704,
+      "learning_rate": 1.2075781248134624e-05,
+      "loss": 0.3308,
+      "step": 8805
+    },
+    {
+      "epoch": 1.7360015772870663,
+      "grad_norm": 0.4725168433278604,
+      "learning_rate": 1.2074264959205167e-05,
+      "loss": 0.3248,
+      "step": 8806
+    },
+    {
+      "epoch": 1.736198738170347,
+      "grad_norm": 0.4685757443962958,
+      "learning_rate": 1.2072748620439816e-05,
+      "loss": 0.3162,
+      "step": 8807
+    },
+    {
+      "epoch": 1.736395899053628,
+      "grad_norm": 0.4944655388806196,
+      "learning_rate": 1.2071232231875017e-05,
+      "loss": 0.3465,
+      "step": 8808
+    },
+    {
+      "epoch": 1.7365930599369084,
+      "grad_norm": 0.48192011814447067,
+      "learning_rate": 1.2069715793547192e-05,
+      "loss": 0.3312,
+      "step": 8809
+    },
+    {
+      "epoch": 1.7367902208201893,
+      "grad_norm": 0.5000553460824155,
+      "learning_rate": 1.2068199305492781e-05,
+      "loss": 0.3439,
+      "step": 8810
+    },
+    {
+      "epoch": 1.73698738170347,
+      "grad_norm": 0.47422303429265467,
+      "learning_rate": 1.2066682767748212e-05,
+      "loss": 0.3176,
+      "step": 8811
+    },
+    {
+      "epoch": 1.7371845425867507,
+      "grad_norm": 0.449422917568121,
+      "learning_rate": 1.2065166180349928e-05,
+      "loss": 0.3241,
+      "step": 8812
+    },
+    {
+      "epoch": 1.7373817034700316,
+      "grad_norm": 0.49595270985884926,
+      "learning_rate": 1.2063649543334364e-05,
+      "loss": 0.3461,
+      "step": 8813
+    },
+    {
+      "epoch": 1.7375788643533123,
+      "grad_norm": 0.44630470527387806,
+      "learning_rate": 1.2062132856737958e-05,
+      "loss": 0.3099,
+      "step": 8814
+    },
+    {
+      "epoch": 1.737776025236593,
+      "grad_norm": 2.0412021740875836,
+      "learning_rate": 1.2060616120597149e-05,
+      "loss": 0.3194,
+      "step": 8815
+    },
+    {
+      "epoch": 1.737973186119874,
+      "grad_norm": 0.47472434744895176,
+      "learning_rate": 1.2059099334948376e-05,
+      "loss": 0.3325,
+      "step": 8816
+    },
+    {
+      "epoch": 1.7381703470031544,
+      "grad_norm": 0.481225491809552,
+      "learning_rate": 1.2057582499828086e-05,
+      "loss": 0.336,
+      "step": 8817
+    },
+    {
+      "epoch": 1.7383675078864353,
+      "grad_norm": 0.44821446660142955,
+      "learning_rate": 1.205606561527272e-05,
+      "loss": 0.3134,
+      "step": 8818
+    },
+    {
+      "epoch": 1.738564668769716,
+      "grad_norm": 0.4813063073960643,
+      "learning_rate": 1.205454868131872e-05,
+      "loss": 0.3246,
+      "step": 8819
+    },
+    {
+      "epoch": 1.7387618296529967,
+      "grad_norm": 0.4753107258555097,
+      "learning_rate": 1.2053031698002533e-05,
+      "loss": 0.333,
+      "step": 8820
+    },
+    {
+      "epoch": 1.7389589905362777,
+      "grad_norm": 0.5613865154278634,
+      "learning_rate": 1.2051514665360606e-05,
+      "loss": 0.3334,
+      "step": 8821
+    },
+    {
+      "epoch": 1.7391561514195584,
+      "grad_norm": 0.5166325707314708,
+      "learning_rate": 1.2049997583429389e-05,
+      "loss": 0.3388,
+      "step": 8822
+    },
+    {
+      "epoch": 1.739353312302839,
+      "grad_norm": 0.4689529203647351,
+      "learning_rate": 1.2048480452245328e-05,
+      "loss": 0.3409,
+      "step": 8823
+    },
+    {
+      "epoch": 1.73955047318612,
+      "grad_norm": 0.5124805778270197,
+      "learning_rate": 1.2046963271844876e-05,
+      "loss": 0.3589,
+      "step": 8824
+    },
+    {
+      "epoch": 1.7397476340694005,
+      "grad_norm": 0.4555862428808988,
+      "learning_rate": 1.2045446042264482e-05,
+      "loss": 0.3218,
+      "step": 8825
+    },
+    {
+      "epoch": 1.7399447949526814,
+      "grad_norm": 0.46880230449353444,
+      "learning_rate": 1.2043928763540598e-05,
+      "loss": 0.3204,
+      "step": 8826
+    },
+    {
+      "epoch": 1.740141955835962,
+      "grad_norm": 0.4661584032901464,
+      "learning_rate": 1.2042411435709683e-05,
+      "loss": 0.3008,
+      "step": 8827
+    },
+    {
+      "epoch": 1.7403391167192428,
+      "grad_norm": 0.49560394022565896,
+      "learning_rate": 1.2040894058808183e-05,
+      "loss": 0.3304,
+      "step": 8828
+    },
+    {
+      "epoch": 1.7405362776025237,
+      "grad_norm": 0.477044636888031,
+      "learning_rate": 1.2039376632872565e-05,
+      "loss": 0.3348,
+      "step": 8829
+    },
+    {
+      "epoch": 1.7407334384858044,
+      "grad_norm": 0.4768628146082518,
+      "learning_rate": 1.2037859157939278e-05,
+      "loss": 0.3517,
+      "step": 8830
+    },
+    {
+      "epoch": 1.7409305993690851,
+      "grad_norm": 0.46710532512756137,
+      "learning_rate": 1.2036341634044785e-05,
+      "loss": 0.3378,
+      "step": 8831
+    },
+    {
+      "epoch": 1.741127760252366,
+      "grad_norm": 0.4838108367315246,
+      "learning_rate": 1.2034824061225545e-05,
+      "loss": 0.3403,
+      "step": 8832
+    },
+    {
+      "epoch": 1.7413249211356467,
+      "grad_norm": 0.4814792464848545,
+      "learning_rate": 1.2033306439518017e-05,
+      "loss": 0.328,
+      "step": 8833
+    },
+    {
+      "epoch": 1.7415220820189274,
+      "grad_norm": 0.4927086846869189,
+      "learning_rate": 1.2031788768958666e-05,
+      "loss": 0.3458,
+      "step": 8834
+    },
+    {
+      "epoch": 1.7417192429022084,
+      "grad_norm": 0.4617442660897952,
+      "learning_rate": 1.203027104958395e-05,
+      "loss": 0.3377,
+      "step": 8835
+    },
+    {
+      "epoch": 1.7419164037854888,
+      "grad_norm": 0.48638088204825836,
+      "learning_rate": 1.2028753281430343e-05,
+      "loss": 0.3269,
+      "step": 8836
+    },
+    {
+      "epoch": 1.7421135646687698,
+      "grad_norm": 0.5049199749667796,
+      "learning_rate": 1.20272354645343e-05,
+      "loss": 0.3766,
+      "step": 8837
+    },
+    {
+      "epoch": 1.7423107255520505,
+      "grad_norm": 0.48188659695790537,
+      "learning_rate": 1.2025717598932293e-05,
+      "loss": 0.3414,
+      "step": 8838
+    },
+    {
+      "epoch": 1.7425078864353312,
+      "grad_norm": 0.4738832139863155,
+      "learning_rate": 1.2024199684660792e-05,
+      "loss": 0.3253,
+      "step": 8839
+    },
+    {
+      "epoch": 1.742705047318612,
+      "grad_norm": 0.477503317970411,
+      "learning_rate": 1.202268172175626e-05,
+      "loss": 0.3344,
+      "step": 8840
+    },
+    {
+      "epoch": 1.7429022082018928,
+      "grad_norm": 0.5304814544412956,
+      "learning_rate": 1.2021163710255173e-05,
+      "loss": 0.334,
+      "step": 8841
+    },
+    {
+      "epoch": 1.7430993690851735,
+      "grad_norm": 0.508157775015286,
+      "learning_rate": 1.2019645650193999e-05,
+      "loss": 0.3512,
+      "step": 8842
+    },
+    {
+      "epoch": 1.7432965299684544,
+      "grad_norm": 0.5017879226850557,
+      "learning_rate": 1.2018127541609212e-05,
+      "loss": 0.333,
+      "step": 8843
+    },
+    {
+      "epoch": 1.743493690851735,
+      "grad_norm": 0.4867632085277509,
+      "learning_rate": 1.2016609384537287e-05,
+      "loss": 0.3366,
+      "step": 8844
+    },
+    {
+      "epoch": 1.7436908517350158,
+      "grad_norm": 0.5105248897104497,
+      "learning_rate": 1.2015091179014696e-05,
+      "loss": 0.3556,
+      "step": 8845
+    },
+    {
+      "epoch": 1.7438880126182965,
+      "grad_norm": 0.4608424473916468,
+      "learning_rate": 1.2013572925077919e-05,
+      "loss": 0.3371,
+      "step": 8846
+    },
+    {
+      "epoch": 1.7440851735015772,
+      "grad_norm": 0.45038670620879856,
+      "learning_rate": 1.2012054622763425e-05,
+      "loss": 0.3052,
+      "step": 8847
+    },
+    {
+      "epoch": 1.7442823343848581,
+      "grad_norm": 0.5081471358396383,
+      "learning_rate": 1.2010536272107706e-05,
+      "loss": 0.3455,
+      "step": 8848
+    },
+    {
+      "epoch": 1.7444794952681388,
+      "grad_norm": 0.4644321538998626,
+      "learning_rate": 1.200901787314723e-05,
+      "loss": 0.3091,
+      "step": 8849
+    },
+    {
+      "epoch": 1.7446766561514195,
+      "grad_norm": 0.48115698751245345,
+      "learning_rate": 1.2007499425918483e-05,
+      "loss": 0.3482,
+      "step": 8850
+    },
+    {
+      "epoch": 1.7448738170347005,
+      "grad_norm": 0.5228016067341045,
+      "learning_rate": 1.2005980930457946e-05,
+      "loss": 0.3275,
+      "step": 8851
+    },
+    {
+      "epoch": 1.745070977917981,
+      "grad_norm": 0.583985733994483,
+      "learning_rate": 1.2004462386802098e-05,
+      "loss": 0.3648,
+      "step": 8852
+    },
+    {
+      "epoch": 1.7452681388012619,
+      "grad_norm": 0.45939749735947777,
+      "learning_rate": 1.2002943794987432e-05,
+      "loss": 0.3389,
+      "step": 8853
+    },
+    {
+      "epoch": 1.7454652996845426,
+      "grad_norm": 0.4648960544103313,
+      "learning_rate": 1.2001425155050423e-05,
+      "loss": 0.3412,
+      "step": 8854
+    },
+    {
+      "epoch": 1.7456624605678233,
+      "grad_norm": 0.4793461007710896,
+      "learning_rate": 1.1999906467027568e-05,
+      "loss": 0.3385,
+      "step": 8855
+    },
+    {
+      "epoch": 1.7458596214511042,
+      "grad_norm": 0.4611557554227105,
+      "learning_rate": 1.1998387730955345e-05,
+      "loss": 0.2932,
+      "step": 8856
+    },
+    {
+      "epoch": 1.7460567823343849,
+      "grad_norm": 0.4679923259603527,
+      "learning_rate": 1.1996868946870252e-05,
+      "loss": 0.3347,
+      "step": 8857
+    },
+    {
+      "epoch": 1.7462539432176656,
+      "grad_norm": 0.47215734774680346,
+      "learning_rate": 1.1995350114808772e-05,
+      "loss": 0.3437,
+      "step": 8858
+    },
+    {
+      "epoch": 1.7464511041009465,
+      "grad_norm": 0.4648472195023899,
+      "learning_rate": 1.1993831234807401e-05,
+      "loss": 0.3623,
+      "step": 8859
+    },
+    {
+      "epoch": 1.746648264984227,
+      "grad_norm": 0.473482682420203,
+      "learning_rate": 1.1992312306902625e-05,
+      "loss": 0.3372,
+      "step": 8860
+    },
+    {
+      "epoch": 1.746845425867508,
+      "grad_norm": 0.46094418222077305,
+      "learning_rate": 1.1990793331130944e-05,
+      "loss": 0.3221,
+      "step": 8861
+    },
+    {
+      "epoch": 1.7470425867507886,
+      "grad_norm": 0.5016161341201105,
+      "learning_rate": 1.1989274307528848e-05,
+      "loss": 0.3539,
+      "step": 8862
+    },
+    {
+      "epoch": 1.7472397476340693,
+      "grad_norm": 0.49099642673007926,
+      "learning_rate": 1.1987755236132839e-05,
+      "loss": 0.3462,
+      "step": 8863
+    },
+    {
+      "epoch": 1.7474369085173502,
+      "grad_norm": 0.48487743130133104,
+      "learning_rate": 1.1986236116979406e-05,
+      "loss": 0.3585,
+      "step": 8864
+    },
+    {
+      "epoch": 1.747634069400631,
+      "grad_norm": 0.46908185404595404,
+      "learning_rate": 1.1984716950105054e-05,
+      "loss": 0.3306,
+      "step": 8865
+    },
+    {
+      "epoch": 1.7478312302839116,
+      "grad_norm": 0.4832688212778634,
+      "learning_rate": 1.1983197735546275e-05,
+      "loss": 0.3631,
+      "step": 8866
+    },
+    {
+      "epoch": 1.7480283911671926,
+      "grad_norm": 0.4602457165492578,
+      "learning_rate": 1.1981678473339576e-05,
+      "loss": 0.3071,
+      "step": 8867
+    },
+    {
+      "epoch": 1.748225552050473,
+      "grad_norm": 0.43528249927314083,
+      "learning_rate": 1.1980159163521454e-05,
+      "loss": 0.3135,
+      "step": 8868
+    },
+    {
+      "epoch": 1.748422712933754,
+      "grad_norm": 0.4695790802517424,
+      "learning_rate": 1.1978639806128416e-05,
+      "loss": 0.331,
+      "step": 8869
+    },
+    {
+      "epoch": 1.7486198738170347,
+      "grad_norm": 0.5280131284138565,
+      "learning_rate": 1.1977120401196963e-05,
+      "loss": 0.3353,
+      "step": 8870
+    },
+    {
+      "epoch": 1.7488170347003154,
+      "grad_norm": 0.4681473144779894,
+      "learning_rate": 1.1975600948763597e-05,
+      "loss": 0.3286,
+      "step": 8871
+    },
+    {
+      "epoch": 1.7490141955835963,
+      "grad_norm": 0.4604810167753171,
+      "learning_rate": 1.197408144886483e-05,
+      "loss": 0.3244,
+      "step": 8872
+    },
+    {
+      "epoch": 1.749211356466877,
+      "grad_norm": 0.46975125282307023,
+      "learning_rate": 1.1972561901537164e-05,
+      "loss": 0.3175,
+      "step": 8873
+    },
+    {
+      "epoch": 1.7494085173501577,
+      "grad_norm": 0.4932348572031116,
+      "learning_rate": 1.1971042306817113e-05,
+      "loss": 0.3538,
+      "step": 8874
+    },
+    {
+      "epoch": 1.7496056782334386,
+      "grad_norm": 0.46969073205555756,
+      "learning_rate": 1.196952266474118e-05,
+      "loss": 0.3453,
+      "step": 8875
+    },
+    {
+      "epoch": 1.749802839116719,
+      "grad_norm": 0.4808934210094641,
+      "learning_rate": 1.1968002975345882e-05,
+      "loss": 0.3415,
+      "step": 8876
+    },
+    {
+      "epoch": 1.749802839116719,
+      "eval_loss": 0.42366865277290344,
+      "eval_runtime": 344.3756,
+      "eval_samples_per_second": 23.608,
+      "eval_steps_per_second": 1.478,
+      "step": 8876
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.48060930083956127,
+      "learning_rate": 1.1966483238667725e-05,
+      "loss": 0.3601,
+      "step": 8877
+    },
+    {
+      "epoch": 1.750197160883281,
+      "grad_norm": 0.4821573663894791,
+      "learning_rate": 1.1964963454743228e-05,
+      "loss": 0.3615,
+      "step": 8878
+    },
+    {
+      "epoch": 1.7503943217665614,
+      "grad_norm": 0.47251836827858207,
+      "learning_rate": 1.1963443623608897e-05,
+      "loss": 0.3372,
+      "step": 8879
+    },
+    {
+      "epoch": 1.7505914826498423,
+      "grad_norm": 0.47137428118391156,
+      "learning_rate": 1.1961923745301256e-05,
+      "loss": 0.3314,
+      "step": 8880
+    },
+    {
+      "epoch": 1.750788643533123,
+      "grad_norm": 0.4957152153915973,
+      "learning_rate": 1.1960403819856815e-05,
+      "loss": 0.3537,
+      "step": 8881
+    },
+    {
+      "epoch": 1.7509858044164037,
+      "grad_norm": 0.4470537419100556,
+      "learning_rate": 1.1958883847312092e-05,
+      "loss": 0.3231,
+      "step": 8882
+    },
+    {
+      "epoch": 1.7511829652996846,
+      "grad_norm": 0.4608113402256386,
+      "learning_rate": 1.1957363827703612e-05,
+      "loss": 0.3211,
+      "step": 8883
+    },
+    {
+      "epoch": 1.7513801261829653,
+      "grad_norm": 0.49969096886092734,
+      "learning_rate": 1.1955843761067886e-05,
+      "loss": 0.337,
+      "step": 8884
+    },
+    {
+      "epoch": 1.751577287066246,
+      "grad_norm": 0.480041688896454,
+      "learning_rate": 1.1954323647441439e-05,
+      "loss": 0.3579,
+      "step": 8885
+    },
+    {
+      "epoch": 1.751774447949527,
+      "grad_norm": 0.45142752291944904,
+      "learning_rate": 1.1952803486860794e-05,
+      "loss": 0.314,
+      "step": 8886
+    },
+    {
+      "epoch": 1.7519716088328074,
+      "grad_norm": 0.48553207420875816,
+      "learning_rate": 1.1951283279362471e-05,
+      "loss": 0.3485,
+      "step": 8887
+    },
+    {
+      "epoch": 1.7521687697160884,
+      "grad_norm": 0.4724742952583676,
+      "learning_rate": 1.1949763024982997e-05,
+      "loss": 0.3496,
+      "step": 8888
+    },
+    {
+      "epoch": 1.752365930599369,
+      "grad_norm": 0.4854477143262649,
+      "learning_rate": 1.1948242723758896e-05,
+      "loss": 0.3421,
+      "step": 8889
+    },
+    {
+      "epoch": 1.7525630914826498,
+      "grad_norm": 0.46735796360977727,
+      "learning_rate": 1.1946722375726694e-05,
+      "loss": 0.3375,
+      "step": 8890
+    },
+    {
+      "epoch": 1.7527602523659307,
+      "grad_norm": 0.4637010683614967,
+      "learning_rate": 1.194520198092292e-05,
+      "loss": 0.3038,
+      "step": 8891
+    },
+    {
+      "epoch": 1.7529574132492114,
+      "grad_norm": 0.4793650345380348,
+      "learning_rate": 1.1943681539384103e-05,
+      "loss": 0.3252,
+      "step": 8892
+    },
+    {
+      "epoch": 1.753154574132492,
+      "grad_norm": 0.4636919118333324,
+      "learning_rate": 1.194216105114677e-05,
+      "loss": 0.3256,
+      "step": 8893
+    },
+    {
+      "epoch": 1.753351735015773,
+      "grad_norm": 0.47895800206523675,
+      "learning_rate": 1.194064051624745e-05,
+      "loss": 0.325,
+      "step": 8894
+    },
+    {
+      "epoch": 1.7535488958990535,
+      "grad_norm": 0.46796297495549594,
+      "learning_rate": 1.1939119934722685e-05,
+      "loss": 0.322,
+      "step": 8895
+    },
+    {
+      "epoch": 1.7537460567823344,
+      "grad_norm": 0.46995032074065873,
+      "learning_rate": 1.1937599306609e-05,
+      "loss": 0.35,
+      "step": 8896
+    },
+    {
+      "epoch": 1.7539432176656151,
+      "grad_norm": 55.908718330694654,
+      "learning_rate": 1.193607863194293e-05,
+      "loss": 0.5092,
+      "step": 8897
+    },
+    {
+      "epoch": 1.7541403785488958,
+      "grad_norm": 0.5121025623411513,
+      "learning_rate": 1.1934557910761013e-05,
+      "loss": 0.3565,
+      "step": 8898
+    },
+    {
+      "epoch": 1.7543375394321767,
+      "grad_norm": 0.45082554071248027,
+      "learning_rate": 1.1933037143099786e-05,
+      "loss": 0.3044,
+      "step": 8899
+    },
+    {
+      "epoch": 1.7545347003154574,
+      "grad_norm": 0.47611001136680525,
+      "learning_rate": 1.1931516328995782e-05,
+      "loss": 0.3135,
+      "step": 8900
+    },
+    {
+      "epoch": 1.7547318611987381,
+      "grad_norm": 0.4768767652243685,
+      "learning_rate": 1.1929995468485545e-05,
+      "loss": 0.3442,
+      "step": 8901
+    },
+    {
+      "epoch": 1.754929022082019,
+      "grad_norm": 0.49669976262352766,
+      "learning_rate": 1.1928474561605612e-05,
+      "loss": 0.3118,
+      "step": 8902
+    },
+    {
+      "epoch": 1.7551261829652995,
+      "grad_norm": 0.4900205596977599,
+      "learning_rate": 1.1926953608392522e-05,
+      "loss": 0.3165,
+      "step": 8903
+    },
+    {
+      "epoch": 1.7553233438485805,
+      "grad_norm": 0.48731950619097536,
+      "learning_rate": 1.1925432608882826e-05,
+      "loss": 0.3524,
+      "step": 8904
+    },
+    {
+      "epoch": 1.7555205047318612,
+      "grad_norm": 0.5192299995573778,
+      "learning_rate": 1.1923911563113053e-05,
+      "loss": 0.3761,
+      "step": 8905
+    },
+    {
+      "epoch": 1.7557176656151419,
+      "grad_norm": 0.4698410536685136,
+      "learning_rate": 1.1922390471119763e-05,
+      "loss": 0.3271,
+      "step": 8906
+    },
+    {
+      "epoch": 1.7559148264984228,
+      "grad_norm": 0.4891636860035787,
+      "learning_rate": 1.1920869332939488e-05,
+      "loss": 0.3466,
+      "step": 8907
+    },
+    {
+      "epoch": 1.7561119873817035,
+      "grad_norm": 0.49905335282315777,
+      "learning_rate": 1.1919348148608782e-05,
+      "loss": 0.3497,
+      "step": 8908
+    },
+    {
+      "epoch": 1.7563091482649842,
+      "grad_norm": 0.5085891428975882,
+      "learning_rate": 1.1917826918164193e-05,
+      "loss": 0.3384,
+      "step": 8909
+    },
+    {
+      "epoch": 1.756506309148265,
+      "grad_norm": 0.49833646762619443,
+      "learning_rate": 1.1916305641642265e-05,
+      "loss": 0.3471,
+      "step": 8910
+    },
+    {
+      "epoch": 1.7567034700315456,
+      "grad_norm": 0.49886395851113163,
+      "learning_rate": 1.1914784319079554e-05,
+      "loss": 0.3834,
+      "step": 8911
+    },
+    {
+      "epoch": 1.7569006309148265,
+      "grad_norm": 0.5085736701155745,
+      "learning_rate": 1.1913262950512605e-05,
+      "loss": 0.348,
+      "step": 8912
+    },
+    {
+      "epoch": 1.7570977917981072,
+      "grad_norm": 0.5032507524362276,
+      "learning_rate": 1.1911741535977972e-05,
+      "loss": 0.3574,
+      "step": 8913
+    },
+    {
+      "epoch": 1.757294952681388,
+      "grad_norm": 0.48298477571501336,
+      "learning_rate": 1.1910220075512213e-05,
+      "loss": 0.318,
+      "step": 8914
+    },
+    {
+      "epoch": 1.7574921135646688,
+      "grad_norm": 0.48667269740164176,
+      "learning_rate": 1.1908698569151877e-05,
+      "loss": 0.3329,
+      "step": 8915
+    },
+    {
+      "epoch": 1.7576892744479495,
+      "grad_norm": 0.47116779180865714,
+      "learning_rate": 1.190717701693352e-05,
+      "loss": 0.3365,
+      "step": 8916
+    },
+    {
+      "epoch": 1.7578864353312302,
+      "grad_norm": 0.4997514621039943,
+      "learning_rate": 1.19056554188937e-05,
+      "loss": 0.3235,
+      "step": 8917
+    },
+    {
+      "epoch": 1.7580835962145112,
+      "grad_norm": 0.48309653739189073,
+      "learning_rate": 1.1904133775068974e-05,
+      "loss": 0.3161,
+      "step": 8918
+    },
+    {
+      "epoch": 1.7582807570977916,
+      "grad_norm": 0.5210440531213502,
+      "learning_rate": 1.1902612085495902e-05,
+      "loss": 0.3631,
+      "step": 8919
+    },
+    {
+      "epoch": 1.7584779179810726,
+      "grad_norm": 0.4921227572826234,
+      "learning_rate": 1.1901090350211037e-05,
+      "loss": 0.3294,
+      "step": 8920
+    },
+    {
+      "epoch": 1.7586750788643533,
+      "grad_norm": 0.49727990636910413,
+      "learning_rate": 1.1899568569250951e-05,
+      "loss": 0.3416,
+      "step": 8921
+    },
+    {
+      "epoch": 1.758872239747634,
+      "grad_norm": 0.5053613903983315,
+      "learning_rate": 1.1898046742652196e-05,
+      "loss": 0.3138,
+      "step": 8922
+    },
+    {
+      "epoch": 1.7590694006309149,
+      "grad_norm": 0.45734876191997076,
+      "learning_rate": 1.1896524870451344e-05,
+      "loss": 0.3403,
+      "step": 8923
+    },
+    {
+      "epoch": 1.7592665615141956,
+      "grad_norm": 0.43524997755300077,
+      "learning_rate": 1.1895002952684952e-05,
+      "loss": 0.3275,
+      "step": 8924
+    },
+    {
+      "epoch": 1.7594637223974763,
+      "grad_norm": 0.46379750213806303,
+      "learning_rate": 1.189348098938959e-05,
+      "loss": 0.3241,
+      "step": 8925
+    },
+    {
+      "epoch": 1.7596608832807572,
+      "grad_norm": 0.5193732674368089,
+      "learning_rate": 1.1891958980601819e-05,
+      "loss": 0.3229,
+      "step": 8926
+    },
+    {
+      "epoch": 1.759858044164038,
+      "grad_norm": 0.5374560365743167,
+      "learning_rate": 1.1890436926358214e-05,
+      "loss": 0.3549,
+      "step": 8927
+    },
+    {
+      "epoch": 1.7600552050473186,
+      "grad_norm": 0.5510070896034697,
+      "learning_rate": 1.1888914826695336e-05,
+      "loss": 0.3427,
+      "step": 8928
+    },
+    {
+      "epoch": 1.7602523659305995,
+      "grad_norm": 0.500101497587192,
+      "learning_rate": 1.1887392681649761e-05,
+      "loss": 0.3519,
+      "step": 8929
+    },
+    {
+      "epoch": 1.76044952681388,
+      "grad_norm": 0.5025929287914618,
+      "learning_rate": 1.1885870491258054e-05,
+      "loss": 0.3415,
+      "step": 8930
+    },
+    {
+      "epoch": 1.760646687697161,
+      "grad_norm": 0.4493713242328926,
+      "learning_rate": 1.1884348255556793e-05,
+      "loss": 0.327,
+      "step": 8931
+    },
+    {
+      "epoch": 1.7608438485804416,
+      "grad_norm": 0.49432274508122315,
+      "learning_rate": 1.1882825974582546e-05,
+      "loss": 0.3382,
+      "step": 8932
+    },
+    {
+      "epoch": 1.7610410094637223,
+      "grad_norm": 0.4970567230452097,
+      "learning_rate": 1.1881303648371889e-05,
+      "loss": 0.3509,
+      "step": 8933
+    },
+    {
+      "epoch": 1.7612381703470033,
+      "grad_norm": 0.47318188032190406,
+      "learning_rate": 1.1879781276961396e-05,
+      "loss": 0.3271,
+      "step": 8934
+    },
+    {
+      "epoch": 1.761435331230284,
+      "grad_norm": 0.5242959226678022,
+      "learning_rate": 1.1878258860387644e-05,
+      "loss": 0.3749,
+      "step": 8935
+    },
+    {
+      "epoch": 1.7616324921135647,
+      "grad_norm": 0.46123437786870813,
+      "learning_rate": 1.1876736398687212e-05,
+      "loss": 0.3346,
+      "step": 8936
+    },
+    {
+      "epoch": 1.7618296529968456,
+      "grad_norm": 0.4597024041990773,
+      "learning_rate": 1.1875213891896676e-05,
+      "loss": 0.3226,
+      "step": 8937
+    },
+    {
+      "epoch": 1.762026813880126,
+      "grad_norm": 0.4708652177872067,
+      "learning_rate": 1.1873691340052615e-05,
+      "loss": 0.3292,
+      "step": 8938
+    },
+    {
+      "epoch": 1.762223974763407,
+      "grad_norm": 0.4520141786068151,
+      "learning_rate": 1.1872168743191613e-05,
+      "loss": 0.3026,
+      "step": 8939
+    },
+    {
+      "epoch": 1.7624211356466877,
+      "grad_norm": 0.45200547022438,
+      "learning_rate": 1.1870646101350247e-05,
+      "loss": 0.3254,
+      "step": 8940
+    },
+    {
+      "epoch": 1.7626182965299684,
+      "grad_norm": 0.4656284736142514,
+      "learning_rate": 1.18691234145651e-05,
+      "loss": 0.3146,
+      "step": 8941
+    },
+    {
+      "epoch": 1.7628154574132493,
+      "grad_norm": 0.45349874844695887,
+      "learning_rate": 1.1867600682872764e-05,
+      "loss": 0.3145,
+      "step": 8942
+    },
+    {
+      "epoch": 1.76301261829653,
+      "grad_norm": 0.4977307339476207,
+      "learning_rate": 1.1866077906309812e-05,
+      "loss": 0.3329,
+      "step": 8943
+    },
+    {
+      "epoch": 1.7632097791798107,
+      "grad_norm": 0.47172564633086056,
+      "learning_rate": 1.1864555084912839e-05,
+      "loss": 0.32,
+      "step": 8944
+    },
+    {
+      "epoch": 1.7634069400630916,
+      "grad_norm": 0.6274170785811407,
+      "learning_rate": 1.1863032218718424e-05,
+      "loss": 0.3686,
+      "step": 8945
+    },
+    {
+      "epoch": 1.763604100946372,
+      "grad_norm": 0.4976279806243257,
+      "learning_rate": 1.1861509307763166e-05,
+      "loss": 0.3433,
+      "step": 8946
+    },
+    {
+      "epoch": 1.763801261829653,
+      "grad_norm": 0.5010044169487304,
+      "learning_rate": 1.1859986352083644e-05,
+      "loss": 0.3308,
+      "step": 8947
+    },
+    {
+      "epoch": 1.7639984227129337,
+      "grad_norm": 0.4709732593585705,
+      "learning_rate": 1.185846335171645e-05,
+      "loss": 0.3243,
+      "step": 8948
+    },
+    {
+      "epoch": 1.7641955835962144,
+      "grad_norm": 0.46672840740361604,
+      "learning_rate": 1.1856940306698182e-05,
+      "loss": 0.3288,
+      "step": 8949
+    },
+    {
+      "epoch": 1.7643927444794953,
+      "grad_norm": 0.4876404296213418,
+      "learning_rate": 1.1855417217065427e-05,
+      "loss": 0.3454,
+      "step": 8950
+    },
+    {
+      "epoch": 1.764589905362776,
+      "grad_norm": 0.7344474694941918,
+      "learning_rate": 1.1853894082854778e-05,
+      "loss": 0.3343,
+      "step": 8951
+    },
+    {
+      "epoch": 1.7647870662460567,
+      "grad_norm": 0.5281718261013864,
+      "learning_rate": 1.185237090410283e-05,
+      "loss": 0.3559,
+      "step": 8952
+    },
+    {
+      "epoch": 1.7649842271293377,
+      "grad_norm": 0.45336944968705656,
+      "learning_rate": 1.1850847680846181e-05,
+      "loss": 0.3367,
+      "step": 8953
+    },
+    {
+      "epoch": 1.7651813880126181,
+      "grad_norm": 0.48515961309525846,
+      "learning_rate": 1.1849324413121424e-05,
+      "loss": 0.35,
+      "step": 8954
+    },
+    {
+      "epoch": 1.765378548895899,
+      "grad_norm": 0.4683396505847219,
+      "learning_rate": 1.184780110096516e-05,
+      "loss": 0.3291,
+      "step": 8955
+    },
+    {
+      "epoch": 1.7655757097791798,
+      "grad_norm": 0.47251430009899725,
+      "learning_rate": 1.1846277744413988e-05,
+      "loss": 0.3143,
+      "step": 8956
+    },
+    {
+      "epoch": 1.7657728706624605,
+      "grad_norm": 0.47184209011286815,
+      "learning_rate": 1.1844754343504503e-05,
+      "loss": 0.3256,
+      "step": 8957
+    },
+    {
+      "epoch": 1.7659700315457414,
+      "grad_norm": 0.4783002948832379,
+      "learning_rate": 1.1843230898273312e-05,
+      "loss": 0.3487,
+      "step": 8958
+    },
+    {
+      "epoch": 1.766167192429022,
+      "grad_norm": 0.5175752592893191,
+      "learning_rate": 1.1841707408757012e-05,
+      "loss": 0.348,
+      "step": 8959
+    },
+    {
+      "epoch": 1.7663643533123028,
+      "grad_norm": 0.48774710368554364,
+      "learning_rate": 1.184018387499221e-05,
+      "loss": 0.3489,
+      "step": 8960
+    },
+    {
+      "epoch": 1.7665615141955837,
+      "grad_norm": 0.5091512140833712,
+      "learning_rate": 1.183866029701551e-05,
+      "loss": 0.3437,
+      "step": 8961
+    },
+    {
+      "epoch": 1.7667586750788642,
+      "grad_norm": 0.4579012245288681,
+      "learning_rate": 1.1837136674863512e-05,
+      "loss": 0.3116,
+      "step": 8962
+    },
+    {
+      "epoch": 1.7669558359621451,
+      "grad_norm": 0.4607106881483517,
+      "learning_rate": 1.1835613008572828e-05,
+      "loss": 0.3182,
+      "step": 8963
+    },
+    {
+      "epoch": 1.7671529968454258,
+      "grad_norm": 0.48460789200010285,
+      "learning_rate": 1.1834089298180062e-05,
+      "loss": 0.3515,
+      "step": 8964
+    },
+    {
+      "epoch": 1.7673501577287065,
+      "grad_norm": 0.5066322063807643,
+      "learning_rate": 1.1832565543721828e-05,
+      "loss": 0.3611,
+      "step": 8965
+    },
+    {
+      "epoch": 1.7675473186119874,
+      "grad_norm": 0.4619333716892967,
+      "learning_rate": 1.1831041745234728e-05,
+      "loss": 0.3183,
+      "step": 8966
+    },
+    {
+      "epoch": 1.7677444794952681,
+      "grad_norm": 0.46313697004860765,
+      "learning_rate": 1.1829517902755375e-05,
+      "loss": 0.322,
+      "step": 8967
+    },
+    {
+      "epoch": 1.7679416403785488,
+      "grad_norm": 0.6763284941264115,
+      "learning_rate": 1.1827994016320381e-05,
+      "loss": 0.3173,
+      "step": 8968
+    },
+    {
+      "epoch": 1.7681388012618298,
+      "grad_norm": 0.47210950027001125,
+      "learning_rate": 1.1826470085966357e-05,
+      "loss": 0.3312,
+      "step": 8969
+    },
+    {
+      "epoch": 1.7683359621451105,
+      "grad_norm": 0.47290245061079667,
+      "learning_rate": 1.1824946111729922e-05,
+      "loss": 0.3043,
+      "step": 8970
+    },
+    {
+      "epoch": 1.7685331230283912,
+      "grad_norm": 0.479790151905134,
+      "learning_rate": 1.1823422093647684e-05,
+      "loss": 0.3469,
+      "step": 8971
+    },
+    {
+      "epoch": 1.768730283911672,
