diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,60382 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9997680083516993,
+  "eval_steps": 500,
+  "global_step": 8620,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00023199164830066117,
+      "grad_norm": 74.39870819089624,
+      "learning_rate": 0.0,
+      "loss": 2.1794,
+      "step": 1
+    },
+    {
+      "epoch": 0.00046398329660132234,
+      "grad_norm": 112.28468562448423,
+      "learning_rate": 3.1041712554031024e-07,
+      "loss": 2.3896,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006959749449019835,
+      "grad_norm": 75.42842213959644,
+      "learning_rate": 4.919995035630432e-07,
+      "loss": 2.0454,
+      "step": 3
+    },
+    {
+      "epoch": 0.0009279665932026447,
+      "grad_norm": 105.0533217529224,
+      "learning_rate": 6.208342510806205e-07,
+      "loss": 2.3266,
+      "step": 4
+    },
+    {
+      "epoch": 0.0011599582415033058,
+      "grad_norm": 148.58005460638825,
+      "learning_rate": 7.207662449262237e-07,
+      "loss": 2.6506,
+      "step": 5
+    },
+    {
+      "epoch": 0.001391949889803967,
+      "grad_norm": 69.77358744845148,
+      "learning_rate": 8.024166291033534e-07,
+      "loss": 1.8992,
+      "step": 6
+    },
+    {
+      "epoch": 0.0016239415381046282,
+      "grad_norm": 114.58129896954524,
+      "learning_rate": 8.71451045276563e-07,
+      "loss": 1.137,
+      "step": 7
+    },
+    {
+      "epoch": 0.0018559331864052894,
+      "grad_norm": 94.46712989635765,
+      "learning_rate": 9.312513766209306e-07,
+      "loss": 0.9458,
+      "step": 8
+    },
+    {
+      "epoch": 0.0020879248347059507,
+      "grad_norm": 137.0968911795173,
+      "learning_rate": 9.839990071260865e-07,
+      "loss": 0.7461,
+      "step": 9
+    },
+    {
+      "epoch": 0.0023199164830066117,
+      "grad_norm": 31.144565059150484,
+      "learning_rate": 1.031183370466534e-06,
+      "loss": 0.9004,
+      "step": 10
+    },
+    {
+      "epoch": 0.002551908131307273,
+      "grad_norm": 32.870107426052726,
+      "learning_rate": 1.0738668190606527e-06,
+      "loss": 0.9956,
+      "step": 11
+    },
+    {
+      "epoch": 0.002783899779607934,
+      "grad_norm": 52.54313022163986,
+      "learning_rate": 1.1128337546436637e-06,
+      "loss": 0.7702,
+      "step": 12
+    },
+    {
+      "epoch": 0.0030158914279085954,
+      "grad_norm": 43.63442634279012,
+      "learning_rate": 1.1486798605405537e-06,
+      "loss": 0.8403,
+      "step": 13
+    },
+    {
+      "epoch": 0.0032478830762092564,
+      "grad_norm": 18.930724925338343,
+      "learning_rate": 1.1818681708168735e-06,
+      "loss": 0.9008,
+      "step": 14
+    },
+    {
+      "epoch": 0.0034798747245099178,
+      "grad_norm": 37.59291122718292,
+      "learning_rate": 1.212765748489267e-06,
+      "loss": 0.8522,
+      "step": 15
+    },
+    {
+      "epoch": 0.0037118663728105787,
+      "grad_norm": 771.0884603354021,
+      "learning_rate": 1.241668502161241e-06,
+      "loss": 0.721,
+      "step": 16
+    },
+    {
+      "epoch": 0.00394385802111124,
+      "grad_norm": 28.570318092553354,
+      "learning_rate": 1.2688184659337598e-06,
+      "loss": 0.7248,
+      "step": 17
+    },
+    {
+      "epoch": 0.0041758496694119015,
+      "grad_norm": 24.426651580387347,
+      "learning_rate": 1.2944161326663965e-06,
+      "loss": 0.8323,
+      "step": 18
+    },
+    {
+      "epoch": 0.004407841317712562,
+      "grad_norm": 19.369835379439408,
+      "learning_rate": 1.3186294482267552e-06,
+      "loss": 0.696,
+      "step": 19
+    },
+    {
+      "epoch": 0.004639832966013223,
+      "grad_norm": 41.14435097126691,
+      "learning_rate": 1.3416004960068441e-06,
+      "loss": 0.7574,
+      "step": 20
+    },
+    {
+      "epoch": 0.004871824614313885,
+      "grad_norm": 13.411501946608672,
+      "learning_rate": 1.3634505488396065e-06,
+      "loss": 0.7499,
+      "step": 21
+    },
+    {
+      "epoch": 0.005103816262614546,
+      "grad_norm": 32.892805474385675,
+      "learning_rate": 1.384283944600963e-06,
+      "loss": 0.6824,
+      "step": 22
+    },
+    {
+      "epoch": 0.005335807910915207,
+      "grad_norm": 39.84552686658576,
+      "learning_rate": 1.4041910996027212e-06,
+      "loss": 0.4729,
+      "step": 23
+    },
+    {
+      "epoch": 0.005567799559215868,
+      "grad_norm": 22.06254779204553,
+      "learning_rate": 1.423250880183974e-06,
+      "loss": 0.6192,
+      "step": 24
+    },
+    {
+      "epoch": 0.0057997912075165295,
+      "grad_norm": 37.344375903194816,
+      "learning_rate": 1.4415324898524475e-06,
+      "loss": 0.7357,
+      "step": 25
+    },
+    {
+      "epoch": 0.006031782855817191,
+      "grad_norm": 32.31772823142612,
+      "learning_rate": 1.459096986080864e-06,
+      "loss": 0.5671,
+      "step": 26
+    },
+    {
+      "epoch": 0.006263774504117851,
+      "grad_norm": 10.92384480057118,
+      "learning_rate": 1.4759985106891297e-06,
+      "loss": 0.6269,
+      "step": 27
+    },
+    {
+      "epoch": 0.006495766152418513,
+      "grad_norm": 35.369564086076316,
+      "learning_rate": 1.4922852963571837e-06,
+      "loss": 0.6397,
+      "step": 28
+    },
+    {
+      "epoch": 0.006727757800719174,
+      "grad_norm": 18.49217287082658,
+      "learning_rate": 1.5080004964369569e-06,
+      "loss": 0.628,
+      "step": 29
+    },
+    {
+      "epoch": 0.0069597494490198355,
+      "grad_norm": 41.13501433825827,
+      "learning_rate": 1.5231828740295771e-06,
+      "loss": 0.6695,
+      "step": 30
+    },
+    {
+      "epoch": 0.007191741097320496,
+      "grad_norm": 54.22576765655488,
+      "learning_rate": 1.5378673780327045e-06,
+      "loss": 0.6157,
+      "step": 31
+    },
+    {
+      "epoch": 0.007423732745621157,
+      "grad_norm": 16.447372540566636,
+      "learning_rate": 1.5520856277015512e-06,
+      "loss": 0.5254,
+      "step": 32
+    },
+    {
+      "epoch": 0.007655724393921819,
+      "grad_norm": 32.63353299704918,
+      "learning_rate": 1.5658663226236957e-06,
+      "loss": 0.6434,
+      "step": 33
+    },
+    {
+      "epoch": 0.00788771604222248,
+      "grad_norm": 24.91733224181655,
+      "learning_rate": 1.57923559147407e-06,
+      "loss": 0.5484,
+      "step": 34
+    },
+    {
+      "epoch": 0.00811970769052314,
+      "grad_norm": 25.808395796187114,
+      "learning_rate": 1.592217290202787e-06,
+      "loss": 0.5112,
+      "step": 35
+    },
+    {
+      "epoch": 0.008351699338823803,
+      "grad_norm": 16.243462702948914,
+      "learning_rate": 1.6048332582067068e-06,
+      "loss": 0.4698,
+      "step": 36
+    },
+    {
+      "epoch": 0.008583690987124463,
+      "grad_norm": 23.97628827091189,
+      "learning_rate": 1.6171035393948334e-06,
+      "loss": 0.5316,
+      "step": 37
+    },
+    {
+      "epoch": 0.008815682635425124,
+      "grad_norm": 27.040198776241745,
+      "learning_rate": 1.6290465737670655e-06,
+      "loss": 0.485,
+      "step": 38
+    },
+    {
+      "epoch": 0.009047674283725786,
+      "grad_norm": 38.2640149689467,
+      "learning_rate": 1.6406793641035967e-06,
+      "loss": 0.5918,
+      "step": 39
+    },
+    {
+      "epoch": 0.009279665932026447,
+      "grad_norm": 22.98109068927263,
+      "learning_rate": 1.6520176215471544e-06,
+      "loss": 0.4683,
+      "step": 40
+    },
+    {
+      "epoch": 0.009511657580327109,
+      "grad_norm": 48.11308667579086,
+      "learning_rate": 1.6630758932062726e-06,
+      "loss": 0.625,
+      "step": 41
+    },
+    {
+      "epoch": 0.00974364922862777,
+      "grad_norm": 33.404519417339436,
+      "learning_rate": 1.6738676743799167e-06,
+      "loss": 0.5841,
+      "step": 42
+    },
+    {
+      "epoch": 0.00997564087692843,
+      "grad_norm": 58.61162129240976,
+      "learning_rate": 1.6844055075753218e-06,
+      "loss": 0.4964,
+      "step": 43
+    },
+    {
+      "epoch": 0.010207632525229092,
+      "grad_norm": 38.349120521370125,
+      "learning_rate": 1.694701070141273e-06,
+      "loss": 0.5981,
+      "step": 44
+    },
+    {
+      "epoch": 0.010439624173529753,
+      "grad_norm": 14.12305267247844,
+      "learning_rate": 1.7047652520523101e-06,
+      "loss": 0.3997,
+      "step": 45
+    },
+    {
+      "epoch": 0.010671615821830413,
+      "grad_norm": 51.54573323050249,
+      "learning_rate": 1.7146082251430314e-06,
+      "loss": 0.5963,
+      "step": 46
+    },
+    {
+      "epoch": 0.010903607470131076,
+      "grad_norm": 39.1682679389499,
+      "learning_rate": 1.7242395048960248e-06,
+      "loss": 0.5537,
+      "step": 47
+    },
+    {
+      "epoch": 0.011135599118431736,
+      "grad_norm": 46.60587716219377,
+      "learning_rate": 1.7336680057242842e-06,
+      "loss": 0.6186,
+      "step": 48
+    },
+    {
+      "epoch": 0.011367590766732398,
+      "grad_norm": 34.28832247434921,
+      "learning_rate": 1.742902090553126e-06,
+      "loss": 0.4981,
+      "step": 49
+    },
+    {
+      "epoch": 0.011599582415033059,
+      "grad_norm": 36.911818320913,
+      "learning_rate": 1.7519496153927577e-06,
+      "loss": 0.5478,
+      "step": 50
+    },
+    {
+      "epoch": 0.01183157406333372,
+      "grad_norm": 19.20406082351278,
+      "learning_rate": 1.760817969496803e-06,
+      "loss": 0.4656,
+      "step": 51
+    },
+    {
+      "epoch": 0.012063565711634382,
+      "grad_norm": 25.02534356933273,
+      "learning_rate": 1.7695141116211742e-06,
+      "loss": 0.3916,
+      "step": 52
+    },
+    {
+      "epoch": 0.012295557359935042,
+      "grad_norm": 18.593964365810773,
+      "learning_rate": 1.7780446028290557e-06,
+      "loss": 0.4645,
+      "step": 53
+    },
+    {
+      "epoch": 0.012527549008235703,
+      "grad_norm": 17.886039648783353,
+      "learning_rate": 1.7864156362294398e-06,
+      "loss": 0.4096,
+      "step": 54
+    },
+    {
+      "epoch": 0.012759540656536365,
+      "grad_norm": 34.88545271696248,
+      "learning_rate": 1.7946330639868765e-06,
+      "loss": 0.642,
+      "step": 55
+    },
+    {
+      "epoch": 0.012991532304837025,
+      "grad_norm": 18.784349156418234,
+      "learning_rate": 1.802702421897494e-06,
+      "loss": 0.4415,
+      "step": 56
+    },
+    {
+      "epoch": 0.013223523953137688,
+      "grad_norm": 37.50634451781366,
+      "learning_rate": 1.8106289517897987e-06,
+      "loss": 0.3873,
+      "step": 57
+    },
+    {
+      "epoch": 0.013455515601438348,
+      "grad_norm": 16.880603630761,
+      "learning_rate": 1.818417621977267e-06,
+      "loss": 0.5032,
+      "step": 58
+    },
+    {
+      "epoch": 0.013687507249739009,
+      "grad_norm": 17.02585172512492,
+      "learning_rate": 1.8260731459625882e-06,
+      "loss": 0.4495,
+      "step": 59
+    },
+    {
+      "epoch": 0.013919498898039671,
+      "grad_norm": 107.35379091818216,
+      "learning_rate": 1.8335999995698872e-06,
+      "loss": 0.696,
+      "step": 60
+    },
+    {
+      "epoch": 0.014151490546340332,
+      "grad_norm": 19.40508069322086,
+      "learning_rate": 1.8410024366608638e-06,
+      "loss": 0.4458,
+      "step": 61
+    },
+    {
+      "epoch": 0.014383482194640992,
+      "grad_norm": 23.044132880786158,
+      "learning_rate": 1.8482845035730148e-06,
+      "loss": 0.5393,
+      "step": 62
+    },
+    {
+      "epoch": 0.014615473842941654,