+      "grad_norm": 0.4581566972384319,
+      "learning_rate": 1.1821898031756265e-05,
+      "loss": 0.33,
+      "step": 8972
+    },
+    {
+      "epoch": 1.7689274447949526,
+      "grad_norm": 0.4721423210829706,
+      "learning_rate": 1.1820373926092274e-05,
+      "loss": 0.3159,
+      "step": 8973
+    },
+    {
+      "epoch": 1.7691246056782335,
+      "grad_norm": 0.49894798175464766,
+      "learning_rate": 1.181884977669234e-05,
+      "loss": 0.3572,
+      "step": 8974
+    },
+    {
+      "epoch": 1.7693217665615142,
+      "grad_norm": 0.49659191839512695,
+      "learning_rate": 1.181732558359307e-05,
+      "loss": 0.334,
+      "step": 8975
+    },
+    {
+      "epoch": 1.7695189274447949,
+      "grad_norm": 0.5171496834529237,
+      "learning_rate": 1.181580134683109e-05,
+      "loss": 0.345,
+      "step": 8976
+    },
+    {
+      "epoch": 1.7697160883280758,
+      "grad_norm": 0.448236524997749,
+      "learning_rate": 1.1814277066443023e-05,
+      "loss": 0.2887,
+      "step": 8977
+    },
+    {
+      "epoch": 1.7699132492113565,
+      "grad_norm": 5.333849505830885,
+      "learning_rate": 1.1812752742465488e-05,
+      "loss": 0.3734,
+      "step": 8978
+    },
+    {
+      "epoch": 1.7701104100946372,
+      "grad_norm": 0.49076446223605413,
+      "learning_rate": 1.1811228374935107e-05,
+      "loss": 0.3396,
+      "step": 8979
+    },
+    {
+      "epoch": 1.7703075709779181,
+      "grad_norm": 0.5459828067722964,
+      "learning_rate": 1.1809703963888506e-05,
+      "loss": 0.3123,
+      "step": 8980
+    },
+    {
+      "epoch": 1.7705047318611986,
+      "grad_norm": 0.48374652525635975,
+      "learning_rate": 1.180817950936231e-05,
+      "loss": 0.3285,
+      "step": 8981
+    },
+    {
+      "epoch": 1.7707018927444795,
+      "grad_norm": 0.4755660557838733,
+      "learning_rate": 1.1806655011393144e-05,
+      "loss": 0.3507,
+      "step": 8982
+    },
+    {
+      "epoch": 1.7708990536277602,
+      "grad_norm": 0.6006920450403271,
+      "learning_rate": 1.1805130470017639e-05,
+      "loss": 0.3379,
+      "step": 8983
+    },
+    {
+      "epoch": 1.771096214511041,
+      "grad_norm": 0.5009897132212682,
+      "learning_rate": 1.180360588527242e-05,
+      "loss": 0.3573,
+      "step": 8984
+    },
+    {
+      "epoch": 1.7712933753943219,
+      "grad_norm": 0.4620563729527647,
+      "learning_rate": 1.1802081257194116e-05,
+      "loss": 0.3262,
+      "step": 8985
+    },
+    {
+      "epoch": 1.7714905362776026,
+      "grad_norm": 0.4769119308524991,
+      "learning_rate": 1.180055658581936e-05,
+      "loss": 0.3184,
+      "step": 8986
+    },
+    {
+      "epoch": 1.7716876971608833,
+      "grad_norm": 0.4487850697936526,
+      "learning_rate": 1.179903187118478e-05,
+      "loss": 0.3179,
+      "step": 8987
+    },
+    {
+      "epoch": 1.7718848580441642,
+      "grad_norm": 0.4677363529785754,
+      "learning_rate": 1.179750711332701e-05,
+      "loss": 0.3009,
+      "step": 8988
+    },
+    {
+      "epoch": 1.7720820189274447,
+      "grad_norm": 0.48154564362378643,
+      "learning_rate": 1.179598231228269e-05,
+      "loss": 0.3517,
+      "step": 8989
+    },
+    {
+      "epoch": 1.7722791798107256,
+      "grad_norm": 0.4608411338811139,
+      "learning_rate": 1.1794457468088443e-05,
+      "loss": 0.3284,
+      "step": 8990
+    },
+    {
+      "epoch": 1.7724763406940063,
+      "grad_norm": 0.5027074279023396,
+      "learning_rate": 1.1792932580780913e-05,
+      "loss": 0.3442,
+      "step": 8991
+    },
+    {
+      "epoch": 1.772673501577287,
+      "grad_norm": 0.4399957421863353,
+      "learning_rate": 1.1791407650396731e-05,
+      "loss": 0.3108,
+      "step": 8992
+    },
+    {
+      "epoch": 1.772870662460568,
+      "grad_norm": 0.45380590331375986,
+      "learning_rate": 1.1789882676972541e-05,
+      "loss": 0.2994,
+      "step": 8993
+    },
+    {
+      "epoch": 1.7730678233438486,
+      "grad_norm": 0.5204613664594859,
+      "learning_rate": 1.1788357660544976e-05,
+      "loss": 0.3543,
+      "step": 8994
+    },
+    {
+      "epoch": 1.7732649842271293,
+      "grad_norm": 0.48732752300245125,
+      "learning_rate": 1.1786832601150677e-05,
+      "loss": 0.3421,
+      "step": 8995
+    },
+    {
+      "epoch": 1.7734621451104102,
+      "grad_norm": 0.473467187951441,
+      "learning_rate": 1.1785307498826288e-05,
+      "loss": 0.3338,
+      "step": 8996
+    },
+    {
+      "epoch": 1.7736593059936907,
+      "grad_norm": 0.48137614570470716,
+      "learning_rate": 1.1783782353608449e-05,
+      "loss": 0.3533,
+      "step": 8997
+    },
+    {
+      "epoch": 1.7738564668769716,
+      "grad_norm": 0.4665674415229966,
+      "learning_rate": 1.1782257165533802e-05,
+      "loss": 0.3309,
+      "step": 8998
+    },
+    {
+      "epoch": 1.7740536277602523,
+      "grad_norm": 0.47639470639484743,
+      "learning_rate": 1.1780731934638992e-05,
+      "loss": 0.3181,
+      "step": 8999
+    },
+    {
+      "epoch": 1.774250788643533,
+      "grad_norm": 0.4888316448783417,
+      "learning_rate": 1.177920666096066e-05,
+      "loss": 0.3469,
+      "step": 9000
+    },
+    {
+      "epoch": 1.774447949526814,
+      "grad_norm": 0.488778406351827,
+      "learning_rate": 1.177768134453546e-05,
+      "loss": 0.3744,
+      "step": 9001
+    },
+    {
+      "epoch": 1.7746451104100947,
+      "grad_norm": 0.44883233582815296,
+      "learning_rate": 1.177615598540003e-05,
+      "loss": 0.3301,
+      "step": 9002
+    },
+    {
+      "epoch": 1.7748422712933754,
+      "grad_norm": 0.4506843011240513,
+      "learning_rate": 1.1774630583591024e-05,
+      "loss": 0.3038,
+      "step": 9003
+    },
+    {
+      "epoch": 1.7750394321766563,
+      "grad_norm": 0.49508062610720943,
+      "learning_rate": 1.1773105139145088e-05,
+      "loss": 0.3583,
+      "step": 9004
+    },
+    {
+      "epoch": 1.7752365930599368,
+      "grad_norm": 0.5164716499986945,
+      "learning_rate": 1.1771579652098874e-05,
+      "loss": 0.382,
+      "step": 9005
+    },
+    {
+      "epoch": 1.7754337539432177,
+      "grad_norm": 0.4882485785032844,
+      "learning_rate": 1.1770054122489031e-05,
+      "loss": 0.337,
+      "step": 9006
+    },
+    {
+      "epoch": 1.7756309148264984,
+      "grad_norm": 0.4663980919945463,
+      "learning_rate": 1.1768528550352216e-05,
+      "loss": 0.3192,
+      "step": 9007
+    },
+    {
+      "epoch": 1.775828075709779,
+      "grad_norm": 0.43682922880005404,
+      "learning_rate": 1.1767002935725076e-05,
+      "loss": 0.3017,
+      "step": 9008
+    },
+    {
+      "epoch": 1.77602523659306,
+      "grad_norm": 0.4867060152007287,
+      "learning_rate": 1.1765477278644264e-05,
+      "loss": 0.3299,
+      "step": 9009
+    },
+    {
+      "epoch": 1.7762223974763407,
+      "grad_norm": 0.5124563775251622,
+      "learning_rate": 1.1763951579146444e-05,
+      "loss": 0.3266,
+      "step": 9010
+    },
+    {
+      "epoch": 1.7764195583596214,
+      "grad_norm": 0.47078691041664955,
+      "learning_rate": 1.1762425837268263e-05,
+      "loss": 0.3356,
+      "step": 9011
+    },
+    {
+      "epoch": 1.7766167192429023,
+      "grad_norm": 0.4916978260239702,
+      "learning_rate": 1.1760900053046386e-05,
+      "loss": 0.3288,
+      "step": 9012
+    },
+    {
+      "epoch": 1.776813880126183,
+      "grad_norm": 0.4956955563685568,
+      "learning_rate": 1.1759374226517464e-05,
+      "loss": 0.3384,
+      "step": 9013
+    },
+    {
+      "epoch": 1.7770110410094637,
+      "grad_norm": 0.4728851012371811,
+      "learning_rate": 1.1757848357718162e-05,
+      "loss": 0.3307,
+      "step": 9014
+    },
+    {
+      "epoch": 1.7772082018927446,
+      "grad_norm": 0.503512410418657,
+      "learning_rate": 1.1756322446685134e-05,
+      "loss": 0.3366,
+      "step": 9015
+    },
+    {
+      "epoch": 1.7774053627760251,
+      "grad_norm": 0.5196421133302855,
+      "learning_rate": 1.1754796493455048e-05,
+      "loss": 0.3633,
+      "step": 9016
+    },
+    {
+      "epoch": 1.777602523659306,
+      "grad_norm": 0.49369062439862366,
+      "learning_rate": 1.1753270498064561e-05,
+      "loss": 0.3453,
+      "step": 9017
+    },
+    {
+      "epoch": 1.7777996845425867,
+      "grad_norm": 0.4653735508627901,
+      "learning_rate": 1.1751744460550338e-05,
+      "loss": 0.3296,
+      "step": 9018
+    },
+    {
+      "epoch": 1.7779968454258674,
+      "grad_norm": 0.5068627718887688,
+      "learning_rate": 1.1750218380949047e-05,
+      "loss": 0.3514,
+      "step": 9019
+    },
+    {
+      "epoch": 1.7781940063091484,
+      "grad_norm": 0.47122281155877815,
+      "learning_rate": 1.1748692259297347e-05,
+      "loss": 0.3461,
+      "step": 9020
+    },
+    {
+      "epoch": 1.778391167192429,
+      "grad_norm": 0.4566301776497934,
+      "learning_rate": 1.174716609563191e-05,
+      "loss": 0.3396,
+      "step": 9021
+    },
+    {
+      "epoch": 1.7785883280757098,
+      "grad_norm": 0.4861860829634978,
+      "learning_rate": 1.1745639889989398e-05,
+      "loss": 0.3213,
+      "step": 9022
+    },
+    {
+      "epoch": 1.7787854889589907,
+      "grad_norm": 0.4783918537273916,
+      "learning_rate": 1.1744113642406483e-05,
+      "loss": 0.3139,
+      "step": 9023
+    },
+    {
+      "epoch": 1.7789826498422712,
+      "grad_norm": 0.4813446133963429,
+      "learning_rate": 1.1742587352919833e-05,
+      "loss": 0.3439,
+      "step": 9024
+    },
+    {
+      "epoch": 1.779179810725552,
+      "grad_norm": 0.5247316124213983,
+      "learning_rate": 1.1741061021566118e-05,
+      "loss": 0.3264,
+      "step": 9025
+    },
+    {
+      "epoch": 1.7793769716088328,
+      "grad_norm": 0.4666782968847508,
+      "learning_rate": 1.173953464838201e-05,
+      "loss": 0.3286,
+      "step": 9026
+    },
+    {
+      "epoch": 1.7795741324921135,
+      "grad_norm": 0.4530221240481721,
+      "learning_rate": 1.1738008233404181e-05,
+      "loss": 0.3428,
+      "step": 9027
+    },
+    {
+      "epoch": 1.7797712933753944,
+      "grad_norm": 0.4647953682965125,
+      "learning_rate": 1.1736481776669307e-05,
+      "loss": 0.3247,
+      "step": 9028
+    },
+    {
+      "epoch": 1.7799684542586751,
+      "grad_norm": 0.49727784750492343,
+      "learning_rate": 1.1734955278214057e-05,
+      "loss": 0.3619,
+      "step": 9029
+    },
+    {
+      "epoch": 1.7801656151419558,
+      "grad_norm": 0.49005891060030987,
+      "learning_rate": 1.1733428738075108e-05,
+      "loss": 0.3562,
+      "step": 9030
+    },
+    {
+      "epoch": 1.7803627760252367,
+      "grad_norm": 0.734206410674185,
+      "learning_rate": 1.1731902156289142e-05,
+      "loss": 0.3285,
+      "step": 9031
+    },
+    {
+      "epoch": 1.7805599369085172,
+      "grad_norm": 0.4681761876719222,
+      "learning_rate": 1.173037553289283e-05,
+      "loss": 0.3409,
+      "step": 9032
+    },
+    {
+      "epoch": 1.7807570977917981,
+      "grad_norm": 0.46203491841893324,
+      "learning_rate": 1.1728848867922853e-05,
+      "loss": 0.3186,
+      "step": 9033
+    },
+    {
+      "epoch": 1.7809542586750788,
+      "grad_norm": 0.4835537269864709,
+      "learning_rate": 1.1727322161415888e-05,
+      "loss": 0.3312,
+      "step": 9034
+    },
+    {
+      "epoch": 1.7811514195583595,
+      "grad_norm": 0.44861177894612153,
+      "learning_rate": 1.1725795413408618e-05,
+      "loss": 0.3016,
+      "step": 9035
+    },
+    {
+      "epoch": 1.7813485804416405,
+      "grad_norm": 0.4566247359785835,
+      "learning_rate": 1.1724268623937725e-05,
+      "loss": 0.3094,
+      "step": 9036
+    },
+    {
+      "epoch": 1.7815457413249212,
+      "grad_norm": 0.5013017232922788,
+      "learning_rate": 1.1722741793039885e-05,
+      "loss": 0.3556,
+      "step": 9037
+    },
+    {
+      "epoch": 1.7817429022082019,
+      "grad_norm": 0.4828140840198569,
+      "learning_rate": 1.172121492075179e-05,
+      "loss": 0.3398,
+      "step": 9038
+    },
+    {
+      "epoch": 1.7819400630914828,
+      "grad_norm": 0.4836084111183976,
+      "learning_rate": 1.171968800711012e-05,
+      "loss": 0.3516,
+      "step": 9039
+    },
+    {
+      "epoch": 1.7821372239747633,
+      "grad_norm": 0.45421673308016836,
+      "learning_rate": 1.1718161052151562e-05,
+      "loss": 0.2943,
+      "step": 9040
+    },
+    {
+      "epoch": 1.7823343848580442,
+      "grad_norm": 0.46771437738564686,
+      "learning_rate": 1.1716634055912796e-05,
+      "loss": 0.358,
+      "step": 9041
+    },
+    {
+      "epoch": 1.7825315457413249,
+      "grad_norm": 0.47666003792216416,
+      "learning_rate": 1.1715107018430522e-05,
+      "loss": 0.3418,
+      "step": 9042
+    },
+    {
+      "epoch": 1.7827287066246056,
+      "grad_norm": 0.4840693365790069,
+      "learning_rate": 1.1713579939741415e-05,
+      "loss": 0.3198,
+      "step": 9043
+    },
+    {
+      "epoch": 1.7829258675078865,
+      "grad_norm": 0.46910784593485333,
+      "learning_rate": 1.1712052819882171e-05,
+      "loss": 0.3262,
+      "step": 9044
+    },
+    {
+      "epoch": 1.7831230283911672,
+      "grad_norm": 0.4896822327451005,
+      "learning_rate": 1.171052565888948e-05,
+      "loss": 0.3433,
+      "step": 9045
+    },
+    {
+      "epoch": 1.783320189274448,
+      "grad_norm": 0.476154410913374,
+      "learning_rate": 1.1708998456800034e-05,
+      "loss": 0.3148,
+      "step": 9046
+    },
+    {
+      "epoch": 1.7835173501577288,
+      "grad_norm": 0.5217716610338343,
+      "learning_rate": 1.170747121365052e-05,
+      "loss": 0.3312,
+      "step": 9047
+    },
+    {
+      "epoch": 1.7837145110410093,
+      "grad_norm": 0.49551436556952305,
+      "learning_rate": 1.1705943929477639e-05,
+      "loss": 0.348,
+      "step": 9048
+    },
+    {
+      "epoch": 1.7839116719242902,
+      "grad_norm": 0.4920600225347391,
+      "learning_rate": 1.170441660431808e-05,
+      "loss": 0.349,
+      "step": 9049
+    },
+    {
+      "epoch": 1.784108832807571,
+      "grad_norm": 0.4378125833985353,
+      "learning_rate": 1.1702889238208539e-05,
+      "loss": 0.312,
+      "step": 9050
+    },
+    {
+      "epoch": 1.7843059936908516,
+      "grad_norm": 0.43674490442544767,
+      "learning_rate": 1.1701361831185714e-05,
+      "loss": 0.2936,
+      "step": 9051
+    },
+    {
+      "epoch": 1.7845031545741326,
+      "grad_norm": 0.4842053898639523,
+      "learning_rate": 1.1699834383286299e-05,
+      "loss": 0.3474,
+      "step": 9052
+    },
+    {
+      "epoch": 1.7847003154574133,
+      "grad_norm": 0.4529779374085598,
+      "learning_rate": 1.1698306894546995e-05,
+      "loss": 0.3128,
+      "step": 9053
+    },
+    {
+      "epoch": 1.784897476340694,
+      "grad_norm": 0.4801304915326945,
+      "learning_rate": 1.16967793650045e-05,
+      "loss": 0.3324,
+      "step": 9054
+    },
+    {
+      "epoch": 1.7850946372239749,
+      "grad_norm": 0.5173598334277061,
+      "learning_rate": 1.1695251794695514e-05,
+      "loss": 0.3415,
+      "step": 9055
+    },
+    {
+      "epoch": 1.7852917981072554,
+      "grad_norm": 0.4654828527782663,
+      "learning_rate": 1.169372418365674e-05,
+      "loss": 0.3172,
+      "step": 9056
+    },
+    {
+      "epoch": 1.7854889589905363,
+      "grad_norm": 0.4563252768971276,
+      "learning_rate": 1.1692196531924877e-05,
+      "loss": 0.3165,
+      "step": 9057
+    },
+    {
+      "epoch": 1.785686119873817,
+      "grad_norm": 0.48839701961063525,
+      "learning_rate": 1.169066883953663e-05,
+      "loss": 0.3321,
+      "step": 9058
+    },
+    {
+      "epoch": 1.7858832807570977,
+      "grad_norm": 0.4668548629020815,
+      "learning_rate": 1.1689141106528703e-05,
+      "loss": 0.3229,
+      "step": 9059
+    },
+    {
+      "epoch": 1.7860804416403786,
+      "grad_norm": 0.48668760035869535,
+      "learning_rate": 1.16876133329378e-05,
+      "loss": 0.3411,
+      "step": 9060
+    },
+    {
+      "epoch": 1.7862776025236593,
+      "grad_norm": 0.4648244886523526,
+      "learning_rate": 1.168608551880063e-05,
+      "loss": 0.337,
+      "step": 9061
+    },
+    {
+      "epoch": 1.78647476340694,
+      "grad_norm": 0.48604760080798454,
+      "learning_rate": 1.1684557664153893e-05,
+      "loss": 0.3144,
+      "step": 9062
+    },
+    {
+      "epoch": 1.786671924290221,
+      "grad_norm": 0.4783043678898303,
+      "learning_rate": 1.1683029769034304e-05,
+      "loss": 0.3225,
+      "step": 9063
+    },
+    {
+      "epoch": 1.7868690851735016,
+      "grad_norm": 0.47691055878103017,
+      "learning_rate": 1.168150183347857e-05,
+      "loss": 0.3317,
+      "step": 9064
+    },
+    {
+      "epoch": 1.7870662460567823,
+      "grad_norm": 0.4868775598571958,
+      "learning_rate": 1.16799738575234e-05,
+      "loss": 0.3326,
+      "step": 9065
+    },
+    {
+      "epoch": 1.7872634069400632,
+      "grad_norm": 0.45774853271321925,
+      "learning_rate": 1.1678445841205506e-05,
+      "loss": 0.3183,
+      "step": 9066
+    },
+    {
+      "epoch": 1.7874605678233437,
+      "grad_norm": 0.9497850183764717,
+      "learning_rate": 1.1676917784561599e-05,
+      "loss": 0.3714,
+      "step": 9067
+    },
+    {
+      "epoch": 1.7876577287066246,
+      "grad_norm": 0.4768647293354841,
+      "learning_rate": 1.1675389687628389e-05,
+      "loss": 0.3462,
+      "step": 9068
+    },
+    {
+      "epoch": 1.7878548895899053,
+      "grad_norm": 0.4783944603727662,
+      "learning_rate": 1.1673861550442596e-05,
+      "loss": 0.3372,
+      "step": 9069
+    },
+    {
+      "epoch": 1.788052050473186,
+      "grad_norm": 0.46436887776659636,
+      "learning_rate": 1.167233337304093e-05,
+      "loss": 0.3318,
+      "step": 9070
+    },
+    {
+      "epoch": 1.788249211356467,
+      "grad_norm": 4.345180913181195,
+      "learning_rate": 1.1670805155460108e-05,
+      "loss": 0.3453,
+      "step": 9071
+    },
+    {
+      "epoch": 1.7884463722397477,
+      "grad_norm": 0.5159025017330271,
+      "learning_rate": 1.1669276897736847e-05,
+      "loss": 0.3449,
+      "step": 9072
+    },
+    {
+      "epoch": 1.7886435331230284,
+      "grad_norm": 0.5595335747614689,
+      "learning_rate": 1.1667748599907864e-05,
+      "loss": 0.3788,
+      "step": 9073
+    },
+    {
+      "epoch": 1.7888406940063093,
+      "grad_norm": 0.5325186373611229,
+      "learning_rate": 1.1666220262009877e-05,
+      "loss": 0.353,
+      "step": 9074
+    },
+    {
+      "epoch": 1.7890378548895898,
+      "grad_norm": 1.2209348375435845,
+      "learning_rate": 1.1664691884079606e-05,
+      "loss": 0.3454,
+      "step": 9075
+    },
+    {
+      "epoch": 1.7892350157728707,
+      "grad_norm": 0.4529621721978549,
+      "learning_rate": 1.1663163466153775e-05,
+      "loss": 0.3219,
+      "step": 9076
+    },
+    {
+      "epoch": 1.7894321766561514,
+      "grad_norm": 0.5786960304033676,
+      "learning_rate": 1.16616350082691e-05,
+      "loss": 0.3458,
+      "step": 9077
+    },
+    {
+      "epoch": 1.789629337539432,
+      "grad_norm": 0.46513659246814454,
+      "learning_rate": 1.1660106510462305e-05,
+      "loss": 0.3205,
+      "step": 9078
+    },
+    {
+      "epoch": 1.789826498422713,
+      "grad_norm": 0.49129392716967757,
+      "learning_rate": 1.1658577972770115e-05,
+      "loss": 0.3643,
+      "step": 9079
+    },
+    {
+      "epoch": 1.7900236593059937,
+      "grad_norm": 0.4946634044459266,
+      "learning_rate": 1.1657049395229255e-05,
+      "loss": 0.3585,
+      "step": 9080
+    },
+    {
+      "epoch": 1.7902208201892744,
+      "grad_norm": 0.4625534608044079,
+      "learning_rate": 1.1655520777876446e-05,
+      "loss": 0.3295,
+      "step": 9081
+    },
+    {
+      "epoch": 1.7904179810725553,
+      "grad_norm": 0.4770599185465994,
+      "learning_rate": 1.1653992120748421e-05,
+      "loss": 0.3232,
+      "step": 9082
+    },
+    {
+      "epoch": 1.7906151419558358,
+      "grad_norm": 0.5732472260639709,
+      "learning_rate": 1.1652463423881898e-05,
+      "loss": 0.3939,
+      "step": 9083
+    },
+    {
+      "epoch": 1.7908123028391167,
+      "grad_norm": 0.5032610006719888,
+      "learning_rate": 1.1650934687313615e-05,
+      "loss": 0.3256,
+      "step": 9084
+    },
+    {
+      "epoch": 1.7910094637223974,
+      "grad_norm": 0.5315497537438166,
+      "learning_rate": 1.1649405911080298e-05,
+      "loss": 0.3533,
+      "step": 9085
+    },
+    {
+      "epoch": 1.7912066246056781,
+      "grad_norm": 0.47972476193122204,
+      "learning_rate": 1.1647877095218671e-05,
+      "loss": 0.3087,
+      "step": 9086
+    },
+    {
+      "epoch": 1.791403785488959,
+      "grad_norm": 0.49904975210777663,
+      "learning_rate": 1.1646348239765475e-05,
+      "loss": 0.333,
+      "step": 9087
+    },
+    {
+      "epoch": 1.7916009463722398,
+      "grad_norm": 0.4982533128788867,
+      "learning_rate": 1.164481934475743e-05,
+      "loss": 0.3569,
+      "step": 9088
+    },
+    {
+      "epoch": 1.7917981072555205,
+      "grad_norm": 0.9122269960563105,
+      "learning_rate": 1.1643290410231282e-05,
+      "loss": 0.3395,
+      "step": 9089
+    },
+    {
+      "epoch": 1.7919952681388014,
+      "grad_norm": 0.4557932362062137,
+      "learning_rate": 1.1641761436223753e-05,
+      "loss": 0.3024,
+      "step": 9090
+    },
+    {
+      "epoch": 1.7921924290220819,
+      "grad_norm": 0.4941263997632442,
+      "learning_rate": 1.1640232422771586e-05,
+      "loss": 0.3476,
+      "step": 9091
+    },
+    {
+      "epoch": 1.7923895899053628,
+      "grad_norm": 0.5155585824775918,
+      "learning_rate": 1.1638703369911517e-05,
+      "loss": 0.3596,
+      "step": 9092
+    },
+    {
+      "epoch": 1.7925867507886435,
+      "grad_norm": 1.4023286561635016,
+      "learning_rate": 1.1637174277680277e-05,
+      "loss": 0.3335,
+      "step": 9093
+    },
+    {
+      "epoch": 1.7927839116719242,
+      "grad_norm": 0.5065005447760449,
+      "learning_rate": 1.1635645146114607e-05,
+      "loss": 0.355,
+      "step": 9094
+    },
+    {
+      "epoch": 1.7929810725552051,
+      "grad_norm": 0.526370677958319,
+      "learning_rate": 1.1634115975251245e-05,
+      "loss": 0.3056,
+      "step": 9095
+    },
+    {
+      "epoch": 1.7931782334384858,
+      "grad_norm": 0.486266087331147,
+      "learning_rate": 1.1632586765126929e-05,
+      "loss": 0.3269,
+      "step": 9096
+    },
+    {
+      "epoch": 1.7933753943217665,
+      "grad_norm": 0.4580937992770503,
+      "learning_rate": 1.1631057515778403e-05,
+      "loss": 0.3137,
+      "step": 9097
+    },
+    {
+      "epoch": 1.7935725552050474,
+      "grad_norm": 0.5054061865722272,
+      "learning_rate": 1.1629528227242408e-05,
+      "loss": 0.3398,
+      "step": 9098
+    },
+    {
+      "epoch": 1.793769716088328,
+      "grad_norm": 0.47624328346608846,
+      "learning_rate": 1.1627998899555684e-05,
+      "loss": 0.3217,
+      "step": 9099
+    },
+    {
+      "epoch": 1.7939668769716088,
+      "grad_norm": 0.47501826137837944,
+      "learning_rate": 1.1626469532754975e-05,
+      "loss": 0.3436,
+      "step": 9100
+    },
+    {
+      "epoch": 1.7941640378548895,
+      "grad_norm": 0.5083570086033906,
+      "learning_rate": 1.1624940126877027e-05,
+      "loss": 0.3342,
+      "step": 9101
+    },
+    {
+      "epoch": 1.7943611987381702,
+      "grad_norm": 0.43834857149607065,
+      "learning_rate": 1.1623410681958583e-05,
+      "loss": 0.3176,
+      "step": 9102
+    },
+    {
+      "epoch": 1.7945583596214512,
+      "grad_norm": 0.46797550187905224,
+      "learning_rate": 1.1621881198036389e-05,
+      "loss": 0.3296,
+      "step": 9103
+    },
+    {
+      "epoch": 1.7947555205047319,
+      "grad_norm": 0.747473707199481,
+      "learning_rate": 1.1620351675147195e-05,
+      "loss": 0.336,
+      "step": 9104
+    },
+    {
+      "epoch": 1.7949526813880126,
+      "grad_norm": 0.4864213421889568,
+      "learning_rate": 1.1618822113327743e-05,
+      "loss": 0.346,
+      "step": 9105
+    },
+    {
+      "epoch": 1.7951498422712935,
+      "grad_norm": 0.48393670272857564,
+      "learning_rate": 1.1617292512614793e-05,
+      "loss": 0.3323,
+      "step": 9106
+    },
+    {
+      "epoch": 1.7953470031545742,
+      "grad_norm": 0.42975326675170367,
+      "learning_rate": 1.161576287304508e-05,
+      "loss": 0.2663,
+      "step": 9107
+    },
+    {
+      "epoch": 1.7955441640378549,
+      "grad_norm": 0.4861731628110563,
+      "learning_rate": 1.1614233194655371e-05,
+      "loss": 0.3257,
+      "step": 9108
+    },
+    {
+      "epoch": 1.7957413249211358,
+      "grad_norm": 0.8570529099252474,
+      "learning_rate": 1.1612703477482403e-05,
+      "loss": 0.32,
+      "step": 9109
+    },
+    {
+      "epoch": 1.7959384858044163,
+      "grad_norm": 0.4868050444217125,
+      "learning_rate": 1.161117372156294e-05,
+      "loss": 0.3324,
+      "step": 9110
+    },
+    {
+      "epoch": 1.7961356466876972,
+      "grad_norm": 0.5166058354822001,
+      "learning_rate": 1.1609643926933727e-05,
+      "loss": 0.3296,
+      "step": 9111
+    },
+    {
+      "epoch": 1.796332807570978,
+      "grad_norm": 0.454694260080299,
+      "learning_rate": 1.1608114093631523e-05,
+      "loss": 0.3057,
+      "step": 9112
+    },
+    {
+      "epoch": 1.7965299684542586,
+      "grad_norm": 0.4873463120948357,
+      "learning_rate": 1.1606584221693084e-05,
+      "loss": 0.3327,
+      "step": 9113
+    },
+    {
+      "epoch": 1.7967271293375395,
+      "grad_norm": 0.46258370737085464,
+      "learning_rate": 1.1605054311155165e-05,
+      "loss": 0.2937,
+      "step": 9114
+    },
+    {
+      "epoch": 1.7969242902208202,
+      "grad_norm": 0.5089951413840461,
+      "learning_rate": 1.1603524362054525e-05,
+      "loss": 0.347,
+      "step": 9115
+    },
+    {
+      "epoch": 1.797121451104101,
+      "grad_norm": 0.49112572937967497,
+      "learning_rate": 1.1601994374427921e-05,
+      "loss": 0.3328,
+      "step": 9116
+    },
+    {
+      "epoch": 1.7973186119873819,
+      "grad_norm": 0.4743944704648095,
+      "learning_rate": 1.160046434831211e-05,
+      "loss": 0.316,
+      "step": 9117
+    },
+    {
+      "epoch": 1.7975157728706623,
+      "grad_norm": 0.5012809900872015,
+      "learning_rate": 1.1598934283743855e-05,
+      "loss": 0.3618,
+      "step": 9118
+    },
+    {
+      "epoch": 1.7977129337539433,
+      "grad_norm": 0.4632607405847242,
+      "learning_rate": 1.1597404180759917e-05,
+      "loss": 0.3282,
+      "step": 9119
+    },
+    {
+      "epoch": 1.797910094637224,
+      "grad_norm": 0.4893372010154669,
+      "learning_rate": 1.1595874039397055e-05,
+      "loss": 0.3256,
+      "step": 9120
+    },
+    {
+      "epoch": 1.7981072555205047,
+      "grad_norm": 0.49627890794355706,
+      "learning_rate": 1.1594343859692037e-05,
+      "loss": 0.3487,
+      "step": 9121
+    },
+    {
+      "epoch": 1.7983044164037856,
+      "grad_norm": 0.49100214178136625,
+      "learning_rate": 1.1592813641681621e-05,
+      "loss": 0.3388,
+      "step": 9122
+    },
+    {
+      "epoch": 1.7985015772870663,
+      "grad_norm": 0.4542961165749031,
+      "learning_rate": 1.1591283385402577e-05,
+      "loss": 0.3299,
+      "step": 9123
+    },
+    {
+      "epoch": 1.798698738170347,
+      "grad_norm": 0.4801501150330402,
+      "learning_rate": 1.1589753090891667e-05,
+      "loss": 0.3323,
+      "step": 9124
+    },
+    {
+      "epoch": 1.798895899053628,
+      "grad_norm": 0.4851771812965912,
+      "learning_rate": 1.158822275818566e-05,
+      "loss": 0.3357,
+      "step": 9125
+    },
+    {
+      "epoch": 1.7990930599369084,
+      "grad_norm": 0.45182412694780133,
+      "learning_rate": 1.158669238732132e-05,
+      "loss": 0.305,
+      "step": 9126
+    },
+    {
+      "epoch": 1.7992902208201893,
+      "grad_norm": 0.47897297293682256,
+      "learning_rate": 1.158516197833542e-05,
+      "loss": 0.3231,
+      "step": 9127
+    },
+    {
+      "epoch": 1.79948738170347,
+      "grad_norm": 0.43389581021319834,
+      "learning_rate": 1.1583631531264723e-05,
+      "loss": 0.3173,
+      "step": 9128
+    },
+    {
+      "epoch": 1.7996845425867507,
+      "grad_norm": 0.47731478351982176,
+      "learning_rate": 1.1582101046146008e-05,
+      "loss": 0.3394,
+      "step": 9129
+    },
+    {
+      "epoch": 1.7998817034700316,
+      "grad_norm": 0.435886477568427,
+      "learning_rate": 1.1580570523016036e-05,
+      "loss": 0.2997,
+      "step": 9130
+    },
+    {
+      "epoch": 1.8000788643533123,
+      "grad_norm": 0.5238566166040188,
+      "learning_rate": 1.1579039961911591e-05,
+      "loss": 0.3432,
+      "step": 9131
+    },
+    {
+      "epoch": 1.800276025236593,
+      "grad_norm": 0.5379605041705956,
+      "learning_rate": 1.1577509362869433e-05,
+      "loss": 0.357,
+      "step": 9132
+    },
+    {
+      "epoch": 1.800473186119874,
+      "grad_norm": 0.449491662574799,
+      "learning_rate": 1.1575978725926347e-05,
+      "loss": 0.3106,
+      "step": 9133
+    },
+    {
+      "epoch": 1.8006703470031544,
+      "grad_norm": 0.5080608887234477,
+      "learning_rate": 1.1574448051119101e-05,
+      "loss": 0.3672,
+      "step": 9134
+    },
+    {
+      "epoch": 1.8008675078864353,
+      "grad_norm": 0.5323631360585213,
+      "learning_rate": 1.1572917338484471e-05,
+      "loss": 0.3742,
+      "step": 9135
+    },
+    {
+      "epoch": 1.801064668769716,
+      "grad_norm": 0.5031109008850254,
+      "learning_rate": 1.1571386588059236e-05,
+      "loss": 0.3406,
+      "step": 9136
+    },
+    {
+      "epoch": 1.8012618296529967,
+      "grad_norm": 0.49254801312177554,
+      "learning_rate": 1.1569855799880174e-05,
+      "loss": 0.3412,
+      "step": 9137
+    },
+    {
+      "epoch": 1.8014589905362777,
+      "grad_norm": 0.4836870483836016,
+      "learning_rate": 1.1568324973984065e-05,
+      "loss": 0.3256,
+      "step": 9138
+    },
+    {
+      "epoch": 1.8016561514195584,
+      "grad_norm": 0.5079333391604857,
+      "learning_rate": 1.1566794110407681e-05,
+      "loss": 0.3129,
+      "step": 9139
+    },
+    {
+      "epoch": 1.801853312302839,
+      "grad_norm": 0.48306063016614514,
+      "learning_rate": 1.156526320918781e-05,
+      "loss": 0.3263,
+      "step": 9140
+    },
+    {
+      "epoch": 1.80205047318612,
+      "grad_norm": 0.46236068282322035,
+      "learning_rate": 1.1563732270361228e-05,
+      "loss": 0.3218,
+      "step": 9141
+    },
+    {
+      "epoch": 1.8022476340694005,
+      "grad_norm": 0.46822948046376184,
+      "learning_rate": 1.1562201293964716e-05,
+      "loss": 0.3159,
+      "step": 9142
+    },
+    {
+      "epoch": 1.8024447949526814,
+      "grad_norm": 0.457275091142312,
+      "learning_rate": 1.1560670280035065e-05,
+      "loss": 0.3422,
+      "step": 9143
+    },
+    {
+      "epoch": 1.802641955835962,
+      "grad_norm": 0.45010333968693195,
+      "learning_rate": 1.155913922860905e-05,
+      "loss": 0.3038,
+      "step": 9144
+    },
+    {
+      "epoch": 1.8028391167192428,
+      "grad_norm": 0.460887214721349,
+      "learning_rate": 1.155760813972346e-05,
+      "loss": 0.3184,
+      "step": 9145
+    },
+    {
+      "epoch": 1.8030362776025237,
+      "grad_norm": 0.47074306673977423,
+      "learning_rate": 1.1556077013415084e-05,
+      "loss": 0.327,
+      "step": 9146
+    },
+    {
+      "epoch": 1.8032334384858044,
+      "grad_norm": 0.44541887094425797,
+      "learning_rate": 1.15545458497207e-05,
+      "loss": 0.314,
+      "step": 9147
+    },
+    {
+      "epoch": 1.8034305993690851,
+      "grad_norm": 0.46761848114178406,
+      "learning_rate": 1.1553014648677104e-05,
+      "loss": 0.3236,
+      "step": 9148
+    },
+    {
+      "epoch": 1.803627760252366,
+      "grad_norm": 0.47178941550054526,
+      "learning_rate": 1.1551483410321075e-05,
+      "loss": 0.3619,
+      "step": 9149
+    },
+    {
+      "epoch": 1.8038249211356467,
+      "grad_norm": 0.5060596355050806,
+      "learning_rate": 1.1549952134689414e-05,
+      "loss": 0.3555,
+      "step": 9150
+    },
+    {
+      "epoch": 1.8040220820189274,
+      "grad_norm": 0.48942752728337136,
+      "learning_rate": 1.1548420821818902e-05,
+      "loss": 0.3512,
+      "step": 9151
+    },
+    {
+      "epoch": 1.8042192429022084,
+      "grad_norm": 0.4656609142013564,
+      "learning_rate": 1.1546889471746333e-05,
+      "loss": 0.3254,
+      "step": 9152
+    },
+    {
+      "epoch": 1.8044164037854888,
+      "grad_norm": 0.5678872974738812,
+      "learning_rate": 1.1545358084508497e-05,
+      "loss": 0.3503,
+      "step": 9153
+    },
+    {
+      "epoch": 1.8046135646687698,
+      "grad_norm": 0.4746836161670529,
+      "learning_rate": 1.154382666014219e-05,
+      "loss": 0.3026,
+      "step": 9154
+    },
+    {
+      "epoch": 1.8048107255520505,
+      "grad_norm": 0.4919723933858667,
+      "learning_rate": 1.1542295198684206e-05,
+      "loss": 0.3567,
+      "step": 9155
+    },
+    {
+      "epoch": 1.8050078864353312,
+      "grad_norm": 0.4658920446396069,
+      "learning_rate": 1.1540763700171334e-05,
+      "loss": 0.3139,
+      "step": 9156
+    },
+    {
+      "epoch": 1.805205047318612,
+      "grad_norm": 0.4998610736336036,
+      "learning_rate": 1.1539232164640378e-05,
+      "loss": 0.3537,
+      "step": 9157
+    },
+    {
+      "epoch": 1.8054022082018928,
+      "grad_norm": 0.4838154631555964,
+      "learning_rate": 1.1537700592128126e-05,
+      "loss": 0.3231,
+      "step": 9158
+    },
+    {
+      "epoch": 1.8055993690851735,
+      "grad_norm": 0.481476456192672,
+      "learning_rate": 1.1536168982671378e-05,
+      "loss": 0.3587,
+      "step": 9159
+    },
+    {
+      "epoch": 1.8057965299684544,
+      "grad_norm": 0.6593255142170363,
+      "learning_rate": 1.1534637336306935e-05,
+      "loss": 0.3702,
+      "step": 9160
+    },
+    {
+      "epoch": 1.805993690851735,
+      "grad_norm": 0.47074519329216513,
+      "learning_rate": 1.1533105653071594e-05,
+      "loss": 0.3292,
+      "step": 9161
+    },
+    {
+      "epoch": 1.8061908517350158,
+      "grad_norm": 0.47104591661023276,
+      "learning_rate": 1.1531573933002156e-05,
+      "loss": 0.3273,
+      "step": 9162
+    },
+    {
+      "epoch": 1.8063880126182965,
+      "grad_norm": 0.4962259307049863,
+      "learning_rate": 1.153004217613542e-05,
+      "loss": 0.3381,
+      "step": 9163
+    },
+    {
+      "epoch": 1.8065851735015772,
+      "grad_norm": 0.47352216159289334,
+      "learning_rate": 1.152851038250819e-05,
+      "loss": 0.3293,
+      "step": 9164
+    },
+    {
+      "epoch": 1.8067823343848581,
+      "grad_norm": 0.4661605793322122,
+      "learning_rate": 1.1526978552157266e-05,
+      "loss": 0.3221,
+      "step": 9165
+    },
+    {
+      "epoch": 1.8069794952681388,
+      "grad_norm": 0.45860831067733454,
+      "learning_rate": 1.1525446685119452e-05,
+      "loss": 0.3341,
+      "step": 9166
+    },
+    {
+      "epoch": 1.8071766561514195,
+      "grad_norm": 0.4704413143236224,
+      "learning_rate": 1.1523914781431555e-05,
+      "loss": 0.3071,
+      "step": 9167
+    },
+    {
+      "epoch": 1.8073738170347005,
+      "grad_norm": 0.47559115700741583,
+      "learning_rate": 1.1522382841130377e-05,
+      "loss": 0.3467,
+      "step": 9168
+    },
+    {
+      "epoch": 1.807570977917981,
+      "grad_norm": 0.4852070321473458,
+      "learning_rate": 1.1520850864252724e-05,
+      "loss": 0.3132,
+      "step": 9169
+    },
+    {
+      "epoch": 1.8077681388012619,
+      "grad_norm": 0.5150731573019411,
+      "learning_rate": 1.1519318850835406e-05,
+      "loss": 0.3426,
+      "step": 9170
+    },
+    {
+      "epoch": 1.8079652996845426,
+      "grad_norm": 0.46442119276118826,
+      "learning_rate": 1.1517786800915229e-05,
+      "loss": 0.318,
+      "step": 9171
+    },
+    {
+      "epoch": 1.8081624605678233,
+      "grad_norm": 0.47760132684470824,
+      "learning_rate": 1.1516254714529001e-05,
+      "loss": 0.3138,
+      "step": 9172
+    },
+    {
+      "epoch": 1.8083596214511042,
+      "grad_norm": 0.5063732452560264,
+      "learning_rate": 1.1514722591713529e-05,
+      "loss": 0.3535,
+      "step": 9173
+    },
+    {
+      "epoch": 1.8085567823343849,
+      "grad_norm": 0.46353551339902077,
+      "learning_rate": 1.1513190432505634e-05,
+      "loss": 0.3085,
+      "step": 9174
+    },
+    {
+      "epoch": 1.8087539432176656,
+      "grad_norm": 0.45193354361169114,
+      "learning_rate": 1.1511658236942114e-05,
+      "loss": 0.2871,
+      "step": 9175
+    },
+    {
+      "epoch": 1.8089511041009465,
+      "grad_norm": 0.5095923560172854,
+      "learning_rate": 1.1510126005059793e-05,
+      "loss": 0.3514,
+      "step": 9176
+    },
+    {
+      "epoch": 1.809148264984227,
+      "grad_norm": 0.45401003471081053,
+      "learning_rate": 1.1508593736895475e-05,
+      "loss": 0.32,
+      "step": 9177
+    },
+    {
+      "epoch": 1.809345425867508,
+      "grad_norm": 0.4670090425518402,
+      "learning_rate": 1.150706143248598e-05,
+      "loss": 0.3338,
+      "step": 9178
+    },
+    {
+      "epoch": 1.8095425867507886,
+      "grad_norm": 0.49260267310440076,
+      "learning_rate": 1.1505529091868117e-05,
+      "loss": 0.3646,
+      "step": 9179
+    },
+    {
+      "epoch": 1.8097397476340693,
+      "grad_norm": 0.4892695211612479,
+      "learning_rate": 1.1503996715078707e-05,
+      "loss": 0.3159,
+      "step": 9180
+    },
+    {
+      "epoch": 1.8099369085173502,
+      "grad_norm": 0.5269511149927203,
+      "learning_rate": 1.1502464302154566e-05,
+      "loss": 0.3476,
+      "step": 9181
+    },
+    {
+      "epoch": 1.810134069400631,
+      "grad_norm": 0.477149389147672,
+      "learning_rate": 1.150093185313251e-05,
+      "loss": 0.3394,
+      "step": 9182
+    },
+    {
+      "epoch": 1.8103312302839116,
+      "grad_norm": 0.4911874732780398,
+      "learning_rate": 1.1499399368049356e-05,
+      "loss": 0.3367,
+      "step": 9183
+    },
+    {
+      "epoch": 1.8105283911671926,
+      "grad_norm": 0.7822176448681528,
+      "learning_rate": 1.1497866846941926e-05,
+      "loss": 0.3192,
+      "step": 9184
+    },
+    {
+      "epoch": 1.810725552050473,
+      "grad_norm": 0.4936248418394106,
+      "learning_rate": 1.1496334289847038e-05,
+      "loss": 0.3429,
+      "step": 9185
+    },
+    {
+      "epoch": 1.810922712933754,
+      "grad_norm": 0.47779103622401925,
+      "learning_rate": 1.1494801696801515e-05,
+      "loss": 0.3391,
+      "step": 9186
+    },
+    {
+      "epoch": 1.8111198738170347,
+      "grad_norm": 0.4911635374833865,
+      "learning_rate": 1.1493269067842175e-05,
+      "loss": 0.3392,
+      "step": 9187
+    },
+    {
+      "epoch": 1.8113170347003154,
+      "grad_norm": 0.4604725274713814,
+      "learning_rate": 1.1491736403005844e-05,
+      "loss": 0.324,
+      "step": 9188
+    },
+    {
+      "epoch": 1.8115141955835963,
+      "grad_norm": 0.5032560937410412,
+      "learning_rate": 1.1490203702329346e-05,
+      "loss": 0.3603,
+      "step": 9189
+    },
+    {
+      "epoch": 1.811711356466877,
+      "grad_norm": 0.6385600610637551,
+      "learning_rate": 1.1488670965849505e-05,
+      "loss": 0.3369,
+      "step": 9190
+    },
+    {
+      "epoch": 1.8119085173501577,
+      "grad_norm": 0.4885866435640685,
+      "learning_rate": 1.1487138193603142e-05,
+      "loss": 0.3309,
+      "step": 9191
+    },
+    {
+      "epoch": 1.8121056782334386,
+      "grad_norm": 0.5183397825685779,
+      "learning_rate": 1.1485605385627088e-05,
+      "loss": 0.3343,
+      "step": 9192
+    },
+    {
+      "epoch": 1.812302839116719,
+      "grad_norm": 0.4989032074820882,
+      "learning_rate": 1.1484072541958167e-05,
+      "loss": 0.3397,
+      "step": 9193
+    },
+    {
+      "epoch": 1.8125,
+      "grad_norm": 0.5033661808175196,
+      "learning_rate": 1.1482539662633208e-05,
+      "loss": 0.3417,
+      "step": 9194
+    },
+    {
+      "epoch": 1.812697160883281,
+      "grad_norm": 5.86204139038015,
+      "learning_rate": 1.1481006747689043e-05,
+      "loss": 0.328,
+      "step": 9195
+    },
+    {
+      "epoch": 1.8128943217665614,
+      "grad_norm": 0.5051699316872136,
+      "learning_rate": 1.1479473797162492e-05,
+      "loss": 0.3173,
+      "step": 9196
+    },
+    {
+      "epoch": 1.8130914826498423,
+      "grad_norm": 0.5070080136339032,
+      "learning_rate": 1.1477940811090398e-05,
+      "loss": 0.3271,
+      "step": 9197
+    },
+    {
+      "epoch": 1.813288643533123,
+      "grad_norm": 0.4811441468791982,
+      "learning_rate": 1.1476407789509583e-05,
+      "loss": 0.3268,
+      "step": 9198
+    },
+    {
+      "epoch": 1.8134858044164037,
+      "grad_norm": 0.4785568333077474,
+      "learning_rate": 1.1474874732456884e-05,
+      "loss": 0.3212,
+      "step": 9199
+    },
+    {
+      "epoch": 1.8136829652996846,
+      "grad_norm": 0.4698877579251847,
+      "learning_rate": 1.147334163996913e-05,
+      "loss": 0.3072,
+      "step": 9200
+    },
+    {
+      "epoch": 1.8138801261829653,
+      "grad_norm": 0.47991128581690556,
+      "learning_rate": 1.1471808512083156e-05,
+      "loss": 0.3416,
+      "step": 9201
+    },
+    {
+      "epoch": 1.814077287066246,
+      "grad_norm": 0.46961744858878407,
+      "learning_rate": 1.1470275348835797e-05,
+      "loss": 0.3187,
+      "step": 9202
+    },
+    {
+      "epoch": 1.814274447949527,
+      "grad_norm": 0.5056112272060148,
+      "learning_rate": 1.146874215026389e-05,
+      "loss": 0.3397,
+      "step": 9203
+    },
+    {
+      "epoch": 1.8144716088328074,
+      "grad_norm": 0.4972247643473326,
+      "learning_rate": 1.1467208916404271e-05,
+      "loss": 0.3404,
+      "step": 9204
+    },
+    {
+      "epoch": 1.8146687697160884,
+      "grad_norm": 0.4723317093599732,
+      "learning_rate": 1.1465675647293772e-05,
+      "loss": 0.304,
+      "step": 9205
+    },
+    {
+      "epoch": 1.814865930599369,
+      "grad_norm": 0.5198317290369266,
+      "learning_rate": 1.1464142342969242e-05,
+      "loss": 0.3535,
+      "step": 9206
+    },
+    {
+      "epoch": 1.8150630914826498,
+      "grad_norm": 2.3677318794574895,
+      "learning_rate": 1.1462609003467508e-05,
+      "loss": 0.4222,
+      "step": 9207
+    },
+    {
+      "epoch": 1.8152602523659307,
+      "grad_norm": 0.4794343279462834,
+      "learning_rate": 1.1461075628825416e-05,
+      "loss": 0.3156,
+      "step": 9208
+    },
+    {
+      "epoch": 1.8154574132492114,
+      "grad_norm": 0.48662660979986266,
+      "learning_rate": 1.1459542219079808e-05,
+      "loss": 0.3432,
+      "step": 9209
+    },
+    {
+      "epoch": 1.815654574132492,
+      "grad_norm": 0.48035488934776704,
+      "learning_rate": 1.1458008774267518e-05,
+      "loss": 0.3383,
+      "step": 9210
+    },
+    {
+      "epoch": 1.815851735015773,
+      "grad_norm": 0.48382873384961195,
+      "learning_rate": 1.1456475294425396e-05,
+      "loss": 0.327,
+      "step": 9211
+    },
+    {
+      "epoch": 1.8160488958990535,
+      "grad_norm": 0.47241477963626444,
+      "learning_rate": 1.1454941779590283e-05,
+      "loss": 0.3271,
+      "step": 9212
+    },
+    {
+      "epoch": 1.8162460567823344,
+      "grad_norm": 0.46721123265648445,
+      "learning_rate": 1.1453408229799017e-05,
+      "loss": 0.3125,
+      "step": 9213
+    },
+    {
+      "epoch": 1.8164432176656151,
+      "grad_norm": 0.4739931333373127,
+      "learning_rate": 1.1451874645088455e-05,
+      "loss": 0.3265,
+      "step": 9214
+    },
+    {
+      "epoch": 1.8166403785488958,
+      "grad_norm": 0.5189118810899088,
+      "learning_rate": 1.145034102549543e-05,
+      "loss": 0.3672,
+      "step": 9215
+    },
+    {
+      "epoch": 1.8168375394321767,
+      "grad_norm": 0.49357938108124044,
+      "learning_rate": 1.1448807371056798e-05,
+      "loss": 0.3343,
+      "step": 9216
+    },
+    {
+      "epoch": 1.8170347003154574,
+      "grad_norm": 0.476806812670295,
+      "learning_rate": 1.14472736818094e-05,
+      "loss": 0.3544,
+      "step": 9217
+    },
+    {
+      "epoch": 1.8172318611987381,
+      "grad_norm": 0.5426318526785301,
+      "learning_rate": 1.1445739957790087e-05,
+      "loss": 0.3388,
+      "step": 9218
+    },
+    {
+      "epoch": 1.817429022082019,
+      "grad_norm": 0.48940437942449916,
+      "learning_rate": 1.1444206199035708e-05,
+      "loss": 0.3436,
+      "step": 9219
+    },
+    {
+      "epoch": 1.8176261829652995,
+      "grad_norm": 0.47285073767586155,
+      "learning_rate": 1.1442672405583109e-05,
+      "loss": 0.342,
+      "step": 9220
+    },
+    {
+      "epoch": 1.8178233438485805,
+      "grad_norm": 0.4668025709102716,
+      "learning_rate": 1.1441138577469147e-05,
+      "loss": 0.327,
+      "step": 9221
+    },
+    {
+      "epoch": 1.8180205047318612,
+      "grad_norm": 0.45875055734217174,
+      "learning_rate": 1.1439604714730666e-05,
+      "loss": 0.3192,
+      "step": 9222
+    },
+    {
+      "epoch": 1.8182176656151419,
+      "grad_norm": 0.4409526335848642,
+      "learning_rate": 1.1438070817404527e-05,
+      "loss": 0.3101,
+      "step": 9223
+    },
+    {
+      "epoch": 1.8184148264984228,
+      "grad_norm": 0.45114553583024647,
+      "learning_rate": 1.1436536885527576e-05,
+      "loss": 0.3342,
+      "step": 9224
+    },
+    {
+      "epoch": 1.8186119873817035,
+      "grad_norm": 0.4806686518283818,
+      "learning_rate": 1.1435002919136671e-05,
+      "loss": 0.3391,
+      "step": 9225
+    },
+    {
+      "epoch": 1.8188091482649842,
+      "grad_norm": 0.5044312994268182,
+      "learning_rate": 1.1433468918268663e-05,
+      "loss": 0.3555,
+      "step": 9226
+    },
+    {
+      "epoch": 1.819006309148265,
+      "grad_norm": 0.49263755874516185,
+      "learning_rate": 1.1431934882960412e-05,
+      "loss": 0.3607,
+      "step": 9227
+    },
+    {
+      "epoch": 1.8192034700315456,
+      "grad_norm": 0.4536998690340305,
+      "learning_rate": 1.1430400813248772e-05,
+      "loss": 0.2918,
+      "step": 9228
+    },
+    {
+      "epoch": 1.8194006309148265,
+      "grad_norm": 0.532141725281099,
+      "learning_rate": 1.1428866709170599e-05,
+      "loss": 0.355,
+      "step": 9229
+    },
+    {
+      "epoch": 1.8195977917981072,
+      "grad_norm": 0.4635187686939784,
+      "learning_rate": 1.1427332570762754e-05,
+      "loss": 0.3211,
+      "step": 9230
+    },
+    {
+      "epoch": 1.819794952681388,
+      "grad_norm": 0.602694804528996,
+      "learning_rate": 1.1425798398062093e-05,
+      "loss": 0.3574,
+      "step": 9231
+    },
+    {
+      "epoch": 1.8199921135646688,
+      "grad_norm": 0.49418051715015165,
+      "learning_rate": 1.1424264191105481e-05,
+      "loss": 0.3526,
+      "step": 9232
+    },
+    {
+      "epoch": 1.8201892744479495,
+      "grad_norm": 0.47528793595220115,
+      "learning_rate": 1.1422729949929772e-05,
+      "loss": 0.3408,
+      "step": 9233
+    },
+    {
+      "epoch": 1.8203864353312302,
+      "grad_norm": 0.5040867868946186,
+      "learning_rate": 1.142119567457183e-05,
+      "loss": 0.3401,
+      "step": 9234
+    },
+    {
+      "epoch": 1.8205835962145112,
+      "grad_norm": 0.4709221541985399,
+      "learning_rate": 1.141966136506852e-05,
+      "loss": 0.322,
+      "step": 9235
+    },
+    {
+      "epoch": 1.8207807570977916,
+      "grad_norm": 0.5048500824063943,
+      "learning_rate": 1.14181270214567e-05,
+      "loss": 0.3449,
+      "step": 9236
+    },
+    {
+      "epoch": 1.8209779179810726,
+      "grad_norm": 0.5520156337084202,
+      "learning_rate": 1.1416592643773236e-05,
+      "loss": 0.3485,
+      "step": 9237
+    },
+    {
+      "epoch": 1.8211750788643533,
+      "grad_norm": 0.4662159013287956,
+      "learning_rate": 1.1415058232054995e-05,
+      "loss": 0.3116,
+      "step": 9238
+    },
+    {
+      "epoch": 1.821372239747634,
+      "grad_norm": 0.4583750639044251,
+      "learning_rate": 1.1413523786338838e-05,
+      "loss": 0.314,
+      "step": 9239
+    },
+    {
+      "epoch": 1.8215694006309149,
+      "grad_norm": 0.47754415155408275,
+      "learning_rate": 1.1411989306661635e-05,
+      "loss": 0.3322,
+      "step": 9240
+    },
+    {
+      "epoch": 1.8217665615141956,
+      "grad_norm": 0.4974506056160547,
+      "learning_rate": 1.1410454793060251e-05,
+      "loss": 0.3507,
+      "step": 9241
+    },
+    {
+      "epoch": 1.8219637223974763,
+      "grad_norm": 0.4734526011420336,
+      "learning_rate": 1.1408920245571558e-05,
+      "loss": 0.3168,
+      "step": 9242
+    },
+    {
+      "epoch": 1.8221608832807572,
+      "grad_norm": 0.46295318137196184,
+      "learning_rate": 1.1407385664232415e-05,
+      "loss": 0.3195,
+      "step": 9243
+    },
+    {
+      "epoch": 1.822358044164038,
+      "grad_norm": 0.4997671967465472,
+      "learning_rate": 1.1405851049079706e-05,
+      "loss": 0.3458,
+      "step": 9244
+    },
+    {
+      "epoch": 1.8225552050473186,
+      "grad_norm": 0.4727316629472926,
+      "learning_rate": 1.1404316400150288e-05,
+      "loss": 0.3358,
+      "step": 9245
+    },
+    {
+      "epoch": 1.8227523659305995,
+      "grad_norm": 0.4600159341304322,
+      "learning_rate": 1.1402781717481042e-05,
+      "loss": 0.3019,
+      "step": 9246
+    },
+    {
+      "epoch": 1.82294952681388,
+      "grad_norm": 0.4832490316233833,
+      "learning_rate": 1.1401247001108828e-05,
+      "loss": 0.3215,
+      "step": 9247
+    },
+    {
+      "epoch": 1.823146687697161,
+      "grad_norm": 0.4552192341859621,
+      "learning_rate": 1.1399712251070532e-05,
+      "loss": 0.3036,
+      "step": 9248
+    },
+    {
+      "epoch": 1.8233438485804416,
+      "grad_norm": 0.45254379685961227,
+      "learning_rate": 1.1398177467403022e-05,
+      "loss": 0.3207,
+      "step": 9249
+    },
+    {
+      "epoch": 1.8235410094637223,
+      "grad_norm": 0.48097837033082497,
+      "learning_rate": 1.1396642650143171e-05,
+      "loss": 0.3383,
+      "step": 9250
+    },
+    {
+      "epoch": 1.8237381703470033,
+      "grad_norm": 2.4890867749317755,
+      "learning_rate": 1.1395107799327856e-05,
+      "loss": 0.3532,
+      "step": 9251
+    },
+    {
+      "epoch": 1.823935331230284,
+      "grad_norm": 0.5014918794110886,
+      "learning_rate": 1.1393572914993954e-05,
+      "loss": 0.3663,
+      "step": 9252
+    },
+    {
+      "epoch": 1.8241324921135647,
+      "grad_norm": 0.5245523336186161,
+      "learning_rate": 1.1392037997178338e-05,
+      "loss": 0.3445,
+      "step": 9253
+    },
+    {
+      "epoch": 1.8243296529968456,
+      "grad_norm": 0.5134602963738575,
+      "learning_rate": 1.1390503045917892e-05,
+      "loss": 0.3631,
+      "step": 9254
+    },
+    {
+      "epoch": 1.824526813880126,
+      "grad_norm": 0.4701559391883266,
+      "learning_rate": 1.1388968061249486e-05,
+      "loss": 0.3283,
+      "step": 9255
+    },
+    {
+      "epoch": 1.824723974763407,
+      "grad_norm": 0.47107381684930266,
+      "learning_rate": 1.1387433043210006e-05,
+      "loss": 0.3278,
+      "step": 9256
+    },
+    {
+      "epoch": 1.8249211356466877,
+      "grad_norm": 0.47653058629900974,
+      "learning_rate": 1.138589799183633e-05,
+      "loss": 0.3238,
+      "step": 9257
+    },
+    {
+      "epoch": 1.8251182965299684,
+      "grad_norm": 0.48085309963347217,
+      "learning_rate": 1.138436290716534e-05,
+      "loss": 0.3355,
+      "step": 9258
+    },
+    {
+      "epoch": 1.8253154574132493,
+      "grad_norm": 0.46243620388093765,
+      "learning_rate": 1.1382827789233912e-05,
+      "loss": 0.3259,
+      "step": 9259
+    },
+    {
+      "epoch": 1.82551261829653,
+      "grad_norm": 0.45802315988824477,
+      "learning_rate": 1.1381292638078935e-05,
+      "loss": 0.3149,
+      "step": 9260
+    },
+    {
+      "epoch": 1.8257097791798107,
+      "grad_norm": 0.4794295563766952,
+      "learning_rate": 1.1379757453737293e-05,
+      "loss": 0.3337,
+      "step": 9261
+    },
+    {
+      "epoch": 1.8259069400630916,
+      "grad_norm": 0.4758846030069048,
+      "learning_rate": 1.1378222236245862e-05,
+      "loss": 0.324,
+      "step": 9262
+    },
+    {
+      "epoch": 1.826104100946372,
+      "grad_norm": 0.4709615758764996,
+      "learning_rate": 1.1376686985641536e-05,
+      "loss": 0.3223,
+      "step": 9263
+    },
+    {
+      "epoch": 1.826301261829653,
+      "grad_norm": 0.49999866958616157,
+      "learning_rate": 1.1375151701961191e-05,
+      "loss": 0.3621,
+      "step": 9264
+    },
+    {
+      "epoch": 1.8264984227129337,
+      "grad_norm": 0.5326305619030082,
+      "learning_rate": 1.1373616385241726e-05,
+      "loss": 0.3455,
+      "step": 9265
+    },
+    {
+      "epoch": 1.8266955835962144,
+      "grad_norm": 0.45691910409746533,
+      "learning_rate": 1.1372081035520015e-05,
+      "loss": 0.3085,
+      "step": 9266
+    },
+    {
+      "epoch": 1.8268927444794953,
+      "grad_norm": 0.4587323635393261,
+      "learning_rate": 1.1370545652832958e-05,
+      "loss": 0.3322,
+      "step": 9267
+    },
+    {
+      "epoch": 1.827089905362776,
+      "grad_norm": 0.4826907976994329,
+      "learning_rate": 1.1369010237217435e-05,
+      "loss": 0.3391,
+      "step": 9268
+    },
+    {
+      "epoch": 1.8272870662460567,
+      "grad_norm": 0.4816253057556002,
+      "learning_rate": 1.1367474788710338e-05,
+      "loss": 0.3408,
+      "step": 9269
+    },
+    {
+      "epoch": 1.8274842271293377,
+      "grad_norm": 0.5249937140583107,
+      "learning_rate": 1.1365939307348559e-05,
+      "loss": 0.3464,
+      "step": 9270
+    },
+    {
+      "epoch": 1.8276813880126181,
+      "grad_norm": 0.48583193536640634,
+      "learning_rate": 1.1364403793168988e-05,
+      "loss": 0.3469,
+      "step": 9271
+    },
+    {
+      "epoch": 1.827878548895899,
+      "grad_norm": 0.5051024554952468,
+      "learning_rate": 1.1362868246208519e-05,
+      "loss": 0.3261,
+      "step": 9272
+    },
+    {
+      "epoch": 1.8280757097791798,
+      "grad_norm": 0.4968850771719625,
+      "learning_rate": 1.1361332666504038e-05,
+      "loss": 0.3272,
+      "step": 9273
+    },
+    {
+      "epoch": 1.8282728706624605,
+      "grad_norm": 0.47294764285242963,
+      "learning_rate": 1.135979705409245e-05,
+      "loss": 0.3247,
+      "step": 9274
+    },
+    {
+      "epoch": 1.8284700315457414,
+      "grad_norm": 0.4606069743822721,
+      "learning_rate": 1.1358261409010636e-05,
+      "loss": 0.3035,
+      "step": 9275
+    },
+    {
+      "epoch": 1.828667192429022,
+      "grad_norm": 0.4771239186952894,
+      "learning_rate": 1.1356725731295501e-05,
+      "loss": 0.332,
+      "step": 9276
+    },
+    {
+      "epoch": 1.8288643533123028,
+      "grad_norm": 0.4478351388311611,
+      "learning_rate": 1.1355190020983937e-05,
+      "loss": 0.3076,
+      "step": 9277
+    },
+    {
+      "epoch": 1.8290615141955837,
+      "grad_norm": 0.46550837164514514,
+      "learning_rate": 1.1353654278112841e-05,
+      "loss": 0.3273,
+      "step": 9278
+    },
+    {
+      "epoch": 1.8292586750788642,
+      "grad_norm": 0.4811875611417182,
+      "learning_rate": 1.1352118502719115e-05,
+      "loss": 0.3424,
+      "step": 9279
+    },
+    {
+      "epoch": 1.8294558359621451,
+      "grad_norm": 0.4600111130971501,
+      "learning_rate": 1.135058269483965e-05,
+      "loss": 0.3362,
+      "step": 9280
+    },
+    {
+      "epoch": 1.8296529968454258,
+      "grad_norm": 0.4475616886590054,
+      "learning_rate": 1.1349046854511347e-05,
+      "loss": 0.3139,
+      "step": 9281
+    },
+    {
+      "epoch": 1.8298501577287065,
+      "grad_norm": 0.4806604678847091,
+      "learning_rate": 1.1347510981771108e-05,
+      "loss": 0.3672,
+      "step": 9282
+    },
+    {
+      "epoch": 1.8300473186119874,
+      "grad_norm": 0.5175341707370579,
+      "learning_rate": 1.1345975076655832e-05,
+      "loss": 0.339,
+      "step": 9283
+    },
+    {
+      "epoch": 1.8302444794952681,
+      "grad_norm": 0.5158084398771197,
+      "learning_rate": 1.134443913920242e-05,
+      "loss": 0.3422,
+      "step": 9284
+    },
+    {
+      "epoch": 1.8304416403785488,
+      "grad_norm": 0.44628621178473027,
+      "learning_rate": 1.1342903169447778e-05,
+      "loss": 0.3123,
+      "step": 9285
+    },
+    {
+      "epoch": 1.8306388012618298,
+      "grad_norm": 0.491244561834142,
+      "learning_rate": 1.1341367167428806e-05,
+      "loss": 0.3356,
+      "step": 9286
+    },
+    {
+      "epoch": 1.8308359621451105,
+      "grad_norm": 0.4618159349464898,
+      "learning_rate": 1.1339831133182405e-05,
+      "loss": 0.3195,
+      "step": 9287
+    },
+    {
+      "epoch": 1.8310331230283912,
+      "grad_norm": 0.46357096939810233,
+      "learning_rate": 1.1338295066745482e-05,
+      "loss": 0.3277,
+      "step": 9288
+    },
+    {
+      "epoch": 1.831230283911672,
+      "grad_norm": 0.44971140458495884,
+      "learning_rate": 1.1336758968154943e-05,
+      "loss": 0.3236,
+      "step": 9289
+    },
+    {
+      "epoch": 1.8314274447949526,
+      "grad_norm": 0.4528759786598488,
+      "learning_rate": 1.1335222837447692e-05,
+      "loss": 0.3091,
+      "step": 9290
+    },
+    {
+      "epoch": 1.8316246056782335,
+      "grad_norm": 0.4797326570038922,
+      "learning_rate": 1.1333686674660643e-05,
+      "loss": 0.3388,
+      "step": 9291
+    },
+    {
+      "epoch": 1.8318217665615142,
+      "grad_norm": 0.9510910814957004,
+      "learning_rate": 1.133215047983069e-05,
+      "loss": 0.3446,
+      "step": 9292
+    },
+    {
+      "epoch": 1.8320189274447949,
+      "grad_norm": 0.47801369290214624,
+      "learning_rate": 1.1330614252994753e-05,
+      "loss": 0.337,
+      "step": 9293
+    },
+    {
+      "epoch": 1.8322160883280758,
+      "grad_norm": 0.4798818841814811,
+      "learning_rate": 1.1329077994189736e-05,
+      "loss": 0.3283,
+      "step": 9294
+    },
+    {
+      "epoch": 1.8324132492113565,
+      "grad_norm": 0.4649250612908643,
+      "learning_rate": 1.132754170345255e-05,
+      "loss": 0.3143,
+      "step": 9295
+    },
+    {
+      "epoch": 1.8326104100946372,
+      "grad_norm": 0.5111682031018355,
+      "learning_rate": 1.1326005380820106e-05,
+      "loss": 0.3489,
+      "step": 9296
+    },
+    {
+      "epoch": 1.8328075709779181,
+      "grad_norm": 0.479500042771139,
+      "learning_rate": 1.1324469026329314e-05,
+      "loss": 0.3467,
+      "step": 9297
+    },
+    {
+      "epoch": 1.8330047318611986,
+      "grad_norm": 0.48760322704258524,
+      "learning_rate": 1.1322932640017087e-05,
+      "loss": 0.338,
+      "step": 9298
+    },
+    {
+      "epoch": 1.8332018927444795,
+      "grad_norm": 0.4615980037814281,
+      "learning_rate": 1.132139622192034e-05,
+      "loss": 0.3211,
+      "step": 9299
+    },
+    {
+      "epoch": 1.8333990536277602,
+      "grad_norm": 0.478582392329709,
+      "learning_rate": 1.1319859772075982e-05,
+      "loss": 0.3437,
+      "step": 9300
+    },
+    {
+      "epoch": 1.833596214511041,
+      "grad_norm": 0.45732711770150825,
+      "learning_rate": 1.1318323290520935e-05,
+      "loss": 0.307,
+      "step": 9301
+    },
+    {
+      "epoch": 1.8337933753943219,
+      "grad_norm": 0.4785336546942506,
+      "learning_rate": 1.1316786777292103e-05,
+      "loss": 0.3467,
+      "step": 9302
+    },
+    {
+      "epoch": 1.8339905362776026,
+      "grad_norm": 0.48770360639670446,
+      "learning_rate": 1.1315250232426411e-05,
+      "loss": 0.3178,
+      "step": 9303
+    },
+    {
+      "epoch": 1.8341876971608833,
+      "grad_norm": 0.4909997945985066,
+      "learning_rate": 1.1313713655960773e-05,
+      "loss": 0.339,
+      "step": 9304
+    },
+    {
+      "epoch": 1.8343848580441642,
+      "grad_norm": 0.46943047019703255,
+      "learning_rate": 1.1312177047932107e-05,
+      "loss": 0.3233,
+      "step": 9305
+    },
+    {
+      "epoch": 1.8345820189274447,
+      "grad_norm": 0.4637968220251219,
+      "learning_rate": 1.1310640408377331e-05,
+      "loss": 0.3354,
+      "step": 9306
+    },
+    {
+      "epoch": 1.8347791798107256,
+      "grad_norm": 0.4811381690374539,
+      "learning_rate": 1.1309103737333363e-05,
+      "loss": 0.3206,
+      "step": 9307
+    },
+    {
+      "epoch": 1.8349763406940063,
+      "grad_norm": 0.4841461834363835,