+      "grad_norm": 28.654552405092105,
+      "learning_rate": 1.8554500524026495e-06,
+      "loss": 0.5256,
+      "step": 63
+    },
+    {
+      "epoch": 0.014847465491242315,
+      "grad_norm": 48.577148536807755,
+      "learning_rate": 1.8625027532418612e-06,
+      "loss": 0.5776,
+      "step": 64
+    },
+    {
+      "epoch": 0.015079457139542977,
+      "grad_norm": 21.245792672313023,
+      "learning_rate": 1.869446105466777e-06,
+      "loss": 0.4652,
+      "step": 65
+    },
+    {
+      "epoch": 0.015311448787843638,
+      "grad_norm": 71.60793690174401,
+      "learning_rate": 1.8762834481640057e-06,
+      "loss": 0.5653,
+      "step": 66
+    },
+    {
+      "epoch": 0.015543440436144298,
+      "grad_norm": 26.128541484591057,
+      "learning_rate": 1.8830179697730491e-06,
+      "loss": 0.535,
+      "step": 67
+    },
+    {
+      "epoch": 0.01577543208444496,
+      "grad_norm": 33.64066732768441,
+      "learning_rate": 1.8896527170143803e-06,
+      "loss": 0.5011,
+      "step": 68
+    },
+    {
+      "epoch": 0.01600742373274562,
+      "grad_norm": 26.88917149800284,
+      "learning_rate": 1.8961906031657644e-06,
+      "loss": 0.5581,
+      "step": 69
+    },
+    {
+      "epoch": 0.01623941538104628,
+      "grad_norm": 25.195762814497893,
+      "learning_rate": 1.9026344157430974e-06,
+      "loss": 0.4721,
+      "step": 70
+    },
+    {
+      "epoch": 0.016471407029346942,
+      "grad_norm": 29.42534723385049,
+      "learning_rate": 1.908986823636446e-06,
+      "loss": 0.5184,
+      "step": 71
+    },
+    {
+      "epoch": 0.016703398677647606,
+      "grad_norm": 16.62919699797701,
+      "learning_rate": 1.9152503837470172e-06,
+      "loss": 0.4352,
+      "step": 72
+    },
+    {
+      "epoch": 0.016935390325948266,
+      "grad_norm": 19.78067075131713,
+      "learning_rate": 1.921427547166354e-06,
+      "loss": 0.4264,
+      "step": 73
+    },
+    {
+      "epoch": 0.017167381974248927,
+      "grad_norm": 26.136176618497988,
+      "learning_rate": 1.927520664935144e-06,
+      "loss": 0.4323,
+      "step": 74
+    },
+    {
+      "epoch": 0.017399373622549587,
+      "grad_norm": 20.42370322133567,
+      "learning_rate": 1.9335319934154905e-06,
+      "loss": 0.4796,
+      "step": 75
+    },
+    {
+      "epoch": 0.017631365270850248,
+      "grad_norm": 16.995153150338616,
+      "learning_rate": 1.939463699307376e-06,
+      "loss": 0.3959,
+      "step": 76
+    },
+    {
+      "epoch": 0.017863356919150912,
+      "grad_norm": 16.339592308878576,
+      "learning_rate": 1.945317864337216e-06,
+      "loss": 0.3985,
+      "step": 77
+    },
+    {
+      "epoch": 0.018095348567451573,
+      "grad_norm": 19.2386041009152,
+      "learning_rate": 1.951096489643907e-06,
+      "loss": 0.4068,
+      "step": 78
+    },
+    {
+      "epoch": 0.018327340215752233,
+      "grad_norm": 23.57524018997069,
+      "learning_rate": 1.9568014998854824e-06,
+      "loss": 0.3625,
+      "step": 79
+    },
+    {
+      "epoch": 0.018559331864052894,
+      "grad_norm": 32.00958631342296,
+      "learning_rate": 1.9624347470874646e-06,
+      "loss": 0.4652,
+      "step": 80
+    },
+    {
+      "epoch": 0.018791323512353554,
+      "grad_norm": 29.277000871097183,
+      "learning_rate": 1.967998014252173e-06,
+      "loss": 0.3404,
+      "step": 81
+    },
+    {
+      "epoch": 0.019023315160654218,
+      "grad_norm": 26.679579757688188,
+      "learning_rate": 1.973493018746583e-06,
+      "loss": 0.5614,
+      "step": 82
+    },
+    {
+      "epoch": 0.01925530680895488,
+      "grad_norm": 34.03519083482874,
+      "learning_rate": 1.9789214154848464e-06,
+      "loss": 0.4527,
+      "step": 83
+    },
+    {
+      "epoch": 0.01948729845725554,
+      "grad_norm": 9.307613839376076,
+      "learning_rate": 1.984284799920227e-06,
+      "loss": 0.3238,
+      "step": 84
+    },
+    {
+      "epoch": 0.0197192901055562,
+      "grad_norm": 28.17577070445086,
+      "learning_rate": 1.989584710859984e-06,
+      "loss": 0.584,
+      "step": 85
+    },
+    {
+      "epoch": 0.01995128175385686,
+      "grad_norm": 32.013482960486684,
+      "learning_rate": 1.994822633115632e-06,
+      "loss": 0.4381,
+      "step": 86
+    },
+    {
+      "epoch": 0.02018327340215752,
+      "grad_norm": 26.285276118944726,
+      "learning_rate": 2e-06,
+      "loss": 0.3722,
+      "step": 87
+    },
+    {
+      "epoch": 0.020415265050458185,
+      "grad_norm": 37.319181949930496,
+      "learning_rate": 2e-06,
+      "loss": 0.4557,
+      "step": 88
+    },
+    {
+      "epoch": 0.020647256698758845,
+      "grad_norm": 22.094760938419988,
+      "learning_rate": 2e-06,
+      "loss": 0.3266,
+      "step": 89
+    },
+    {
+      "epoch": 0.020879248347059506,
+      "grad_norm": 25.245767362643576,
+      "learning_rate": 2e-06,
+      "loss": 0.3231,
+      "step": 90
+    },
+    {
+      "epoch": 0.021111239995360166,
+      "grad_norm": 21.175594320556343,
+      "learning_rate": 2e-06,
+      "loss": 0.4793,
+      "step": 91
+    },
+    {
+      "epoch": 0.021343231643660827,
+      "grad_norm": 34.0650426085687,
+      "learning_rate": 2e-06,
+      "loss": 0.3412,
+      "step": 92
+    },
+    {
+      "epoch": 0.02157522329196149,
+      "grad_norm": 18.241508100008726,
+      "learning_rate": 2e-06,
+      "loss": 0.4224,
+      "step": 93
+    },
+    {
+      "epoch": 0.02180721494026215,
+      "grad_norm": 60.461910401186756,
+      "learning_rate": 2e-06,
+      "loss": 0.3686,
+      "step": 94
+    },
+    {
+      "epoch": 0.022039206588562812,
+      "grad_norm": 30.838743569237803,
+      "learning_rate": 2e-06,
+      "loss": 0.4954,
+      "step": 95
+    },
+    {
+      "epoch": 0.022271198236863472,
+      "grad_norm": 18.870087576100286,
+      "learning_rate": 2e-06,
+      "loss": 0.4342,
+      "step": 96
+    },
+    {
+      "epoch": 0.022503189885164133,
+      "grad_norm": 17.830426935702803,
+      "learning_rate": 2e-06,
+      "loss": 0.3998,
+      "step": 97
+    },
+    {
+      "epoch": 0.022735181533464797,
+      "grad_norm": 22.72931222482498,
+      "learning_rate": 2e-06,
+      "loss": 0.4001,
+      "step": 98
+    },
+    {
+      "epoch": 0.022967173181765457,
+      "grad_norm": 28.15915823308931,
+      "learning_rate": 2e-06,
+      "loss": 0.3818,
+      "step": 99
+    },
+    {
+      "epoch": 0.023199164830066118,
+      "grad_norm": 19.509920544774893,
+      "learning_rate": 2e-06,
+      "loss": 0.3897,
+      "step": 100
+    },
+    {
+      "epoch": 0.02343115647836678,
+      "grad_norm": 22.200505904185203,
+      "learning_rate": 2e-06,
+      "loss": 0.3631,
+      "step": 101
+    },
+    {
+      "epoch": 0.02366314812666744,
+      "grad_norm": 46.8337504969018,
+      "learning_rate": 2e-06,
+      "loss": 0.4703,
+      "step": 102
+    },
+    {
+      "epoch": 0.023895139774968103,
+      "grad_norm": 30.52389628493483,
+      "learning_rate": 2e-06,
+      "loss": 0.4831,
+      "step": 103
+    },
+    {
+      "epoch": 0.024127131423268763,
+      "grad_norm": 33.49480618399563,
+      "learning_rate": 2e-06,
+      "loss": 0.4041,
+      "step": 104
+    },
+    {
+      "epoch": 0.024359123071569424,
+      "grad_norm": 50.81297884729123,
+      "learning_rate": 2e-06,
+      "loss": 0.6375,
+      "step": 105
+    },
+    {
+      "epoch": 0.024591114719870084,
+      "grad_norm": 17.464920058894066,
+      "learning_rate": 2e-06,
+      "loss": 0.3552,
+      "step": 106
+    },
+    {
+      "epoch": 0.024823106368170745,
+      "grad_norm": 24.45022223936699,
+      "learning_rate": 2e-06,
+      "loss": 0.4592,
+      "step": 107
+    },
+    {
+      "epoch": 0.025055098016471405,
+      "grad_norm": 19.966546967508155,
+      "learning_rate": 2e-06,
+      "loss": 0.4533,
+      "step": 108
+    },
+    {
+      "epoch": 0.02528708966477207,
+      "grad_norm": 28.450205806954816,
+      "learning_rate": 2e-06,
+      "loss": 0.4727,
+      "step": 109
+    },
+    {
+      "epoch": 0.02551908131307273,
+      "grad_norm": 41.13916073966549,
+      "learning_rate": 2e-06,
+      "loss": 0.497,
+      "step": 110
+    },
+    {
+      "epoch": 0.02575107296137339,
+      "grad_norm": 16.112754647057017,
+      "learning_rate": 2e-06,
+      "loss": 0.3844,
+      "step": 111
+    },
+    {
+      "epoch": 0.02598306460967405,
+      "grad_norm": 28.162277990599502,
+      "learning_rate": 2e-06,
+      "loss": 0.4953,
+      "step": 112
+    },
+    {
+      "epoch": 0.02621505625797471,
+      "grad_norm": 23.758505464030357,
+      "learning_rate": 2e-06,
+      "loss": 0.3308,
+      "step": 113
+    },
+    {
+      "epoch": 0.026447047906275375,
+      "grad_norm": 32.06452268639564,
+      "learning_rate": 2e-06,
+      "loss": 0.4046,
+      "step": 114
+    },
+    {
+      "epoch": 0.026679039554576036,
+      "grad_norm": 17.30981532939412,
+      "learning_rate": 2e-06,
+      "loss": 0.3985,
+      "step": 115
+    },
+    {
+      "epoch": 0.026911031202876697,
+      "grad_norm": 33.43471126599173,
+      "learning_rate": 2e-06,
+      "loss": 0.4519,
+      "step": 116
+    },
+    {
+      "epoch": 0.027143022851177357,
+      "grad_norm": 27.51227040876881,
+      "learning_rate": 2e-06,
+      "loss": 0.4136,
+      "step": 117
+    },
+    {
+      "epoch": 0.027375014499478018,
+      "grad_norm": 23.974504782175565,
+      "learning_rate": 2e-06,
+      "loss": 0.477,
+      "step": 118
+    },
+    {
+      "epoch": 0.02760700614777868,
+      "grad_norm": 30.741717997138643,
+      "learning_rate": 2e-06,
+      "loss": 0.4243,
+      "step": 119
+    },
+    {
+      "epoch": 0.027838997796079342,
+      "grad_norm": 11.972532620974395,
+      "learning_rate": 2e-06,
+      "loss": 0.2537,
+      "step": 120
+    },
+    {
+      "epoch": 0.028070989444380003,
+      "grad_norm": 11.808199219768877,
+      "learning_rate": 2e-06,
+      "loss": 0.3055,
+      "step": 121
+    },
+    {
+      "epoch": 0.028302981092680663,
+      "grad_norm": 41.16658965347815,
+      "learning_rate": 2e-06,
+      "loss": 0.446,
+      "step": 122
+    },
+    {
+      "epoch": 0.028534972740981324,
+      "grad_norm": 36.30280128025487,
+      "learning_rate": 2e-06,
+      "loss": 0.3894,
+      "step": 123
+    },
+    {
+      "epoch": 0.028766964389281984,
+      "grad_norm": 38.95823329601204,
+      "learning_rate": 2e-06,
+      "loss": 0.4648,
+      "step": 124
+    },
+    {
+      "epoch": 0.028998956037582648,
+      "grad_norm": 22.228812125565018,
+      "learning_rate": 2e-06,
+      "loss": 0.4194,
+      "step": 125
+    },
+    {
+      "epoch": 0.02923094768588331,
+      "grad_norm": 15.979110407781146,
+      "learning_rate": 2e-06,
+      "loss": 0.4041,
+      "step": 126
+    },
+    {
+      "epoch": 0.02946293933418397,
+      "grad_norm": 37.18105571848787,
+      "learning_rate": 2e-06,
+      "loss": 0.4889,
+      "step": 127
+    },
+    {
+      "epoch": 0.02969493098248463,
+      "grad_norm": 30.857636960509943,
+      "learning_rate": 2e-06,
+      "loss": 0.4519,
+      "step": 128
+    },
+    {
+      "epoch": 0.02992692263078529,
+      "grad_norm": 39.61185636984746,
+      "learning_rate": 2e-06,
+      "loss": 0.3358,
+      "step": 129
+    },
+    {
+      "epoch": 0.030158914279085954,
+      "grad_norm": 42.498140804000236,
+      "learning_rate": 2e-06,
+      "loss": 0.5417,
+      "step": 130
+    },
+    {
+      "epoch": 0.030390905927386615,
+      "grad_norm": 10.908948081927463,
+      "learning_rate": 2e-06,
+      "loss": 0.3374,
+      "step": 131
+    },
+    {
+      "epoch": 0.030622897575687275,
+      "grad_norm": 14.806615866444728,
+      "learning_rate": 2e-06,
+      "loss": 0.3544,
+      "step": 132
+    },
+    {
+      "epoch": 0.030854889223987936,
+      "grad_norm": 39.602376233592565,
+      "learning_rate": 2e-06,
+      "loss": 0.5983,
+      "step": 133
+    },
+    {
+      "epoch": 0.031086880872288596,
+      "grad_norm": 25.46432735816376,
+      "learning_rate": 2e-06,
+      "loss": 0.3938,
+      "step": 134
+    },
+    {
+      "epoch": 0.03131887252058926,
+      "grad_norm": 23.611133494637883,
+      "learning_rate": 2e-06,
+      "loss": 0.4045,
+      "step": 135
+    },
+    {
+      "epoch": 0.03155086416888992,
+      "grad_norm": 20.360702542186612,
+      "learning_rate": 2e-06,
+      "loss": 0.3803,
+      "step": 136
+    },
+    {
+      "epoch": 0.03178285581719058,
+      "grad_norm": 18.853473807665043,
+      "learning_rate": 2e-06,
+      "loss": 0.3924,
+      "step": 137
+    },
+    {
+      "epoch": 0.03201484746549124,
+      "grad_norm": 17.706059410717437,
+      "learning_rate": 2e-06,
+      "loss": 0.3616,
+      "step": 138
+    },
+    {
+      "epoch": 0.032246839113791906,
+      "grad_norm": 25.910709281503944,
+      "learning_rate": 2e-06,
+      "loss": 0.4097,
+      "step": 139
+    },
+    {
+      "epoch": 0.03247883076209256,
+      "grad_norm": 18.78408165355993,
+      "learning_rate": 2e-06,
+      "loss": 0.4219,
+      "step": 140
+    },
+    {
+      "epoch": 0.03271082241039323,
+      "grad_norm": 17.592827541949497,
+      "learning_rate": 2e-06,
+      "loss": 0.4167,
+      "step": 141
+    },
+    {
+      "epoch": 0.032942814058693884,
+      "grad_norm": 20.38480421675272,
+      "learning_rate": 2e-06,
+      "loss": 0.3778,
+      "step": 142
+    },
+    {
+      "epoch": 0.03317480570699455,
+      "grad_norm": 26.556214891507235,
+      "learning_rate": 2e-06,
+      "loss": 0.5647,
+      "step": 143
+    },
+    {
+      "epoch": 0.03340679735529521,
+      "grad_norm": 10.280063069710177,
+      "learning_rate": 2e-06,
+      "loss": 0.2938,
+      "step": 144
+    },
+    {
+      "epoch": 0.03363878900359587,
+      "grad_norm": 35.07937942331207,
+      "learning_rate": 2e-06,
+      "loss": 0.4143,
+      "step": 145
+    },
+    {
+      "epoch": 0.03387078065189653,
+      "grad_norm": 28.537691552524954,
+      "learning_rate": 2e-06,
+      "loss": 0.4831,
+      "step": 146
+    },
+    {
+      "epoch": 0.03410277230019719,
+      "grad_norm": 16.60729525087867,
+      "learning_rate": 2e-06,
+      "loss": 0.4095,
+      "step": 147
+    },
+    {
+      "epoch": 0.034334763948497854,
+      "grad_norm": 12.328955527513653,
+      "learning_rate": 2e-06,
+      "loss": 0.3769,
+      "step": 148
+    },
+    {
+      "epoch": 0.03456675559679852,
+      "grad_norm": 18.092020051851453,
+      "learning_rate": 2e-06,
+      "loss": 0.3555,
+      "step": 149
+    },
+    {
+      "epoch": 0.034798747245099175,
+      "grad_norm": 44.18114865674143,
+      "learning_rate": 2e-06,
+      "loss": 0.5052,
+      "step": 150
+    },
+    {
+      "epoch": 0.03503073889339984,
+      "grad_norm": 16.552193439071686,
+      "learning_rate": 2e-06,
+      "loss": 0.3389,
+      "step": 151
+    },
+    {
+      "epoch": 0.035262730541700496,
+      "grad_norm": 28.697477777252256,
+      "learning_rate": 2e-06,
+      "loss": 0.3826,
+      "step": 152
+    },
+    {
+      "epoch": 0.03549472219000116,
+      "grad_norm": 9.627827612017324,
+      "learning_rate": 2e-06,
+      "loss": 0.3534,
+      "step": 153
+    },
+    {
+      "epoch": 0.035726713838301824,
+      "grad_norm": 18.495594499790304,
+      "learning_rate": 2e-06,
+      "loss": 0.2883,
+      "step": 154
+    },
+    {
+      "epoch": 0.03595870548660248,
+      "grad_norm": 14.819731470172004,
+      "learning_rate": 2e-06,
+      "loss": 0.3543,
+      "step": 155
+    },
+    {
+      "epoch": 0.036190697134903145,
+      "grad_norm": 30.084238573946536,
+      "learning_rate": 2e-06,
+      "loss": 0.3936,
+      "step": 156
+    },
+    {
+      "epoch": 0.0364226887832038,
+      "grad_norm": 38.32707194245598,
+      "learning_rate": 2e-06,
+      "loss": 0.5518,
+      "step": 157
+    },
+    {
+      "epoch": 0.036654680431504466,
+      "grad_norm": 30.99662633417196,
+      "learning_rate": 2e-06,
+      "loss": 0.3367,
+      "step": 158
+    },
+    {
+      "epoch": 0.03688667207980513,
+      "grad_norm": 29.788479638237202,
+      "learning_rate": 2e-06,
+      "loss": 0.44,
+      "step": 159
+    },
+    {
+      "epoch": 0.03711866372810579,
+      "grad_norm": 17.494933531167206,
+      "learning_rate": 2e-06,
+      "loss": 0.3427,
+      "step": 160
+    },
+    {
+      "epoch": 0.03735065537640645,
+      "grad_norm": 33.72980029025054,
+      "learning_rate": 2e-06,
+      "loss": 0.3467,
+      "step": 161
+    },
+    {
+      "epoch": 0.03758264702470711,
+      "grad_norm": 13.379171231265387,
+      "learning_rate": 2e-06,
+      "loss": 0.3747,
+      "step": 162
+    },
+    {
+      "epoch": 0.03781463867300777,
+      "grad_norm": 20.70832977530479,
+      "learning_rate": 2e-06,
+      "loss": 0.3986,
+      "step": 163
+    },
+    {
+      "epoch": 0.038046630321308436,
+      "grad_norm": 28.74074583084915,
+      "learning_rate": 2e-06,
+      "loss": 0.3806,
+      "step": 164
+    },
+    {
+      "epoch": 0.03827862196960909,
+      "grad_norm": 52.2052448308457,
+      "learning_rate": 2e-06,
+      "loss": 0.4445,
+      "step": 165
+    },
+    {
+      "epoch": 0.03851061361790976,
+      "grad_norm": 31.048472582638617,
+      "learning_rate": 2e-06,
+      "loss": 0.4167,
+      "step": 166
+    },
+    {
+      "epoch": 0.038742605266210414,
+      "grad_norm": 21.227802893594617,
+      "learning_rate": 2e-06,
+      "loss": 0.4363,
+      "step": 167
+    },
+    {
+      "epoch": 0.03897459691451108,
+      "grad_norm": 22.483410466187955,
+      "learning_rate": 2e-06,
+      "loss": 0.3964,
+      "step": 168
+    },
+    {
+      "epoch": 0.03920658856281174,
+      "grad_norm": 15.724111111556514,
+      "learning_rate": 2e-06,
+      "loss": 0.3786,
+      "step": 169
+    },
+    {
+      "epoch": 0.0394385802111124,
+      "grad_norm": 17.28514355448206,
+      "learning_rate": 2e-06,
+      "loss": 0.3596,
+      "step": 170
+    },
+    {
+      "epoch": 0.03967057185941306,
+      "grad_norm": 13.689374632434935,
+      "learning_rate": 2e-06,
+      "loss": 0.283,
+      "step": 171
+    },
+    {
+      "epoch": 0.03990256350771372,
+      "grad_norm": 14.577661385441322,
+      "learning_rate": 2e-06,
+      "loss": 0.3097,
+      "step": 172
+    },
+    {
+      "epoch": 0.040134555156014384,
+      "grad_norm": 24.861537338293093,
+      "learning_rate": 2e-06,
+      "loss": 0.4145,
+      "step": 173
+    },
+    {
+      "epoch": 0.04036654680431504,
+      "grad_norm": 33.91806599011101,
+      "learning_rate": 2e-06,
+      "loss": 0.4137,
+      "step": 174
+    },
+    {
+      "epoch": 0.040598538452615705,
+      "grad_norm": 37.330205558404835,
+      "learning_rate": 2e-06,
+      "loss": 0.4001,
+      "step": 175
+    },
+    {
+      "epoch": 0.04083053010091637,
+      "grad_norm": 10.75490224381776,
+      "learning_rate": 2e-06,
+      "loss": 0.3346,
+      "step": 176
+    },
+    {
+      "epoch": 0.041062521749217026,
+      "grad_norm": 17.920632030971035,
+      "learning_rate": 2e-06,
+      "loss": 0.3957,
+      "step": 177
+    },
+    {
+      "epoch": 0.04129451339751769,
+      "grad_norm": 35.06077517129045,
+      "learning_rate": 2e-06,
+      "loss": 0.5442,
+      "step": 178
+    },
+    {
+      "epoch": 0.04152650504581835,
+      "grad_norm": 37.373576118845364,
+      "learning_rate": 2e-06,
+      "loss": 0.3909,
+      "step": 179
+    },
+    {
+      "epoch": 0.04175849669411901,
+      "grad_norm": 21.774127565938404,
+      "learning_rate": 2e-06,
+      "loss": 0.3612,
+      "step": 180
+    },
+    {
+      "epoch": 0.041990488342419675,
+      "grad_norm": 22.310572412193864,
+      "learning_rate": 2e-06,
+      "loss": 0.3541,
+      "step": 181
+    },
+    {
+      "epoch": 0.04222247999072033,
+      "grad_norm": 19.91238122930712,
+      "learning_rate": 2e-06,
+      "loss": 0.3518,
+      "step": 182
+    },
+    {
+      "epoch": 0.042454471639020996,
+      "grad_norm": 38.82299720583173,
+      "learning_rate": 2e-06,
+      "loss": 0.4455,
+      "step": 183
+    },
+    {
+      "epoch": 0.04268646328732165,
+      "grad_norm": 81.78612014143567,
+      "learning_rate": 2e-06,
+      "loss": 0.4527,
+      "step": 184
+    },
+    {
+      "epoch": 0.04291845493562232,
+      "grad_norm": 38.671646262718454,
+      "learning_rate": 2e-06,
+      "loss": 0.4736,
+      "step": 185
+    },
+    {
+      "epoch": 0.04315044658392298,
+      "grad_norm": 28.292078981601982,
+      "learning_rate": 2e-06,
+      "loss": 0.3465,
+      "step": 186
+    },
+    {
+      "epoch": 0.04338243823222364,
+      "grad_norm": 28.201025845524587,
+      "learning_rate": 2e-06,
+      "loss": 0.3274,
+      "step": 187
+    },
+    {
+      "epoch": 0.0436144298805243,
+      "grad_norm": 16.502642381982742,
+      "learning_rate": 2e-06,
+      "loss": 0.3552,
+      "step": 188
+    },
+    {
+      "epoch": 0.04384642152882496,
+      "grad_norm": 21.502109491213528,
+      "learning_rate": 2e-06,
+      "loss": 0.4031,
+      "step": 189
+    },
+    {
+      "epoch": 0.044078413177125624,
+      "grad_norm": 45.395445806829784,
+      "learning_rate": 2e-06,
+      "loss": 0.3891,
+      "step": 190
+    },
+    {
+      "epoch": 0.04431040482542629,
+      "grad_norm": 16.772245712643286,
+      "learning_rate": 2e-06,
+      "loss": 0.3811,
+      "step": 191
+    },
+    {
+      "epoch": 0.044542396473726945,
+      "grad_norm": 16.484095337228272,
+      "learning_rate": 2e-06,
+      "loss": 0.3893,
+      "step": 192
+    },
+    {
+      "epoch": 0.04477438812202761,
+      "grad_norm": 19.347435512156647,
+      "learning_rate": 2e-06,
+      "loss": 0.3649,
+      "step": 193
+    },
+    {
+      "epoch": 0.045006379770328266,
+      "grad_norm": 24.383625612454,
+      "learning_rate": 2e-06,
+      "loss": 0.3976,
+      "step": 194
+    },
+    {
+      "epoch": 0.04523837141862893,
+      "grad_norm": 16.780375260685737,
+      "learning_rate": 2e-06,
+      "loss": 0.2944,
+      "step": 195
+    },
+    {
+      "epoch": 0.045470363066929594,
+      "grad_norm": 16.538828083347266,
+      "learning_rate": 2e-06,
+      "loss": 0.3967,
+      "step": 196
+    },
+    {
+      "epoch": 0.04570235471523025,
+      "grad_norm": 25.831467483102088,
+      "learning_rate": 2e-06,
+      "loss": 0.4292,
+      "step": 197
+    },
+    {
+      "epoch": 0.045934346363530915,
+      "grad_norm": 29.933106472676695,
+      "learning_rate": 2e-06,
+      "loss": 0.3907,
+      "step": 198
+    },
+    {
+      "epoch": 0.04616633801183157,
+      "grad_norm": 34.722557394383415,
+      "learning_rate": 2e-06,
+      "loss": 0.4854,
+      "step": 199
+    },
+    {