+      "learning_rate": 1.1307567034837123e-05,
+      "loss": 0.3355,
+      "step": 9308
+    },
+    {
+      "epoch": 1.835173501577287,
+      "grad_norm": 0.5075047264571099,
+      "learning_rate": 1.1306030300925531e-05,
+      "loss": 0.3364,
+      "step": 9309
+    },
+    {
+      "epoch": 1.835370662460568,
+      "grad_norm": 0.48761090026693843,
+      "learning_rate": 1.1304493535635512e-05,
+      "loss": 0.334,
+      "step": 9310
+    },
+    {
+      "epoch": 1.8355678233438486,
+      "grad_norm": 0.4707017965699666,
+      "learning_rate": 1.1302956739003981e-05,
+      "loss": 0.3206,
+      "step": 9311
+    },
+    {
+      "epoch": 1.8357649842271293,
+      "grad_norm": 0.47390782968886325,
+      "learning_rate": 1.1301419911067871e-05,
+      "loss": 0.3581,
+      "step": 9312
+    },
+    {
+      "epoch": 1.8359621451104102,
+      "grad_norm": 0.5179730801274194,
+      "learning_rate": 1.1299883051864095e-05,
+      "loss": 0.3345,
+      "step": 9313
+    },
+    {
+      "epoch": 1.8361593059936907,
+      "grad_norm": 0.4565817457776229,
+      "learning_rate": 1.1298346161429585e-05,
+      "loss": 0.324,
+      "step": 9314
+    },
+    {
+      "epoch": 1.8363564668769716,
+      "grad_norm": 0.4743170511252832,
+      "learning_rate": 1.1296809239801258e-05,
+      "loss": 0.3257,
+      "step": 9315
+    },
+    {
+      "epoch": 1.8365536277602523,
+      "grad_norm": 0.5062603340279239,
+      "learning_rate": 1.129527228701605e-05,
+      "loss": 0.3397,
+      "step": 9316
+    },
+    {
+      "epoch": 1.836750788643533,
+      "grad_norm": 0.4627524681098276,
+      "learning_rate": 1.129373530311088e-05,
+      "loss": 0.3008,
+      "step": 9317
+    },
+    {
+      "epoch": 1.836947949526814,
+      "grad_norm": 0.45366614630787505,
+      "learning_rate": 1.1292198288122678e-05,
+      "loss": 0.3239,
+      "step": 9318
+    },
+    {
+      "epoch": 1.8371451104100947,
+      "grad_norm": 0.49553583595035205,
+      "learning_rate": 1.1290661242088373e-05,
+      "loss": 0.3335,
+      "step": 9319
+    },
+    {
+      "epoch": 1.8373422712933754,
+      "grad_norm": 0.4656239756890831,
+      "learning_rate": 1.1289124165044889e-05,
+      "loss": 0.3343,
+      "step": 9320
+    },
+    {
+      "epoch": 1.8375394321766563,
+      "grad_norm": 0.4592207478297998,
+      "learning_rate": 1.1287587057029164e-05,
+      "loss": 0.31,
+      "step": 9321
+    },
+    {
+      "epoch": 1.8377365930599368,
+      "grad_norm": 0.48719866177332166,
+      "learning_rate": 1.1286049918078118e-05,
+      "loss": 0.3601,
+      "step": 9322
+    },
+    {
+      "epoch": 1.8379337539432177,
+      "grad_norm": 0.4609017660568222,
+      "learning_rate": 1.1284512748228686e-05,
+      "loss": 0.3532,
+      "step": 9323
+    },
+    {
+      "epoch": 1.8381309148264984,
+      "grad_norm": 0.471845332576064,
+      "learning_rate": 1.1282975547517805e-05,
+      "loss": 0.3428,
+      "step": 9324
+    },
+    {
+      "epoch": 1.838328075709779,
+      "grad_norm": 0.45141311669092293,
+      "learning_rate": 1.1281438315982403e-05,
+      "loss": 0.3313,
+      "step": 9325
+    },
+    {
+      "epoch": 1.83852523659306,
+      "grad_norm": 0.4846124283156861,
+      "learning_rate": 1.127990105365941e-05,
+      "loss": 0.3437,
+      "step": 9326
+    },
+    {
+      "epoch": 1.8387223974763407,
+      "grad_norm": 0.4706449746560727,
+      "learning_rate": 1.1278363760585767e-05,
+      "loss": 0.3239,
+      "step": 9327
+    },
+    {
+      "epoch": 1.8389195583596214,
+      "grad_norm": 0.46592281748659253,
+      "learning_rate": 1.1276826436798406e-05,
+      "loss": 0.2927,
+      "step": 9328
+    },
+    {
+      "epoch": 1.8391167192429023,
+      "grad_norm": 0.4506091488970719,
+      "learning_rate": 1.1275289082334257e-05,
+      "loss": 0.3299,
+      "step": 9329
+    },
+    {
+      "epoch": 1.839313880126183,
+      "grad_norm": 0.49247797496055923,
+      "learning_rate": 1.1273751697230262e-05,
+      "loss": 0.3177,
+      "step": 9330
+    },
+    {
+      "epoch": 1.8395110410094637,
+      "grad_norm": 0.4823856031434032,
+      "learning_rate": 1.1272214281523359e-05,
+      "loss": 0.3369,
+      "step": 9331
+    },
+    {
+      "epoch": 1.8397082018927446,
+      "grad_norm": 0.44703802630892214,
+      "learning_rate": 1.127067683525048e-05,
+      "loss": 0.3009,
+      "step": 9332
+    },
+    {
+      "epoch": 1.8399053627760251,
+      "grad_norm": 0.4352564519007158,
+      "learning_rate": 1.1269139358448573e-05,
+      "loss": 0.2883,
+      "step": 9333
+    },
+    {
+      "epoch": 1.840102523659306,
+      "grad_norm": 0.4897196424734099,
+      "learning_rate": 1.1267601851154569e-05,
+      "loss": 0.3292,
+      "step": 9334
+    },
+    {
+      "epoch": 1.8402996845425867,
+      "grad_norm": 0.47061110718447907,
+      "learning_rate": 1.1266064313405404e-05,
+      "loss": 0.3198,
+      "step": 9335
+    },
+    {
+      "epoch": 1.8404968454258674,
+      "grad_norm": 0.4725374845188905,
+      "learning_rate": 1.1264526745238032e-05,
+      "loss": 0.3303,
+      "step": 9336
+    },
+    {
+      "epoch": 1.8406940063091484,
+      "grad_norm": 0.46014350127590964,
+      "learning_rate": 1.1262989146689378e-05,
+      "loss": 0.3047,
+      "step": 9337
+    },
+    {
+      "epoch": 1.840891167192429,
+      "grad_norm": 0.48716019997900245,
+      "learning_rate": 1.12614515177964e-05,
+      "loss": 0.3345,
+      "step": 9338
+    },
+    {
+      "epoch": 1.8410883280757098,
+      "grad_norm": 0.462419364697719,
+      "learning_rate": 1.125991385859603e-05,
+      "loss": 0.3345,
+      "step": 9339
+    },
+    {
+      "epoch": 1.8412854889589907,
+      "grad_norm": 0.4838141067825107,
+      "learning_rate": 1.1258376169125218e-05,
+      "loss": 0.3418,
+      "step": 9340
+    },
+    {
+      "epoch": 1.8414826498422712,
+      "grad_norm": 0.4638010611429982,
+      "learning_rate": 1.1256838449420902e-05,
+      "loss": 0.3397,
+      "step": 9341
+    },
+    {
+      "epoch": 1.841679810725552,
+      "grad_norm": 0.4742578303058376,
+      "learning_rate": 1.125530069952003e-05,
+      "loss": 0.3369,
+      "step": 9342
+    },
+    {
+      "epoch": 1.8418769716088328,
+      "grad_norm": 0.6012207890435203,
+      "learning_rate": 1.1253762919459548e-05,
+      "loss": 0.3922,
+      "step": 9343
+    },
+    {
+      "epoch": 1.8420741324921135,
+      "grad_norm": 0.4640385924151846,
+      "learning_rate": 1.1252225109276404e-05,
+      "loss": 0.3276,
+      "step": 9344
+    },
+    {
+      "epoch": 1.8422712933753944,
+      "grad_norm": 0.46720201438034464,
+      "learning_rate": 1.1250687269007544e-05,
+      "loss": 0.3231,
+      "step": 9345
+    },
+    {
+      "epoch": 1.8424684542586751,
+      "grad_norm": 0.4809439170465585,
+      "learning_rate": 1.1249149398689912e-05,
+      "loss": 0.3359,
+      "step": 9346
+    },
+    {
+      "epoch": 1.8426656151419558,
+      "grad_norm": 0.46092655938380556,
+      "learning_rate": 1.1247611498360463e-05,
+      "loss": 0.3288,
+      "step": 9347
+    },
+    {
+      "epoch": 1.8428627760252367,
+      "grad_norm": 0.46354010282488706,
+      "learning_rate": 1.124607356805614e-05,
+      "loss": 0.3085,
+      "step": 9348
+    },
+    {
+      "epoch": 1.8430599369085172,
+      "grad_norm": 0.4580081566598378,
+      "learning_rate": 1.1244535607813898e-05,
+      "loss": 0.3361,
+      "step": 9349
+    },
+    {
+      "epoch": 1.8432570977917981,
+      "grad_norm": 0.49486686537167013,
+      "learning_rate": 1.1242997617670685e-05,
+      "loss": 0.3191,
+      "step": 9350
+    },
+    {
+      "epoch": 1.8434542586750788,
+      "grad_norm": 0.4738563965909639,
+      "learning_rate": 1.1241459597663453e-05,
+      "loss": 0.358,
+      "step": 9351
+    },
+    {
+      "epoch": 1.8436514195583595,
+      "grad_norm": 0.46422380906100263,
+      "learning_rate": 1.1239921547829156e-05,
+      "loss": 0.3178,
+      "step": 9352
+    },
+    {
+      "epoch": 1.8438485804416405,
+      "grad_norm": 0.4863236006835687,
+      "learning_rate": 1.1238383468204744e-05,
+      "loss": 0.3352,
+      "step": 9353
+    },
+    {
+      "epoch": 1.8440457413249212,
+      "grad_norm": 0.5448224040290974,
+      "learning_rate": 1.1236845358827174e-05,
+      "loss": 0.3508,
+      "step": 9354
+    },
+    {
+      "epoch": 1.8442429022082019,
+      "grad_norm": 0.5119358198476269,
+      "learning_rate": 1.1235307219733396e-05,
+      "loss": 0.3626,
+      "step": 9355
+    },
+    {
+      "epoch": 1.8444400630914828,
+      "grad_norm": 0.48695813930072723,
+      "learning_rate": 1.1233769050960366e-05,
+      "loss": 0.3276,
+      "step": 9356
+    },
+    {
+      "epoch": 1.8446372239747633,
+      "grad_norm": 0.45412864850612233,
+      "learning_rate": 1.1232230852545042e-05,
+      "loss": 0.3367,
+      "step": 9357
+    },
+    {
+      "epoch": 1.8448343848580442,
+      "grad_norm": 0.41684857678310244,
+      "learning_rate": 1.1230692624524379e-05,
+      "loss": 0.2956,
+      "step": 9358
+    },
+    {
+      "epoch": 1.8450315457413249,
+      "grad_norm": 0.4765022219720706,
+      "learning_rate": 1.1229154366935337e-05,
+      "loss": 0.3451,
+      "step": 9359
+    },
+    {
+      "epoch": 1.8452287066246056,
+      "grad_norm": 0.4451789441382438,
+      "learning_rate": 1.1227616079814869e-05,
+      "loss": 0.3078,
+      "step": 9360
+    },
+    {
+      "epoch": 1.8454258675078865,
+      "grad_norm": 0.4783197235034446,
+      "learning_rate": 1.1226077763199941e-05,
+      "loss": 0.3355,
+      "step": 9361
+    },
+    {
+      "epoch": 1.8456230283911672,
+      "grad_norm": 0.45356257014627016,
+      "learning_rate": 1.12245394171275e-05,
+      "loss": 0.3263,
+      "step": 9362
+    },
+    {
+      "epoch": 1.845820189274448,
+      "grad_norm": 0.6426674681188173,
+      "learning_rate": 1.1223001041634517e-05,
+      "loss": 0.3478,
+      "step": 9363
+    },
+    {
+      "epoch": 1.8460173501577288,
+      "grad_norm": 0.5256663639141104,
+      "learning_rate": 1.122146263675795e-05,
+      "loss": 0.3536,
+      "step": 9364
+    },
+    {
+      "epoch": 1.8462145110410093,
+      "grad_norm": 0.49322701169852573,
+      "learning_rate": 1.121992420253476e-05,
+      "loss": 0.3507,
+      "step": 9365
+    },
+    {
+      "epoch": 1.8464116719242902,
+      "grad_norm": 0.49133913419542524,
+      "learning_rate": 1.1218385739001908e-05,
+      "loss": 0.3297,
+      "step": 9366
+    },
+    {
+      "epoch": 1.846608832807571,
+      "grad_norm": 0.4936609617307454,
+      "learning_rate": 1.1216847246196356e-05,
+      "loss": 0.3569,
+      "step": 9367
+    },
+    {
+      "epoch": 1.8468059936908516,
+      "grad_norm": 0.4662341419677541,
+      "learning_rate": 1.121530872415507e-05,
+      "loss": 0.2912,
+      "step": 9368
+    },
+    {
+      "epoch": 1.8470031545741326,
+      "grad_norm": 0.4643026216620009,
+      "learning_rate": 1.1213770172915012e-05,
+      "loss": 0.3205,
+      "step": 9369
+    },
+    {
+      "epoch": 1.8472003154574133,
+      "grad_norm": 0.4620634065880125,
+      "learning_rate": 1.121223159251315e-05,
+      "loss": 0.3324,
+      "step": 9370
+    },
+    {
+      "epoch": 1.847397476340694,
+      "grad_norm": 0.45144218730120467,
+      "learning_rate": 1.1210692982986447e-05,
+      "loss": 0.3085,
+      "step": 9371
+    },
+    {
+      "epoch": 1.8475946372239749,
+      "grad_norm": 0.481324283568681,
+      "learning_rate": 1.120915434437187e-05,
+      "loss": 0.3384,
+      "step": 9372
+    },
+    {
+      "epoch": 1.8477917981072554,
+      "grad_norm": 0.48469219856799683,
+      "learning_rate": 1.1207615676706387e-05,
+      "loss": 0.3237,
+      "step": 9373
+    },
+    {
+      "epoch": 1.8479889589905363,
+      "grad_norm": 0.43868531841936526,
+      "learning_rate": 1.1206076980026963e-05,
+      "loss": 0.331,
+      "step": 9374
+    },
+    {
+      "epoch": 1.848186119873817,
+      "grad_norm": 0.49905037754094533,
+      "learning_rate": 1.120453825437057e-05,
+      "loss": 0.3504,
+      "step": 9375
+    },
+    {
+      "epoch": 1.8483832807570977,
+      "grad_norm": 0.47057721365819305,
+      "learning_rate": 1.1202999499774174e-05,
+      "loss": 0.3341,
+      "step": 9376
+    },
+    {
+      "epoch": 1.8485804416403786,
+      "grad_norm": 0.5115538854447896,
+      "learning_rate": 1.1201460716274745e-05,
+      "loss": 0.3557,
+      "step": 9377
+    },
+    {
+      "epoch": 1.8487776025236593,
+      "grad_norm": 0.49106393680197397,
+      "learning_rate": 1.1199921903909258e-05,
+      "loss": 0.3348,
+      "step": 9378
+    },
+    {
+      "epoch": 1.84897476340694,
+      "grad_norm": 0.4783550963001248,
+      "learning_rate": 1.119838306271468e-05,
+      "loss": 0.3447,
+      "step": 9379
+    },
+    {
+      "epoch": 1.849171924290221,
+      "grad_norm": 15.982642688070285,
+      "learning_rate": 1.1196844192727984e-05,
+      "loss": 0.3471,
+      "step": 9380
+    },
+    {
+      "epoch": 1.8493690851735016,
+      "grad_norm": 0.49460635668203545,
+      "learning_rate": 1.119530529398614e-05,
+      "loss": 0.3547,
+      "step": 9381
+    },
+    {
+      "epoch": 1.8495662460567823,
+      "grad_norm": 0.4767109854012262,
+      "learning_rate": 1.1193766366526128e-05,
+      "loss": 0.3347,
+      "step": 9382
+    },
+    {
+      "epoch": 1.8497634069400632,
+      "grad_norm": 0.4344678704925985,
+      "learning_rate": 1.1192227410384915e-05,
+      "loss": 0.2973,
+      "step": 9383
+    },
+    {
+      "epoch": 1.8499605678233437,
+      "grad_norm": 0.4942642287727857,
+      "learning_rate": 1.1190688425599478e-05,
+      "loss": 0.3555,
+      "step": 9384
+    },
+    {
+      "epoch": 1.8501577287066246,
+      "grad_norm": 0.45936308831216105,
+      "learning_rate": 1.1189149412206795e-05,
+      "loss": 0.3256,
+      "step": 9385
+    },
+    {
+      "epoch": 1.8503548895899053,
+      "grad_norm": 0.48263042540453904,
+      "learning_rate": 1.1187610370243837e-05,
+      "loss": 0.337,
+      "step": 9386
+    },
+    {
+      "epoch": 1.850552050473186,
+      "grad_norm": 0.48546115028194037,
+      "learning_rate": 1.1186071299747588e-05,
+      "loss": 0.3529,
+      "step": 9387
+    },
+    {
+      "epoch": 1.850749211356467,
+      "grad_norm": 1.2555124942050995,
+      "learning_rate": 1.1184532200755017e-05,
+      "loss": 0.3508,
+      "step": 9388
+    },
+    {
+      "epoch": 1.8509463722397477,
+      "grad_norm": 0.517094897963285,
+      "learning_rate": 1.1182993073303107e-05,
+      "loss": 0.3507,
+      "step": 9389
+    },
+    {
+      "epoch": 1.8511435331230284,
+      "grad_norm": 0.4931959158526763,
+      "learning_rate": 1.1181453917428835e-05,
+      "loss": 0.3404,
+      "step": 9390
+    },
+    {
+      "epoch": 1.8513406940063093,
+      "grad_norm": 0.4720258856465021,
+      "learning_rate": 1.117991473316918e-05,
+      "loss": 0.3271,
+      "step": 9391
+    },
+    {
+      "epoch": 1.8515378548895898,
+      "grad_norm": 0.4663809318007513,
+      "learning_rate": 1.1178375520561126e-05,
+      "loss": 0.3231,
+      "step": 9392
+    },
+    {
+      "epoch": 1.8517350157728707,
+      "grad_norm": 0.4791754199299042,
+      "learning_rate": 1.1176836279641649e-05,
+      "loss": 0.3495,
+      "step": 9393
+    },
+    {
+      "epoch": 1.8519321766561514,
+      "grad_norm": 0.44613918375908324,
+      "learning_rate": 1.1175297010447734e-05,
+      "loss": 0.3234,
+      "step": 9394
+    },
+    {
+      "epoch": 1.852129337539432,
+      "grad_norm": 0.5340353989096874,
+      "learning_rate": 1.1173757713016362e-05,
+      "loss": 0.3315,
+      "step": 9395
+    },
+    {
+      "epoch": 1.852326498422713,
+      "grad_norm": 0.46860519180264953,
+      "learning_rate": 1.1172218387384517e-05,
+      "loss": 0.3234,
+      "step": 9396
+    },
+    {
+      "epoch": 1.8525236593059937,
+      "grad_norm": 0.46923350765977884,
+      "learning_rate": 1.117067903358918e-05,
+      "loss": 0.3152,
+      "step": 9397
+    },
+    {
+      "epoch": 1.8527208201892744,
+      "grad_norm": 0.4989908576360559,
+      "learning_rate": 1.1169139651667334e-05,
+      "loss": 0.3334,
+      "step": 9398
+    },
+    {
+      "epoch": 1.8529179810725553,
+      "grad_norm": 0.48850368752631385,
+      "learning_rate": 1.1167600241655969e-05,
+      "loss": 0.3168,
+      "step": 9399
+    },
+    {
+      "epoch": 1.8531151419558358,
+      "grad_norm": 0.506462305386158,
+      "learning_rate": 1.116606080359207e-05,
+      "loss": 0.3354,
+      "step": 9400
+    },
+    {
+      "epoch": 1.8533123028391167,
+      "grad_norm": 0.4783276972037197,
+      "learning_rate": 1.1164521337512618e-05,
+      "loss": 0.3444,
+      "step": 9401
+    },
+    {
+      "epoch": 1.8535094637223974,
+      "grad_norm": 0.49486017910579383,
+      "learning_rate": 1.1162981843454603e-05,
+      "loss": 0.322,
+      "step": 9402
+    },
+    {
+      "epoch": 1.8537066246056781,
+      "grad_norm": 0.48872781997990056,
+      "learning_rate": 1.1161442321455013e-05,
+      "loss": 0.3376,
+      "step": 9403
+    },
+    {
+      "epoch": 1.853903785488959,
+      "grad_norm": 0.49644332416860143,
+      "learning_rate": 1.1159902771550836e-05,
+      "loss": 0.3282,
+      "step": 9404
+    },
+    {
+      "epoch": 1.8541009463722398,
+      "grad_norm": 0.483535918659441,
+      "learning_rate": 1.115836319377906e-05,
+      "loss": 0.3394,
+      "step": 9405
+    },
+    {
+      "epoch": 1.8542981072555205,
+      "grad_norm": 0.45628473409890424,
+      "learning_rate": 1.115682358817668e-05,
+      "loss": 0.3345,
+      "step": 9406
+    },
+    {
+      "epoch": 1.8544952681388014,
+      "grad_norm": 0.5002627802730854,
+      "learning_rate": 1.1155283954780676e-05,
+      "loss": 0.3342,
+      "step": 9407
+    },
+    {
+      "epoch": 1.8546924290220819,
+      "grad_norm": 0.46363502477326907,
+      "learning_rate": 1.1153744293628049e-05,
+      "loss": 0.341,
+      "step": 9408
+    },
+    {
+      "epoch": 1.8548895899053628,
+      "grad_norm": 0.45754159982180126,
+      "learning_rate": 1.115220460475578e-05,
+      "loss": 0.322,
+      "step": 9409
+    },
+    {
+      "epoch": 1.8550867507886435,
+      "grad_norm": 0.4811557071074235,
+      "learning_rate": 1.1150664888200874e-05,
+      "loss": 0.3448,
+      "step": 9410
+    },
+    {
+      "epoch": 1.8552839116719242,
+      "grad_norm": 0.45695254954143394,
+      "learning_rate": 1.1149125144000315e-05,
+      "loss": 0.3172,
+      "step": 9411
+    },
+    {
+      "epoch": 1.8554810725552051,
+      "grad_norm": 0.4611849986321203,
+      "learning_rate": 1.1147585372191099e-05,
+      "loss": 0.3324,
+      "step": 9412
+    },
+    {
+      "epoch": 1.8556782334384858,
+      "grad_norm": 0.4951261883208246,
+      "learning_rate": 1.114604557281022e-05,
+      "loss": 0.3467,
+      "step": 9413
+    },
+    {
+      "epoch": 1.8558753943217665,
+      "grad_norm": 0.45894758304595423,
+      "learning_rate": 1.1144505745894674e-05,
+      "loss": 0.3354,
+      "step": 9414
+    },
+    {
+      "epoch": 1.8560725552050474,
+      "grad_norm": 0.47361962578122113,
+      "learning_rate": 1.1142965891481456e-05,
+      "loss": 0.338,
+      "step": 9415
+    },
+    {
+      "epoch": 1.856269716088328,
+      "grad_norm": 0.46717084409335563,
+      "learning_rate": 1.1141426009607562e-05,
+      "loss": 0.3181,
+      "step": 9416
+    },
+    {
+      "epoch": 1.8564668769716088,
+      "grad_norm": 0.46910516420489423,
+      "learning_rate": 1.1139886100309987e-05,
+      "loss": 0.3114,
+      "step": 9417
+    },
+    {
+      "epoch": 1.8566640378548895,
+      "grad_norm": 0.45627899747872586,
+      "learning_rate": 1.1138346163625732e-05,
+      "loss": 0.3362,
+      "step": 9418
+    },
+    {
+      "epoch": 1.8568611987381702,
+      "grad_norm": 0.488764972259954,
+      "learning_rate": 1.1136806199591794e-05,
+      "loss": 0.318,
+      "step": 9419
+    },
+    {
+      "epoch": 1.8570583596214512,
+      "grad_norm": 0.48207706868844363,
+      "learning_rate": 1.1135266208245173e-05,
+      "loss": 0.3252,
+      "step": 9420
+    },
+    {
+      "epoch": 1.8572555205047319,
+      "grad_norm": 0.4533304622987771,
+      "learning_rate": 1.1133726189622865e-05,
+      "loss": 0.3134,
+      "step": 9421
+    },
+    {
+      "epoch": 1.8574526813880126,
+      "grad_norm": 0.5209819124401526,
+      "learning_rate": 1.1132186143761872e-05,
+      "loss": 0.3225,
+      "step": 9422
+    },
+    {
+      "epoch": 1.8576498422712935,
+      "grad_norm": 0.4764148070256464,
+      "learning_rate": 1.1130646070699196e-05,
+      "loss": 0.335,
+      "step": 9423
+    },
+    {
+      "epoch": 1.8578470031545742,
+      "grad_norm": 0.46903416952858623,
+      "learning_rate": 1.1129105970471836e-05,
+      "loss": 0.3372,
+      "step": 9424
+    },
+    {
+      "epoch": 1.8580441640378549,
+      "grad_norm": 0.4386714428559711,
+      "learning_rate": 1.1127565843116798e-05,
+      "loss": 0.2949,
+      "step": 9425
+    },
+    {
+      "epoch": 1.8582413249211358,
+      "grad_norm": 0.4453659320932465,
+      "learning_rate": 1.1126025688671081e-05,
+      "loss": 0.3091,
+      "step": 9426
+    },
+    {
+      "epoch": 1.8584384858044163,
+      "grad_norm": 0.6218845162516824,
+      "learning_rate": 1.1124485507171691e-05,
+      "loss": 0.367,
+      "step": 9427
+    },
+    {
+      "epoch": 1.8586356466876972,
+      "grad_norm": 0.5032362384746675,
+      "learning_rate": 1.112294529865563e-05,
+      "loss": 0.3386,
+      "step": 9428
+    },
+    {
+      "epoch": 1.858832807570978,
+      "grad_norm": 0.46700700293384795,
+      "learning_rate": 1.1121405063159906e-05,
+      "loss": 0.344,
+      "step": 9429
+    },
+    {
+      "epoch": 1.8590299684542586,
+      "grad_norm": 0.4650433144492055,
+      "learning_rate": 1.111986480072152e-05,
+      "loss": 0.3257,
+      "step": 9430
+    },
+    {
+      "epoch": 1.8592271293375395,
+      "grad_norm": 0.470873212417909,
+      "learning_rate": 1.1118324511377482e-05,
+      "loss": 0.3012,
+      "step": 9431
+    },
+    {
+      "epoch": 1.8594242902208202,
+      "grad_norm": 0.4694211550117268,
+      "learning_rate": 1.1116784195164797e-05,
+      "loss": 0.3423,
+      "step": 9432
+    },
+    {
+      "epoch": 1.859621451104101,
+      "grad_norm": 0.46797333442001665,
+      "learning_rate": 1.1115243852120472e-05,
+      "loss": 0.3196,
+      "step": 9433
+    },
+    {
+      "epoch": 1.8598186119873819,
+      "grad_norm": 0.4847132759132295,
+      "learning_rate": 1.1113703482281515e-05,
+      "loss": 0.3374,
+      "step": 9434
+    },
+    {
+      "epoch": 1.8600157728706623,
+      "grad_norm": 0.48758924974039297,
+      "learning_rate": 1.1112163085684935e-05,
+      "loss": 0.3437,
+      "step": 9435
+    },
+    {
+      "epoch": 1.8602129337539433,
+      "grad_norm": 0.48133455329307273,
+      "learning_rate": 1.111062266236774e-05,
+      "loss": 0.3175,
+      "step": 9436
+    },
+    {
+      "epoch": 1.860410094637224,
+      "grad_norm": 0.4906755288935251,
+      "learning_rate": 1.1109082212366944e-05,
+      "loss": 0.318,
+      "step": 9437
+    },
+    {
+      "epoch": 1.8606072555205047,
+      "grad_norm": 0.4633813435313994,
+      "learning_rate": 1.1107541735719554e-05,
+      "loss": 0.3103,
+      "step": 9438
+    },
+    {
+      "epoch": 1.8608044164037856,
+      "grad_norm": 0.5163900036338083,
+      "learning_rate": 1.110600123246258e-05,
+      "loss": 0.3263,
+      "step": 9439
+    },
+    {
+      "epoch": 1.8610015772870663,
+      "grad_norm": 0.5956374108250188,
+      "learning_rate": 1.1104460702633038e-05,
+      "loss": 0.3441,
+      "step": 9440
+    },
+    {
+      "epoch": 1.861198738170347,
+      "grad_norm": 0.4763548264026657,
+      "learning_rate": 1.1102920146267938e-05,
+      "loss": 0.3032,
+      "step": 9441
+    },
+    {
+      "epoch": 1.861395899053628,
+      "grad_norm": 0.4958429310254217,
+      "learning_rate": 1.1101379563404291e-05,
+      "loss": 0.3171,
+      "step": 9442
+    },
+    {
+      "epoch": 1.8615930599369084,
+      "grad_norm": 0.4894185337405908,
+      "learning_rate": 1.1099838954079117e-05,
+      "loss": 0.3349,
+      "step": 9443
+    },
+    {
+      "epoch": 1.8617902208201893,
+      "grad_norm": 0.48658157622850023,
+      "learning_rate": 1.1098298318329421e-05,
+      "loss": 0.3516,
+      "step": 9444
+    },
+    {
+      "epoch": 1.86198738170347,
+      "grad_norm": 0.477421882843568,
+      "learning_rate": 1.1096757656192226e-05,
+      "loss": 0.3442,
+      "step": 9445
+    },
+    {
+      "epoch": 1.8621845425867507,
+      "grad_norm": 0.4587814387324469,
+      "learning_rate": 1.1095216967704548e-05,
+      "loss": 0.3155,
+      "step": 9446
+    },
+    {
+      "epoch": 1.8623817034700316,
+      "grad_norm": 0.46345456592813855,
+      "learning_rate": 1.1093676252903395e-05,
+      "loss": 0.3245,
+      "step": 9447
+    },
+    {
+      "epoch": 1.8625788643533123,
+      "grad_norm": 0.49258102378140683,
+      "learning_rate": 1.1092135511825795e-05,
+      "loss": 0.3385,
+      "step": 9448
+    },
+    {
+      "epoch": 1.862776025236593,
+      "grad_norm": 0.4885376518923072,
+      "learning_rate": 1.1090594744508754e-05,
+      "loss": 0.3429,
+      "step": 9449
+    },
+    {
+      "epoch": 1.862973186119874,
+      "grad_norm": 0.5003966905526008,
+      "learning_rate": 1.1089053950989301e-05,
+      "loss": 0.3572,
+      "step": 9450
+    },
+    {
+      "epoch": 1.8631703470031544,
+      "grad_norm": 0.48522433245566304,
+      "learning_rate": 1.1087513131304446e-05,
+      "loss": 0.3174,
+      "step": 9451
+    },
+    {
+      "epoch": 1.8633675078864353,
+      "grad_norm": 0.5072726660997204,
+      "learning_rate": 1.1085972285491213e-05,
+      "loss": 0.3142,
+      "step": 9452
+    },
+    {
+      "epoch": 1.863564668769716,
+      "grad_norm": 0.4492624933232977,
+      "learning_rate": 1.1084431413586625e-05,
+      "loss": 0.3264,
+      "step": 9453
+    },
+    {
+      "epoch": 1.8637618296529967,
+      "grad_norm": 0.47440212617966615,
+      "learning_rate": 1.1082890515627696e-05,
+      "loss": 0.3221,
+      "step": 9454
+    },
+    {
+      "epoch": 1.8639589905362777,
+      "grad_norm": 0.4449072907644378,
+      "learning_rate": 1.108134959165145e-05,
+      "loss": 0.3093,
+      "step": 9455
+    },
+    {
+      "epoch": 1.8641561514195584,
+      "grad_norm": 0.4454938083744686,
+      "learning_rate": 1.1079808641694909e-05,
+      "loss": 0.3215,
+      "step": 9456
+    },
+    {
+      "epoch": 1.864353312302839,
+      "grad_norm": 0.472087185708321,
+      "learning_rate": 1.10782676657951e-05,
+      "loss": 0.3397,
+      "step": 9457
+    },
+    {
+      "epoch": 1.86455047318612,
+      "grad_norm": 0.4678074204749738,
+      "learning_rate": 1.1076726663989037e-05,
+      "loss": 0.3284,
+      "step": 9458
+    },
+    {
+      "epoch": 1.8647476340694005,
+      "grad_norm": 0.46656067923498784,
+      "learning_rate": 1.107518563631375e-05,
+      "loss": 0.3283,
+      "step": 9459
+    },
+    {
+      "epoch": 1.8649447949526814,
+      "grad_norm": 0.4635770298520586,
+      "learning_rate": 1.1073644582806263e-05,
+      "loss": 0.3187,
+      "step": 9460
+    },
+    {
+      "epoch": 1.865141955835962,
+      "grad_norm": 0.44301184098285906,
+      "learning_rate": 1.1072103503503599e-05,
+      "loss": 0.3063,
+      "step": 9461
+    },
+    {
+      "epoch": 1.8653391167192428,
+      "grad_norm": 0.4628837166980266,
+      "learning_rate": 1.1070562398442789e-05,
+      "loss": 0.3335,
+      "step": 9462
+    },
+    {
+      "epoch": 1.8655362776025237,
+      "grad_norm": 0.6943912518558549,
+      "learning_rate": 1.106902126766085e-05,
+      "loss": 0.3459,
+      "step": 9463
+    },
+    {
+      "epoch": 1.8657334384858044,
+      "grad_norm": 0.5689959265430884,
+      "learning_rate": 1.1067480111194817e-05,
+      "loss": 0.3639,
+      "step": 9464
+    },
+    {
+      "epoch": 1.8659305993690851,
+      "grad_norm": 0.478559291667436,
+      "learning_rate": 1.1065938929081714e-05,
+      "loss": 0.3357,
+      "step": 9465
+    },
+    {
+      "epoch": 1.866127760252366,
+      "grad_norm": 0.4676687566999173,
+      "learning_rate": 1.1064397721358571e-05,
+      "loss": 0.3022,
+      "step": 9466
+    },
+    {
+      "epoch": 1.8663249211356467,
+      "grad_norm": 0.46290577274262135,
+      "learning_rate": 1.1062856488062414e-05,
+      "loss": 0.3096,
+      "step": 9467
+    },
+    {
+      "epoch": 1.8665220820189274,
+      "grad_norm": 0.5040641995883025,
+      "learning_rate": 1.1061315229230276e-05,
+      "loss": 0.3394,
+      "step": 9468
+    },
+    {
+      "epoch": 1.8667192429022084,
+      "grad_norm": 0.45321672662297996,
+      "learning_rate": 1.1059773944899183e-05,
+      "loss": 0.3369,
+      "step": 9469
+    },
+    {
+      "epoch": 1.8669164037854888,
+      "grad_norm": 0.4487604982696012,
+      "learning_rate": 1.1058232635106167e-05,
+      "loss": 0.3296,
+      "step": 9470
+    },
+    {
+      "epoch": 1.8671135646687698,
+      "grad_norm": 0.5159266623498804,
+      "learning_rate": 1.1056691299888262e-05,
+      "loss": 0.3446,
+      "step": 9471
+    },
+    {
+      "epoch": 1.8673107255520505,
+      "grad_norm": 0.48017706589184767,
+      "learning_rate": 1.1055149939282497e-05,
+      "loss": 0.3296,
+      "step": 9472
+    },
+    {
+      "epoch": 1.8675078864353312,
+      "grad_norm": 0.47518034239431195,
+      "learning_rate": 1.1053608553325901e-05,
+      "loss": 0.3314,
+      "step": 9473
+    },
+    {
+      "epoch": 1.867705047318612,
+      "grad_norm": 0.4352311407076608,
+      "learning_rate": 1.1052067142055516e-05,
+      "loss": 0.3109,
+      "step": 9474
+    },
+    {
+      "epoch": 1.8679022082018928,
+      "grad_norm": 0.47370145361968125,
+      "learning_rate": 1.1050525705508369e-05,
+      "loss": 0.3371,
+      "step": 9475
+    },
+    {
+      "epoch": 1.8680993690851735,
+      "grad_norm": 0.5177165653942762,
+      "learning_rate": 1.1048984243721496e-05,
+      "loss": 0.353,
+      "step": 9476
+    },
+    {
+      "epoch": 1.8682965299684544,
+      "grad_norm": 0.4685075059578066,
+      "learning_rate": 1.104744275673193e-05,
+      "loss": 0.3392,
+      "step": 9477
+    },
+    {
+      "epoch": 1.868493690851735,
+      "grad_norm": 0.4547513508412821,
+      "learning_rate": 1.1045901244576713e-05,
+      "loss": 0.3096,
+      "step": 9478
+    },
+    {
+      "epoch": 1.8686908517350158,
+      "grad_norm": 0.4715660539717716,
+      "learning_rate": 1.104435970729287e-05,
+      "loss": 0.3337,
+      "step": 9479
+    },
+    {
+      "epoch": 1.8688880126182965,
+      "grad_norm": 0.5173676222111224,
+      "learning_rate": 1.1042818144917449e-05,
+      "loss": 0.3429,
+      "step": 9480
+    },
+    {
+      "epoch": 1.8690851735015772,
+      "grad_norm": 0.4748763088025496,
+      "learning_rate": 1.1041276557487482e-05,
+      "loss": 0.3285,
+      "step": 9481
+    },
+    {
+      "epoch": 1.8692823343848581,
+      "grad_norm": 0.4570571040060861,
+      "learning_rate": 1.1039734945040004e-05,
+      "loss": 0.3291,
+      "step": 9482
+    },
+    {
+      "epoch": 1.8694794952681388,
+      "grad_norm": 0.5076175558770766,
+      "learning_rate": 1.103819330761206e-05,
+      "loss": 0.3466,
+      "step": 9483
+    },
+    {
+      "epoch": 1.8696766561514195,
+      "grad_norm": 0.43967840426291815,
+      "learning_rate": 1.1036651645240683e-05,
+      "loss": 0.3072,
+      "step": 9484
+    },
+    {
+      "epoch": 1.8698738170347005,
+      "grad_norm": 0.488582403780876,
+      "learning_rate": 1.1035109957962918e-05,
+      "loss": 0.3156,
+      "step": 9485
+    },
+    {
+      "epoch": 1.870070977917981,
+      "grad_norm": 0.47909357015758164,
+      "learning_rate": 1.10335682458158e-05,
+      "loss": 0.3425,
+      "step": 9486
+    },
+    {
+      "epoch": 1.8702681388012619,
+      "grad_norm": 0.47912178648751735,
+      "learning_rate": 1.1032026508836376e-05,
+      "loss": 0.3053,
+      "step": 9487
+    },
+    {
+      "epoch": 1.8704652996845426,
+      "grad_norm": 0.480061395371812,
+      "learning_rate": 1.103048474706168e-05,
+      "loss": 0.364,
+      "step": 9488
+    },
+    {
+      "epoch": 1.8706624605678233,
+      "grad_norm": 0.46424946405738576,
+      "learning_rate": 1.102894296052876e-05,
+      "loss": 0.3152,
+      "step": 9489
+    },
+    {
+      "epoch": 1.8708596214511042,
+      "grad_norm": 0.46382382583955356,
+      "learning_rate": 1.1027401149274658e-05,
+      "loss": 0.331,
+      "step": 9490
+    },
+    {
+      "epoch": 1.8710567823343849,
+      "grad_norm": 0.4707474802471004,
+      "learning_rate": 1.1025859313336415e-05,
+      "loss": 0.3308,
+      "step": 9491
+    },
+    {
+      "epoch": 1.8712539432176656,
+      "grad_norm": 0.46377381491584824,
+      "learning_rate": 1.1024317452751076e-05,
+      "loss": 0.327,
+      "step": 9492
+    },
+    {
+      "epoch": 1.8714511041009465,
+      "grad_norm": 0.44780835687333587,
+      "learning_rate": 1.1022775567555686e-05,
+      "loss": 0.3347,
+      "step": 9493
+    },
+    {
+      "epoch": 1.871648264984227,
+      "grad_norm": 0.45613465092557437,
+      "learning_rate": 1.1021233657787285e-05,
+      "loss": 0.3223,
+      "step": 9494
+    },
+    {
+      "epoch": 1.871845425867508,
+      "grad_norm": 0.4667856099355164,
+      "learning_rate": 1.1019691723482928e-05,
+      "loss": 0.3262,
+      "step": 9495
+    },
+    {
+      "epoch": 1.8720425867507886,
+      "grad_norm": 0.5087693131824845,
+      "learning_rate": 1.1018149764679653e-05,
+      "loss": 0.3565,
+      "step": 9496
+    },
+    {
+      "epoch": 1.8722397476340693,
+      "grad_norm": 0.47261955625623386,
+      "learning_rate": 1.1016607781414514e-05,
+      "loss": 0.3387,
+      "step": 9497
+    },
+    {
+      "epoch": 1.8724369085173502,
+      "grad_norm": 0.4507061778033932,
+      "learning_rate": 1.101506577372455e-05,
+      "loss": 0.3378,
+      "step": 9498
+    },
+    {
+      "epoch": 1.872634069400631,
+      "grad_norm": 0.4914038805520105,
+      "learning_rate": 1.1013523741646817e-05,
+      "loss": 0.3635,
+      "step": 9499
+    },
+    {
+      "epoch": 1.8728312302839116,
+      "grad_norm": 0.46523334735695693,
+      "learning_rate": 1.1011981685218355e-05,
+      "loss": 0.3163,
+      "step": 9500
+    },
+    {
+      "epoch": 1.8730283911671926,
+      "grad_norm": 0.461046105271434,
+      "learning_rate": 1.1010439604476222e-05,
+      "loss": 0.3027,
+      "step": 9501
+    },
+    {
+      "epoch": 1.873225552050473,
+      "grad_norm": 0.4917301931134718,
+      "learning_rate": 1.1008897499457466e-05,
+      "loss": 0.3536,
+      "step": 9502
+    },
+    {
+      "epoch": 1.873422712933754,
+      "grad_norm": 0.4931865848981832,
+      "learning_rate": 1.100735537019913e-05,
+      "loss": 0.338,
+      "step": 9503
+    },
+    {
+      "epoch": 1.8736198738170347,
+      "grad_norm": 0.465842711087265,
+      "learning_rate": 1.1005813216738273e-05,
+      "loss": 0.3327,
+      "step": 9504
+    },
+    {
+      "epoch": 1.8738170347003154,
+      "grad_norm": 0.46721071233696176,
+      "learning_rate": 1.1004271039111943e-05,
+      "loss": 0.3349,
+      "step": 9505
+    },
+    {
+      "epoch": 1.8740141955835963,
+      "grad_norm": 0.47405778733750703,
+      "learning_rate": 1.1002728837357192e-05,
+      "loss": 0.3388,
+      "step": 9506
+    },
+    {
+      "epoch": 1.874211356466877,
+      "grad_norm": 0.4734967427166212,
+      "learning_rate": 1.1001186611511071e-05,
+      "loss": 0.3385,
+      "step": 9507
+    },
+    {
+      "epoch": 1.8744085173501577,
+      "grad_norm": 0.5255895043873378,
+      "learning_rate": 1.099964436161064e-05,
+      "loss": 0.3372,
+      "step": 9508
+    },
+    {
+      "epoch": 1.8746056782334386,
+      "grad_norm": 0.4787785522800639,
+      "learning_rate": 1.0998102087692946e-05,
+      "loss": 0.3355,
+      "step": 9509
+    },
+    {
+      "epoch": 1.874802839116719,
+      "grad_norm": 0.5023739258469361,
+      "learning_rate": 1.0996559789795045e-05,
+      "loss": 0.3478,
+      "step": 9510
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.46464061305129783,
+      "learning_rate": 1.0995017467953994e-05,
+      "loss": 0.3331,
+      "step": 9511
+    },
+    {
+      "epoch": 1.875197160883281,
+      "grad_norm": 0.4995781916354229,
+      "learning_rate": 1.0993475122206846e-05,
+      "loss": 0.3407,
+      "step": 9512
+    },
+    {
+      "epoch": 1.8753943217665614,
+      "grad_norm": 0.4733291385390516,
+      "learning_rate": 1.0991932752590657e-05,
+      "loss": 0.3197,
+      "step": 9513
+    },
+    {
+      "epoch": 1.8755914826498423,
+      "grad_norm": 0.5105493418105221,
+      "learning_rate": 1.0990390359142488e-05,
+      "loss": 0.3507,
+      "step": 9514
+    },
+    {
+      "epoch": 1.875788643533123,
+      "grad_norm": 0.4942055666455541,
+      "learning_rate": 1.098884794189939e-05,
+      "loss": 0.3403,
+      "step": 9515
+    },
+    {
+      "epoch": 1.8759858044164037,
+      "grad_norm": 0.4900283583301513,
+      "learning_rate": 1.0987305500898427e-05,
+      "loss": 0.3239,
+      "step": 9516
+    },
+    {
+      "epoch": 1.8761829652996846,
+      "grad_norm": 0.4555388435909809,
+      "learning_rate": 1.0985763036176648e-05,
+      "loss": 0.3254,
+      "step": 9517
+    },
+    {
+      "epoch": 1.8763801261829653,
+      "grad_norm": 0.521257593570195,
+      "learning_rate": 1.0984220547771127e-05,
+      "loss": 0.3489,
+      "step": 9518
+    },
+    {
+      "epoch": 1.876577287066246,
+      "grad_norm": 0.45372277802554656,
+      "learning_rate": 1.098267803571891e-05,
+      "loss": 0.2968,
+      "step": 9519
+    },
+    {
+      "epoch": 1.876774447949527,
+      "grad_norm": 0.4731885774415591,
+      "learning_rate": 1.098113550005706e-05,
+      "loss": 0.348,
+      "step": 9520
+    },
+    {
+      "epoch": 1.8769716088328074,
+      "grad_norm": 0.4704727543410908,
+      "learning_rate": 1.0979592940822643e-05,
+      "loss": 0.3237,
+      "step": 9521
+    },
+    {
+      "epoch": 1.8771687697160884,
+      "grad_norm": 0.4716211417273583,
+      "learning_rate": 1.0978050358052715e-05,
+      "loss": 0.3332,
+      "step": 9522
+    },
+    {
+      "epoch": 1.877365930599369,
+      "grad_norm": 0.4656934460308149,
+      "learning_rate": 1.0976507751784343e-05,
+      "loss": 0.3253,
+      "step": 9523
+    },
+    {
+      "epoch": 1.8775630914826498,
+      "grad_norm": 0.45429803312317435,
+      "learning_rate": 1.097496512205458e-05,
+      "loss": 0.3213,
+      "step": 9524
+    },
+    {
+      "epoch": 1.8777602523659307,
+      "grad_norm": 0.4941603848877281,
+      "learning_rate": 1.0973422468900498e-05,
+      "loss": 0.3314,
+      "step": 9525
+    },
+    {
+      "epoch": 1.8779574132492114,
+      "grad_norm": 0.4921863038345052,
+      "learning_rate": 1.0971879792359154e-05,
+      "loss": 0.3161,
+      "step": 9526
+    },
+    {
+      "epoch": 1.878154574132492,
+      "grad_norm": 0.46662564704980725,
+      "learning_rate": 1.097033709246762e-05,
+      "loss": 0.3451,
+      "step": 9527
+    },
+    {
+      "epoch": 1.878351735015773,
+      "grad_norm": 0.4434045432752366,
+      "learning_rate": 1.0968794369262954e-05,
+      "loss": 0.3125,
+      "step": 9528
+    },
+    {
+      "epoch": 1.8785488958990535,
+      "grad_norm": 0.4809233182201705,
+      "learning_rate": 1.0967251622782223e-05,
+      "loss": 0.3559,
+      "step": 9529
+    },
+    {
+      "epoch": 1.8787460567823344,
+      "grad_norm": 0.4744798298419412,
+      "learning_rate": 1.0965708853062493e-05,
+      "loss": 0.3364,
+      "step": 9530
+    },
+    {
+      "epoch": 1.8789432176656151,
+      "grad_norm": 0.47368018073228213,
+      "learning_rate": 1.0964166060140831e-05,
+      "loss": 0.3171,
+      "step": 9531
+    },
+    {
+      "epoch": 1.8791403785488958,
+      "grad_norm": 0.5127652861424508,
+      "learning_rate": 1.0962623244054302e-05,
+      "loss": 0.3484,
+      "step": 9532
+    },
+    {
+      "epoch": 1.8793375394321767,
+      "grad_norm": 0.4858636735208188,
+      "learning_rate": 1.0961080404839974e-05,
+      "loss": 0.3506,
+      "step": 9533
+    },
+    {
+      "epoch": 1.8795347003154574,
+      "grad_norm": 0.4957252696098338,
+      "learning_rate": 1.0959537542534916e-05,
+      "loss": 0.3296,
+      "step": 9534
+    },
+    {
+      "epoch": 1.8797318611987381,
+      "grad_norm": 0.5147145847410192,
+      "learning_rate": 1.0957994657176197e-05,
+      "loss": 0.3467,
+      "step": 9535
+    },
+    {
+      "epoch": 1.879929022082019,
+      "grad_norm": 0.4335412701687705,
+      "learning_rate": 1.0956451748800883e-05,
+      "loss": 0.3149,
+      "step": 9536
+    },
+    {
+      "epoch": 1.8801261829652995,
+      "grad_norm": 0.46479740344238335,
+      "learning_rate": 1.0954908817446047e-05,
+      "loss": 0.3318,
+      "step": 9537
+    },
+    {
+      "epoch": 1.8803233438485805,
+      "grad_norm": 0.4683628747889899,
+      "learning_rate": 1.0953365863148757e-05,
+      "loss": 0.3287,
+      "step": 9538
+    },
+    {
+      "epoch": 1.8805205047318612,
+      "grad_norm": 0.472736232015758,
+      "learning_rate": 1.0951822885946084e-05,
+      "loss": 0.3136,
+      "step": 9539
+    },
+    {
+      "epoch": 1.8807176656151419,
+      "grad_norm": 0.4788330670775191,
+      "learning_rate": 1.0950279885875098e-05,
+      "loss": 0.3368,
+      "step": 9540
+    },
+    {
+      "epoch": 1.8809148264984228,
+      "grad_norm": 0.4760770871051563,
+      "learning_rate": 1.0948736862972873e-05,
+      "loss": 0.3262,
+      "step": 9541
+    },
+    {
+      "epoch": 1.8811119873817035,
+      "grad_norm": 0.456235107469069,
+      "learning_rate": 1.0947193817276485e-05,
+      "loss": 0.3283,
+      "step": 9542
+    },
+    {
+      "epoch": 1.8813091482649842,
+      "grad_norm": 0.4949947731416457,
+      "learning_rate": 1.0945650748822998e-05,
+      "loss": 0.3504,
+      "step": 9543
+    },
+    {
+      "epoch": 1.881506309148265,
+      "grad_norm": 0.5538356410729359,
+      "learning_rate": 1.0944107657649494e-05,
+      "loss": 0.319,
+      "step": 9544
+    },
+    {
+      "epoch": 1.8817034700315456,
+      "grad_norm": 0.5051031777149498,
+      "learning_rate": 1.0942564543793039e-05,
+      "loss": 0.3364,
+      "step": 9545
+    },
+    {
+      "epoch": 1.8819006309148265,
+      "grad_norm": 0.4953979151686171,
+      "learning_rate": 1.0941021407290717e-05,
+      "loss": 0.3856,
+      "step": 9546
+    },
+    {
+      "epoch": 1.8820977917981072,
+      "grad_norm": 0.4632524126131711,
+      "learning_rate": 1.0939478248179594e-05,
+      "loss": 0.3225,
+      "step": 9547
+    },
+    {
+      "epoch": 1.882294952681388,
+      "grad_norm": 0.46099411952055647,
+      "learning_rate": 1.093793506649675e-05,
+      "loss": 0.3221,
+      "step": 9548
+    },
+    {
+      "epoch": 1.8824921135646688,
+      "grad_norm": 0.48987986252693555,
+      "learning_rate": 1.093639186227926e-05,
+      "loss": 0.3414,
+      "step": 9549
+    },
+    {
+      "epoch": 1.8826892744479495,
+      "grad_norm": 0.5892412407244659,
+      "learning_rate": 1.0934848635564203e-05,
+      "loss": 0.3317,
+      "step": 9550
+    },
+    {
+      "epoch": 1.8828864353312302,
+      "grad_norm": 0.464294311971776,
+      "learning_rate": 1.0933305386388656e-05,
+      "loss": 0.3274,
+      "step": 9551
+    },
+    {
+      "epoch": 1.8830835962145112,
+      "grad_norm": 0.48576088684391133,
+      "learning_rate": 1.0931762114789695e-05,
+      "loss": 0.3357,
+      "step": 9552
+    },
+    {
+      "epoch": 1.8832807570977916,
+      "grad_norm": 0.44811752905575575,
+      "learning_rate": 1.0930218820804398e-05,
+      "loss": 0.3113,
+      "step": 9553
+    },
+    {
+      "epoch": 1.8834779179810726,
+      "grad_norm": 0.4927473006999295,
+      "learning_rate": 1.0928675504469843e-05,
+      "loss": 0.3513,
+      "step": 9554
+    },
+    {
+      "epoch": 1.8836750788643533,
+      "grad_norm": 0.45362869939341455,
+      "learning_rate": 1.0927132165823113e-05,
+      "loss": 0.3272,
+      "step": 9555
+    },
+    {
+      "epoch": 1.883872239747634,
+      "grad_norm": 0.45381446754200727,
+      "learning_rate": 1.0925588804901286e-05,
+      "loss": 0.3233,
+      "step": 9556
+    },
+    {
+      "epoch": 1.8840694006309149,
+      "grad_norm": 0.44466594957868893,
+      "learning_rate": 1.0924045421741442e-05,
+      "loss": 0.3331,
+      "step": 9557
+    },
+    {
+      "epoch": 1.8842665615141956,
+      "grad_norm": 0.5061065779984183,
+      "learning_rate": 1.0922502016380663e-05,
+      "loss": 0.3372,
+      "step": 9558
+    },
+    {
+      "epoch": 1.8844637223974763,
+      "grad_norm": 0.5421030190329682,
+      "learning_rate": 1.092095858885603e-05,
+      "loss": 0.3578,
+      "step": 9559
+    },
+    {
+      "epoch": 1.8846608832807572,
+      "grad_norm": 0.5002948397548355,
+      "learning_rate": 1.0919415139204625e-05,
+      "loss": 0.3225,
+      "step": 9560
+    },
+    {
+      "epoch": 1.884858044164038,
+      "grad_norm": 0.49365892934666894,
+      "learning_rate": 1.0917871667463533e-05,
+      "loss": 0.3349,
+      "step": 9561
+    },
+    {
+      "epoch": 1.8850552050473186,
+      "grad_norm": 0.4803469375107623,
+      "learning_rate": 1.091632817366983e-05,
+      "loss": 0.3515,
+      "step": 9562
+    },
+    {
+      "epoch": 1.8852523659305995,
+      "grad_norm": 0.49730327018911846,
+      "learning_rate": 1.091478465786061e-05,
+      "loss": 0.3453,
+      "step": 9563
+    },
+    {
+      "epoch": 1.88544952681388,
+      "grad_norm": 0.5162410884212086,
+      "learning_rate": 1.0913241120072947e-05,
+      "loss": 0.34,
+      "step": 9564
+    },
+    {
+      "epoch": 1.885646687697161,
+      "grad_norm": 0.4919385394499981,
+      "learning_rate": 1.0911697560343937e-05,
+      "loss": 0.347,
+      "step": 9565
+    },
+    {
+      "epoch": 1.8858438485804416,
+      "grad_norm": 0.48694966659079764,
+      "learning_rate": 1.0910153978710654e-05,
+      "loss": 0.3522,
+      "step": 9566
+    },
+    {
+      "epoch": 1.8860410094637223,
+      "grad_norm": 0.488293991846791,
+      "learning_rate": 1.0908610375210193e-05,
+      "loss": 0.3287,
+      "step": 9567
+    },
+    {
+      "epoch": 1.8862381703470033,
+      "grad_norm": 0.48621217336734124,
+      "learning_rate": 1.0907066749879632e-05,
+      "loss": 0.3401,
+      "step": 9568
+    },
+    {
+      "epoch": 1.886435331230284,
+      "grad_norm": 0.47778447085942694,
+      "learning_rate": 1.0905523102756061e-05,
+      "loss": 0.3199,
+      "step": 9569
+    },
+    {
+      "epoch": 1.8866324921135647,
+      "grad_norm": 0.47090656319170815,
+      "learning_rate": 1.0903979433876573e-05,
+      "loss": 0.3117,
+      "step": 9570
+    },
+    {
+      "epoch": 1.8868296529968456,
+      "grad_norm": 0.4595803156099776,
+      "learning_rate": 1.0902435743278248e-05,
+      "loss": 0.3295,
+      "step": 9571
+    },
+    {
+      "epoch": 1.887026813880126,
+      "grad_norm": 0.441046356319071,
+      "learning_rate": 1.0900892030998181e-05,
+      "loss": 0.2994,
+      "step": 9572
+    },
+    {
+      "epoch": 1.887223974763407,
+      "grad_norm": 0.46255381742264307,
+      "learning_rate": 1.089934829707345e-05,
+      "loss": 0.3106,
+      "step": 9573
+    },
+    {
+      "epoch": 1.8874211356466877,
+      "grad_norm": 0.48070659599098814,
+      "learning_rate": 1.0897804541541159e-05,
+      "loss": 0.3268,
+      "step": 9574
+    },
+    {
+      "epoch": 1.8876182965299684,
+      "grad_norm": 0.47064959203140777,
+      "learning_rate": 1.0896260764438387e-05,
+      "loss": 0.3443,
+      "step": 9575
+    },
+    {
+      "epoch": 1.8878154574132493,
+      "grad_norm": 0.48437387551779304,
+      "learning_rate": 1.089471696580223e-05,
+      "loss": 0.3274,
+      "step": 9576
+    },
+    {
+      "epoch": 1.88801261829653,
+      "grad_norm": 0.4936834752237707,
+      "learning_rate": 1.0893173145669777e-05,
+      "loss": 0.354,
+      "step": 9577
+    },
+    {
+      "epoch": 1.8882097791798107,
+      "grad_norm": 0.47759344988792984,
+      "learning_rate": 1.089162930407812e-05,
+      "loss": 0.3548,
+      "step": 9578
+    },
+    {
+      "epoch": 1.8884069400630916,
+      "grad_norm": 0.4673787356484052,
+      "learning_rate": 1.089008544106435e-05,
+      "loss": 0.3184,
+      "step": 9579
+    },
+    {
+      "epoch": 1.888604100946372,
+      "grad_norm": 5.043034026594802,
+      "learning_rate": 1.0888541556665562e-05,
+      "loss": 0.3314,
+      "step": 9580
+    },
+    {
+      "epoch": 1.888801261829653,
+      "grad_norm": 0.49514263389138563,
+      "learning_rate": 1.0886997650918848e-05,
+      "loss": 0.3212,
+      "step": 9581
+    },
+    {
+      "epoch": 1.8889984227129337,
+      "grad_norm": 0.45933092936924486,
+      "learning_rate": 1.08854537238613e-05,
+      "loss": 0.3102,
+      "step": 9582
+    },
+    {
+      "epoch": 1.8891955835962144,
+      "grad_norm": 0.4590562261492261,
+      "learning_rate": 1.0883909775530013e-05,
+      "loss": 0.3114,
+      "step": 9583
+    },
+    {
+      "epoch": 1.8893927444794953,
+      "grad_norm": 0.4834863213201024,
+      "learning_rate": 1.0882365805962083e-05,
+      "loss": 0.3437,
+      "step": 9584
+    },
+    {
+      "epoch": 1.889589905362776,
+      "grad_norm": 0.4659483454174267,
+      "learning_rate": 1.0880821815194602e-05,
+      "loss": 0.316,
+      "step": 9585
+    },
+    {
+      "epoch": 1.8897870662460567,
+      "grad_norm": 0.4768864743117075,
+      "learning_rate": 1.087927780326467e-05,
+      "loss": 0.3356,
+      "step": 9586
+    },
+    {
+      "epoch": 1.8899842271293377,
+      "grad_norm": 0.459781264277085,
+      "learning_rate": 1.087773377020938e-05,
+      "loss": 0.3245,
+      "step": 9587
+    },
+    {
+      "epoch": 1.8901813880126181,
+      "grad_norm": 0.4583408374138778,
+      "learning_rate": 1.0876189716065825e-05,
+      "loss": 0.323,
+      "step": 9588
+    },
+    {
+      "epoch": 1.890378548895899,
+      "grad_norm": 0.4504538132773227,
+      "learning_rate": 1.0874645640871114e-05,
+      "loss": 0.3059,
+      "step": 9589
+    },
+    {
+      "epoch": 1.8905757097791798,
+      "grad_norm": 0.47695654088042766,
+      "learning_rate": 1.087310154466233e-05,
+      "loss": 0.3227,
+      "step": 9590
+    },
+    {
+      "epoch": 1.8907728706624605,
+      "grad_norm": 0.4482267647409884,
+      "learning_rate": 1.0871557427476585e-05,
+      "loss": 0.3133,
+      "step": 9591
+    },
+    {
+      "epoch": 1.8909700315457414,
+      "grad_norm": 0.46554460153463734,
+      "learning_rate": 1.0870013289350964e-05,
+      "loss": 0.3353,
+      "step": 9592
+    },
+    {
+      "epoch": 1.891167192429022,
+      "grad_norm": 0.4672677533066259,
+      "learning_rate": 1.0868469130322581e-05,
+      "loss": 0.314,
+      "step": 9593
+    },
+    {
+      "epoch": 1.8913643533123028,
+      "grad_norm": 0.6158071510944071,
+      "learning_rate": 1.086692495042852e-05,
+      "loss": 0.3526,
+      "step": 9594
+    },
+    {
+      "epoch": 1.8915615141955837,
+      "grad_norm": 0.43646887186476924,
+      "learning_rate": 1.0865380749705892e-05,
+      "loss": 0.3137,
+      "step": 9595
+    },
+    {
+      "epoch": 1.8917586750788642,
+      "grad_norm": 0.5134052948284018,
+      "learning_rate": 1.0863836528191795e-05,
+      "loss": 0.3521,
+      "step": 9596
+    },
+    {
+      "epoch": 1.8919558359621451,
+      "grad_norm": 0.4882031174942994,
+      "learning_rate": 1.0862292285923331e-05,
+      "loss": 0.3404,
+      "step": 9597
+    },
+    {
+      "epoch": 1.8921529968454258,
+      "grad_norm": 0.44241922335608497,
+      "learning_rate": 1.08607480229376e-05,
+      "loss": 0.3076,
+      "step": 9598
+    },
+    {
+      "epoch": 1.8923501577287065,
+      "grad_norm": 0.48941048003891413,
+      "learning_rate": 1.0859203739271702e-05,
+      "loss": 0.3379,
+      "step": 9599
+    },
+    {
+      "epoch": 1.8925473186119874,
+      "grad_norm": 0.5021082128605233,
+      "learning_rate": 1.0857659434962744e-05,
+      "loss": 0.3502,
+      "step": 9600
+    },
+    {
+      "epoch": 1.8927444794952681,
+      "grad_norm": 0.49462878367308655,
+      "learning_rate": 1.0856115110047829e-05,
+      "loss": 0.344,
+      "step": 9601
+    },
+    {
+      "epoch": 1.8929416403785488,
+      "grad_norm": 0.5710152300950999,
+      "learning_rate": 1.0854570764564057e-05,
+      "loss": 0.3439,
+      "step": 9602
+    },
+    {
+      "epoch": 1.8931388012618298,
+      "grad_norm": 0.4611516368668063,
+      "learning_rate": 1.0853026398548535e-05,
+      "loss": 0.3331,
+      "step": 9603
+    },
+    {
+      "epoch": 1.8933359621451105,
+      "grad_norm": 0.45617109895197894,
+      "learning_rate": 1.0851482012038366e-05,
+      "loss": 0.299,
+      "step": 9604
+    },
+    {
+      "epoch": 1.8935331230283912,
+      "grad_norm": 0.4710318513048329,
+      "learning_rate": 1.0849937605070658e-05,
+      "loss": 0.3069,
+      "step": 9605
+    },
+    {
+      "epoch": 1.893730283911672,
+      "grad_norm": 0.5515208008448891,
+      "learning_rate": 1.0848393177682513e-05,
+      "loss": 0.3783,
+      "step": 9606
+    },
+    {
+      "epoch": 1.8939274447949526,
+      "grad_norm": 0.48486227843023466,
+      "learning_rate": 1.0846848729911037e-05,
+      "loss": 0.337,
+      "step": 9607
+    },
+    {
+      "epoch": 1.8941246056782335,
+      "grad_norm": 0.4820429170759435,
+      "learning_rate": 1.084530426179334e-05,
+      "loss": 0.3312,
+      "step": 9608
+    },
+    {
+      "epoch": 1.8943217665615142,
+      "grad_norm": 0.5600154595642701,
+      "learning_rate": 1.0843759773366526e-05,
+      "loss": 0.3339,
+      "step": 9609
+    },
+    {
+      "epoch": 1.8945189274447949,
+      "grad_norm": 0.48414804671739764,
+      "learning_rate": 1.0842215264667708e-05,
+      "loss": 0.3515,
+      "step": 9610
+    },
+    {
+      "epoch": 1.8947160883280758,
+      "grad_norm": 0.46300778683793087,
+      "learning_rate": 1.0840670735733984e-05,
+      "loss": 0.3089,
+      "step": 9611
+    },
+    {
+      "epoch": 1.8949132492113565,
+      "grad_norm": 0.4933738100516299,
+      "learning_rate": 1.0839126186602475e-05,
+      "loss": 0.3337,
+      "step": 9612
+    },
+    {
+      "epoch": 1.8951104100946372,
+      "grad_norm": 0.4854372334064325,
+      "learning_rate": 1.0837581617310279e-05,
+      "loss": 0.3506,
+      "step": 9613
+    },
+    {
+      "epoch": 1.8953075709779181,
+      "grad_norm": 0.458378537894229,
+      "learning_rate": 1.0836037027894515e-05,
+      "loss": 0.3205,
+      "step": 9614
+    },
+    {
+      "epoch": 1.8955047318611986,
+      "grad_norm": 0.4564763917102619,
+      "learning_rate": 1.0834492418392281e-05,
+      "loss": 0.3117,
+      "step": 9615
+    },
+    {
+      "epoch": 1.8957018927444795,
+      "grad_norm": 0.508560600760082,
+      "learning_rate": 1.0832947788840699e-05,
+      "loss": 0.3454,
+      "step": 9616
+    },
+    {
+      "epoch": 1.8958990536277602,
+      "grad_norm": 0.4692816776007284,
+      "learning_rate": 1.0831403139276875e-05,
+      "loss": 0.3364,
+      "step": 9617
+    },
+    {
+      "epoch": 1.896096214511041,
+      "grad_norm": 0.4606352560271506,
+      "learning_rate": 1.0829858469737921e-05,
+      "loss": 0.3357,
+      "step": 9618
+    },
+    {
+      "epoch": 1.8962933753943219,
+      "grad_norm": 0.4547120875924028,
+      "learning_rate": 1.082831378026095e-05,
+      "loss": 0.3286,
+      "step": 9619
+    },
+    {
+      "epoch": 1.8964905362776026,
+      "grad_norm": 0.5162498136227341,
+      "learning_rate": 1.0826769070883073e-05,
+      "loss": 0.3436,
+      "step": 9620
+    },
+    {
+      "epoch": 1.8966876971608833,
+      "grad_norm": 0.4804924944515462,
+      "learning_rate": 1.0825224341641403e-05,
+      "loss": 0.3225,
+      "step": 9621
+    },
+    {
+      "epoch": 1.8968848580441642,
+      "grad_norm": 0.4755043731674942,
+      "learning_rate": 1.0823679592573052e-05,
+      "loss": 0.3484,
+      "step": 9622
+    },
+    {
+      "epoch": 1.8970820189274447,
+      "grad_norm": 0.47229749756388667,
+      "learning_rate": 1.0822134823715139e-05,
+      "loss": 0.3348,
+      "step": 9623
+    },
+    {
+      "epoch": 1.8972791798107256,
+      "grad_norm": 0.443728175982202,
+      "learning_rate": 1.0820590035104773e-05,
+      "loss": 0.3156,
+      "step": 9624
+    },
+    {
+      "epoch": 1.8974763406940063,
+      "grad_norm": 0.45201673400276804,
+      "learning_rate": 1.0819045226779071e-05,
+      "loss": 0.3132,
+      "step": 9625
+    },
+    {
+      "epoch": 1.897673501577287,
+      "grad_norm": 0.46420972816236206,
+      "learning_rate": 1.0817500398775147e-05,
+      "loss": 0.3285,
+      "step": 9626
+    },
+    {
+      "epoch": 1.897870662460568,
+      "grad_norm": 0.4878624846038825,
+      "learning_rate": 1.0815955551130117e-05,
+      "loss": 0.3621,
+      "step": 9627
+    },
+    {
+      "epoch": 1.8980678233438486,
+      "grad_norm": 0.46796009590591403,
+      "learning_rate": 1.0814410683881098e-05,
+      "loss": 0.3341,
+      "step": 9628
+    },
+    {
+      "epoch": 1.8982649842271293,
+      "grad_norm": 0.46658974852311247,
+      "learning_rate": 1.0812865797065209e-05,
+      "loss": 0.3476,
+      "step": 9629
+    },
+    {
+      "epoch": 1.8984621451104102,
+      "grad_norm": 0.48330294725808115,
+      "learning_rate": 1.0811320890719558e-05,
+      "loss": 0.3436,
+      "step": 9630
+    },
+    {
+      "epoch": 1.8986593059936907,
+      "grad_norm": 0.47789134303226777,
+      "learning_rate": 1.0809775964881278e-05,
+      "loss": 0.3399,
+      "step": 9631
+    },
+    {
+      "epoch": 1.8988564668769716,
+      "grad_norm": 0.462828839034776,
+      "learning_rate": 1.0808231019587472e-05,
+      "loss": 0.3406,
+      "step": 9632
+    },
+    {
+      "epoch": 1.8990536277602523,
+      "grad_norm": 0.4498197895663266,
+      "learning_rate": 1.0806686054875268e-05,
+      "loss": 0.3139,
+      "step": 9633
+    },
+    {
+      "epoch": 1.899250788643533,
+      "grad_norm": 0.5032026325434863,
+      "learning_rate": 1.080514107078178e-05,
+      "loss": 0.333,
+      "step": 9634
+    },
+    {
+      "epoch": 1.899447949526814,
+      "grad_norm": 0.4669772787243432,
+      "learning_rate": 1.0803596067344134e-05,
+      "loss": 0.319,
+      "step": 9635
+    },
+    {
+      "epoch": 1.8996451104100947,
+      "grad_norm": 0.4506661711957322,
+      "learning_rate": 1.0802051044599441e-05,
+      "loss": 0.3223,
+      "step": 9636
+    },
+    {
+      "epoch": 1.8998422712933754,
+      "grad_norm": 0.47399636404681555,
+      "learning_rate": 1.0800506002584825e-05,
+      "loss": 0.3244,
+      "step": 9637
+    },
+    {
+      "epoch": 1.9000394321766563,
+      "grad_norm": 0.4508527663354465,
+      "learning_rate": 1.0798960941337411e-05,
+      "loss": 0.3369,
+      "step": 9638
+    },
+    {
+      "epoch": 1.9002365930599368,
+      "grad_norm": 0.5017850647875128,
+      "learning_rate": 1.0797415860894313e-05,
+      "loss": 0.3421,
+      "step": 9639
+    },
+    {
+      "epoch": 1.9004337539432177,
+      "grad_norm": 0.4883584730902664,
+      "learning_rate": 1.0795870761292661e-05,
+      "loss": 0.3313,
+      "step": 9640
+    },
+    {
+      "epoch": 1.9006309148264984,
+      "grad_norm": 0.4761447190397066,
+      "learning_rate": 1.079432564256957e-05,
+      "loss": 0.3124,
+      "step": 9641
+    },
+    {
+      "epoch": 1.900828075709779,
+      "grad_norm": 0.44783562310151726,
+      "learning_rate": 1.0792780504762168e-05,
+      "loss": 0.2974,
+      "step": 9642
+    },
+    {
+      "epoch": 1.90102523659306,
+      "grad_norm": 0.47770652935997016,
+      "learning_rate": 1.0791235347907573e-05,
+      "loss": 0.3256,
+      "step": 9643
+    },
+    {
+      "epoch": 1.9012223974763407,
+      "grad_norm": 0.5877815822517798,