+      "epoch": 0.046398329660132236,
+      "grad_norm": 23.307072911383692,
+      "learning_rate": 2e-06,
+      "loss": 0.3485,
+      "step": 200
+    },
+    {
+      "epoch": 0.0466303213084329,
+      "grad_norm": 44.83081493918162,
+      "learning_rate": 2e-06,
+      "loss": 0.3677,
+      "step": 201
+    },
+    {
+      "epoch": 0.04686231295673356,
+      "grad_norm": 17.456496077989364,
+      "learning_rate": 2e-06,
+      "loss": 0.2696,
+      "step": 202
+    },
+    {
+      "epoch": 0.04709430460503422,
+      "grad_norm": 47.09280879290812,
+      "learning_rate": 2e-06,
+      "loss": 0.4629,
+      "step": 203
+    },
+    {
+      "epoch": 0.04732629625333488,
+      "grad_norm": 28.03807838243593,
+      "learning_rate": 2e-06,
+      "loss": 0.4508,
+      "step": 204
+    },
+    {
+      "epoch": 0.04755828790163554,
+      "grad_norm": 22.95656727807171,
+      "learning_rate": 2e-06,
+      "loss": 0.5087,
+      "step": 205
+    },
+    {
+      "epoch": 0.047790279549936206,
+      "grad_norm": 19.916633250635453,
+      "learning_rate": 2e-06,
+      "loss": 0.3502,
+      "step": 206
+    },
+    {
+      "epoch": 0.04802227119823686,
+      "grad_norm": 18.429230451297414,
+      "learning_rate": 2e-06,
+      "loss": 0.3992,
+      "step": 207
+    },
+    {
+      "epoch": 0.04825426284653753,
+      "grad_norm": 19.146947435265297,
+      "learning_rate": 2e-06,
+      "loss": 0.37,
+      "step": 208
+    },
+    {
+      "epoch": 0.048486254494838184,
+      "grad_norm": 19.635638653510593,
+      "learning_rate": 2e-06,
+      "loss": 0.4428,
+      "step": 209
+    },
+    {
+      "epoch": 0.04871824614313885,
+      "grad_norm": 16.65678239968762,
+      "learning_rate": 2e-06,
+      "loss": 0.3539,
+      "step": 210
+    },
+    {
+      "epoch": 0.048950237791439505,
+      "grad_norm": 12.811312162663347,
+      "learning_rate": 2e-06,
+      "loss": 0.3111,
+      "step": 211
+    },
+    {
+      "epoch": 0.04918222943974017,
+      "grad_norm": 26.544754694919302,
+      "learning_rate": 2e-06,
+      "loss": 0.3996,
+      "step": 212
+    },
+    {
+      "epoch": 0.04941422108804083,
+      "grad_norm": 27.038312574981394,
+      "learning_rate": 2e-06,
+      "loss": 0.4816,
+      "step": 213
+    },
+    {
+      "epoch": 0.04964621273634149,
+      "grad_norm": 19.110034604846202,
+      "learning_rate": 2e-06,
+      "loss": 0.3396,
+      "step": 214
+    },
+    {
+      "epoch": 0.049878204384642154,
+      "grad_norm": 17.23770447640579,
+      "learning_rate": 2e-06,
+      "loss": 0.4118,
+      "step": 215
+    },
+    {
+      "epoch": 0.05011019603294281,
+      "grad_norm": 25.7082676561769,
+      "learning_rate": 2e-06,
+      "loss": 0.3486,
+      "step": 216
+    },
+    {
+      "epoch": 0.050342187681243475,
+      "grad_norm": 29.31762312050998,
+      "learning_rate": 2e-06,
+      "loss": 0.4206,
+      "step": 217
+    },
+    {
+      "epoch": 0.05057417932954414,
+      "grad_norm": 28.316750519747554,
+      "learning_rate": 2e-06,
+      "loss": 0.3658,
+      "step": 218
+    },
+    {
+      "epoch": 0.050806170977844796,
+      "grad_norm": 18.295835928970263,
+      "learning_rate": 2e-06,
+      "loss": 0.3191,
+      "step": 219
+    },
+    {
+      "epoch": 0.05103816262614546,
+      "grad_norm": 22.15390958004895,
+      "learning_rate": 2e-06,
+      "loss": 0.3842,
+      "step": 220
+    },
+    {
+      "epoch": 0.05127015427444612,
+      "grad_norm": 31.55727431986,
+      "learning_rate": 2e-06,
+      "loss": 0.4348,
+      "step": 221
+    },
+    {
+      "epoch": 0.05150214592274678,
+      "grad_norm": 18.142365032470167,
+      "learning_rate": 2e-06,
+      "loss": 0.3251,
+      "step": 222
+    },
+    {
+      "epoch": 0.051734137571047445,
+      "grad_norm": 14.63675771579077,
+      "learning_rate": 2e-06,
+      "loss": 0.3872,
+      "step": 223
+    },
+    {
+      "epoch": 0.0519661292193481,
+      "grad_norm": 15.783701599136487,
+      "learning_rate": 2e-06,
+      "loss": 0.4255,
+      "step": 224
+    },
+    {
+      "epoch": 0.052198120867648766,
+      "grad_norm": 21.07727010080486,
+      "learning_rate": 2e-06,
+      "loss": 0.3722,
+      "step": 225
+    },
+    {
+      "epoch": 0.05243011251594942,
+      "grad_norm": 19.40775575437732,
+      "learning_rate": 2e-06,
+      "loss": 0.3341,
+      "step": 226
+    },
+    {
+      "epoch": 0.05266210416425009,
+      "grad_norm": 27.236675635773192,
+      "learning_rate": 2e-06,
+      "loss": 0.2782,
+      "step": 227
+    },
+    {
+      "epoch": 0.05289409581255075,
+      "grad_norm": 24.8485428044272,
+      "learning_rate": 2e-06,
+      "loss": 0.4178,
+      "step": 228
+    },
+    {
+      "epoch": 0.05312608746085141,
+      "grad_norm": 25.44002982309577,
+      "learning_rate": 2e-06,
+      "loss": 0.4114,
+      "step": 229
+    },
+    {
+      "epoch": 0.05335807910915207,
+      "grad_norm": 32.39078707431755,
+      "learning_rate": 2e-06,
+      "loss": 0.3624,
+      "step": 230
+    },
+    {
+      "epoch": 0.05359007075745273,
+      "grad_norm": 20.898925225007876,
+      "learning_rate": 2e-06,
+      "loss": 0.3823,
+      "step": 231
+    },
+    {
+      "epoch": 0.05382206240575339,
+      "grad_norm": 14.481012835578381,
+      "learning_rate": 2e-06,
+      "loss": 0.2783,
+      "step": 232
+    },
+    {
+      "epoch": 0.05405405405405406,
+      "grad_norm": 33.345472766269374,
+      "learning_rate": 2e-06,
+      "loss": 0.4105,
+      "step": 233
+    },
+    {
+      "epoch": 0.054286045702354714,
+      "grad_norm": 27.313411652481147,
+      "learning_rate": 2e-06,
+      "loss": 0.3728,
+      "step": 234
+    },
+    {
+      "epoch": 0.05451803735065538,
+      "grad_norm": 57.409672982284185,
+      "learning_rate": 2e-06,
+      "loss": 0.2871,
+      "step": 235
+    },
+    {
+      "epoch": 0.054750028998956035,
+      "grad_norm": 16.72238049877021,
+      "learning_rate": 2e-06,
+      "loss": 0.3353,
+      "step": 236
+    },
+    {
+      "epoch": 0.0549820206472567,
+      "grad_norm": 29.3975251341519,
+      "learning_rate": 2e-06,
+      "loss": 0.4285,
+      "step": 237
+    },
+    {
+      "epoch": 0.05521401229555736,
+      "grad_norm": 16.557186007773907,
+      "learning_rate": 2e-06,
+      "loss": 0.3779,
+      "step": 238
+    },
+    {
+      "epoch": 0.05544600394385802,
+      "grad_norm": 14.576894842730137,
+      "learning_rate": 2e-06,
+      "loss": 0.2714,
+      "step": 239
+    },
+    {
+      "epoch": 0.055677995592158684,
+      "grad_norm": 33.24881434167466,
+      "learning_rate": 2e-06,
+      "loss": 0.4713,
+      "step": 240
+    },
+    {
+      "epoch": 0.05590998724045934,
+      "grad_norm": 30.214762606934197,
+      "learning_rate": 2e-06,
+      "loss": 0.3491,
+      "step": 241
+    },
+    {
+      "epoch": 0.056141978888760005,
+      "grad_norm": 24.468541329934542,
+      "learning_rate": 2e-06,
+      "loss": 0.3431,
+      "step": 242
+    },
+    {
+      "epoch": 0.05637397053706067,
+      "grad_norm": 16.08552961180912,
+      "learning_rate": 2e-06,
+      "loss": 0.4033,
+      "step": 243
+    },
+    {
+      "epoch": 0.056605962185361326,
+      "grad_norm": 14.354582913537627,
+      "learning_rate": 2e-06,
+      "loss": 0.3266,
+      "step": 244
+    },
+    {
+      "epoch": 0.05683795383366199,
+      "grad_norm": 19.620651208208127,
+      "learning_rate": 2e-06,
+      "loss": 0.3277,
+      "step": 245
+    },
+    {
+      "epoch": 0.05706994548196265,
+      "grad_norm": 15.691928454576992,
+      "learning_rate": 2e-06,
+      "loss": 0.3291,
+      "step": 246
+    },
+    {
+      "epoch": 0.05730193713026331,
+      "grad_norm": 17.42590199984151,
+      "learning_rate": 2e-06,
+      "loss": 0.3284,
+      "step": 247
+    },
+    {
+      "epoch": 0.05753392877856397,
+      "grad_norm": 11.225459850807889,
+      "learning_rate": 2e-06,
+      "loss": 0.2378,
+      "step": 248
+    },
+    {
+      "epoch": 0.05776592042686463,
+      "grad_norm": 23.956123325800576,
+      "learning_rate": 2e-06,
+      "loss": 0.4609,
+      "step": 249
+    },
+    {
+      "epoch": 0.057997912075165296,
+      "grad_norm": 30.195231126761364,
+      "learning_rate": 2e-06,
+      "loss": 0.3564,
+      "step": 250
+    },
+    {
+      "epoch": 0.05822990372346595,
+      "grad_norm": 16.841385249087196,
+      "learning_rate": 2e-06,
+      "loss": 0.4339,
+      "step": 251
+    },
+    {
+      "epoch": 0.05846189537176662,
+      "grad_norm": 18.80534936760663,
+      "learning_rate": 2e-06,
+      "loss": 0.3906,
+      "step": 252
+    },
+    {
+      "epoch": 0.058693887020067274,
+      "grad_norm": 21.267472826987586,
+      "learning_rate": 2e-06,
+      "loss": 0.4437,
+      "step": 253
+    },
+    {
+      "epoch": 0.05892587866836794,
+      "grad_norm": 16.367714475333194,
+      "learning_rate": 2e-06,
+      "loss": 0.3078,
+      "step": 254
+    },
+    {
+      "epoch": 0.0591578703166686,
+      "grad_norm": 16.808024362000612,
+      "learning_rate": 2e-06,
+      "loss": 0.4284,
+      "step": 255
+    },
+    {
+      "epoch": 0.05938986196496926,
+      "grad_norm": 14.47203839492185,
+      "learning_rate": 2e-06,
+      "loss": 0.3487,
+      "step": 256
+    },
+    {
+      "epoch": 0.05962185361326992,
+      "grad_norm": 14.422808168299367,
+      "learning_rate": 2e-06,
+      "loss": 0.3002,
+      "step": 257
+    },
+    {
+      "epoch": 0.05985384526157058,
+      "grad_norm": 17.241932663553175,
+      "learning_rate": 2e-06,
+      "loss": 0.3285,
+      "step": 258
+    },
+    {
+      "epoch": 0.060085836909871244,
+      "grad_norm": 10.136726922616239,
+      "learning_rate": 2e-06,
+      "loss": 0.2544,
+      "step": 259
+    },
+    {
+      "epoch": 0.06031782855817191,
+      "grad_norm": 31.901966098793842,
+      "learning_rate": 2e-06,
+      "loss": 0.3934,
+      "step": 260
+    },
+    {
+      "epoch": 0.060549820206472565,
+      "grad_norm": 10.203419621588298,
+      "learning_rate": 2e-06,
+      "loss": 0.2739,
+      "step": 261
+    },
+    {
+      "epoch": 0.06078181185477323,
+      "grad_norm": 30.516359662869256,
+      "learning_rate": 2e-06,
+      "loss": 0.4177,
+      "step": 262
+    },
+    {
+      "epoch": 0.061013803503073887,
+      "grad_norm": 23.490412449083408,
+      "learning_rate": 2e-06,
+      "loss": 0.3069,
+      "step": 263
+    },
+    {
+      "epoch": 0.06124579515137455,
+      "grad_norm": 26.962705108860312,
+      "learning_rate": 2e-06,
+      "loss": 0.488,
+      "step": 264
+    },
+    {
+      "epoch": 0.061477786799675214,
+      "grad_norm": 21.8606732946384,
+      "learning_rate": 2e-06,
+      "loss": 0.3353,
+      "step": 265
+    },
+    {
+      "epoch": 0.06170977844797587,
+      "grad_norm": 11.84435585441604,
+      "learning_rate": 2e-06,
+      "loss": 0.2634,
+      "step": 266
+    },
+    {
+      "epoch": 0.061941770096276536,
+      "grad_norm": 18.759386751316498,
+      "learning_rate": 2e-06,
+      "loss": 0.335,
+      "step": 267
+    },
+    {
+      "epoch": 0.06217376174457719,
+      "grad_norm": 10.860341407594653,
+      "learning_rate": 2e-06,
+      "loss": 0.269,
+      "step": 268
+    },
+    {
+      "epoch": 0.06240575339287786,
+      "grad_norm": 22.756283905059107,
+      "learning_rate": 2e-06,