+      "learning_rate": 1.0789690172042912e-05,
+      "loss": 0.2918,
+      "step": 9644
+    },
+    {
+      "epoch": 1.9014195583596214,
+      "grad_norm": 0.4726070261592881,
+      "learning_rate": 1.078814497720531e-05,
+      "loss": 0.321,
+      "step": 9645
+    },
+    {
+      "epoch": 1.9016167192429023,
+      "grad_norm": 0.4907105853731056,
+      "learning_rate": 1.0786599763431891e-05,
+      "loss": 0.3557,
+      "step": 9646
+    },
+    {
+      "epoch": 1.901813880126183,
+      "grad_norm": 0.4597358486182623,
+      "learning_rate": 1.078505453075978e-05,
+      "loss": 0.3013,
+      "step": 9647
+    },
+    {
+      "epoch": 1.9020110410094637,
+      "grad_norm": 0.5048614261322799,
+      "learning_rate": 1.0783509279226099e-05,
+      "loss": 0.3458,
+      "step": 9648
+    },
+    {
+      "epoch": 1.9022082018927446,
+      "grad_norm": 0.4938247930425652,
+      "learning_rate": 1.0781964008867979e-05,
+      "loss": 0.3517,
+      "step": 9649
+    },
+    {
+      "epoch": 1.9024053627760251,
+      "grad_norm": 0.4798368494115209,
+      "learning_rate": 1.0780418719722544e-05,
+      "loss": 0.3387,
+      "step": 9650
+    },
+    {
+      "epoch": 1.902602523659306,
+      "grad_norm": 0.4594536938455119,
+      "learning_rate": 1.0778873411826918e-05,
+      "loss": 0.3178,
+      "step": 9651
+    },
+    {
+      "epoch": 1.9027996845425867,
+      "grad_norm": 0.4600549598776842,
+      "learning_rate": 1.0777328085218232e-05,
+      "loss": 0.3114,
+      "step": 9652
+    },
+    {
+      "epoch": 1.9029968454258674,
+      "grad_norm": 0.6555278594200854,
+      "learning_rate": 1.0775782739933614e-05,
+      "loss": 0.3139,
+      "step": 9653
+    },
+    {
+      "epoch": 1.9031940063091484,
+      "grad_norm": 0.47716562507209187,
+      "learning_rate": 1.077423737601019e-05,
+      "loss": 0.3301,
+      "step": 9654
+    },
+    {
+      "epoch": 1.903391167192429,
+      "grad_norm": 0.47582609005457543,
+      "learning_rate": 1.0772691993485091e-05,
+      "loss": 0.3196,
+      "step": 9655
+    },
+    {
+      "epoch": 1.9035883280757098,
+      "grad_norm": 0.4654364416610627,
+      "learning_rate": 1.0771146592395443e-05,
+      "loss": 0.3211,
+      "step": 9656
+    },
+    {
+      "epoch": 1.9037854889589907,
+      "grad_norm": 0.46468189188527265,
+      "learning_rate": 1.0769601172778379e-05,
+      "loss": 0.3355,
+      "step": 9657
+    },
+    {
+      "epoch": 1.9039826498422712,
+      "grad_norm": 0.45410022068436584,
+      "learning_rate": 1.0768055734671023e-05,
+      "loss": 0.3109,
+      "step": 9658
+    },
+    {
+      "epoch": 1.904179810725552,
+      "grad_norm": 0.48221225731467116,
+      "learning_rate": 1.0766510278110514e-05,
+      "loss": 0.3421,
+      "step": 9659
+    },
+    {
+      "epoch": 1.9043769716088328,
+      "grad_norm": 0.47483955562907587,
+      "learning_rate": 1.0764964803133975e-05,
+      "loss": 0.324,
+      "step": 9660
+    },
+    {
+      "epoch": 1.9045741324921135,
+      "grad_norm": 0.5016946982181942,
+      "learning_rate": 1.0763419309778544e-05,
+      "loss": 0.3586,
+      "step": 9661
+    },
+    {
+      "epoch": 1.9047712933753944,
+      "grad_norm": 0.4816207167555914,
+      "learning_rate": 1.0761873798081343e-05,
+      "loss": 0.3425,
+      "step": 9662
+    },
+    {
+      "epoch": 1.9049684542586751,
+      "grad_norm": 0.46646055121198043,
+      "learning_rate": 1.0760328268079517e-05,
+      "loss": 0.3387,
+      "step": 9663
+    },
+    {
+      "epoch": 1.9051656151419558,
+      "grad_norm": 0.4558246279347243,
+      "learning_rate": 1.075878271981019e-05,
+      "loss": 0.3035,
+      "step": 9664
+    },
+    {
+      "epoch": 1.9053627760252367,
+      "grad_norm": 0.49067026661890945,
+      "learning_rate": 1.0757237153310496e-05,
+      "loss": 0.3373,
+      "step": 9665
+    },
+    {
+      "epoch": 1.9055599369085172,
+      "grad_norm": 0.4654982629990961,
+      "learning_rate": 1.0755691568617573e-05,
+      "loss": 0.3214,
+      "step": 9666
+    },
+    {
+      "epoch": 1.9057570977917981,
+      "grad_norm": 0.48044344209508094,
+      "learning_rate": 1.0754145965768548e-05,
+      "loss": 0.3276,
+      "step": 9667
+    },
+    {
+      "epoch": 1.9059542586750788,
+      "grad_norm": 0.4639232573098766,
+      "learning_rate": 1.075260034480056e-05,
+      "loss": 0.3367,
+      "step": 9668
+    },
+    {
+      "epoch": 1.9061514195583595,
+      "grad_norm": 0.4610967414683846,
+      "learning_rate": 1.0751054705750744e-05,
+      "loss": 0.3195,
+      "step": 9669
+    },
+    {
+      "epoch": 1.9063485804416405,
+      "grad_norm": 0.4456558119416812,
+      "learning_rate": 1.0749509048656231e-05,
+      "loss": 0.3163,
+      "step": 9670
+    },
+    {
+      "epoch": 1.9065457413249212,
+      "grad_norm": 0.4679939791043013,
+      "learning_rate": 1.074796337355416e-05,
+      "loss": 0.3354,
+      "step": 9671
+    },
+    {
+      "epoch": 1.9067429022082019,
+      "grad_norm": 0.4886824586751215,
+      "learning_rate": 1.074641768048167e-05,
+      "loss": 0.3325,
+      "step": 9672
+    },
+    {
+      "epoch": 1.9069400630914828,
+      "grad_norm": 0.6283610239404007,
+      "learning_rate": 1.074487196947589e-05,
+      "loss": 0.3256,
+      "step": 9673
+    },
+    {
+      "epoch": 1.9071372239747633,
+      "grad_norm": 0.47479516642440406,
+      "learning_rate": 1.0743326240573964e-05,
+      "loss": 0.3345,
+      "step": 9674
+    },
+    {
+      "epoch": 1.9073343848580442,
+      "grad_norm": 0.48102011928532773,
+      "learning_rate": 1.0741780493813025e-05,
+      "loss": 0.3687,
+      "step": 9675
+    },
+    {
+      "epoch": 1.9075315457413249,
+      "grad_norm": 0.44787874825005314,
+      "learning_rate": 1.0740234729230213e-05,
+      "loss": 0.3302,
+      "step": 9676
+    },
+    {
+      "epoch": 1.9077287066246056,
+      "grad_norm": 0.4786425290897722,
+      "learning_rate": 1.0738688946862661e-05,
+      "loss": 0.3335,
+      "step": 9677
+    },
+    {
+      "epoch": 1.9079258675078865,
+      "grad_norm": 0.47312252556976,
+      "learning_rate": 1.073714314674752e-05,
+      "loss": 0.3166,
+      "step": 9678
+    },
+    {
+      "epoch": 1.9081230283911672,
+      "grad_norm": 0.4382653645073557,
+      "learning_rate": 1.0735597328921914e-05,
+      "loss": 0.3034,
+      "step": 9679
+    },
+    {
+      "epoch": 1.908320189274448,
+      "grad_norm": 0.44778499820386447,
+      "learning_rate": 1.0734051493422996e-05,
+      "loss": 0.3082,
+      "step": 9680
+    },
+    {
+      "epoch": 1.9085173501577288,
+      "grad_norm": 0.46994341815466767,
+      "learning_rate": 1.0732505640287895e-05,
+      "loss": 0.3271,
+      "step": 9681
+    },
+    {
+      "epoch": 1.9087145110410093,
+      "grad_norm": 0.5035019762587363,
+      "learning_rate": 1.0730959769553762e-05,
+      "loss": 0.3197,
+      "step": 9682
+    },
+    {
+      "epoch": 1.9089116719242902,
+      "grad_norm": 0.4658601369964348,
+      "learning_rate": 1.0729413881257725e-05,
+      "loss": 0.3325,
+      "step": 9683
+    },
+    {
+      "epoch": 1.909108832807571,
+      "grad_norm": 0.48330429263892694,
+      "learning_rate": 1.0727867975436936e-05,
+      "loss": 0.322,
+      "step": 9684
+    },
+    {
+      "epoch": 1.9093059936908516,
+      "grad_norm": 0.489211526633787,
+      "learning_rate": 1.072632205212853e-05,
+      "loss": 0.3395,
+      "step": 9685
+    },
+    {
+      "epoch": 1.9095031545741326,
+      "grad_norm": 0.49250668046451535,
+      "learning_rate": 1.0724776111369654e-05,
+      "loss": 0.3657,
+      "step": 9686
+    },
+    {
+      "epoch": 1.9097003154574133,
+      "grad_norm": 0.4740052263296321,
+      "learning_rate": 1.072323015319745e-05,
+      "loss": 0.3284,
+      "step": 9687
+    },
+    {
+      "epoch": 1.909897476340694,
+      "grad_norm": 0.46842386445090717,
+      "learning_rate": 1.0721684177649056e-05,
+      "loss": 0.3339,
+      "step": 9688
+    },
+    {
+      "epoch": 1.9100946372239749,
+      "grad_norm": 0.46625362508947077,
+      "learning_rate": 1.0720138184761621e-05,
+      "loss": 0.342,
+      "step": 9689
+    },
+    {
+      "epoch": 1.9102917981072554,
+      "grad_norm": 0.4848740749436997,
+      "learning_rate": 1.0718592174572285e-05,
+      "loss": 0.3375,
+      "step": 9690
+    },
+    {
+      "epoch": 1.9104889589905363,
+      "grad_norm": 0.48490516898151786,
+      "learning_rate": 1.0717046147118193e-05,
+      "loss": 0.3189,
+      "step": 9691
+    },
+    {
+      "epoch": 1.910686119873817,
+      "grad_norm": 0.46286777382849786,
+      "learning_rate": 1.071550010243649e-05,
+      "loss": 0.3207,
+      "step": 9692
+    },
+    {
+      "epoch": 1.9108832807570977,
+      "grad_norm": 0.4509100884863312,
+      "learning_rate": 1.071395404056432e-05,
+      "loss": 0.3283,
+      "step": 9693
+    },
+    {
+      "epoch": 1.9110804416403786,
+      "grad_norm": 0.4959185650773511,
+      "learning_rate": 1.071240796153883e-05,
+      "loss": 0.3451,
+      "step": 9694
+    },
+    {
+      "epoch": 1.9112776025236593,
+      "grad_norm": 0.4708104478931626,
+      "learning_rate": 1.0710861865397166e-05,
+      "loss": 0.348,
+      "step": 9695
+    },
+    {
+      "epoch": 1.91147476340694,
+      "grad_norm": 0.48697087265632166,
+      "learning_rate": 1.0709315752176472e-05,
+      "loss": 0.3357,
+      "step": 9696
+    },
+    {
+      "epoch": 1.911671924290221,
+      "grad_norm": 0.49229875756280933,
+      "learning_rate": 1.0707769621913897e-05,
+      "loss": 0.3522,
+      "step": 9697
+    },
+    {
+      "epoch": 1.9118690851735016,
+      "grad_norm": 0.4787942182967307,
+      "learning_rate": 1.0706223474646581e-05,
+      "loss": 0.3469,
+      "step": 9698
+    },
+    {
+      "epoch": 1.9120662460567823,
+      "grad_norm": 0.46221031697674925,
+      "learning_rate": 1.0704677310411686e-05,
+      "loss": 0.3215,
+      "step": 9699
+    },
+    {
+      "epoch": 1.9122634069400632,
+      "grad_norm": 0.49657974351122547,
+      "learning_rate": 1.0703131129246347e-05,
+      "loss": 0.3449,
+      "step": 9700
+    },
+    {
+      "epoch": 1.9124605678233437,
+      "grad_norm": 0.4700103591052916,
+      "learning_rate": 1.070158493118772e-05,
+      "loss": 0.3143,
+      "step": 9701
+    },
+    {
+      "epoch": 1.9126577287066246,
+      "grad_norm": 0.5045654021745932,
+      "learning_rate": 1.0700038716272944e-05,
+      "loss": 0.3246,
+      "step": 9702
+    },
+    {
+      "epoch": 1.9128548895899053,
+      "grad_norm": 0.4708499061368683,
+      "learning_rate": 1.0698492484539178e-05,
+      "loss": 0.3436,
+      "step": 9703
+    },
+    {
+      "epoch": 1.913052050473186,
+      "grad_norm": 0.4808287774080232,
+      "learning_rate": 1.0696946236023566e-05,
+      "loss": 0.3474,
+      "step": 9704
+    },
+    {
+      "epoch": 1.913249211356467,
+      "grad_norm": 0.4919946738804181,
+      "learning_rate": 1.0695399970763258e-05,
+      "loss": 0.331,
+      "step": 9705
+    },
+    {
+      "epoch": 1.9134463722397477,
+      "grad_norm": 0.48323751424237105,
+      "learning_rate": 1.069385368879541e-05,
+      "loss": 0.3436,
+      "step": 9706
+    },
+    {
+      "epoch": 1.9136435331230284,
+      "grad_norm": 0.4694330210445866,
+      "learning_rate": 1.0692307390157164e-05,
+      "loss": 0.3144,
+      "step": 9707
+    },
+    {
+      "epoch": 1.9138406940063093,
+      "grad_norm": 0.4689921139270199,
+      "learning_rate": 1.069076107488568e-05,
+      "loss": 0.3196,
+      "step": 9708
+    },
+    {
+      "epoch": 1.9140378548895898,
+      "grad_norm": 0.46415715567906024,
+      "learning_rate": 1.0689214743018102e-05,
+      "loss": 0.3288,
+      "step": 9709
+    },
+    {
+      "epoch": 1.9142350157728707,
+      "grad_norm": 0.515132913417703,
+      "learning_rate": 1.0687668394591586e-05,
+      "loss": 0.3284,
+      "step": 9710
+    },
+    {
+      "epoch": 1.9144321766561514,
+      "grad_norm": 0.4513787042971108,
+      "learning_rate": 1.068612202964328e-05,
+      "loss": 0.311,
+      "step": 9711
+    },
+    {
+      "epoch": 1.914629337539432,
+      "grad_norm": 0.502933214732867,
+      "learning_rate": 1.0684575648210343e-05,
+      "loss": 0.3317,
+      "step": 9712
+    },
+    {
+      "epoch": 1.914826498422713,
+      "grad_norm": 0.45248819432377035,
+      "learning_rate": 1.0683029250329924e-05,
+      "loss": 0.3317,
+      "step": 9713
+    },
+    {
+      "epoch": 1.9150236593059937,
+      "grad_norm": 0.46741383909252043,
+      "learning_rate": 1.0681482836039176e-05,
+      "loss": 0.3067,
+      "step": 9714
+    },
+    {
+      "epoch": 1.9152208201892744,
+      "grad_norm": 0.47201782201403564,
+      "learning_rate": 1.0679936405375255e-05,
+      "loss": 0.3541,
+      "step": 9715
+    },
+    {
+      "epoch": 1.9154179810725553,
+      "grad_norm": 0.4945029413774007,
+      "learning_rate": 1.0678389958375316e-05,
+      "loss": 0.3423,
+      "step": 9716
+    },
+    {
+      "epoch": 1.9156151419558358,
+      "grad_norm": 0.5418610271930345,
+      "learning_rate": 1.067684349507651e-05,
+      "loss": 0.3696,
+      "step": 9717
+    },
+    {
+      "epoch": 1.9158123028391167,
+      "grad_norm": 0.5060226118055992,
+      "learning_rate": 1.0675297015515993e-05,
+      "loss": 0.3504,
+      "step": 9718
+    },
+    {
+      "epoch": 1.9160094637223974,
+      "grad_norm": 0.44874325788032515,
+      "learning_rate": 1.0673750519730923e-05,
+      "loss": 0.2836,
+      "step": 9719
+    },
+    {
+      "epoch": 1.9162066246056781,
+      "grad_norm": 0.4762929005302375,
+      "learning_rate": 1.0672204007758453e-05,
+      "loss": 0.3362,
+      "step": 9720
+    },
+    {
+      "epoch": 1.916403785488959,
+      "grad_norm": 0.4643277801877159,
+      "learning_rate": 1.0670657479635742e-05,
+      "loss": 0.3236,
+      "step": 9721
+    },
+    {
+      "epoch": 1.9166009463722398,
+      "grad_norm": 0.49866980334934874,
+      "learning_rate": 1.0669110935399944e-05,
+      "loss": 0.3474,
+      "step": 9722
+    },
+    {
+      "epoch": 1.9167981072555205,
+      "grad_norm": 0.46636862046761957,
+      "learning_rate": 1.0667564375088218e-05,
+      "loss": 0.3371,
+      "step": 9723
+    },
+    {
+      "epoch": 1.9169952681388014,
+      "grad_norm": 0.4624733590198138,
+      "learning_rate": 1.066601779873772e-05,
+      "loss": 0.3149,
+      "step": 9724
+    },
+    {
+      "epoch": 1.9171924290220819,
+      "grad_norm": 0.4477193303841462,
+      "learning_rate": 1.0664471206385607e-05,
+      "loss": 0.2978,
+      "step": 9725
+    },
+    {
+      "epoch": 1.9173895899053628,
+      "grad_norm": 0.4436196530416114,
+      "learning_rate": 1.0662924598069035e-05,
+      "loss": 0.302,
+      "step": 9726
+    },
+    {
+      "epoch": 1.9175867507886435,
+      "grad_norm": 0.4980769406665461,
+      "learning_rate": 1.0661377973825173e-05,
+      "loss": 0.355,
+      "step": 9727
+    },
+    {
+      "epoch": 1.9177839116719242,
+      "grad_norm": 0.4692953470690748,
+      "learning_rate": 1.0659831333691166e-05,
+      "loss": 0.3222,
+      "step": 9728
+    },
+    {
+      "epoch": 1.9179810725552051,
+      "grad_norm": 0.4855708808826588,
+      "learning_rate": 1.0658284677704187e-05,
+      "loss": 0.3387,
+      "step": 9729
+    },
+    {
+      "epoch": 1.9181782334384858,
+      "grad_norm": 0.4926783442889073,
+      "learning_rate": 1.0656738005901382e-05,
+      "loss": 0.3263,
+      "step": 9730
+    },
+    {
+      "epoch": 1.9183753943217665,
+      "grad_norm": 0.49008026881045197,
+      "learning_rate": 1.0655191318319921e-05,
+      "loss": 0.3788,
+      "step": 9731
+    },
+    {
+      "epoch": 1.9185725552050474,
+      "grad_norm": 0.47996998660573686,
+      "learning_rate": 1.0653644614996958e-05,
+      "loss": 0.3424,
+      "step": 9732
+    },
+    {
+      "epoch": 1.918769716088328,
+      "grad_norm": 0.6127908630549663,
+      "learning_rate": 1.0652097895969657e-05,
+      "loss": 0.3558,
+      "step": 9733
+    },
+    {
+      "epoch": 1.9189668769716088,
+      "grad_norm": 0.4627932710115664,
+      "learning_rate": 1.0650551161275182e-05,
+      "loss": 0.3155,
+      "step": 9734
+    },
+    {
+      "epoch": 1.9191640378548895,
+      "grad_norm": 0.45413739024853905,
+      "learning_rate": 1.064900441095069e-05,
+      "loss": 0.3031,
+      "step": 9735
+    },
+    {
+      "epoch": 1.9193611987381702,
+      "grad_norm": 0.4495957982119002,
+      "learning_rate": 1.0647457645033343e-05,
+      "loss": 0.3198,
+      "step": 9736
+    },
+    {
+      "epoch": 1.9195583596214512,
+      "grad_norm": 0.44672878440578423,
+      "learning_rate": 1.0645910863560306e-05,
+      "loss": 0.3065,
+      "step": 9737
+    },
+    {
+      "epoch": 1.9197555205047319,
+      "grad_norm": 0.4540120569214132,
+      "learning_rate": 1.0644364066568742e-05,
+      "loss": 0.316,
+      "step": 9738
+    },
+    {
+      "epoch": 1.9199526813880126,
+      "grad_norm": 1.8452643662313766,
+      "learning_rate": 1.0642817254095809e-05,
+      "loss": 0.3175,
+      "step": 9739
+    },
+    {
+      "epoch": 1.9201498422712935,
+      "grad_norm": 0.5187412295956796,
+      "learning_rate": 1.0641270426178677e-05,
+      "loss": 0.3445,
+      "step": 9740
+    },
+    {
+      "epoch": 1.9203470031545742,
+      "grad_norm": 0.4665160701342525,
+      "learning_rate": 1.0639723582854505e-05,
+      "loss": 0.3309,
+      "step": 9741
+    },
+    {
+      "epoch": 1.9205441640378549,
+      "grad_norm": 0.4551439613909609,
+      "learning_rate": 1.0638176724160458e-05,
+      "loss": 0.3291,
+      "step": 9742
+    },
+    {
+      "epoch": 1.9207413249211358,
+      "grad_norm": 0.5160208376705452,
+      "learning_rate": 1.0636629850133705e-05,
+      "loss": 0.3391,
+      "step": 9743
+    },
+    {
+      "epoch": 1.9209384858044163,
+      "grad_norm": 0.49989041787707916,
+      "learning_rate": 1.0635082960811403e-05,
+      "loss": 0.3444,
+      "step": 9744
+    },
+    {
+      "epoch": 1.9211356466876972,
+      "grad_norm": 0.4964849127683154,
+      "learning_rate": 1.063353605623072e-05,
+      "loss": 0.3492,
+      "step": 9745
+    },
+    {
+      "epoch": 1.921332807570978,
+      "grad_norm": 0.6049573542199426,
+      "learning_rate": 1.0631989136428828e-05,
+      "loss": 0.3631,
+      "step": 9746
+    },
+    {
+      "epoch": 1.9215299684542586,
+      "grad_norm": 0.4756893237472388,
+      "learning_rate": 1.0630442201442884e-05,
+      "loss": 0.3283,
+      "step": 9747
+    },
+    {
+      "epoch": 1.9217271293375395,
+      "grad_norm": 0.6792395870038784,
+      "learning_rate": 1.0628895251310063e-05,
+      "loss": 0.3448,
+      "step": 9748
+    },
+    {
+      "epoch": 1.9219242902208202,
+      "grad_norm": 0.4749099206929661,
+      "learning_rate": 1.0627348286067521e-05,
+      "loss": 0.3263,
+      "step": 9749
+    },
+    {
+      "epoch": 1.922121451104101,
+      "grad_norm": 0.46009586758698623,
+      "learning_rate": 1.0625801305752436e-05,
+      "loss": 0.3141,
+      "step": 9750
+    },
+    {
+      "epoch": 1.9223186119873819,
+      "grad_norm": 0.46687228371318157,
+      "learning_rate": 1.062425431040197e-05,
+      "loss": 0.3375,
+      "step": 9751
+    },
+    {
+      "epoch": 1.9225157728706623,
+      "grad_norm": 0.5157651333274518,
+      "learning_rate": 1.062270730005329e-05,
+      "loss": 0.3194,
+      "step": 9752
+    },
+    {
+      "epoch": 1.9227129337539433,
+      "grad_norm": 0.5008517739542978,
+      "learning_rate": 1.0621160274743564e-05,
+      "loss": 0.3596,
+      "step": 9753
+    },
+    {
+      "epoch": 1.922910094637224,
+      "grad_norm": 0.44671709049249914,
+      "learning_rate": 1.0619613234509967e-05,
+      "loss": 0.3119,
+      "step": 9754
+    },
+    {
+      "epoch": 1.9231072555205047,
+      "grad_norm": 0.4313404095262679,
+      "learning_rate": 1.0618066179389663e-05,
+      "loss": 0.2903,
+      "step": 9755
+    },
+    {
+      "epoch": 1.9233044164037856,
+      "grad_norm": 0.44712186986789765,
+      "learning_rate": 1.0616519109419815e-05,
+      "loss": 0.31,
+      "step": 9756
+    },
+    {
+      "epoch": 1.9235015772870663,
+      "grad_norm": 0.48107891788084883,
+      "learning_rate": 1.0614972024637606e-05,
+      "loss": 0.3371,
+      "step": 9757
+    },
+    {
+      "epoch": 1.923698738170347,
+      "grad_norm": 0.4302137346468184,
+      "learning_rate": 1.0613424925080194e-05,
+      "loss": 0.2886,
+      "step": 9758
+    },
+    {
+      "epoch": 1.923895899053628,
+      "grad_norm": 0.4463000909131128,
+      "learning_rate": 1.0611877810784756e-05,
+      "loss": 0.3051,
+      "step": 9759
+    },
+    {
+      "epoch": 1.9240930599369084,
+      "grad_norm": 0.46877366951404054,
+      "learning_rate": 1.061033068178846e-05,
+      "loss": 0.3319,
+      "step": 9760
+    },
+    {
+      "epoch": 1.9242902208201893,
+      "grad_norm": 0.45058760214666865,
+      "learning_rate": 1.0608783538128479e-05,
+      "loss": 0.3294,
+      "step": 9761
+    },
+    {
+      "epoch": 1.92448738170347,
+      "grad_norm": 0.4673506711450003,
+      "learning_rate": 1.0607236379841984e-05,
+      "loss": 0.3131,
+      "step": 9762
+    },
+    {
+      "epoch": 1.9246845425867507,
+      "grad_norm": 0.4448969256754839,
+      "learning_rate": 1.0605689206966145e-05,
+      "loss": 0.3046,
+      "step": 9763
+    },
+    {
+      "epoch": 1.9248817034700316,
+      "grad_norm": 0.44398749817741634,
+      "learning_rate": 1.0604142019538135e-05,
+      "loss": 0.3036,
+      "step": 9764
+    },
+    {
+      "epoch": 1.9250788643533123,
+      "grad_norm": 0.4662281069913097,
+      "learning_rate": 1.0602594817595126e-05,
+      "loss": 0.3327,
+      "step": 9765
+    },
+    {
+      "epoch": 1.925276025236593,
+      "grad_norm": 0.6123896643925617,
+      "learning_rate": 1.060104760117429e-05,
+      "loss": 0.2994,
+      "step": 9766
+    },
+    {
+      "epoch": 1.925473186119874,
+      "grad_norm": 0.4800059909800257,
+      "learning_rate": 1.0599500370312805e-05,
+      "loss": 0.332,
+      "step": 9767
+    },
+    {
+      "epoch": 1.9256703470031544,
+      "grad_norm": 0.47073997673558676,
+      "learning_rate": 1.0597953125047839e-05,
+      "loss": 0.3173,
+      "step": 9768
+    },
+    {
+      "epoch": 1.9258675078864353,
+      "grad_norm": 0.486613027526376,
+      "learning_rate": 1.0596405865416569e-05,
+      "loss": 0.3397,
+      "step": 9769
+    },
+    {
+      "epoch": 1.926064668769716,
+      "grad_norm": 0.4412020035394467,
+      "learning_rate": 1.0594858591456166e-05,
+      "loss": 0.3088,
+      "step": 9770
+    },
+    {
+      "epoch": 1.9262618296529967,
+      "grad_norm": 0.4645088549635914,
+      "learning_rate": 1.0593311303203806e-05,
+      "loss": 0.3399,
+      "step": 9771
+    },
+    {
+      "epoch": 1.9264589905362777,
+      "grad_norm": 0.5045226039115558,
+      "learning_rate": 1.0591764000696665e-05,
+      "loss": 0.3695,
+      "step": 9772
+    },
+    {
+      "epoch": 1.9266561514195584,
+      "grad_norm": 0.4783611486444141,
+      "learning_rate": 1.0590216683971915e-05,
+      "loss": 0.334,
+      "step": 9773
+    },
+    {
+      "epoch": 1.926853312302839,
+      "grad_norm": 0.47330458094671435,
+      "learning_rate": 1.0588669353066739e-05,
+      "loss": 0.3499,
+      "step": 9774
+    },
+    {
+      "epoch": 1.92705047318612,
+      "grad_norm": 0.4625886872385459,
+      "learning_rate": 1.0587122008018303e-05,
+      "loss": 0.3267,
+      "step": 9775
+    },
+    {
+      "epoch": 1.9272476340694005,
+      "grad_norm": 0.4856730238683575,
+      "learning_rate": 1.058557464886379e-05,
+      "loss": 0.3312,
+      "step": 9776
+    },
+    {
+      "epoch": 1.9274447949526814,
+      "grad_norm": 0.46470351748252176,
+      "learning_rate": 1.0584027275640372e-05,
+      "loss": 0.3281,
+      "step": 9777
+    },
+    {
+      "epoch": 1.927641955835962,
+      "grad_norm": 0.4805088560192612,
+      "learning_rate": 1.0582479888385233e-05,
+      "loss": 0.3373,
+      "step": 9778
+    },
+    {
+      "epoch": 1.9278391167192428,
+      "grad_norm": 0.4876012478543451,
+      "learning_rate": 1.0580932487135541e-05,
+      "loss": 0.3425,
+      "step": 9779
+    },
+    {
+      "epoch": 1.9280362776025237,
+      "grad_norm": 0.5473660280771162,
+      "learning_rate": 1.057938507192848e-05,
+      "loss": 0.351,
+      "step": 9780
+    },
+    {
+      "epoch": 1.9282334384858044,
+      "grad_norm": 0.4684135144874953,
+      "learning_rate": 1.0577837642801227e-05,
+      "loss": 0.338,
+      "step": 9781
+    },
+    {
+      "epoch": 1.9284305993690851,
+      "grad_norm": 0.48445875938977345,
+      "learning_rate": 1.0576290199790959e-05,
+      "loss": 0.3315,
+      "step": 9782
+    },
+    {
+      "epoch": 1.928627760252366,
+      "grad_norm": 0.5000034658687103,
+      "learning_rate": 1.0574742742934853e-05,
+      "loss": 0.3623,
+      "step": 9783
+    },
+    {
+      "epoch": 1.9288249211356467,
+      "grad_norm": 0.4536757092944629,
+      "learning_rate": 1.0573195272270091e-05,
+      "loss": 0.3083,
+      "step": 9784
+    },
+    {
+      "epoch": 1.9290220820189274,
+      "grad_norm": 0.47051487862294444,
+      "learning_rate": 1.0571647787833853e-05,
+      "loss": 0.3294,
+      "step": 9785
+    },
+    {
+      "epoch": 1.9292192429022084,
+      "grad_norm": 0.48487850881869465,
+      "learning_rate": 1.0570100289663314e-05,
+      "loss": 0.3444,
+      "step": 9786
+    },
+    {
+      "epoch": 1.9294164037854888,
+      "grad_norm": 0.4587817967805745,
+      "learning_rate": 1.0568552777795657e-05,
+      "loss": 0.3369,
+      "step": 9787
+    },
+    {
+      "epoch": 1.9296135646687698,
+      "grad_norm": 0.4489966707521766,
+      "learning_rate": 1.0567005252268063e-05,
+      "loss": 0.3023,
+      "step": 9788
+    },
+    {
+      "epoch": 1.9298107255520505,
+      "grad_norm": 0.5396452051256785,
+      "learning_rate": 1.056545771311771e-05,
+      "loss": 0.3587,
+      "step": 9789
+    },
+    {
+      "epoch": 1.9300078864353312,
+      "grad_norm": 0.45895425892398584,
+      "learning_rate": 1.056391016038178e-05,
+      "loss": 0.3135,
+      "step": 9790
+    },
+    {
+      "epoch": 1.930205047318612,
+      "grad_norm": 0.45902487475066067,
+      "learning_rate": 1.0562362594097456e-05,
+      "loss": 0.3256,
+      "step": 9791
+    },
+    {
+      "epoch": 1.9304022082018928,
+      "grad_norm": 0.48459698795680317,
+      "learning_rate": 1.0560815014301916e-05,
+      "loss": 0.3315,
+      "step": 9792
+    },
+    {
+      "epoch": 1.9305993690851735,
+      "grad_norm": 0.48790563977946905,
+      "learning_rate": 1.0559267421032345e-05,
+      "loss": 0.3464,
+      "step": 9793
+    },
+    {
+      "epoch": 1.9307965299684544,
+      "grad_norm": 0.45850742759324065,
+      "learning_rate": 1.055771981432592e-05,
+      "loss": 0.3014,
+      "step": 9794
+    },
+    {
+      "epoch": 1.930993690851735,
+      "grad_norm": 0.5039083825372765,
+      "learning_rate": 1.0556172194219831e-05,
+      "loss": 0.3541,
+      "step": 9795
+    },
+    {
+      "epoch": 1.9311908517350158,
+      "grad_norm": 0.48911414228696054,
+      "learning_rate": 1.0554624560751254e-05,
+      "loss": 0.3467,
+      "step": 9796
+    },
+    {
+      "epoch": 1.9313880126182965,
+      "grad_norm": 0.4929694442801919,
+      "learning_rate": 1.0553076913957381e-05,
+      "loss": 0.3387,
+      "step": 9797
+    },
+    {
+      "epoch": 1.9315851735015772,
+      "grad_norm": 0.47700645459297586,
+      "learning_rate": 1.0551529253875383e-05,
+      "loss": 0.3645,
+      "step": 9798
+    },
+    {
+      "epoch": 1.9317823343848581,
+      "grad_norm": 0.45061951666064487,
+      "learning_rate": 1.0549981580542457e-05,
+      "loss": 0.2962,
+      "step": 9799
+    },
+    {
+      "epoch": 1.9319794952681388,
+      "grad_norm": 0.47286203795175763,
+      "learning_rate": 1.0548433893995775e-05,
+      "loss": 0.3213,
+      "step": 9800
+    },
+    {
+      "epoch": 1.9321766561514195,
+      "grad_norm": 0.486492356786951,
+      "learning_rate": 1.054688619427253e-05,
+      "loss": 0.3414,
+      "step": 9801
+    },
+    {
+      "epoch": 1.9323738170347005,
+      "grad_norm": 0.48260028863528787,
+      "learning_rate": 1.0545338481409903e-05,
+      "loss": 0.3489,
+      "step": 9802
+    },
+    {
+      "epoch": 1.932570977917981,
+      "grad_norm": 0.6276466016632191,
+      "learning_rate": 1.054379075544508e-05,
+      "loss": 0.3597,
+      "step": 9803
+    },
+    {
+      "epoch": 1.9327681388012619,
+      "grad_norm": 0.44166593615676375,
+      "learning_rate": 1.0542243016415248e-05,
+      "loss": 0.3116,
+      "step": 9804
+    },
+    {
+      "epoch": 1.9329652996845426,
+      "grad_norm": 6.057606271885772,
+      "learning_rate": 1.0540695264357587e-05,
+      "loss": 0.3798,
+      "step": 9805
+    },
+    {
+      "epoch": 1.9331624605678233,
+      "grad_norm": 0.4682081351995988,
+      "learning_rate": 1.053914749930929e-05,
+      "loss": 0.3357,
+      "step": 9806
+    },
+    {
+      "epoch": 1.9333596214511042,
+      "grad_norm": 0.45711029196728215,
+      "learning_rate": 1.0537599721307538e-05,
+      "loss": 0.3182,
+      "step": 9807
+    },
+    {
+      "epoch": 1.9335567823343849,
+      "grad_norm": 0.477630148881728,
+      "learning_rate": 1.0536051930389522e-05,
+      "loss": 0.3151,
+      "step": 9808
+    },
+    {
+      "epoch": 1.9337539432176656,
+      "grad_norm": 0.45025864365539675,
+      "learning_rate": 1.0534504126592426e-05,
+      "loss": 0.321,
+      "step": 9809
+    },
+    {
+      "epoch": 1.9339511041009465,
+      "grad_norm": 0.45003521165597976,
+      "learning_rate": 1.0532956309953437e-05,
+      "loss": 0.3104,
+      "step": 9810
+    },
+    {
+      "epoch": 1.934148264984227,
+      "grad_norm": 0.5519090372934141,
+      "learning_rate": 1.0531408480509744e-05,
+      "loss": 0.3511,
+      "step": 9811
+    },
+    {
+      "epoch": 1.934345425867508,
+      "grad_norm": 0.45552387144099366,
+      "learning_rate": 1.0529860638298535e-05,
+      "loss": 0.3169,
+      "step": 9812
+    },
+    {
+      "epoch": 1.9345425867507886,
+      "grad_norm": 0.48604649269003025,
+      "learning_rate": 1.0528312783356998e-05,
+      "loss": 0.3348,
+      "step": 9813
+    },
+    {
+      "epoch": 1.9347397476340693,
+      "grad_norm": 0.48778714572885906,
+      "learning_rate": 1.0526764915722319e-05,
+      "loss": 0.3564,
+      "step": 9814
+    },
+    {
+      "epoch": 1.9349369085173502,
+      "grad_norm": 0.46707720374812933,
+      "learning_rate": 1.0525217035431687e-05,
+      "loss": 0.3261,
+      "step": 9815
+    },
+    {
+      "epoch": 1.935134069400631,
+      "grad_norm": 0.5104817179313064,
+      "learning_rate": 1.0523669142522296e-05,
+      "loss": 0.3757,
+      "step": 9816
+    },
+    {
+      "epoch": 1.9353312302839116,
+      "grad_norm": 0.4760419839422579,
+      "learning_rate": 1.0522121237031331e-05,
+      "loss": 0.3202,
+      "step": 9817
+    },
+    {
+      "epoch": 1.9355283911671926,
+      "grad_norm": 0.44494383811455845,
+      "learning_rate": 1.0520573318995986e-05,
+      "loss": 0.32,
+      "step": 9818
+    },
+    {
+      "epoch": 1.935725552050473,
+      "grad_norm": 0.46156901553051594,
+      "learning_rate": 1.051902538845344e-05,
+      "loss": 0.3212,
+      "step": 9819
+    },
+    {
+      "epoch": 1.935922712933754,
+      "grad_norm": 0.4658238149819304,
+      "learning_rate": 1.0517477445440898e-05,
+      "loss": 0.321,
+      "step": 9820
+    },
+    {
+      "epoch": 1.9361198738170347,
+      "grad_norm": 0.44452316579492956,
+      "learning_rate": 1.0515929489995544e-05,
+      "loss": 0.3043,
+      "step": 9821
+    },
+    {
+      "epoch": 1.9363170347003154,
+      "grad_norm": 0.45081267801444497,
+      "learning_rate": 1.0514381522154563e-05,
+      "loss": 0.3149,
+      "step": 9822
+    },
+    {
+      "epoch": 1.9365141955835963,
+      "grad_norm": 0.5245331794257556,
+      "learning_rate": 1.0512833541955158e-05,
+      "loss": 0.3809,
+      "step": 9823
+    },
+    {
+      "epoch": 1.936711356466877,
+      "grad_norm": 0.4930837487343131,
+      "learning_rate": 1.0511285549434509e-05,
+      "loss": 0.3181,
+      "step": 9824
+    },
+    {
+      "epoch": 1.9369085173501577,
+      "grad_norm": 0.4545151283658732,
+      "learning_rate": 1.0509737544629817e-05,
+      "loss": 0.3117,
+      "step": 9825
+    },
+    {
+      "epoch": 1.9371056782334386,
+      "grad_norm": 0.47452766613882447,
+      "learning_rate": 1.0508189527578268e-05,
+      "loss": 0.3187,
+      "step": 9826
+    },
+    {
+      "epoch": 1.937302839116719,
+      "grad_norm": 0.44914864022298373,
+      "learning_rate": 1.0506641498317056e-05,
+      "loss": 0.3108,
+      "step": 9827
+    },
+    {
+      "epoch": 1.9375,
+      "grad_norm": 6.606550108751509,
+      "learning_rate": 1.0505093456883373e-05,
+      "loss": 0.3925,
+      "step": 9828
+    },
+    {
+      "epoch": 1.937697160883281,
+      "grad_norm": 0.48300218271218115,
+      "learning_rate": 1.0503545403314414e-05,
+      "loss": 0.3362,
+      "step": 9829
+    },
+    {
+      "epoch": 1.9378943217665614,
+      "grad_norm": 0.4709845468244718,
+      "learning_rate": 1.0501997337647372e-05,
+      "loss": 0.3292,
+      "step": 9830
+    },
+    {
+      "epoch": 1.9380914826498423,
+      "grad_norm": 0.4845184212039963,
+      "learning_rate": 1.050044925991944e-05,
+      "loss": 0.3554,
+      "step": 9831
+    },
+    {
+      "epoch": 1.938288643533123,
+      "grad_norm": 0.4879693795170068,
+      "learning_rate": 1.049890117016781e-05,
+      "loss": 0.3615,
+      "step": 9832
+    },
+    {
+      "epoch": 1.9384858044164037,
+      "grad_norm": 0.46307402255271723,
+      "learning_rate": 1.0497353068429678e-05,
+      "loss": 0.3401,
+      "step": 9833
+    },
+    {
+      "epoch": 1.9386829652996846,
+      "grad_norm": 0.5531223341102023,
+      "learning_rate": 1.049580495474224e-05,
+      "loss": 0.3652,
+      "step": 9834
+    },
+    {
+      "epoch": 1.9388801261829653,
+      "grad_norm": 0.4687758624727122,
+      "learning_rate": 1.0494256829142687e-05,
+      "loss": 0.3233,
+      "step": 9835
+    },
+    {
+      "epoch": 1.939077287066246,
+      "grad_norm": 0.4923244502736517,
+      "learning_rate": 1.0492708691668216e-05,
+      "loss": 0.3246,
+      "step": 9836
+    },
+    {
+      "epoch": 1.939274447949527,
+      "grad_norm": 0.44535709840076704,
+      "learning_rate": 1.049116054235602e-05,
+      "loss": 0.3239,
+      "step": 9837
+    },
+    {
+      "epoch": 1.9394716088328074,
+      "grad_norm": 0.473698052381602,
+      "learning_rate": 1.0489612381243299e-05,
+      "loss": 0.3464,
+      "step": 9838
+    },
+    {
+      "epoch": 1.9396687697160884,
+      "grad_norm": 0.4778541195055896,
+      "learning_rate": 1.0488064208367246e-05,
+      "loss": 0.3427,
+      "step": 9839
+    },
+    {
+      "epoch": 1.939865930599369,
+      "grad_norm": 0.46660859837931956,
+      "learning_rate": 1.0486516023765057e-05,
+      "loss": 0.3334,
+      "step": 9840
+    },
+    {
+      "epoch": 1.9400630914826498,
+      "grad_norm": 0.4609030291063781,
+      "learning_rate": 1.0484967827473927e-05,
+      "loss": 0.3233,
+      "step": 9841
+    },
+    {
+      "epoch": 1.9402602523659307,
+      "grad_norm": 0.46978565070370426,
+      "learning_rate": 1.0483419619531057e-05,
+      "loss": 0.3397,
+      "step": 9842
+    },
+    {
+      "epoch": 1.9404574132492114,
+      "grad_norm": 0.446991858909842,
+      "learning_rate": 1.0481871399973638e-05,
+      "loss": 0.3048,
+      "step": 9843
+    },
+    {
+      "epoch": 1.940654574132492,
+      "grad_norm": 0.4538700492840836,
+      "learning_rate": 1.0480323168838876e-05,
+      "loss": 0.3104,
+      "step": 9844
+    },
+    {
+      "epoch": 1.940851735015773,
+      "grad_norm": 0.4799559035841734,
+      "learning_rate": 1.0478774926163957e-05,
+      "loss": 0.3292,
+      "step": 9845
+    },
+    {
+      "epoch": 1.9410488958990535,
+      "grad_norm": 0.49610549652225705,
+      "learning_rate": 1.0477226671986089e-05,
+      "loss": 0.3407,
+      "step": 9846
+    },
+    {
+      "epoch": 1.9412460567823344,
+      "grad_norm": 0.4691565602936203,
+      "learning_rate": 1.0475678406342462e-05,
+      "loss": 0.3249,
+      "step": 9847
+    },
+    {
+      "epoch": 1.9414432176656151,
+      "grad_norm": 0.4548964647773779,
+      "learning_rate": 1.0474130129270281e-05,
+      "loss": 0.3158,
+      "step": 9848
+    },
+    {
+      "epoch": 1.9416403785488958,
+      "grad_norm": 0.46442569328638555,
+      "learning_rate": 1.0472581840806742e-05,
+      "loss": 0.3275,
+      "step": 9849
+    },
+    {
+      "epoch": 1.9418375394321767,
+      "grad_norm": 0.47409430097779603,
+      "learning_rate": 1.0471033540989044e-05,
+      "loss": 0.3487,
+      "step": 9850
+    },
+    {
+      "epoch": 1.9420347003154574,
+      "grad_norm": 0.4714341339076567,
+      "learning_rate": 1.0469485229854383e-05,
+      "loss": 0.3247,
+      "step": 9851
+    },
+    {
+      "epoch": 1.9422318611987381,
+      "grad_norm": 0.4471157743782902,
+      "learning_rate": 1.0467936907439966e-05,
+      "loss": 0.3133,
+      "step": 9852
+    },
+    {
+      "epoch": 1.942429022082019,
+      "grad_norm": 0.5320875040440245,
+      "learning_rate": 1.0466388573782984e-05,
+      "loss": 0.3648,
+      "step": 9853
+    },
+    {
+      "epoch": 1.9426261829652995,
+      "grad_norm": 0.4714349073262386,
+      "learning_rate": 1.0464840228920643e-05,
+      "loss": 0.3159,
+      "step": 9854
+    },
+    {
+      "epoch": 1.9428233438485805,
+      "grad_norm": 0.4907557101702487,
+      "learning_rate": 1.046329187289014e-05,
+      "loss": 0.3466,
+      "step": 9855
+    },
+    {
+      "epoch": 1.9430205047318612,
+      "grad_norm": 0.4838478550153892,
+      "learning_rate": 1.046174350572868e-05,
+      "loss": 0.3141,
+      "step": 9856
+    },
+    {
+      "epoch": 1.9432176656151419,
+      "grad_norm": 0.49718799827485166,
+      "learning_rate": 1.0460195127473456e-05,
+      "loss": 0.3363,
+      "step": 9857
+    },
+    {
+      "epoch": 1.9434148264984228,
+      "grad_norm": 0.4538672203134272,
+      "learning_rate": 1.0458646738161676e-05,
+      "loss": 0.318,
+      "step": 9858
+    },
+    {
+      "epoch": 1.9436119873817035,
+      "grad_norm": 0.4718009026198612,
+      "learning_rate": 1.0457098337830536e-05,
+      "loss": 0.3121,
+      "step": 9859
+    },
+    {
+      "epoch": 1.9438091482649842,
+      "grad_norm": 0.4863092292989189,
+      "learning_rate": 1.0455549926517243e-05,
+      "loss": 0.3575,
+      "step": 9860
+    },
+    {
+      "epoch": 1.944006309148265,
+      "grad_norm": 0.45207391928372037,
+      "learning_rate": 1.0454001504258994e-05,
+      "loss": 0.3208,
+      "step": 9861
+    },
+    {
+      "epoch": 1.9442034700315456,
+      "grad_norm": 0.49540411046222155,
+      "learning_rate": 1.0452453071092993e-05,
+      "loss": 0.3148,
+      "step": 9862
+    },
+    {
+      "epoch": 1.9444006309148265,
+      "grad_norm": 0.4902146796712335,
+      "learning_rate": 1.0450904627056446e-05,
+      "loss": 0.354,
+      "step": 9863
+    },
+    {
+      "epoch": 1.9445977917981072,
+      "grad_norm": 0.4302996025002257,
+      "learning_rate": 1.0449356172186548e-05,
+      "loss": 0.2963,
+      "step": 9864
+    },
+    {
+      "epoch": 1.944794952681388,
+      "grad_norm": 0.47578474259947834,
+      "learning_rate": 1.0447807706520513e-05,
+      "loss": 0.3117,
+      "step": 9865
+    },
+    {
+      "epoch": 1.9449921135646688,
+      "grad_norm": 0.4504023483184343,
+      "learning_rate": 1.0446259230095531e-05,
+      "loss": 0.3401,
+      "step": 9866
+    },
+    {
+      "epoch": 1.9451892744479495,
+      "grad_norm": 0.4757062498349321,
+      "learning_rate": 1.0444710742948814e-05,
+      "loss": 0.3366,
+      "step": 9867
+    },
+    {
+      "epoch": 1.9453864353312302,
+      "grad_norm": 0.4819233484036111,
+      "learning_rate": 1.0443162245117562e-05,
+      "loss": 0.3342,
+      "step": 9868
+    },
+    {
+      "epoch": 1.9455835962145112,
+      "grad_norm": 0.5213874575567724,
+      "learning_rate": 1.044161373663898e-05,
+      "loss": 0.3172,
+      "step": 9869
+    },
+    {
+      "epoch": 1.9457807570977916,
+      "grad_norm": 0.44488837497482253,
+      "learning_rate": 1.0440065217550273e-05,
+      "loss": 0.3125,
+      "step": 9870
+    },
+    {
+      "epoch": 1.9459779179810726,
+      "grad_norm": 0.460970808844026,
+      "learning_rate": 1.0438516687888645e-05,
+      "loss": 0.3258,
+      "step": 9871
+    },
+    {
+      "epoch": 1.9461750788643533,
+      "grad_norm": 0.4810006442638309,
+      "learning_rate": 1.04369681476913e-05,
+      "loss": 0.3219,
+      "step": 9872
+    },
+    {
+      "epoch": 1.946372239747634,
+      "grad_norm": 0.4838450735864227,
+      "learning_rate": 1.0435419596995444e-05,
+      "loss": 0.3655,
+      "step": 9873
+    },
+    {
+      "epoch": 1.9465694006309149,
+      "grad_norm": 0.47505078585700605,
+      "learning_rate": 1.0433871035838283e-05,
+      "loss": 0.3257,
+      "step": 9874
+    },
+    {
+      "epoch": 1.9467665615141956,
+      "grad_norm": 0.47990929946296423,
+      "learning_rate": 1.0432322464257019e-05,
+      "loss": 0.3165,
+      "step": 9875
+    },
+    {
+      "epoch": 1.9469637223974763,
+      "grad_norm": 0.4824430346339747,
+      "learning_rate": 1.0430773882288859e-05,
+      "loss": 0.3493,
+      "step": 9876
+    },
+    {
+      "epoch": 1.9471608832807572,
+      "grad_norm": 0.47127696634444705,
+      "learning_rate": 1.042922528997101e-05,
+      "loss": 0.3265,
+      "step": 9877
+    },
+    {
+      "epoch": 1.947358044164038,
+      "grad_norm": 0.45182006106255146,
+      "learning_rate": 1.0427676687340678e-05,
+      "loss": 0.3219,
+      "step": 9878
+    },
+    {
+      "epoch": 1.9475552050473186,
+      "grad_norm": 0.46337044334398714,
+      "learning_rate": 1.0426128074435068e-05,
+      "loss": 0.3319,
+      "step": 9879
+    },
+    {
+      "epoch": 1.9477523659305995,
+      "grad_norm": 0.4555312238661632,
+      "learning_rate": 1.0424579451291393e-05,
+      "loss": 0.3119,
+      "step": 9880
+    },
+    {
+      "epoch": 1.94794952681388,
+      "grad_norm": 0.5027627443388722,
+      "learning_rate": 1.042303081794685e-05,
+      "loss": 0.3372,
+      "step": 9881
+    },
+    {
+      "epoch": 1.948146687697161,
+      "grad_norm": 0.46830494008681434,
+      "learning_rate": 1.042148217443865e-05,
+      "loss": 0.3491,
+      "step": 9882
+    },
+    {
+      "epoch": 1.9483438485804416,
+      "grad_norm": 0.4738873885085205,
+      "learning_rate": 1.0419933520804002e-05,
+      "loss": 0.3493,
+      "step": 9883
+    },
+    {
+      "epoch": 1.9485410094637223,
+      "grad_norm": 0.4965510696033795,
+      "learning_rate": 1.0418384857080118e-05,
+      "loss": 0.336,
+      "step": 9884
+    },
+    {
+      "epoch": 1.9487381703470033,
+      "grad_norm": 0.4779020902575847,
+      "learning_rate": 1.0416836183304198e-05,
+      "loss": 0.3348,
+      "step": 9885
+    },
+    {
+      "epoch": 1.948935331230284,
+      "grad_norm": 0.4533557633530483,
+      "learning_rate": 1.0415287499513452e-05,
+      "loss": 0.3072,
+      "step": 9886
+    },
+    {
+      "epoch": 1.9491324921135647,
+      "grad_norm": 0.46478354483577006,
+      "learning_rate": 1.0413738805745089e-05,
+      "loss": 0.3269,
+      "step": 9887
+    },
+    {
+      "epoch": 1.9493296529968456,
+      "grad_norm": 0.4553262713575033,
+      "learning_rate": 1.0412190102036317e-05,
+      "loss": 0.329,
+      "step": 9888
+    },
+    {
+      "epoch": 1.949526813880126,
+      "grad_norm": 0.5236120156557359,
+      "learning_rate": 1.041064138842435e-05,
+      "loss": 0.3589,
+      "step": 9889
+    },
+    {
+      "epoch": 1.949723974763407,
+      "grad_norm": 0.46404257823966893,
+      "learning_rate": 1.0409092664946388e-05,
+      "loss": 0.3165,
+      "step": 9890
+    },
+    {
+      "epoch": 1.9499211356466877,
+      "grad_norm": 0.46119544702732324,
+      "learning_rate": 1.040754393163965e-05,
+      "loss": 0.3198,
+      "step": 9891
+    },
+    {
+      "epoch": 1.9501182965299684,
+      "grad_norm": 0.48159542350926215,
+      "learning_rate": 1.0405995188541336e-05,
+      "loss": 0.3373,
+      "step": 9892
+    },
+    {
+      "epoch": 1.9503154574132493,
+      "grad_norm": 0.4397391980222927,
+      "learning_rate": 1.0404446435688665e-05,
+      "loss": 0.3109,
+      "step": 9893
+    },
+    {
+      "epoch": 1.95051261829653,
+      "grad_norm": 0.4668066514467705,
+      "learning_rate": 1.040289767311884e-05,
+      "loss": 0.345,
+      "step": 9894
+    },
+    {
+      "epoch": 1.9507097791798107,
+      "grad_norm": 0.4880251587331396,
+      "learning_rate": 1.0401348900869073e-05,
+      "loss": 0.3402,
+      "step": 9895
+    },
+    {
+      "epoch": 1.9509069400630916,
+      "grad_norm": 0.4909340031989673,
+      "learning_rate": 1.0399800118976577e-05,
+      "loss": 0.3561,
+      "step": 9896
+    },
+    {
+      "epoch": 1.951104100946372,
+      "grad_norm": 0.46507313249997145,
+      "learning_rate": 1.0398251327478561e-05,
+      "loss": 0.3226,
+      "step": 9897
+    },
+    {
+      "epoch": 1.951301261829653,
+      "grad_norm": 0.45585618774916764,
+      "learning_rate": 1.0396702526412237e-05,
+      "loss": 0.2961,
+      "step": 9898
+    },
+    {
+      "epoch": 1.9514984227129337,
+      "grad_norm": 0.5000463843480156,
+      "learning_rate": 1.0395153715814816e-05,
+      "loss": 0.3413,
+      "step": 9899
+    },
+    {
+      "epoch": 1.9516955835962144,
+      "grad_norm": 0.45225941628046795,
+      "learning_rate": 1.0393604895723509e-05,
+      "loss": 0.3243,
+      "step": 9900
+    },
+    {
+      "epoch": 1.9518927444794953,
+      "grad_norm": 0.4853464256835092,
+      "learning_rate": 1.0392056066175524e-05,
+      "loss": 0.3294,
+      "step": 9901
+    },
+    {
+      "epoch": 1.952089905362776,
+      "grad_norm": 0.48505362928524964,
+      "learning_rate": 1.039050722720808e-05,
+      "loss": 0.3592,
+      "step": 9902
+    },
+    {
+      "epoch": 1.9522870662460567,
+      "grad_norm": 0.48062154733831464,
+      "learning_rate": 1.0388958378858383e-05,
+      "loss": 0.3211,
+      "step": 9903
+    },
+    {
+      "epoch": 1.9524842271293377,
+      "grad_norm": 0.5171535755363212,
+      "learning_rate": 1.038740952116365e-05,
+      "loss": 0.3546,
+      "step": 9904
+    },
+    {
+      "epoch": 1.9526813880126181,
+      "grad_norm": 0.46818506771167984,
+      "learning_rate": 1.0385860654161088e-05,
+      "loss": 0.3284,
+      "step": 9905
+    },
+    {
+      "epoch": 1.952878548895899,
+      "grad_norm": 0.4915580603272551,
+      "learning_rate": 1.0384311777887916e-05,
+      "loss": 0.3658,
+      "step": 9906
+    },
+    {
+      "epoch": 1.9530757097791798,
+      "grad_norm": 0.46935304552685453,
+      "learning_rate": 1.0382762892381342e-05,
+      "loss": 0.3332,
+      "step": 9907
+    },
+    {
+      "epoch": 1.9532728706624605,
+      "grad_norm": 0.45363492810935857,
+      "learning_rate": 1.0381213997678582e-05,
+      "loss": 0.3388,
+      "step": 9908
+    },
+    {
+      "epoch": 1.9534700315457414,
+      "grad_norm": 0.44641220154213385,
+      "learning_rate": 1.0379665093816848e-05,
+      "loss": 0.3035,
+      "step": 9909
+    },
+    {
+      "epoch": 1.953667192429022,
+      "grad_norm": 0.45206940818234764,
+      "learning_rate": 1.0378116180833357e-05,
+      "loss": 0.3222,
+      "step": 9910
+    },
+    {
+      "epoch": 1.9538643533123028,
+      "grad_norm": 0.4575547617482853,
+      "learning_rate": 1.0376567258765316e-05,
+      "loss": 0.3193,
+      "step": 9911
+    },
+    {
+      "epoch": 1.9540615141955837,
+      "grad_norm": 0.4590418732960005,
+      "learning_rate": 1.0375018327649948e-05,
+      "loss": 0.3023,
+      "step": 9912
+    },
+    {
+      "epoch": 1.9542586750788642,
+      "grad_norm": 0.4543796513395991,
+      "learning_rate": 1.037346938752446e-05,
+      "loss": 0.2872,
+      "step": 9913
+    },
+    {
+      "epoch": 1.9544558359621451,
+      "grad_norm": 0.4805196349855143,
+      "learning_rate": 1.037192043842607e-05,
+      "loss": 0.325,
+      "step": 9914
+    },
+    {
+      "epoch": 1.9546529968454258,
+      "grad_norm": 0.46446662799687205,
+      "learning_rate": 1.037037148039199e-05,
+      "loss": 0.3449,
+      "step": 9915
+    },
+    {
+      "epoch": 1.9548501577287065,
+      "grad_norm": 0.7024562863012095,
+      "learning_rate": 1.036882251345944e-05,
+      "loss": 0.3346,
+      "step": 9916
+    },
+    {
+      "epoch": 1.9550473186119874,
+      "grad_norm": 0.4792762668817157,
+      "learning_rate": 1.036727353766563e-05,
+      "loss": 0.3535,
+      "step": 9917
+    },
+    {
+      "epoch": 1.9552444794952681,
+      "grad_norm": 0.46573884017043454,
+      "learning_rate": 1.0365724553047778e-05,
+      "loss": 0.3114,
+      "step": 9918
+    },
+    {
+      "epoch": 1.9554416403785488,
+      "grad_norm": 0.45556820668060394,
+      "learning_rate": 1.03641755596431e-05,
+      "loss": 0.3101,
+      "step": 9919
+    },
+    {
+      "epoch": 1.9556388012618298,
+      "grad_norm": 0.5061371270375018,
+      "learning_rate": 1.0362626557488811e-05,
+      "loss": 0.3451,
+      "step": 9920
+    },
+    {
+      "epoch": 1.9558359621451105,
+      "grad_norm": 0.45544222128631295,
+      "learning_rate": 1.0361077546622125e-05,
+      "loss": 0.3414,
+      "step": 9921
+    },
+    {
+      "epoch": 1.9560331230283912,
+      "grad_norm": 0.49708322112827474,
+      "learning_rate": 1.0359528527080263e-05,
+      "loss": 0.3501,
+      "step": 9922
+    },
+    {
+      "epoch": 1.956230283911672,
+      "grad_norm": 0.4777705719722766,
+      "learning_rate": 1.0357979498900436e-05,
+      "loss": 0.3216,
+      "step": 9923
+    },
+    {
+      "epoch": 1.9564274447949526,
+      "grad_norm": 0.4774834671431984,
+      "learning_rate": 1.0356430462119865e-05,
+      "loss": 0.3045,
+      "step": 9924
+    },
+    {
+      "epoch": 1.9566246056782335,
+      "grad_norm": 0.47057412024018935,
+      "learning_rate": 1.0354881416775765e-05,
+      "loss": 0.3218,
+      "step": 9925
+    },
+    {
+      "epoch": 1.9568217665615142,
+      "grad_norm": 0.4991382437521349,
+      "learning_rate": 1.0353332362905351e-05,
+      "loss": 0.3485,
+      "step": 9926
+    },
+    {
+      "epoch": 1.9570189274447949,
+      "grad_norm": 0.4561117034939663,
+      "learning_rate": 1.0351783300545843e-05,
+      "loss": 0.333,
+      "step": 9927
+    },
+    {
+      "epoch": 1.9572160883280758,
+      "grad_norm": 0.47236910408393734,
+      "learning_rate": 1.0350234229734459e-05,
+      "loss": 0.341,
+      "step": 9928
+    },
+    {
+      "epoch": 1.9574132492113565,
+      "grad_norm": 0.46739069562356134,
+      "learning_rate": 1.0348685150508417e-05,
+      "loss": 0.316,
+      "step": 9929
+    },
+    {
+      "epoch": 1.9576104100946372,
+      "grad_norm": 0.4726019967619267,
+      "learning_rate": 1.034713606290493e-05,
+      "loss": 0.3164,
+      "step": 9930
+    },
+    {
+      "epoch": 1.9578075709779181,
+      "grad_norm": 0.46230340086353033,
+      "learning_rate": 1.0345586966961223e-05,
+      "loss": 0.3286,
+      "step": 9931
+    },
+    {
+      "epoch": 1.9580047318611986,
+      "grad_norm": 0.4729940247692541,
+      "learning_rate": 1.0344037862714506e-05,
+      "loss": 0.3364,
+      "step": 9932
+    },
+    {
+      "epoch": 1.9582018927444795,
+      "grad_norm": 0.49856429925552365,
+      "learning_rate": 1.034248875020201e-05,
+      "loss": 0.3196,
+      "step": 9933
+    },
+    {
+      "epoch": 1.9583990536277602,
+      "grad_norm": 0.4701007265414603,
+      "learning_rate": 1.0340939629460938e-05,
+      "loss": 0.3181,
+      "step": 9934
+    },
+    {
+      "epoch": 1.958596214511041,
+      "grad_norm": 0.4658639064265051,
+      "learning_rate": 1.0339390500528523e-05,
+      "loss": 0.3145,
+      "step": 9935
+    },
+    {
+      "epoch": 1.9587933753943219,
+      "grad_norm": 0.49004170327794266,
+      "learning_rate": 1.0337841363441973e-05,
+      "loss": 0.336,
+      "step": 9936
+    },
+    {
+      "epoch": 1.9589905362776026,
+      "grad_norm": 0.44921738293967783,
+      "learning_rate": 1.0336292218238514e-05,
+      "loss": 0.3133,
+      "step": 9937
+    },
+    {
+      "epoch": 1.9591876971608833,
+      "grad_norm": 0.4823452316355001,
+      "learning_rate": 1.0334743064955367e-05,
+      "loss": 0.3471,
+      "step": 9938
+    },
+    {
+      "epoch": 1.9593848580441642,
+      "grad_norm": 0.4507059368884933,
+      "learning_rate": 1.0333193903629743e-05,
+      "loss": 0.295,
+      "step": 9939
+    },
+    {
+      "epoch": 1.9595820189274447,
+      "grad_norm": 0.46683494943121584,
+      "learning_rate": 1.0331644734298874e-05,
+      "loss": 0.3325,
+      "step": 9940
+    },
+    {
+      "epoch": 1.9597791798107256,
+      "grad_norm": 0.4744162384642641,
+      "learning_rate": 1.0330095556999966e-05,
+      "loss": 0.32,
+      "step": 9941
+    },
+    {
+      "epoch": 1.9599763406940063,
+      "grad_norm": 0.4617197712445625,
+      "learning_rate": 1.0328546371770249e-05,
+      "loss": 0.3618,
+      "step": 9942
+    },
+    {
+      "epoch": 1.960173501577287,
+      "grad_norm": 0.5244555576526214,
+      "learning_rate": 1.0326997178646941e-05,
+      "loss": 0.3248,
+      "step": 9943
+    },
+    {
+      "epoch": 1.960370662460568,
+      "grad_norm": 0.4799968339049716,
+      "learning_rate": 1.0325447977667262e-05,
+      "loss": 0.3388,
+      "step": 9944
+    },
+    {
+      "epoch": 1.9605678233438486,
+      "grad_norm": 0.44066352020859745,
+      "learning_rate": 1.0323898768868434e-05,
+      "loss": 0.3199,
+      "step": 9945
+    },
+    {
+      "epoch": 1.9607649842271293,
+      "grad_norm": 0.5245090776857998,
+      "learning_rate": 1.0322349552287676e-05,
+      "loss": 0.3169,
+      "step": 9946
+    },
+    {
+      "epoch": 1.9609621451104102,
+      "grad_norm": 0.4716152433206715,
+      "learning_rate": 1.0320800327962212e-05,
+      "loss": 0.3217,
+      "step": 9947
+    },
+    {
+      "epoch": 1.9611593059936907,
+      "grad_norm": 0.4712223481043858,
+      "learning_rate": 1.0319251095929262e-05,
+      "loss": 0.3397,
+      "step": 9948
+    },
+    {
+      "epoch": 1.9613564668769716,
+      "grad_norm": 0.43783005573851125,
+      "learning_rate": 1.0317701856226045e-05,
+      "loss": 0.3187,
+      "step": 9949
+    },
+    {
+      "epoch": 1.9615536277602523,
+      "grad_norm": 0.4490127963028393,
+      "learning_rate": 1.0316152608889787e-05,
+      "loss": 0.3312,
+      "step": 9950
+    },
+    {
+      "epoch": 1.961750788643533,
+      "grad_norm": 0.47018160199287945,
+      "learning_rate": 1.0314603353957709e-05,
+      "loss": 0.3378,
+      "step": 9951
+    },
+    {
+      "epoch": 1.961947949526814,
+      "grad_norm": 0.45982647728871295,
+      "learning_rate": 1.031305409146703e-05,
+      "loss": 0.3053,
+      "step": 9952
+    },
+    {
+      "epoch": 1.9621451104100947,
+      "grad_norm": 0.47793598964818645,
+      "learning_rate": 1.0311504821454973e-05,
+      "loss": 0.3528,
+      "step": 9953
+    },
+    {
+      "epoch": 1.9623422712933754,
+      "grad_norm": 0.5116582725353143,
+      "learning_rate": 1.0309955543958765e-05,
+      "loss": 0.3343,
+      "step": 9954
+    },
+    {
+      "epoch": 1.9625394321766563,
+      "grad_norm": 0.4623753000651675,
+      "learning_rate": 1.0308406259015624e-05,
+      "loss": 0.345,
+      "step": 9955
+    },
+    {
+      "epoch": 1.9627365930599368,
+      "grad_norm": 0.45920787227072735,
+      "learning_rate": 1.0306856966662776e-05,
+      "loss": 0.3179,
+      "step": 9956
+    },
+    {
+      "epoch": 1.9629337539432177,
+      "grad_norm": 0.48812471718237876,
+      "learning_rate": 1.0305307666937441e-05,
+      "loss": 0.3162,
+      "step": 9957
+    },
+    {
+      "epoch": 1.9631309148264984,
+      "grad_norm": 0.4525568703068494,
+      "learning_rate": 1.0303758359876841e-05,
+      "loss": 0.3335,
+      "step": 9958
+    },
+    {
+      "epoch": 1.963328075709779,
+      "grad_norm": 0.4797900640643521,
+      "learning_rate": 1.0302209045518206e-05,
+      "loss": 0.3557,
+      "step": 9959
+    },
+    {
+      "epoch": 1.96352523659306,
+      "grad_norm": 0.46653885535240164,
+      "learning_rate": 1.0300659723898752e-05,
+      "loss": 0.3313,
+      "step": 9960
+    },
+    {
+      "epoch": 1.9637223974763407,
+      "grad_norm": 0.5002009677410206,
+      "learning_rate": 1.029911039505571e-05,
+      "loss": 0.3303,
+      "step": 9961
+    },
+    {
+      "epoch": 1.9639195583596214,
+      "grad_norm": 0.4853690102944126,
+      "learning_rate": 1.0297561059026293e-05,
+      "loss": 0.3165,
+      "step": 9962
+    },
+    {
+      "epoch": 1.9641167192429023,
+      "grad_norm": 0.4855838797298784,
+      "learning_rate": 1.0296011715847738e-05,
+      "loss": 0.3383,
+      "step": 9963
+    },
+    {
+      "epoch": 1.964313880126183,
+      "grad_norm": 0.48686860386638087,
+      "learning_rate": 1.029446236555726e-05,
+      "loss": 0.3265,
+      "step": 9964
+    },
+    {
+      "epoch": 1.9645110410094637,
+      "grad_norm": 0.519576709109696,
+      "learning_rate": 1.0292913008192088e-05,
+      "loss": 0.3623,
+      "step": 9965
+    },
+    {
+      "epoch": 1.9647082018927446,
+      "grad_norm": 0.5048904976816195,
+      "learning_rate": 1.0291363643789445e-05,
+      "loss": 0.3572,
+      "step": 9966
+    },
+    {
+      "epoch": 1.9649053627760251,
+      "grad_norm": 0.42587004814801377,
+      "learning_rate": 1.0289814272386556e-05,
+      "loss": 0.292,
+      "step": 9967
+    },
+    {
+      "epoch": 1.965102523659306,
+      "grad_norm": 0.500466873158178,
+      "learning_rate": 1.0288264894020646e-05,
+      "loss": 0.333,
+      "step": 9968
+    },
+    {
+      "epoch": 1.9652996845425867,
+      "grad_norm": 0.465182567762671,
+      "learning_rate": 1.0286715508728937e-05,
+      "loss": 0.3157,
+      "step": 9969
+    },
+    {
+      "epoch": 1.9654968454258674,
+      "grad_norm": 0.4617031591725659,
+      "learning_rate": 1.0285166116548662e-05,
+      "loss": 0.3202,
+      "step": 9970
+    },
+    {
+      "epoch": 1.9656940063091484,
+      "grad_norm": 0.47956736857907367,
+      "learning_rate": 1.0283616717517037e-05,
+      "loss": 0.3549,
+      "step": 9971
+    },
+    {
+      "epoch": 1.965891167192429,
+      "grad_norm": 0.45146021950685644,
+      "learning_rate": 1.0282067311671293e-05,
+      "loss": 0.3241,
+      "step": 9972
+    },
+    {
+      "epoch": 1.9660883280757098,
+      "grad_norm": 0.4674887113548766,
+      "learning_rate": 1.0280517899048657e-05,
+      "loss": 0.3395,
+      "step": 9973
+    },
+    {
+      "epoch": 1.9662854889589907,
+      "grad_norm": 0.47145112139932266,
+      "learning_rate": 1.027896847968635e-05,
+      "loss": 0.3287,
+      "step": 9974
+    },
+    {
+      "epoch": 1.9664826498422712,
+      "grad_norm": 0.4565740445948791,
+      "learning_rate": 1.0277419053621602e-05,
+      "loss": 0.3196,
+      "step": 9975
+    },
+    {
+      "epoch": 1.966679810725552,
+      "grad_norm": 0.48138321162232844,
+      "learning_rate": 1.0275869620891637e-05,
+      "loss": 0.3403,
+      "step": 9976
+    },
+    {
+      "epoch": 1.9668769716088328,
+      "grad_norm": 0.484975445459746,
+      "learning_rate": 1.0274320181533681e-05,
+      "loss": 0.3283,
+      "step": 9977
+    },
+    {
+      "epoch": 1.9670741324921135,
+      "grad_norm": 0.4387256747272943,
+      "learning_rate": 1.0272770735584966e-05,
+      "loss": 0.3262,
+      "step": 9978
+    },
+    {
+      "epoch": 1.9672712933753944,
+      "grad_norm": 0.471596315685171,
+      "learning_rate": 1.0271221283082709e-05,
+      "loss": 0.3276,
+      "step": 9979
+    },
+    {
+      "epoch": 1.9674684542586751,
+      "grad_norm": 0.4545253122593284,