+      "loss": 0.4431,
+      "step": 269
+    },
+    {
+      "epoch": 0.06263774504117851,
+      "grad_norm": 18.857691106467716,
+      "learning_rate": 2e-06,
+      "loss": 0.4064,
+      "step": 270
+    },
+    {
+      "epoch": 0.06286973668947918,
+      "grad_norm": 13.71142172349342,
+      "learning_rate": 2e-06,
+      "loss": 0.3452,
+      "step": 271
+    },
+    {
+      "epoch": 0.06310172833777984,
+      "grad_norm": 28.59068887733619,
+      "learning_rate": 2e-06,
+      "loss": 0.4787,
+      "step": 272
+    },
+    {
+      "epoch": 0.0633337199860805,
+      "grad_norm": 34.9471101130114,
+      "learning_rate": 2e-06,
+      "loss": 0.4529,
+      "step": 273
+    },
+    {
+      "epoch": 0.06356571163438116,
+      "grad_norm": 15.432357523010165,
+      "learning_rate": 2e-06,
+      "loss": 0.302,
+      "step": 274
+    },
+    {
+      "epoch": 0.06379770328268182,
+      "grad_norm": 21.176022226286484,
+      "learning_rate": 2e-06,
+      "loss": 0.3524,
+      "step": 275
+    },
+    {
+      "epoch": 0.06402969493098248,
+      "grad_norm": 20.398745690881963,
+      "learning_rate": 2e-06,
+      "loss": 0.3602,
+      "step": 276
+    },
+    {
+      "epoch": 0.06426168657928315,
+      "grad_norm": 45.61620524073223,
+      "learning_rate": 2e-06,
+      "loss": 0.4475,
+      "step": 277
+    },
+    {
+      "epoch": 0.06449367822758381,
+      "grad_norm": 11.778603437990974,
+      "learning_rate": 2e-06,
+      "loss": 0.2865,
+      "step": 278
+    },
+    {
+      "epoch": 0.06472566987588446,
+      "grad_norm": 16.05408430375991,
+      "learning_rate": 2e-06,
+      "loss": 0.3002,
+      "step": 279
+    },
+    {
+      "epoch": 0.06495766152418513,
+      "grad_norm": 24.413991600204813,
+      "learning_rate": 2e-06,
+      "loss": 0.4776,
+      "step": 280
+    },
+    {
+      "epoch": 0.06518965317248579,
+      "grad_norm": 20.658726900866288,
+      "learning_rate": 2e-06,
+      "loss": 0.4699,
+      "step": 281
+    },
+    {
+      "epoch": 0.06542164482078645,
+      "grad_norm": 20.214349439391693,
+      "learning_rate": 2e-06,
+      "loss": 0.3172,
+      "step": 282
+    },
+    {
+      "epoch": 0.06565363646908712,
+      "grad_norm": 18.253321100048833,
+      "learning_rate": 2e-06,
+      "loss": 0.3053,
+      "step": 283
+    },
+    {
+      "epoch": 0.06588562811738777,
+      "grad_norm": 25.27398761329044,
+      "learning_rate": 2e-06,
+      "loss": 0.4358,
+      "step": 284
+    },
+    {
+      "epoch": 0.06611761976568843,
+      "grad_norm": 23.617098432941468,
+      "learning_rate": 2e-06,
+      "loss": 0.3992,
+      "step": 285
+    },
+    {
+      "epoch": 0.0663496114139891,
+      "grad_norm": 8.904262369286501,
+      "learning_rate": 2e-06,
+      "loss": 0.2789,
+      "step": 286
+    },
+    {
+      "epoch": 0.06658160306228976,
+      "grad_norm": 24.24410486299016,
+      "learning_rate": 2e-06,
+      "loss": 0.2762,
+      "step": 287
+    },
+    {
+      "epoch": 0.06681359471059042,
+      "grad_norm": 18.83345601805634,
+      "learning_rate": 2e-06,
+      "loss": 0.3888,
+      "step": 288
+    },
+    {
+      "epoch": 0.06704558635889107,
+      "grad_norm": 14.660856467155357,
+      "learning_rate": 2e-06,
+      "loss": 0.2764,
+      "step": 289
+    },
+    {
+      "epoch": 0.06727757800719174,
+      "grad_norm": 33.45569066237724,
+      "learning_rate": 2e-06,
+      "loss": 0.4414,
+      "step": 290
+    },
+    {
+      "epoch": 0.0675095696554924,
+      "grad_norm": 21.82220797728714,
+      "learning_rate": 2e-06,
+      "loss": 0.3749,
+      "step": 291
+    },
+    {
+      "epoch": 0.06774156130379307,
+      "grad_norm": 26.735744320549216,
+      "learning_rate": 2e-06,
+      "loss": 0.3758,
+      "step": 292
+    },
+    {
+      "epoch": 0.06797355295209373,
+      "grad_norm": 19.018016510002802,
+      "learning_rate": 2e-06,
+      "loss": 0.3474,
+      "step": 293
+    },
+    {
+      "epoch": 0.06820554460039438,
+      "grad_norm": 20.773279197240708,
+      "learning_rate": 2e-06,
+      "loss": 0.292,
+      "step": 294
+    },
+    {
+      "epoch": 0.06843753624869504,
+      "grad_norm": 21.652882354360884,
+      "learning_rate": 2e-06,
+      "loss": 0.4562,
+      "step": 295
+    },
+    {
+      "epoch": 0.06866952789699571,
+      "grad_norm": 23.18839543070114,
+      "learning_rate": 2e-06,
+      "loss": 0.4007,
+      "step": 296
+    },
+    {
+      "epoch": 0.06890151954529637,
+      "grad_norm": 20.933515687339632,
+      "learning_rate": 2e-06,
+      "loss": 0.3702,
+      "step": 297
+    },
+    {
+      "epoch": 0.06913351119359704,
+      "grad_norm": 17.835747475594196,
+      "learning_rate": 2e-06,
+      "loss": 0.2801,
+      "step": 298
+    },
+    {
+      "epoch": 0.06936550284189769,
+      "grad_norm": 17.448326380657978,
+      "learning_rate": 2e-06,
+      "loss": 0.3179,
+      "step": 299
+    },
+    {
+      "epoch": 0.06959749449019835,
+      "grad_norm": 26.014972961257733,
+      "learning_rate": 2e-06,
+      "loss": 0.3853,
+      "step": 300
+    },
+    {
+      "epoch": 0.06982948613849901,
+      "grad_norm": 25.590774728334065,
+      "learning_rate": 2e-06,
+      "loss": 0.3799,
+      "step": 301
+    },
+    {
+      "epoch": 0.07006147778679968,
+      "grad_norm": 27.64512941308947,
+      "learning_rate": 2e-06,
+      "loss": 0.4052,
+      "step": 302
+    },
+    {
+      "epoch": 0.07029346943510034,
+      "grad_norm": 30.313882980750208,
+      "learning_rate": 2e-06,
+      "loss": 0.4532,
+      "step": 303
+    },
+    {
+      "epoch": 0.07052546108340099,
+      "grad_norm": 6.010401569300229,
+      "learning_rate": 2e-06,
+      "loss": 0.265,
+      "step": 304
+    },
+    {
+      "epoch": 0.07075745273170166,
+      "grad_norm": 13.149785684164643,
+      "learning_rate": 2e-06,
+      "loss": 0.3219,
+      "step": 305
+    },
+    {
+      "epoch": 0.07098944438000232,
+      "grad_norm": 11.675290474453474,
+      "learning_rate": 2e-06,
+      "loss": 0.2681,
+      "step": 306
+    },
+    {
+      "epoch": 0.07122143602830298,
+      "grad_norm": 23.30421897740761,
+      "learning_rate": 2e-06,
+      "loss": 0.4489,
+      "step": 307
+    },
+    {
+      "epoch": 0.07145342767660365,
+      "grad_norm": 17.335889262546875,
+      "learning_rate": 2e-06,
+      "loss": 0.3924,
+      "step": 308
+    },
+    {
+      "epoch": 0.0716854193249043,
+      "grad_norm": 19.63995512405009,
+      "learning_rate": 2e-06,
+      "loss": 0.3371,
+      "step": 309
+    },
+    {
+      "epoch": 0.07191741097320496,
+      "grad_norm": 24.18936339860376,
+      "learning_rate": 2e-06,
+      "loss": 0.5017,
+      "step": 310
+    },
+    {
+      "epoch": 0.07214940262150563,
+      "grad_norm": 19.796051503412293,
+      "learning_rate": 2e-06,
+      "loss": 0.3462,
+      "step": 311
+    },
+    {
+      "epoch": 0.07238139426980629,
+      "grad_norm": 24.245217740356853,
+      "learning_rate": 2e-06,
+      "loss": 0.3896,
+      "step": 312
+    },
+    {
+      "epoch": 0.07261338591810695,
+      "grad_norm": 12.613025856749827,
+      "learning_rate": 2e-06,
+      "loss": 0.3432,
+      "step": 313
+    },
+    {
+      "epoch": 0.0728453775664076,
+      "grad_norm": 13.186064582017874,
+      "learning_rate": 2e-06,
+      "loss": 0.3445,
+      "step": 314
+    },
+    {
+      "epoch": 0.07307736921470827,
+      "grad_norm": 19.879322945226193,
+      "learning_rate": 2e-06,
+      "loss": 0.3519,
+      "step": 315
+    },
+    {
+      "epoch": 0.07330936086300893,
+      "grad_norm": 15.234452095839453,
+      "learning_rate": 2e-06,
+      "loss": 0.3882,
+      "step": 316
+    },
+    {
+      "epoch": 0.0735413525113096,
+      "grad_norm": 19.154427840014012,
+      "learning_rate": 2e-06,
+      "loss": 0.4501,
+      "step": 317
+    },
+    {
+      "epoch": 0.07377334415961026,
+      "grad_norm": 31.599294049661268,
+      "learning_rate": 2e-06,
+      "loss": 0.4159,
+      "step": 318
+    },
+    {
+      "epoch": 0.07400533580791091,
+      "grad_norm": 12.57362027981612,
+      "learning_rate": 2e-06,
+      "loss": 0.3952,
+      "step": 319
+    },
+    {
+      "epoch": 0.07423732745621157,
+      "grad_norm": 16.12107163904157,
+      "learning_rate": 2e-06,
+      "loss": 0.3671,
+      "step": 320
+    },
+    {
+      "epoch": 0.07446931910451224,
+      "grad_norm": 18.572850404526363,
+      "learning_rate": 2e-06,
+      "loss": 0.3022,
+      "step": 321
+    },
+    {
+      "epoch": 0.0747013107528129,
+      "grad_norm": 34.59231978461037,
+      "learning_rate": 2e-06,
+      "loss": 0.4849,
+      "step": 322
+    },
+    {
+      "epoch": 0.07493330240111357,
+      "grad_norm": 13.285785318659888,
+      "learning_rate": 2e-06,
+      "loss": 0.2325,
+      "step": 323
+    },
+    {
+      "epoch": 0.07516529404941422,
+      "grad_norm": 15.365050538063457,
+      "learning_rate": 2e-06,
+      "loss": 0.3808,
+      "step": 324
+    },
+    {
+      "epoch": 0.07539728569771488,
+      "grad_norm": 19.091634443388465,
+      "learning_rate": 2e-06,
+      "loss": 0.3394,
+      "step": 325
+    },
+    {
+      "epoch": 0.07562927734601554,
+      "grad_norm": 18.766369082786838,
+      "learning_rate": 2e-06,
+      "loss": 0.4084,
+      "step": 326
+    },
+    {
+      "epoch": 0.07586126899431621,
+      "grad_norm": 22.93179092378503,
+      "learning_rate": 2e-06,
+      "loss": 0.3995,
+      "step": 327
+    },
+    {
+      "epoch": 0.07609326064261687,
+      "grad_norm": 26.769794964089836,
+      "learning_rate": 2e-06,
+      "loss": 0.3826,
+      "step": 328
+    },
+    {
+      "epoch": 0.07632525229091752,
+      "grad_norm": 17.51489111550706,
+      "learning_rate": 2e-06,
+      "loss": 0.416,
+      "step": 329
+    },
+    {
+      "epoch": 0.07655724393921819,
+      "grad_norm": 16.635578775167737,
+      "learning_rate": 2e-06,
+      "loss": 0.4193,
+      "step": 330
+    },
+    {
+      "epoch": 0.07678923558751885,
+      "grad_norm": 22.91519385563069,
+      "learning_rate": 2e-06,
+      "loss": 0.4024,
+      "step": 331
+    },
+    {
+      "epoch": 0.07702122723581951,
+      "grad_norm": 24.359791173757518,
+      "learning_rate": 2e-06,
+      "loss": 0.418,
+      "step": 332
+    },
+    {
+      "epoch": 0.07725321888412018,
+      "grad_norm": 17.329821601312467,
+      "learning_rate": 2e-06,
+      "loss": 0.3709,
+      "step": 333
+    },
+    {
+      "epoch": 0.07748521053242083,
+      "grad_norm": 27.673581294656348,
+      "learning_rate": 2e-06,
+      "loss": 0.5115,
+      "step": 334
+    },
+    {
+      "epoch": 0.07771720218072149,
+      "grad_norm": 27.328269842064092,
+      "learning_rate": 2e-06,
+      "loss": 0.4333,
+      "step": 335
+    },
+    {
+      "epoch": 0.07794919382902216,
+      "grad_norm": 17.986657057309625,
+      "learning_rate": 2e-06,
+      "loss": 0.3317,
+      "step": 336
+    },
+    {
+      "epoch": 0.07818118547732282,
+      "grad_norm": 15.710760855716062,
+      "learning_rate": 2e-06,
+      "loss": 0.3436,
+      "step": 337
+    },
+    {
+      "epoch": 0.07841317712562348,
+      "grad_norm": 28.54747304199656,
+      "learning_rate": 2e-06,
+      "loss": 0.3698,
+      "step": 338
+    },
+    {
+      "epoch": 0.07864516877392413,
+      "grad_norm": 28.406446425076776,
+      "learning_rate": 2e-06,
+      "loss": 0.4491,
+      "step": 339
+    },
+    {
+      "epoch": 0.0788771604222248,
+      "grad_norm": 13.00998113821747,
+      "learning_rate": 2e-06,
+      "loss": 0.302,
+      "step": 340
+    },
+    {
+      "epoch": 0.07910915207052546,
+      "grad_norm": 24.394710444728887,
+      "learning_rate": 2e-06,
+      "loss": 0.3499,
+      "step": 341
+    },
+    {
+      "epoch": 0.07934114371882613,
+      "grad_norm": 16.848986324981123,
+      "learning_rate": 2e-06,
+      "loss": 0.3152,
+      "step": 342
+    },
+    {
+      "epoch": 0.07957313536712678,
+      "grad_norm": 31.12501067169636,
+      "learning_rate": 2e-06,
+      "loss": 0.4136,
+      "step": 343
+    },
+    {
+      "epoch": 0.07980512701542744,
+      "grad_norm": 33.14948891851159,
+      "learning_rate": 2e-06,
+      "loss": 0.4752,
+      "step": 344
+    },
+    {
+      "epoch": 0.0800371186637281,
+      "grad_norm": 27.33198419147624,
+      "learning_rate": 2e-06,
+      "loss": 0.4415,
+      "step": 345
+    },
+    {
+      "epoch": 0.08026911031202877,
+      "grad_norm": 23.586083476624456,
+      "learning_rate": 2e-06,
+      "loss": 0.3476,
+      "step": 346
+    },
+    {
+      "epoch": 0.08050110196032943,
+      "grad_norm": 20.590998680061357,
+      "learning_rate": 2e-06,
+      "loss": 0.2976,
+      "step": 347
+    },
+    {
+      "epoch": 0.08073309360863008,
+      "grad_norm": 24.375154903581354,
+      "learning_rate": 2e-06,
+      "loss": 0.4577,
+      "step": 348
+    },
+    {
+      "epoch": 0.08096508525693075,
+      "grad_norm": 28.988118815723574,
+      "learning_rate": 2e-06,
+      "loss": 0.4254,
+      "step": 349
+    },
+    {
+      "epoch": 0.08119707690523141,
+      "grad_norm": 16.422024111545927,
+      "learning_rate": 2e-06,
+      "loss": 0.3159,
+      "step": 350
+    },
+    {
+      "epoch": 0.08142906855353207,
+      "grad_norm": 28.466863498914023,
+      "learning_rate": 2e-06,
+      "loss": 0.4772,
+      "step": 351
+    },
+    {
+      "epoch": 0.08166106020183274,
+      "grad_norm": 24.154986228673167,
+      "learning_rate": 2e-06,
+      "loss": 0.3228,
+      "step": 352
+    },
+    {
+      "epoch": 0.08189305185013339,
+      "grad_norm": 10.822885849051195,
+      "learning_rate": 2e-06,
+      "loss": 0.3191,
+      "step": 353
+    },
+    {
+      "epoch": 0.08212504349843405,
+      "grad_norm": 12.170005974274169,
+      "learning_rate": 2e-06,
+      "loss": 0.3375,
+      "step": 354
+    },
+    {
+      "epoch": 0.08235703514673472,
+      "grad_norm": 16.35142729634425,
+      "learning_rate": 2e-06,
+      "loss": 0.3974,
+      "step": 355
+    },
+    {
+      "epoch": 0.08258902679503538,
+      "grad_norm": 20.3269200243023,
+      "learning_rate": 2e-06,
+      "loss": 0.3734,
+      "step": 356
+    },
+    {
+      "epoch": 0.08282101844333604,
+      "grad_norm": 18.34613766392723,
+      "learning_rate": 2e-06,
+      "loss": 0.464,
+      "step": 357
+    },
+    {
+      "epoch": 0.0830530100916367,
+      "grad_norm": 18.712920482784153,
+      "learning_rate": 2e-06,
+      "loss": 0.3485,
+      "step": 358
+    },
+    {
+      "epoch": 0.08328500173993736,
+      "grad_norm": 14.590189583694684,
+      "learning_rate": 2e-06,
+      "loss": 0.4094,
+      "step": 359
+    },
+    {
+      "epoch": 0.08351699338823802,
+      "grad_norm": 16.316678168602866,
+      "learning_rate": 2e-06,
+      "loss": 0.3666,
+      "step": 360
+    },
+    {
+      "epoch": 0.08374898503653869,
+      "grad_norm": 21.600786547305564,
+      "learning_rate": 2e-06,
+      "loss": 0.3957,
+      "step": 361
+    },
+    {
+      "epoch": 0.08398097668483935,
+      "grad_norm": 6.130256185688987,
+      "learning_rate": 2e-06,
+      "loss": 0.2955,
+      "step": 362
+    },
+    {
+      "epoch": 0.08421296833314,
+      "grad_norm": 15.583176633847687,
+      "learning_rate": 2e-06,
+      "loss": 0.3424,
+      "step": 363
+    },
+    {
+      "epoch": 0.08444495998144066,
+      "grad_norm": 10.731580878972554,
+      "learning_rate": 2e-06,
+      "loss": 0.3268,
+      "step": 364
+    },
+    {
+      "epoch": 0.08467695162974133,
+      "grad_norm": 16.475523855583248,
+      "learning_rate": 2e-06,
+      "loss": 0.3199,
+      "step": 365
+    },
+    {
+      "epoch": 0.08490894327804199,
+      "grad_norm": 19.190504550314518,
+      "learning_rate": 2e-06,
+      "loss": 0.3784,
+      "step": 366
+    },
+    {
+      "epoch": 0.08514093492634266,
+      "grad_norm": 22.69383405935886,
+      "learning_rate": 2e-06,
+      "loss": 0.3864,
+      "step": 367
+    },
+    {
+      "epoch": 0.0853729265746433,
+      "grad_norm": 6.1048447030660125,
+      "learning_rate": 2e-06,
+      "loss": 0.3483,
+      "step": 368
+    },
+    {
+      "epoch": 0.08560491822294397,
+      "grad_norm": 25.73071880587071,
+      "learning_rate": 2e-06,
+      "loss": 0.4592,
+      "step": 369
+    },
+    {
+      "epoch": 0.08583690987124463,
+      "grad_norm": 18.351202040668635,
+      "learning_rate": 2e-06,
+      "loss": 0.3476,
+      "step": 370
+    },
+    {
+      "epoch": 0.0860689015195453,
+      "grad_norm": 16.26909477918099,
+      "learning_rate": 2e-06,
+      "loss": 0.3124,
+      "step": 371
+    },
+    {
+      "epoch": 0.08630089316784596,
+      "grad_norm": 14.241063324741525,
+      "learning_rate": 2e-06,
+      "loss": 0.3076,
+      "step": 372
+    },
+    {
+      "epoch": 0.08653288481614661,
+      "grad_norm": 14.506215185214822,
+      "learning_rate": 2e-06,
+      "loss": 0.3984,
+      "step": 373
+    },
+    {
+      "epoch": 0.08676487646444728,
+      "grad_norm": 9.382376932185698,
+      "learning_rate": 2e-06,
+      "loss": 0.2976,
+      "step": 374
+    },
+    {
+      "epoch": 0.08699686811274794,
+      "grad_norm": 50.59490449016764,
+      "learning_rate": 2e-06,
+      "loss": 0.3454,
+      "step": 375
+    },
+    {
+      "epoch": 0.0872288597610486,
+      "grad_norm": 27.547531076602287,
+      "learning_rate": 2e-06,
+      "loss": 0.4362,
+      "step": 376
+    },
+    {
+      "epoch": 0.08746085140934927,
+      "grad_norm": 16.082596257621272,
+      "learning_rate": 2e-06,
+      "loss": 0.3319,
+      "step": 377
+    },
+    {
+      "epoch": 0.08769284305764992,
+      "grad_norm": 31.876537236032785,
+      "learning_rate": 2e-06,
+      "loss": 0.4915,
+      "step": 378
+    },
+    {
+      "epoch": 0.08792483470595058,
+      "grad_norm": 22.54072299553908,
+      "learning_rate": 2e-06,
+      "loss": 0.3151,
+      "step": 379
+    },
+    {
+      "epoch": 0.08815682635425125,
+      "grad_norm": 17.614758213456692,
+      "learning_rate": 2e-06,
+      "loss": 0.3949,
+      "step": 380
+    },
+    {
+      "epoch": 0.08838881800255191,
+      "grad_norm": 14.148275945729827,
+      "learning_rate": 2e-06,
+      "loss": 0.3013,
+      "step": 381
+    },
+    {
+      "epoch": 0.08862080965085258,
+      "grad_norm": 11.314995508339132,
+      "learning_rate": 2e-06,
+      "loss": 0.285,
+      "step": 382
+    },
+    {
+      "epoch": 0.08885280129915323,
+      "grad_norm": 17.241195116605457,
+      "learning_rate": 2e-06,
+      "loss": 0.3807,
+      "step": 383
+    },
+    {
+      "epoch": 0.08908479294745389,
+      "grad_norm": 15.385646501389344,
+      "learning_rate": 2e-06,
+      "loss": 0.3809,
+      "step": 384
+    },
+    {
+      "epoch": 0.08931678459575455,
+      "grad_norm": 17.684525111155942,
+      "learning_rate": 2e-06,
+      "loss": 0.1991,
+      "step": 385
+    },
+    {
+      "epoch": 0.08954877624405522,
+      "grad_norm": 24.501559406718627,
+      "learning_rate": 2e-06,
+      "loss": 0.4261,
+      "step": 386
+    },
+    {
+      "epoch": 0.08978076789235588,
+      "grad_norm": 18.143176673818,
+      "learning_rate": 2e-06,
+      "loss": 0.2896,
+      "step": 387
+    },
+    {
+      "epoch": 0.09001275954065653,
+      "grad_norm": 27.499625471928624,
+      "learning_rate": 2e-06,
+      "loss": 0.3139,
+      "step": 388
+    },
+    {
+      "epoch": 0.0902447511889572,
+      "grad_norm": 13.029164508292824,
+      "learning_rate": 2e-06,
+      "loss": 0.3339,
+      "step": 389
+    },
+    {
+      "epoch": 0.09047674283725786,
+      "grad_norm": 22.06911169822742,
+      "learning_rate": 2e-06,
+      "loss": 0.4013,
+      "step": 390
+    },
+    {
+      "epoch": 0.09070873448555852,
+      "grad_norm": 25.660629113538942,
+      "learning_rate": 2e-06,
+      "loss": 0.3766,
+      "step": 391
+    },
+    {
+      "epoch": 0.09094072613385919,
+      "grad_norm": 12.330854489315392,
+      "learning_rate": 2e-06,
+      "loss": 0.3189,
+      "step": 392
+    },
+    {
+      "epoch": 0.09117271778215984,
+      "grad_norm": 12.26820205350865,
+      "learning_rate": 2e-06,
+      "loss": 0.3068,
+      "step": 393
+    },
+    {
+      "epoch": 0.0914047094304605,
+      "grad_norm": 19.489121042795496,
+      "learning_rate": 2e-06,
+      "loss": 0.315,
+      "step": 394
+    },
+    {
+      "epoch": 0.09163670107876117,
+      "grad_norm": 17.211896639231988,
+      "learning_rate": 2e-06,
+      "loss": 0.2837,
+      "step": 395
+    },
+    {
+      "epoch": 0.09186869272706183,
+      "grad_norm": 12.236343790554217,
+      "learning_rate": 2e-06,
+      "loss": 0.2436,
+      "step": 396
+    },
+    {
+      "epoch": 0.0921006843753625,
+      "grad_norm": 12.287095537549003,
+      "learning_rate": 2e-06,
+      "loss": 0.3363,
+      "step": 397
+    },
+    {
+      "epoch": 0.09233267602366314,
+      "grad_norm": 36.20698132473805,
+      "learning_rate": 2e-06,
+      "loss": 0.3322,
+      "step": 398
+    },
+    {
+      "epoch": 0.09256466767196381,
+      "grad_norm": 30.012469417565942,
+      "learning_rate": 2e-06,
+      "loss": 0.3058,
+      "step": 399
+    },
+    {
+      "epoch": 0.09279665932026447,
+      "grad_norm": 18.70204786894355,
+      "learning_rate": 2e-06,
+      "loss": 0.28,
+      "step": 400
+    },
+    {
+      "epoch": 0.09302865096856514,
+      "grad_norm": 23.040397243965792,
+      "learning_rate": 2e-06,
+      "loss": 0.3506,
+      "step": 401
+    },
+    {
+      "epoch": 0.0932606426168658,
+      "grad_norm": 14.503689982051851,
+      "learning_rate": 2e-06,
+      "loss": 0.3534,
+      "step": 402
+    },
+    {
+      "epoch": 0.09349263426516645,
+      "grad_norm": 19.06995983751695,
+      "learning_rate": 2e-06,
+      "loss": 0.3877,
+      "step": 403
+    },
+    {
+      "epoch": 0.09372462591346711,
+      "grad_norm": 15.528379690183382,
+      "learning_rate": 2e-06,
+      "loss": 0.2853,
+      "step": 404
+    },
+    {
+      "epoch": 0.09395661756176778,
+      "grad_norm": 12.183328931435046,
+      "learning_rate": 2e-06,
+      "loss": 0.2328,
+      "step": 405
+    },
+    {
+      "epoch": 0.09418860921006844,
+      "grad_norm": 10.249081656164702,
+      "learning_rate": 2e-06,
+      "loss": 0.3154,
+      "step": 406
+    },
+    {
+      "epoch": 0.0944206008583691,
+      "grad_norm": 34.79842136298589,
+      "learning_rate": 2e-06,
+      "loss": 0.392,
+      "step": 407
+    },
+    {
+      "epoch": 0.09465259250666976,
+      "grad_norm": 27.944587538361805,
+      "learning_rate": 2e-06,
+      "loss": 0.4147,
+      "step": 408
+    },
+    {