+      "learning_rate": 1.0269671824064146e-05,
+      "loss": 0.3234,
+      "step": 9980
+    },
+    {
+      "epoch": 1.9676656151419558,
+      "grad_norm": 0.45641111253411415,
+      "learning_rate": 1.0268122358566496e-05,
+      "loss": 0.3175,
+      "step": 9981
+    },
+    {
+      "epoch": 1.9678627760252367,
+      "grad_norm": 0.44841650898946805,
+      "learning_rate": 1.0266572886626997e-05,
+      "loss": 0.2985,
+      "step": 9982
+    },
+    {
+      "epoch": 1.9680599369085172,
+      "grad_norm": 0.43192291651712367,
+      "learning_rate": 1.0265023408282866e-05,
+      "loss": 0.2932,
+      "step": 9983
+    },
+    {
+      "epoch": 1.9682570977917981,
+      "grad_norm": 0.5012210582692301,
+      "learning_rate": 1.0263473923571334e-05,
+      "loss": 0.329,
+      "step": 9984
+    },
+    {
+      "epoch": 1.9684542586750788,
+      "grad_norm": 1.1443330318609166,
+      "learning_rate": 1.0261924432529629e-05,
+      "loss": 0.3461,
+      "step": 9985
+    },
+    {
+      "epoch": 1.9686514195583595,
+      "grad_norm": 0.5004253208337783,
+      "learning_rate": 1.0260374935194979e-05,
+      "loss": 0.3069,
+      "step": 9986
+    },
+    {
+      "epoch": 1.9688485804416405,
+      "grad_norm": 0.4696360012631041,
+      "learning_rate": 1.025882543160461e-05,
+      "loss": 0.3157,
+      "step": 9987
+    },
+    {
+      "epoch": 1.9690457413249212,
+      "grad_norm": 0.4826746914030759,
+      "learning_rate": 1.0257275921795756e-05,
+      "loss": 0.326,
+      "step": 9988
+    },
+    {
+      "epoch": 1.9692429022082019,
+      "grad_norm": 1.3597036785857965,
+      "learning_rate": 1.0255726405805637e-05,
+      "loss": 0.3169,
+      "step": 9989
+    },
+    {
+      "epoch": 1.9694400630914828,
+      "grad_norm": 0.4967882690519761,
+      "learning_rate": 1.0254176883671485e-05,
+      "loss": 0.335,
+      "step": 9990
+    },
+    {
+      "epoch": 1.9696372239747633,
+      "grad_norm": 0.502620609744957,
+      "learning_rate": 1.0252627355430532e-05,
+      "loss": 0.3477,
+      "step": 9991
+    },
+    {
+      "epoch": 1.9698343848580442,
+      "grad_norm": 0.4484287142620489,
+      "learning_rate": 1.0251077821119998e-05,
+      "loss": 0.3114,
+      "step": 9992
+    },
+    {
+      "epoch": 1.9700315457413249,
+      "grad_norm": 0.4747578081774963,
+      "learning_rate": 1.0249528280777121e-05,
+      "loss": 0.3072,
+      "step": 9993
+    },
+    {
+      "epoch": 1.9702287066246056,
+      "grad_norm": 0.4973932701653613,
+      "learning_rate": 1.0247978734439127e-05,
+      "loss": 0.3364,
+      "step": 9994
+    },
+    {
+      "epoch": 1.9704258675078865,
+      "grad_norm": 0.5356785346512383,
+      "learning_rate": 1.0246429182143241e-05,
+      "loss": 0.325,
+      "step": 9995
+    },
+    {
+      "epoch": 1.9706230283911672,
+      "grad_norm": 0.46472104839260125,
+      "learning_rate": 1.0244879623926698e-05,
+      "loss": 0.3337,
+      "step": 9996
+    },
+    {
+      "epoch": 1.970820189274448,
+      "grad_norm": 0.47337341572225305,
+      "learning_rate": 1.0243330059826724e-05,
+      "loss": 0.3333,
+      "step": 9997
+    },
+    {
+      "epoch": 1.9710173501577288,
+      "grad_norm": 0.48176261552874655,
+      "learning_rate": 1.0241780489880546e-05,
+      "loss": 0.338,
+      "step": 9998
+    },
+    {
+      "epoch": 1.9712145110410093,
+      "grad_norm": 0.4816284206757202,
+      "learning_rate": 1.0240230914125401e-05,
+      "loss": 0.3312,
+      "step": 9999
+    },
+    {
+      "epoch": 1.9714116719242902,
+      "grad_norm": 0.4632673744732919,
+      "learning_rate": 1.0238681332598512e-05,
+      "loss": 0.3249,
+      "step": 10000
+    },
+    {
+      "epoch": 1.971608832807571,
+      "grad_norm": 0.48018702662774365,
+      "learning_rate": 1.0237131745337117e-05,
+      "loss": 0.3549,
+      "step": 10001
+    },
+    {
+      "epoch": 1.9718059936908516,
+      "grad_norm": 0.468079428043231,
+      "learning_rate": 1.0235582152378435e-05,
+      "loss": 0.3089,
+      "step": 10002
+    },
+    {
+      "epoch": 1.9720031545741326,
+      "grad_norm": 0.4966800301788038,
+      "learning_rate": 1.0234032553759707e-05,
+      "loss": 0.3486,
+      "step": 10003
+    },
+    {
+      "epoch": 1.9722003154574133,
+      "grad_norm": 0.44884608245777396,
+      "learning_rate": 1.0232482949518157e-05,
+      "loss": 0.2966,
+      "step": 10004
+    },
+    {
+      "epoch": 1.972397476340694,
+      "grad_norm": 8.64090741552038,
+      "learning_rate": 1.0230933339691014e-05,
+      "loss": 0.3888,
+      "step": 10005
+    },
+    {
+      "epoch": 1.9725946372239749,
+      "grad_norm": 0.48715703305183966,
+      "learning_rate": 1.0229383724315516e-05,
+      "loss": 0.3216,
+      "step": 10006
+    },
+    {
+      "epoch": 1.9727917981072554,
+      "grad_norm": 0.44760735526085277,
+      "learning_rate": 1.0227834103428884e-05,
+      "loss": 0.3165,
+      "step": 10007
+    },
+    {
+      "epoch": 1.9729889589905363,
+      "grad_norm": 0.46751672698359104,
+      "learning_rate": 1.022628447706836e-05,
+      "loss": 0.3247,
+      "step": 10008
+    },
+    {
+      "epoch": 1.973186119873817,
+      "grad_norm": 0.5038928323160687,
+      "learning_rate": 1.0224734845271163e-05,
+      "loss": 0.3713,
+      "step": 10009
+    },
+    {
+      "epoch": 1.9733832807570977,
+      "grad_norm": 0.7157608221471825,
+      "learning_rate": 1.0223185208074538e-05,
+      "loss": 0.3501,
+      "step": 10010
+    },
+    {
+      "epoch": 1.9735804416403786,
+      "grad_norm": 0.46267508667900453,
+      "learning_rate": 1.0221635565515699e-05,
+      "loss": 0.3268,
+      "step": 10011
+    },
+    {
+      "epoch": 1.9737776025236593,
+      "grad_norm": 0.47469784854807734,
+      "learning_rate": 1.0220085917631894e-05,
+      "loss": 0.3323,
+      "step": 10012
+    },
+    {
+      "epoch": 1.97397476340694,
+      "grad_norm": 0.4807980595553942,
+      "learning_rate": 1.0218536264460346e-05,
+      "loss": 0.3215,
+      "step": 10013
+    },
+    {
+      "epoch": 1.974171924290221,
+      "grad_norm": 0.4659523222268142,
+      "learning_rate": 1.0216986606038288e-05,
+      "loss": 0.3288,
+      "step": 10014
+    },
+    {
+      "epoch": 1.9743690851735016,
+      "grad_norm": 0.46518600281925643,
+      "learning_rate": 1.0215436942402952e-05,
+      "loss": 0.3337,
+      "step": 10015
+    },
+    {
+      "epoch": 1.9745662460567823,
+      "grad_norm": 0.49142919292843934,
+      "learning_rate": 1.0213887273591573e-05,
+      "loss": 0.3318,
+      "step": 10016
+    },
+    {
+      "epoch": 1.9747634069400632,
+      "grad_norm": 0.46949418334164145,
+      "learning_rate": 1.0212337599641376e-05,
+      "loss": 0.3245,
+      "step": 10017
+    },
+    {
+      "epoch": 1.9749605678233437,
+      "grad_norm": 0.4508227196884441,
+      "learning_rate": 1.0210787920589598e-05,
+      "loss": 0.3187,
+      "step": 10018
+    },
+    {
+      "epoch": 1.9751577287066246,
+      "grad_norm": 0.48656630600989037,
+      "learning_rate": 1.0209238236473472e-05,
+      "loss": 0.3362,
+      "step": 10019
+    },
+    {
+      "epoch": 1.9753548895899053,
+      "grad_norm": 0.4759349272246813,
+      "learning_rate": 1.0207688547330225e-05,
+      "loss": 0.3247,
+      "step": 10020
+    },
+    {
+      "epoch": 1.975552050473186,
+      "grad_norm": 0.5742771214058191,
+      "learning_rate": 1.0206138853197098e-05,
+      "loss": 0.3278,
+      "step": 10021
+    },
+    {
+      "epoch": 1.975749211356467,
+      "grad_norm": 0.5031639152587618,
+      "learning_rate": 1.0204589154111318e-05,
+      "loss": 0.3604,
+      "step": 10022
+    },
+    {
+      "epoch": 1.9759463722397477,
+      "grad_norm": 0.48247181846875936,
+      "learning_rate": 1.0203039450110117e-05,
+      "loss": 0.3441,
+      "step": 10023
+    },
+    {
+      "epoch": 1.9761435331230284,
+      "grad_norm": 0.43458711087785906,
+      "learning_rate": 1.020148974123073e-05,
+      "loss": 0.2844,
+      "step": 10024
+    },
+    {
+      "epoch": 1.9763406940063093,
+      "grad_norm": 0.47496966219956804,
+      "learning_rate": 1.0199940027510392e-05,
+      "loss": 0.3473,
+      "step": 10025
+    },
+    {
+      "epoch": 1.9765378548895898,
+      "grad_norm": 0.5215472264487935,
+      "learning_rate": 1.0198390308986328e-05,
+      "loss": 0.3459,
+      "step": 10026
+    },
+    {
+      "epoch": 1.9767350157728707,
+      "grad_norm": 0.44281471612530954,
+      "learning_rate": 1.0196840585695785e-05,
+      "loss": 0.3057,
+      "step": 10027
+    },
+    {
+      "epoch": 1.9769321766561514,
+      "grad_norm": 0.4818228906154598,
+      "learning_rate": 1.0195290857675982e-05,
+      "loss": 0.3293,
+      "step": 10028
+    },
+    {
+      "epoch": 1.977129337539432,
+      "grad_norm": 0.460334518749255,
+      "learning_rate": 1.0193741124964164e-05,
+      "loss": 0.3286,
+      "step": 10029
+    },
+    {
+      "epoch": 1.977326498422713,
+      "grad_norm": 0.46235057626123255,
+      "learning_rate": 1.0192191387597554e-05,
+      "loss": 0.3202,
+      "step": 10030
+    },
+    {
+      "epoch": 1.9775236593059937,
+      "grad_norm": 0.44454174849526384,
+      "learning_rate": 1.0190641645613397e-05,
+      "loss": 0.3184,
+      "step": 10031
+    },
+    {
+      "epoch": 1.9777208201892744,
+      "grad_norm": 0.46226996398858433,
+      "learning_rate": 1.0189091899048914e-05,
+      "loss": 0.3438,
+      "step": 10032
+    },
+    {
+      "epoch": 1.9779179810725553,
+      "grad_norm": 0.5082875526155569,
+      "learning_rate": 1.0187542147941352e-05,
+      "loss": 0.3517,
+      "step": 10033
+    },
+    {
+      "epoch": 1.9781151419558358,
+      "grad_norm": 0.43241743937186994,
+      "learning_rate": 1.0185992392327936e-05,
+      "loss": 0.3188,
+      "step": 10034
+    },
+    {
+      "epoch": 1.9783123028391167,
+      "grad_norm": 0.4476815196740319,
+      "learning_rate": 1.0184442632245905e-05,
+      "loss": 0.3252,
+      "step": 10035
+    },
+    {
+      "epoch": 1.9785094637223974,
+      "grad_norm": 0.45862822746483506,
+      "learning_rate": 1.018289286773249e-05,
+      "loss": 0.325,
+      "step": 10036
+    },
+    {
+      "epoch": 1.9787066246056781,
+      "grad_norm": 0.4483776581846178,
+      "learning_rate": 1.0181343098824928e-05,
+      "loss": 0.3299,
+      "step": 10037
+    },
+    {
+      "epoch": 1.978903785488959,
+      "grad_norm": 0.46878007068368455,
+      "learning_rate": 1.017979332556045e-05,
+      "loss": 0.3291,
+      "step": 10038
+    },
+    {
+      "epoch": 1.9791009463722398,
+      "grad_norm": 0.4916705031341869,
+      "learning_rate": 1.0178243547976293e-05,
+      "loss": 0.3253,
+      "step": 10039
+    },
+    {
+      "epoch": 1.9792981072555205,
+      "grad_norm": 0.5025149474174093,
+      "learning_rate": 1.017669376610969e-05,
+      "loss": 0.3469,
+      "step": 10040
+    },
+    {
+      "epoch": 1.9794952681388014,
+      "grad_norm": 0.4592700151321055,
+      "learning_rate": 1.0175143979997878e-05,
+      "loss": 0.3341,
+      "step": 10041
+    },
+    {
+      "epoch": 1.9796924290220819,
+      "grad_norm": 0.4918000568041209,
+      "learning_rate": 1.0173594189678093e-05,
+      "loss": 0.3598,
+      "step": 10042
+    },
+    {
+      "epoch": 1.9798895899053628,
+      "grad_norm": 0.4479364044121577,
+      "learning_rate": 1.0172044395187566e-05,
+      "loss": 0.3027,
+      "step": 10043
+    },
+    {
+      "epoch": 1.9800867507886435,
+      "grad_norm": 0.47799619307194813,
+      "learning_rate": 1.0170494596563533e-05,
+      "loss": 0.3433,
+      "step": 10044
+    },
+    {
+      "epoch": 1.9802839116719242,
+      "grad_norm": 0.4523062427056764,
+      "learning_rate": 1.016894479384323e-05,
+      "loss": 0.305,
+      "step": 10045
+    },
+    {
+      "epoch": 1.9804810725552051,
+      "grad_norm": 0.4759998481341186,
+      "learning_rate": 1.0167394987063894e-05,
+      "loss": 0.3446,
+      "step": 10046
+    },
+    {
+      "epoch": 1.9806782334384858,
+      "grad_norm": 0.4869635858753714,
+      "learning_rate": 1.0165845176262757e-05,
+      "loss": 0.3478,
+      "step": 10047
+    },
+    {
+      "epoch": 1.9808753943217665,
+      "grad_norm": 0.45645594907618603,
+      "learning_rate": 1.016429536147706e-05,
+      "loss": 0.3201,
+      "step": 10048
+    },
+    {
+      "epoch": 1.9810725552050474,
+      "grad_norm": 0.46823112725914234,
+      "learning_rate": 1.0162745542744028e-05,
+      "loss": 0.305,
+      "step": 10049
+    },
+    {
+      "epoch": 1.981269716088328,
+      "grad_norm": 0.46841106024358004,
+      "learning_rate": 1.016119572010091e-05,
+      "loss": 0.3323,
+      "step": 10050
+    },
+    {
+      "epoch": 1.9814668769716088,
+      "grad_norm": 0.44805736427909854,
+      "learning_rate": 1.015964589358493e-05,
+      "loss": 0.3241,
+      "step": 10051
+    },
+    {
+      "epoch": 1.9816640378548895,
+      "grad_norm": 0.46092527228178226,
+      "learning_rate": 1.015809606323333e-05,
+      "loss": 0.325,
+      "step": 10052
+    },
+    {
+      "epoch": 1.9818611987381702,
+      "grad_norm": 0.4813889305251375,
+      "learning_rate": 1.0156546229083346e-05,
+      "loss": 0.3346,
+      "step": 10053
+    },
+    {
+      "epoch": 1.9820583596214512,
+      "grad_norm": 0.48424701481984905,
+      "learning_rate": 1.0154996391172211e-05,
+      "loss": 0.3176,
+      "step": 10054
+    },
+    {
+      "epoch": 1.9822555205047319,
+      "grad_norm": 0.5052195601955368,
+      "learning_rate": 1.0153446549537164e-05,
+      "loss": 0.3738,
+      "step": 10055
+    },
+    {
+      "epoch": 1.9824526813880126,
+      "grad_norm": 0.5264958898669171,
+      "learning_rate": 1.0151896704215441e-05,
+      "loss": 0.3509,
+      "step": 10056
+    },
+    {
+      "epoch": 1.9826498422712935,
+      "grad_norm": 0.4424729969934259,
+      "learning_rate": 1.015034685524428e-05,
+      "loss": 0.3016,
+      "step": 10057
+    },
+    {
+      "epoch": 1.9828470031545742,
+      "grad_norm": 0.45230293654878495,
+      "learning_rate": 1.0148797002660909e-05,
+      "loss": 0.3415,
+      "step": 10058
+    },
+    {
+      "epoch": 1.9830441640378549,
+      "grad_norm": 0.4296621647967024,
+      "learning_rate": 1.0147247146502573e-05,
+      "loss": 0.3229,
+      "step": 10059
+    },
+    {
+      "epoch": 1.9832413249211358,
+      "grad_norm": 0.4736201795007952,
+      "learning_rate": 1.0145697286806505e-05,
+      "loss": 0.3193,
+      "step": 10060
+    },
+    {
+      "epoch": 1.9834384858044163,
+      "grad_norm": 0.4483471743167527,
+      "learning_rate": 1.0144147423609942e-05,
+      "loss": 0.3078,
+      "step": 10061
+    },
+    {
+      "epoch": 1.9836356466876972,
+      "grad_norm": 0.4795452273103116,
+      "learning_rate": 1.0142597556950123e-05,
+      "loss": 0.314,
+      "step": 10062
+    },
+    {
+      "epoch": 1.983832807570978,
+      "grad_norm": 0.4651739245868516,
+      "learning_rate": 1.014104768686428e-05,
+      "loss": 0.3245,
+      "step": 10063
+    },
+    {
+      "epoch": 1.9840299684542586,
+      "grad_norm": 0.4678656723303721,
+      "learning_rate": 1.0139497813389654e-05,
+      "loss": 0.3262,
+      "step": 10064
+    },
+    {
+      "epoch": 1.9842271293375395,
+      "grad_norm": 0.4663884375595269,
+      "learning_rate": 1.0137947936563481e-05,
+      "loss": 0.3314,
+      "step": 10065
+    },
+    {
+      "epoch": 1.9844242902208202,
+      "grad_norm": 0.4991990370561572,
+      "learning_rate": 1.0136398056422995e-05,
+      "loss": 0.3662,
+      "step": 10066
+    },
+    {
+      "epoch": 1.984621451104101,
+      "grad_norm": 0.4630237349751292,
+      "learning_rate": 1.013484817300544e-05,
+      "loss": 0.3131,
+      "step": 10067
+    },
+    {
+      "epoch": 1.9848186119873819,
+      "grad_norm": 0.4760005186639651,
+      "learning_rate": 1.0133298286348046e-05,
+      "loss": 0.3309,
+      "step": 10068
+    },
+    {
+      "epoch": 1.9850157728706623,
+      "grad_norm": 0.47490621093107355,
+      "learning_rate": 1.0131748396488057e-05,
+      "loss": 0.3426,
+      "step": 10069
+    },
+    {
+      "epoch": 1.9852129337539433,
+      "grad_norm": 0.4490036193130323,
+      "learning_rate": 1.0130198503462705e-05,
+      "loss": 0.3244,
+      "step": 10070
+    },
+    {
+      "epoch": 1.985410094637224,
+      "grad_norm": 0.43921214974034223,
+      "learning_rate": 1.0128648607309228e-05,
+      "loss": 0.3108,
+      "step": 10071
+    },
+    {
+      "epoch": 1.9856072555205047,
+      "grad_norm": 0.47746525475723234,
+      "learning_rate": 1.0127098708064866e-05,
+      "loss": 0.3631,
+      "step": 10072
+    },
+    {
+      "epoch": 1.9858044164037856,
+      "grad_norm": 0.44863839819394496,
+      "learning_rate": 1.0125548805766852e-05,
+      "loss": 0.3203,
+      "step": 10073
+    },
+    {
+      "epoch": 1.9860015772870663,
+      "grad_norm": 0.43643215846559164,
+      "learning_rate": 1.0123998900452431e-05,
+      "loss": 0.3055,
+      "step": 10074
+    },
+    {
+      "epoch": 1.986198738170347,
+      "grad_norm": 0.4722620454489846,
+      "learning_rate": 1.0122448992158834e-05,
+      "loss": 0.3485,
+      "step": 10075
+    },
+    {
+      "epoch": 1.986395899053628,
+      "grad_norm": 0.4686910381456083,
+      "learning_rate": 1.0120899080923306e-05,
+      "loss": 0.3494,
+      "step": 10076
+    },
+    {
+      "epoch": 1.9865930599369084,
+      "grad_norm": 0.5235470374582382,
+      "learning_rate": 1.0119349166783073e-05,
+      "loss": 0.3158,
+      "step": 10077
+    },
+    {
+      "epoch": 1.9867902208201893,
+      "grad_norm": 0.47775678927346366,
+      "learning_rate": 1.0117799249775387e-05,
+      "loss": 0.3274,
+      "step": 10078
+    },
+    {
+      "epoch": 1.98698738170347,
+      "grad_norm": 0.4703367415865138,
+      "learning_rate": 1.0116249329937474e-05,
+      "loss": 0.3429,
+      "step": 10079
+    },
+    {
+      "epoch": 1.9871845425867507,
+      "grad_norm": 0.4475956538792856,
+      "learning_rate": 1.0114699407306576e-05,
+      "loss": 0.3144,
+      "step": 10080
+    },
+    {
+      "epoch": 1.9873817034700316,
+      "grad_norm": 0.4401013553494955,
+      "learning_rate": 1.0113149481919938e-05,
+      "loss": 0.3181,
+      "step": 10081
+    },
+    {
+      "epoch": 1.9875788643533123,
+      "grad_norm": 0.4383130137426036,
+      "learning_rate": 1.0111599553814788e-05,
+      "loss": 0.3046,
+      "step": 10082
+    },
+    {
+      "epoch": 1.987776025236593,
+      "grad_norm": 0.4776571024905826,
+      "learning_rate": 1.0110049623028371e-05,
+      "loss": 0.3395,
+      "step": 10083
+    },
+    {
+      "epoch": 1.987973186119874,
+      "grad_norm": 0.4672666220627933,
+      "learning_rate": 1.0108499689597924e-05,
+      "loss": 0.3152,
+      "step": 10084
+    },
+    {
+      "epoch": 1.9881703470031544,
+      "grad_norm": 0.45203351381336665,
+      "learning_rate": 1.0106949753560682e-05,
+      "loss": 0.312,
+      "step": 10085
+    },
+    {
+      "epoch": 1.9883675078864353,
+      "grad_norm": 0.4774637834527684,
+      "learning_rate": 1.0105399814953889e-05,
+      "loss": 0.3212,
+      "step": 10086
+    },
+    {
+      "epoch": 1.988564668769716,
+      "grad_norm": 0.4517907516822674,
+      "learning_rate": 1.010384987381478e-05,
+      "loss": 0.2959,
+      "step": 10087
+    },
+    {
+      "epoch": 1.9887618296529967,
+      "grad_norm": 0.47886943448952884,
+      "learning_rate": 1.0102299930180592e-05,
+      "loss": 0.3485,
+      "step": 10088
+    },
+    {
+      "epoch": 1.9889589905362777,
+      "grad_norm": 0.6789498721589519,
+      "learning_rate": 1.0100749984088567e-05,
+      "loss": 0.3312,
+      "step": 10089
+    },
+    {
+      "epoch": 1.9891561514195584,
+      "grad_norm": 0.48039651499274616,
+      "learning_rate": 1.0099200035575943e-05,
+      "loss": 0.3376,
+      "step": 10090
+    },
+    {
+      "epoch": 1.989353312302839,
+      "grad_norm": 0.49344206103676497,
+      "learning_rate": 1.0097650084679957e-05,
+      "loss": 0.3566,
+      "step": 10091
+    },
+    {
+      "epoch": 1.98955047318612,
+      "grad_norm": 0.455626086749079,
+      "learning_rate": 1.0096100131437851e-05,
+      "loss": 0.3167,
+      "step": 10092
+    },
+    {
+      "epoch": 1.9897476340694005,
+      "grad_norm": 0.471900481186439,
+      "learning_rate": 1.009455017588686e-05,
+      "loss": 0.3215,
+      "step": 10093
+    },
+    {
+      "epoch": 1.9899447949526814,
+      "grad_norm": 0.43095008957712944,
+      "learning_rate": 1.0093000218064224e-05,
+      "loss": 0.3152,
+      "step": 10094
+    },
+    {
+      "epoch": 1.990141955835962,
+      "grad_norm": 0.47850763704179666,
+      "learning_rate": 1.0091450258007188e-05,
+      "loss": 0.3285,
+      "step": 10095
+    },
+    {
+      "epoch": 1.9903391167192428,
+      "grad_norm": 0.46902329428111766,
+      "learning_rate": 1.008990029575298e-05,
+      "loss": 0.327,
+      "step": 10096
+    },
+    {
+      "epoch": 1.9905362776025237,
+      "grad_norm": 0.4635116114920501,
+      "learning_rate": 1.008835033133885e-05,
+      "loss": 0.3245,
+      "step": 10097
+    },
+    {
+      "epoch": 1.9907334384858044,
+      "grad_norm": 0.4899631136805948,
+      "learning_rate": 1.0086800364802028e-05,
+      "loss": 0.3454,
+      "step": 10098
+    },
+    {
+      "epoch": 1.9909305993690851,
+      "grad_norm": 0.4424518951849392,
+      "learning_rate": 1.008525039617976e-05,
+      "loss": 0.2982,
+      "step": 10099
+    },
+    {
+      "epoch": 1.991127760252366,
+      "grad_norm": 0.4843796306436317,
+      "learning_rate": 1.008370042550928e-05,
+      "loss": 0.3226,
+      "step": 10100
+    },
+    {
+      "epoch": 1.9913249211356467,
+      "grad_norm": 0.47515439701906687,
+      "learning_rate": 1.0082150452827832e-05,
+      "loss": 0.3647,
+      "step": 10101
+    },
+    {
+      "epoch": 1.9915220820189274,
+      "grad_norm": 0.4984861179086406,
+      "learning_rate": 1.0080600478172653e-05,
+      "loss": 0.3689,
+      "step": 10102
+    },
+    {
+      "epoch": 1.9917192429022084,
+      "grad_norm": 0.46711955198863625,
+      "learning_rate": 1.0079050501580983e-05,
+      "loss": 0.3347,
+      "step": 10103
+    },
+    {
+      "epoch": 1.9919164037854888,
+      "grad_norm": 0.5227568346359791,
+      "learning_rate": 1.0077500523090058e-05,
+      "loss": 0.3125,
+      "step": 10104
+    },
+    {
+      "epoch": 1.9921135646687698,
+      "grad_norm": 0.4491627976049431,
+      "learning_rate": 1.0075950542737123e-05,
+      "loss": 0.3253,
+      "step": 10105
+    },
+    {
+      "epoch": 1.9923107255520505,
+      "grad_norm": 0.47884782845375357,
+      "learning_rate": 1.0074400560559416e-05,
+      "loss": 0.3402,
+      "step": 10106
+    },
+    {
+      "epoch": 1.9925078864353312,
+      "grad_norm": 0.45076597168659827,
+      "learning_rate": 1.0072850576594175e-05,
+      "loss": 0.321,
+      "step": 10107
+    },
+    {
+      "epoch": 1.992705047318612,
+      "grad_norm": 0.45577225933101717,
+      "learning_rate": 1.0071300590878639e-05,
+      "loss": 0.3216,
+      "step": 10108
+    },
+    {
+      "epoch": 1.9929022082018928,
+      "grad_norm": 0.4654018460101875,
+      "learning_rate": 1.006975060345005e-05,
+      "loss": 0.3336,
+      "step": 10109
+    },
+    {
+      "epoch": 1.9930993690851735,
+      "grad_norm": 0.4833827723808277,
+      "learning_rate": 1.0068200614345647e-05,
+      "loss": 0.3326,
+      "step": 10110
+    },
+    {
+      "epoch": 1.9932965299684544,
+      "grad_norm": 0.45506045092199265,
+      "learning_rate": 1.0066650623602667e-05,
+      "loss": 0.2972,
+      "step": 10111
+    },
+    {
+      "epoch": 1.993493690851735,
+      "grad_norm": 0.47117744890482555,
+      "learning_rate": 1.0065100631258356e-05,
+      "loss": 0.3359,
+      "step": 10112
+    },
+    {
+      "epoch": 1.9936908517350158,
+      "grad_norm": 0.4155693551302444,
+      "learning_rate": 1.0063550637349946e-05,
+      "loss": 0.3016,
+      "step": 10113
+    },
+    {
+      "epoch": 1.9938880126182965,
+      "grad_norm": 0.4796712431778576,
+      "learning_rate": 1.0062000641914683e-05,
+      "loss": 0.3038,
+      "step": 10114
+    },
+    {
+      "epoch": 1.9940851735015772,
+      "grad_norm": 0.4460608776935147,
+      "learning_rate": 1.00604506449898e-05,
+      "loss": 0.3156,
+      "step": 10115
+    },
+    {
+      "epoch": 1.9942823343848581,
+      "grad_norm": 0.4636295632848529,
+      "learning_rate": 1.0058900646612548e-05,
+      "loss": 0.34,
+      "step": 10116
+    },
+    {
+      "epoch": 1.9944794952681388,
+      "grad_norm": 0.4419540655820689,
+      "learning_rate": 1.0057350646820157e-05,
+      "loss": 0.3153,
+      "step": 10117
+    },
+    {
+      "epoch": 1.9946766561514195,
+      "grad_norm": 0.4595608276786376,
+      "learning_rate": 1.0055800645649874e-05,
+      "loss": 0.3308,
+      "step": 10118
+    },
+    {
+      "epoch": 1.9948738170347005,
+      "grad_norm": 0.46822641410388405,
+      "learning_rate": 1.0054250643138931e-05,
+      "loss": 0.3351,
+      "step": 10119
+    },
+    {
+      "epoch": 1.995070977917981,
+      "grad_norm": 0.4571949677373604,
+      "learning_rate": 1.0052700639324574e-05,
+      "loss": 0.3409,
+      "step": 10120
+    },
+    {
+      "epoch": 1.9952681388012619,
+      "grad_norm": 0.4565357948126349,
+      "learning_rate": 1.0051150634244042e-05,
+      "loss": 0.3246,
+      "step": 10121
+    },
+    {
+      "epoch": 1.9954652996845426,
+      "grad_norm": 0.4715824788374287,
+      "learning_rate": 1.0049600627934576e-05,
+      "loss": 0.3266,
+      "step": 10122
+    },
+    {
+      "epoch": 1.9956624605678233,
+      "grad_norm": 36.0684477965843,
+      "learning_rate": 1.0048050620433415e-05,
+      "loss": 0.53,
+      "step": 10123
+    },
+    {
+      "epoch": 1.9958596214511042,
+      "grad_norm": 0.4920636243774622,
+      "learning_rate": 1.0046500611777799e-05,
+      "loss": 0.3162,
+      "step": 10124
+    },
+    {
+      "epoch": 1.9960567823343849,
+      "grad_norm": 0.4865650474889048,
+      "learning_rate": 1.004495060200497e-05,
+      "loss": 0.3395,
+      "step": 10125
+    },
+    {
+      "epoch": 1.9962539432176656,
+      "grad_norm": 0.4730171891025844,
+      "learning_rate": 1.0043400591152162e-05,
+      "loss": 0.3264,
+      "step": 10126
+    },
+    {
+      "epoch": 1.9964511041009465,
+      "grad_norm": 0.49951458846201335,
+      "learning_rate": 1.0041850579256623e-05,
+      "loss": 0.3605,
+      "step": 10127
+    },
+    {
+      "epoch": 1.996648264984227,
+      "grad_norm": 0.4595162916836803,
+      "learning_rate": 1.0040300566355588e-05,
+      "loss": 0.3143,
+      "step": 10128
+    },
+    {
+      "epoch": 1.996845425867508,
+      "grad_norm": 0.48676470289418966,
+      "learning_rate": 1.00387505524863e-05,
+      "loss": 0.3381,
+      "step": 10129
+    },
+    {
+      "epoch": 1.9970425867507886,
+      "grad_norm": 0.4953287911408293,
+      "learning_rate": 1.0037200537686001e-05,
+      "loss": 0.3395,
+      "step": 10130
+    },
+    {
+      "epoch": 1.9972397476340693,
+      "grad_norm": 0.4507924402166945,
+      "learning_rate": 1.0035650521991927e-05,
+      "loss": 0.3162,
+      "step": 10131
+    },
+    {
+      "epoch": 1.9974369085173502,
+      "grad_norm": 0.4651561295109272,
+      "learning_rate": 1.0034100505441322e-05,
+      "loss": 0.3225,
+      "step": 10132
+    },
+    {
+      "epoch": 1.997634069400631,
+      "grad_norm": 0.45225944154125214,
+      "learning_rate": 1.0032550488071424e-05,
+      "loss": 0.3214,
+      "step": 10133
+    },
+    {
+      "epoch": 1.9978312302839116,
+      "grad_norm": 0.46650198815431315,
+      "learning_rate": 1.0031000469919474e-05,
+      "loss": 0.3524,
+      "step": 10134
+    },
+    {
+      "epoch": 1.9980283911671926,
+      "grad_norm": 0.4685090100993953,
+      "learning_rate": 1.0029450451022713e-05,
+      "loss": 0.3366,
+      "step": 10135
+    },
+    {
+      "epoch": 1.998225552050473,
+      "grad_norm": 0.46959485471009443,
+      "learning_rate": 1.002790043141838e-05,
+      "loss": 0.3227,
+      "step": 10136
+    },
+    {
+      "epoch": 1.998422712933754,
+      "grad_norm": 0.4875933386597677,
+      "learning_rate": 1.0026350411143719e-05,
+      "loss": 0.337,
+      "step": 10137
+    },
+    {
+      "epoch": 1.9986198738170347,
+      "grad_norm": 0.4640353105289818,
+      "learning_rate": 1.0024800390235967e-05,
+      "loss": 0.3296,
+      "step": 10138
+    },
+    {
+      "epoch": 1.9988170347003154,
+      "grad_norm": 0.4365597644010865,
+      "learning_rate": 1.0023250368732367e-05,
+      "loss": 0.3246,
+      "step": 10139
+    },
+    {
+      "epoch": 1.9990141955835963,
+      "grad_norm": 0.463634992887384,
+      "learning_rate": 1.0021700346670156e-05,
+      "loss": 0.3476,
+      "step": 10140
+    },
+    {
+      "epoch": 1.999211356466877,
+      "grad_norm": 0.4488592770922772,
+      "learning_rate": 1.0020150324086575e-05,
+      "loss": 0.3288,
+      "step": 10141
+    },
+    {
+      "epoch": 1.9994085173501577,
+      "grad_norm": 0.4474284368558747,
+      "learning_rate": 1.0018600301018873e-05,
+      "loss": 0.3236,
+      "step": 10142
+    },
+    {
+      "epoch": 1.9996056782334386,
+      "grad_norm": 0.47845080896837267,
+      "learning_rate": 1.0017050277504276e-05,
+      "loss": 0.3214,
+      "step": 10143
+    },
+    {
+      "epoch": 1.999802839116719,
+      "grad_norm": 0.47724597287669274,
+      "learning_rate": 1.0015500253580039e-05,
+      "loss": 0.3339,
+      "step": 10144
+    },
+    {
+      "epoch": 1.999802839116719,
+      "eval_loss": 0.41522932052612305,
+      "eval_runtime": 344.5021,
+      "eval_samples_per_second": 23.599,
+      "eval_steps_per_second": 1.477,
+      "step": 10144
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 20288,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 5072,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.323382503112704e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}