+      "epoch": 0.09488458415497042,
+      "grad_norm": 20.30318182509188,
+      "learning_rate": 2e-06,
+      "loss": 0.4244,
+      "step": 409
+    },
+    {
+      "epoch": 0.09511657580327108,
+      "grad_norm": 43.576920841913044,
+      "learning_rate": 2e-06,
+      "loss": 0.49,
+      "step": 410
+    },
+    {
+      "epoch": 0.09534856745157175,
+      "grad_norm": 25.236131980802266,
+      "learning_rate": 2e-06,
+      "loss": 0.4109,
+      "step": 411
+    },
+    {
+      "epoch": 0.09558055909987241,
+      "grad_norm": 29.170644729481666,
+      "learning_rate": 2e-06,
+      "loss": 0.5216,
+      "step": 412
+    },
+    {
+      "epoch": 0.09581255074817306,
+      "grad_norm": 19.516436049903138,
+      "learning_rate": 2e-06,
+      "loss": 0.4077,
+      "step": 413
+    },
+    {
+      "epoch": 0.09604454239647373,
+      "grad_norm": 17.128714650282106,
+      "learning_rate": 2e-06,
+      "loss": 0.3999,
+      "step": 414
+    },
+    {
+      "epoch": 0.09627653404477439,
+      "grad_norm": 12.006511406067414,
+      "learning_rate": 2e-06,
+      "loss": 0.2462,
+      "step": 415
+    },
+    {
+      "epoch": 0.09650852569307505,
+      "grad_norm": 16.275965696790472,
+      "learning_rate": 2e-06,
+      "loss": 0.302,
+      "step": 416
+    },
+    {
+      "epoch": 0.0967405173413757,
+      "grad_norm": 14.7500730370023,
+      "learning_rate": 2e-06,
+      "loss": 0.2836,
+      "step": 417
+    },
+    {
+      "epoch": 0.09697250898967637,
+      "grad_norm": 18.667197603654824,
+      "learning_rate": 2e-06,
+      "loss": 0.3598,
+      "step": 418
+    },
+    {
+      "epoch": 0.09720450063797703,
+      "grad_norm": 17.955626925134116,
+      "learning_rate": 2e-06,
+      "loss": 0.2975,
+      "step": 419
+    },
+    {
+      "epoch": 0.0974364922862777,
+      "grad_norm": 21.36474071148061,
+      "learning_rate": 2e-06,
+      "loss": 0.3349,
+      "step": 420
+    },
+    {
+      "epoch": 0.09766848393457836,
+      "grad_norm": 37.371825648292855,
+      "learning_rate": 2e-06,
+      "loss": 0.3935,
+      "step": 421
+    },
+    {
+      "epoch": 0.09790047558287901,
+      "grad_norm": 25.24655195690179,
+      "learning_rate": 2e-06,
+      "loss": 0.3842,
+      "step": 422
+    },
+    {
+      "epoch": 0.09813246723117967,
+      "grad_norm": 18.403246533860493,
+      "learning_rate": 2e-06,
+      "loss": 0.4077,
+      "step": 423
+    },
+    {
+      "epoch": 0.09836445887948034,
+      "grad_norm": 23.049463332646905,
+      "learning_rate": 2e-06,
+      "loss": 0.4219,
+      "step": 424
+    },
+    {
+      "epoch": 0.098596450527781,
+      "grad_norm": 22.791123792015163,
+      "learning_rate": 2e-06,
+      "loss": 0.386,
+      "step": 425
+    },
+    {
+      "epoch": 0.09882844217608167,
+      "grad_norm": 20.05570413178967,
+      "learning_rate": 2e-06,
+      "loss": 0.4471,
+      "step": 426
+    },
+    {
+      "epoch": 0.09906043382438232,
+      "grad_norm": 24.374099834079217,
+      "learning_rate": 2e-06,
+      "loss": 0.3568,
+      "step": 427
+    },
+    {
+      "epoch": 0.09929242547268298,
+      "grad_norm": 26.877154115175962,
+      "learning_rate": 2e-06,
+      "loss": 0.4576,
+      "step": 428
+    },
+    {
+      "epoch": 0.09952441712098364,
+      "grad_norm": 24.697932991796872,
+      "learning_rate": 2e-06,
+      "loss": 0.2783,
+      "step": 429
+    },
+    {
+      "epoch": 0.09975640876928431,
+      "grad_norm": 60.2313041091182,
+      "learning_rate": 2e-06,
+      "loss": 0.3714,
+      "step": 430
+    },
+    {
+      "epoch": 0.09998840041758497,
+      "grad_norm": 13.60633605845247,
+      "learning_rate": 2e-06,
+      "loss": 0.3332,
+      "step": 431
+    },
+    {
+      "epoch": 0.10022039206588562,
+      "grad_norm": 42.72261628492271,
+      "learning_rate": 2e-06,
+      "loss": 0.4944,
+      "step": 432
+    },
+    {
+      "epoch": 0.10045238371418629,
+      "grad_norm": 12.503324964724692,
+      "learning_rate": 2e-06,
+      "loss": 0.3633,
+      "step": 433
+    },
+    {
+      "epoch": 0.10068437536248695,
+      "grad_norm": 22.839555311492852,
+      "learning_rate": 2e-06,
+      "loss": 0.4408,
+      "step": 434
+    },
+    {
+      "epoch": 0.10091636701078761,
+      "grad_norm": 28.45242700473296,
+      "learning_rate": 2e-06,
+      "loss": 0.4241,
+      "step": 435
+    },
+    {
+      "epoch": 0.10114835865908828,
+      "grad_norm": 25.2907973397533,
+      "learning_rate": 2e-06,
+      "loss": 0.3094,
+      "step": 436
+    },
+    {
+      "epoch": 0.10138035030738893,
+      "grad_norm": 20.645973630360405,
+      "learning_rate": 2e-06,
+      "loss": 0.365,
+      "step": 437
+    },
+    {
+      "epoch": 0.10161234195568959,
+      "grad_norm": 13.503907309499892,
+      "learning_rate": 2e-06,
+      "loss": 0.3956,
+      "step": 438
+    },
+    {
+      "epoch": 0.10184433360399026,
+      "grad_norm": 26.58537180570347,
+      "learning_rate": 2e-06,
+      "loss": 0.3936,
+      "step": 439
+    },
+    {
+      "epoch": 0.10207632525229092,
+      "grad_norm": 14.061720828202937,
+      "learning_rate": 2e-06,
+      "loss": 0.399,
+      "step": 440
+    },
+    {
+      "epoch": 0.10230831690059158,
+      "grad_norm": 21.802081089998484,
+      "learning_rate": 2e-06,
+      "loss": 0.4034,
+      "step": 441
+    },
+    {
+      "epoch": 0.10254030854889223,
+      "grad_norm": 16.007298216847893,
+      "learning_rate": 2e-06,
+      "loss": 0.227,
+      "step": 442
+    },
+    {
+      "epoch": 0.1027723001971929,
+      "grad_norm": 13.543803794607616,
+      "learning_rate": 2e-06,
+      "loss": 0.2848,
+      "step": 443
+    },
+    {
+      "epoch": 0.10300429184549356,
+      "grad_norm": 11.917643036483089,
+      "learning_rate": 2e-06,
+      "loss": 0.3229,
+      "step": 444
+    },
+    {
+      "epoch": 0.10323628349379423,
+      "grad_norm": 10.540211537982008,
+      "learning_rate": 2e-06,
+      "loss": 0.3835,
+      "step": 445
+    },
+    {
+      "epoch": 0.10346827514209489,
+      "grad_norm": 13.057127763815975,
+      "learning_rate": 2e-06,
+      "loss": 0.294,
+      "step": 446
+    },
+    {
+      "epoch": 0.10370026679039554,
+      "grad_norm": 40.636902358293355,
+      "learning_rate": 2e-06,
+      "loss": 0.5286,
+      "step": 447
+    },
+    {
+      "epoch": 0.1039322584386962,
+      "grad_norm": 9.295471134140202,
+      "learning_rate": 2e-06,
+      "loss": 0.2626,
+      "step": 448
+    },
+    {
+      "epoch": 0.10416425008699687,
+      "grad_norm": 16.71420553457328,
+      "learning_rate": 2e-06,
+      "loss": 0.4241,
+      "step": 449
+    },
+    {
+      "epoch": 0.10439624173529753,
+      "grad_norm": 8.626134436732368,
+      "learning_rate": 2e-06,
+      "loss": 0.3212,
+      "step": 450
+    },
+    {
+      "epoch": 0.1046282333835982,
+      "grad_norm": 30.773166312699107,
+      "learning_rate": 2e-06,
+      "loss": 0.4788,
+      "step": 451
+    },
+    {
+      "epoch": 0.10486022503189885,
+      "grad_norm": 13.090372218449275,
+      "learning_rate": 2e-06,
+      "loss": 0.3331,
+      "step": 452
+    },
+    {
+      "epoch": 0.10509221668019951,
+      "grad_norm": 13.979905174034451,
+      "learning_rate": 2e-06,
+      "loss": 0.2949,
+      "step": 453
+    },
+    {
+      "epoch": 0.10532420832850017,
+      "grad_norm": 42.856080673219175,
+      "learning_rate": 2e-06,
+      "loss": 0.4169,
+      "step": 454
+    },
+    {
+      "epoch": 0.10555619997680084,
+      "grad_norm": 23.273704420042517,
+      "learning_rate": 2e-06,
+      "loss": 0.3922,
+      "step": 455
+    },
+    {
+      "epoch": 0.1057881916251015,
+      "grad_norm": 12.605371417868731,
+      "learning_rate": 2e-06,
+      "loss": 0.2395,
+      "step": 456
+    },
+    {
+      "epoch": 0.10602018327340215,
+      "grad_norm": 11.582289734117069,
+      "learning_rate": 2e-06,
+      "loss": 0.3205,
+      "step": 457
+    },
+    {
+      "epoch": 0.10625217492170282,
+      "grad_norm": 13.586535826870845,
+      "learning_rate": 2e-06,
+      "loss": 0.3675,
+      "step": 458
+    },
+    {
+      "epoch": 0.10648416657000348,
+      "grad_norm": 17.080354404774585,
+      "learning_rate": 2e-06,
+      "loss": 0.3486,
+      "step": 459
+    },
+    {
+      "epoch": 0.10671615821830414,
+      "grad_norm": 17.925684598773802,
+      "learning_rate": 2e-06,
+      "loss": 0.3022,
+      "step": 460
+    },
+    {
+      "epoch": 0.10694814986660481,
+      "grad_norm": 32.39272785312845,
+      "learning_rate": 2e-06,
+      "loss": 0.3359,
+      "step": 461
+    },
+    {
+      "epoch": 0.10718014151490546,
+      "grad_norm": 18.254480374694175,
+      "learning_rate": 2e-06,
+      "loss": 0.3263,
+      "step": 462
+    },
+    {
+      "epoch": 0.10741213316320612,
+      "grad_norm": 23.591719076716867,
+      "learning_rate": 2e-06,
+      "loss": 0.4446,
+      "step": 463
+    },
+    {
+      "epoch": 0.10764412481150679,
+      "grad_norm": 19.865109921119632,
+      "learning_rate": 2e-06,
+      "loss": 0.4048,
+      "step": 464
+    },
+    {
+      "epoch": 0.10787611645980745,
+      "grad_norm": 8.884190876686526,
+      "learning_rate": 2e-06,
+      "loss": 0.2853,
+      "step": 465
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 19.61323282534642,
+      "learning_rate": 2e-06,
+      "loss": 0.3167,
+      "step": 466
+    },
+    {
+      "epoch": 0.10834009975640876,
+      "grad_norm": 12.695565418193748,
+      "learning_rate": 2e-06,
+      "loss": 0.369,
+      "step": 467
+    },
+    {
+      "epoch": 0.10857209140470943,
+      "grad_norm": 10.243463772330834,
+      "learning_rate": 2e-06,
+      "loss": 0.3189,
+      "step": 468
+    },
+    {
+      "epoch": 0.10880408305301009,
+      "grad_norm": 10.293892595757947,
+      "learning_rate": 2e-06,
+      "loss": 0.3302,
+      "step": 469
+    },
+    {
+      "epoch": 0.10903607470131076,
+      "grad_norm": 21.389390468394783,
+      "learning_rate": 2e-06,
+      "loss": 0.3741,
+      "step": 470
+    },
+    {
+      "epoch": 0.10926806634961142,
+      "grad_norm": 11.940062834217704,
+      "learning_rate": 2e-06,
+      "loss": 0.3796,
+      "step": 471
+    },
+    {
+      "epoch": 0.10950005799791207,
+      "grad_norm": 9.774522293402262,
+      "learning_rate": 2e-06,
+      "loss": 0.2699,
+      "step": 472
+    },
+    {
+      "epoch": 0.10973204964621273,
+      "grad_norm": 11.12192006208675,
+      "learning_rate": 2e-06,
+      "loss": 0.3056,
+      "step": 473
+    },
+    {
+      "epoch": 0.1099640412945134,
+      "grad_norm": 17.91288583179564,
+      "learning_rate": 2e-06,
+      "loss": 0.3468,
+      "step": 474
+    },
+    {
+      "epoch": 0.11019603294281406,
+      "grad_norm": 20.480071149104823,
+      "learning_rate": 2e-06,
+      "loss": 0.3443,
+      "step": 475
+    },
+    {
+      "epoch": 0.11042802459111473,
+      "grad_norm": 24.143986404076905,
+      "learning_rate": 2e-06,
+      "loss": 0.4692,
+      "step": 476
+    },
+    {
+      "epoch": 0.11066001623941538,
+      "grad_norm": 14.08740260271415,
+      "learning_rate": 2e-06,
+      "loss": 0.3726,
+      "step": 477
+    },
+    {
+      "epoch": 0.11089200788771604,
+      "grad_norm": 10.616284079876811,
+      "learning_rate": 2e-06,
+      "loss": 0.3315,
+      "step": 478
+    },
+    {
+      "epoch": 0.1111239995360167,
+      "grad_norm": 12.99385356040717,
+      "learning_rate": 2e-06,
+      "loss": 0.2818,
+      "step": 479
+    },
+    {
+      "epoch": 0.11135599118431737,
+      "grad_norm": 6.6027185157360035,
+      "learning_rate": 2e-06,
+      "loss": 0.2372,
+      "step": 480
+    },
+    {
+      "epoch": 0.11158798283261803,
+      "grad_norm": 17.59288186274003,
+      "learning_rate": 2e-06,
+      "loss": 0.3019,
+      "step": 481
+    },
+    {
+      "epoch": 0.11181997448091868,
+      "grad_norm": 18.112113143400485,
+      "learning_rate": 2e-06,
+      "loss": 0.3168,
+      "step": 482
+    },
+    {
+      "epoch": 0.11205196612921935,
+      "grad_norm": 18.58516521806112,
+      "learning_rate": 2e-06,
+      "loss": 0.2882,
+      "step": 483
+    },
+    {
+      "epoch": 0.11228395777752001,
+      "grad_norm": 20.018829935182424,
+      "learning_rate": 2e-06,
+      "loss": 0.326,
+      "step": 484
+    },
+    {
+      "epoch": 0.11251594942582067,
+      "grad_norm": 20.822972874843753,
+      "learning_rate": 2e-06,
+      "loss": 0.4031,
+      "step": 485
+    },
+    {
+      "epoch": 0.11274794107412134,
+      "grad_norm": 21.934178884656355,
+      "learning_rate": 2e-06,
+      "loss": 0.3179,
+      "step": 486
+    },
+    {
+      "epoch": 0.11297993272242199,
+      "grad_norm": 21.66633969901522,
+      "learning_rate": 2e-06,
+      "loss": 0.3193,
+      "step": 487
+    },
+    {
+      "epoch": 0.11321192437072265,
+      "grad_norm": 16.374257943413088,
+      "learning_rate": 2e-06,
+      "loss": 0.374,
+      "step": 488
+    },
+    {
+      "epoch": 0.11344391601902332,
+      "grad_norm": 19.504755329296714,
+      "learning_rate": 2e-06,
+      "loss": 0.3048,
+      "step": 489
+    },
+    {
+      "epoch": 0.11367590766732398,
+      "grad_norm": 14.763736020821334,
+      "learning_rate": 2e-06,
+      "loss": 0.3575,
+      "step": 490
+    },
+    {
+      "epoch": 0.11390789931562464,
+      "grad_norm": 10.60113072500508,
+      "learning_rate": 2e-06,
+      "loss": 0.2901,
+      "step": 491
+    },
+    {
+      "epoch": 0.1141398909639253,
+      "grad_norm": 15.115397184387295,
+      "learning_rate": 2e-06,
+      "loss": 0.3871,
+      "step": 492
+    },
+    {
+      "epoch": 0.11437188261222596,
+      "grad_norm": 17.2191940468864,
+      "learning_rate": 2e-06,
+      "loss": 0.4249,
+      "step": 493
+    },
+    {
+      "epoch": 0.11460387426052662,
+      "grad_norm": 12.074212003251892,
+      "learning_rate": 2e-06,
+      "loss": 0.2801,
+      "step": 494
+    },
+    {
+      "epoch": 0.11483586590882729,
+      "grad_norm": 35.686595454103376,
+      "learning_rate": 2e-06,
+      "loss": 0.3694,
+      "step": 495
+    },
+    {
+      "epoch": 0.11506785755712794,
+      "grad_norm": 25.21616405090084,
+      "learning_rate": 2e-06,
+      "loss": 0.3483,
+      "step": 496
+    },
+    {
+      "epoch": 0.1152998492054286,
+      "grad_norm": 17.78054548198579,
+      "learning_rate": 2e-06,
+      "loss": 0.3453,
+      "step": 497
+    },
+    {
+      "epoch": 0.11553184085372926,
+      "grad_norm": 16.1860008387927,
+      "learning_rate": 2e-06,
+      "loss": 0.3795,
+      "step": 498
+    },
+    {
+      "epoch": 0.11576383250202993,
+      "grad_norm": 10.73184589288529,
+      "learning_rate": 2e-06,
+      "loss": 0.2779,
+      "step": 499
+    },
+    {
+      "epoch": 0.11599582415033059,
+      "grad_norm": 21.798651290466267,
+      "learning_rate": 2e-06,
+      "loss": 0.3505,
+      "step": 500
+    },
+    {
+      "epoch": 0.11622781579863124,
+      "grad_norm": 20.404407291282162,
+      "learning_rate": 2e-06,
+      "loss": 0.3475,
+      "step": 501
+    },
+    {
+      "epoch": 0.1164598074469319,
+      "grad_norm": 21.873321256737448,
+      "learning_rate": 2e-06,
+      "loss": 0.4091,
+      "step": 502
+    },
+    {
+      "epoch": 0.11669179909523257,
+      "grad_norm": 11.859969122962756,
+      "learning_rate": 2e-06,
+      "loss": 0.3084,
+      "step": 503
+    },
+    {
+      "epoch": 0.11692379074353323,
+      "grad_norm": 26.627204119044684,
+      "learning_rate": 2e-06,
+      "loss": 0.422,
+      "step": 504
+    },
+    {
+      "epoch": 0.1171557823918339,
+      "grad_norm": 8.92584627620991,
+      "learning_rate": 2e-06,
+      "loss": 0.3135,
+      "step": 505
+    },
+    {
+      "epoch": 0.11738777404013455,
+      "grad_norm": 11.993139051388196,
+      "learning_rate": 2e-06,
+      "loss": 0.3289,
+      "step": 506
+    },
+    {
+      "epoch": 0.11761976568843521,
+      "grad_norm": 12.897442540293113,
+      "learning_rate": 2e-06,
+      "loss": 0.2981,
+      "step": 507
+    },
+    {
+      "epoch": 0.11785175733673588,
+      "grad_norm": 14.110845460131074,
+      "learning_rate": 2e-06,
+      "loss": 0.2998,
+      "step": 508
+    },
+    {
+      "epoch": 0.11808374898503654,
+      "grad_norm": 25.33181859269643,
+      "learning_rate": 2e-06,
+      "loss": 0.4741,
+      "step": 509
+    },
+    {
+      "epoch": 0.1183157406333372,
+      "grad_norm": 13.834992576536884,
+      "learning_rate": 2e-06,
+      "loss": 0.3232,
+      "step": 510
+    },
+    {
+      "epoch": 0.11854773228163785,
+      "grad_norm": 14.893252752827362,
+      "learning_rate": 2e-06,
+      "loss": 0.3444,
+      "step": 511
+    },
+    {
+      "epoch": 0.11877972392993852,
+      "grad_norm": 15.761119502374642,
+      "learning_rate": 2e-06,
+      "loss": 0.3421,
+      "step": 512
+    },
+    {
+      "epoch": 0.11901171557823918,
+      "grad_norm": 12.35069424882054,
+      "learning_rate": 2e-06,
+      "loss": 0.3332,
+      "step": 513
+    },
+    {
+      "epoch": 0.11924370722653985,
+      "grad_norm": 19.600361404784426,
+      "learning_rate": 2e-06,
+      "loss": 0.377,
+      "step": 514
+    },
+    {
+      "epoch": 0.11947569887484051,
+      "grad_norm": 24.217849992141204,
+      "learning_rate": 2e-06,
+      "loss": 0.4162,
+      "step": 515
+    },
+    {
+      "epoch": 0.11970769052314116,
+      "grad_norm": 17.814790081649537,
+      "learning_rate": 2e-06,
+      "loss": 0.3611,
+      "step": 516
+    },
+    {
+      "epoch": 0.11993968217144182,
+      "grad_norm": 11.464981304657163,
+      "learning_rate": 2e-06,
+      "loss": 0.2793,
+      "step": 517
+    },
+    {
+      "epoch": 0.12017167381974249,
+      "grad_norm": 26.99997443522923,
+      "learning_rate": 2e-06,
+      "loss": 0.3731,
+      "step": 518
+    },
+    {
+      "epoch": 0.12040366546804315,
+      "grad_norm": 17.150784636126104,
+      "learning_rate": 2e-06,
+      "loss": 0.4064,
+      "step": 519
+    },
+    {
+      "epoch": 0.12063565711634382,
+      "grad_norm": 17.853860712732363,
+      "learning_rate": 2e-06,
+      "loss": 0.3889,
+      "step": 520
+    },
+    {
+      "epoch": 0.12086764876464447,
+      "grad_norm": 14.08571749826114,
+      "learning_rate": 2e-06,
+      "loss": 0.2729,
+      "step": 521
+    },
+    {
+      "epoch": 0.12109964041294513,
+      "grad_norm": 20.754236483960852,
+      "learning_rate": 2e-06,
+      "loss": 0.4262,
+      "step": 522
+    },
+    {
+      "epoch": 0.1213316320612458,
+      "grad_norm": 16.55873039914546,
+      "learning_rate": 2e-06,
+      "loss": 0.3337,
+      "step": 523
+    },
+    {
+      "epoch": 0.12156362370954646,
+      "grad_norm": 32.91313558109254,
+      "learning_rate": 2e-06,
+      "loss": 0.4844,
+      "step": 524
+    },
+    {
+      "epoch": 0.12179561535784712,
+      "grad_norm": 11.950163381560618,
+      "learning_rate": 2e-06,
+      "loss": 0.3711,
+      "step": 525
+    },
+    {
+      "epoch": 0.12202760700614777,
+      "grad_norm": 34.77644420590345,
+      "learning_rate": 2e-06,
+      "loss": 0.5912,
+      "step": 526
+    },
+    {
+      "epoch": 0.12225959865444844,
+      "grad_norm": 18.000554624080966,
+      "learning_rate": 2e-06,
+      "loss": 0.4028,
+      "step": 527
+    },
+    {
+      "epoch": 0.1224915903027491,
+      "grad_norm": 12.121073135087734,
+      "learning_rate": 2e-06,
+      "loss": 0.3123,
+      "step": 528
+    },
+    {
+      "epoch": 0.12272358195104976,
+      "grad_norm": 26.90471471621084,
+      "learning_rate": 2e-06,
+      "loss": 0.427,
+      "step": 529
+    },
+    {
+      "epoch": 0.12295557359935043,
+      "grad_norm": 21.990425205897445,
+      "learning_rate": 2e-06,
+      "loss": 0.4281,
+      "step": 530
+    },
+    {
+      "epoch": 0.12318756524765108,
+      "grad_norm": 19.60118632192807,
+      "learning_rate": 2e-06,
+      "loss": 0.3642,
+      "step": 531
+    },
+    {
+      "epoch": 0.12341955689595174,
+      "grad_norm": 11.736726359770271,
+      "learning_rate": 2e-06,
+      "loss": 0.2567,
+      "step": 532
+    },
+    {
+      "epoch": 0.12365154854425241,
+      "grad_norm": 12.09972833385718,
+      "learning_rate": 2e-06,
+      "loss": 0.3409,
+      "step": 533
+    },
+    {
+      "epoch": 0.12388354019255307,
+      "grad_norm": 9.957374867715385,
+      "learning_rate": 2e-06,
+      "loss": 0.2945,
+      "step": 534
+    },
+    {
+      "epoch": 0.12411553184085374,
+      "grad_norm": 10.561712491251779,
+      "learning_rate": 2e-06,
+      "loss": 0.3149,
+      "step": 535
+    },
+    {
+      "epoch": 0.12434752348915439,
+      "grad_norm": 22.962972593305583,
+      "learning_rate": 2e-06,
+      "loss": 0.3786,
+      "step": 536
+    },
+    {
+      "epoch": 0.12457951513745505,
+      "grad_norm": 21.81855834354691,
+      "learning_rate": 2e-06,
+      "loss": 0.4408,
+      "step": 537
+    },
+    {
+      "epoch": 0.12481150678575571,
+      "grad_norm": 43.577518805619206,
+      "learning_rate": 2e-06,
+      "loss": 0.5206,
+      "step": 538
+    },
+    {
+      "epoch": 0.12504349843405638,
+      "grad_norm": 21.43613203738653,
+      "learning_rate": 2e-06,
+      "loss": 0.4016,
+      "step": 539
+    },
+    {
+      "epoch": 0.12527549008235703,
+      "grad_norm": 8.521408142301265,
+      "learning_rate": 2e-06,
+      "loss": 0.3127,
+      "step": 540
+    },
+    {
+      "epoch": 0.1255074817306577,
+      "grad_norm": 21.675160395032297,
+      "learning_rate": 2e-06,
+      "loss": 0.4071,
+      "step": 541
+    },
+    {
+      "epoch": 0.12573947337895836,
+      "grad_norm": 10.01476995805297,
+      "learning_rate": 2e-06,
+      "loss": 0.3015,
+      "step": 542
+    },
+    {
+      "epoch": 0.125971465027259,
+      "grad_norm": 12.745527693576305,
+      "learning_rate": 2e-06,
+      "loss": 0.3208,
+      "step": 543
+    },
+    {
+      "epoch": 0.12620345667555968,
+      "grad_norm": 10.28277249296884,
+      "learning_rate": 2e-06,
+      "loss": 0.2952,
+      "step": 544
+    },
+    {
+      "epoch": 0.12643544832386033,
+      "grad_norm": 24.837971502781365,
+      "learning_rate": 2e-06,
+      "loss": 0.3971,
+      "step": 545
+    },
+    {
+      "epoch": 0.126667439972161,
+      "grad_norm": 10.839451211684377,
+      "learning_rate": 2e-06,
+      "loss": 0.336,
+      "step": 546
+    },
+    {
+      "epoch": 0.12689943162046166,
+      "grad_norm": 15.188607097212989,
+      "learning_rate": 2e-06,
+      "loss": 0.2531,
+      "step": 547
+    },
+    {
+      "epoch": 0.1271314232687623,
+      "grad_norm": 11.852306795226589,
+      "learning_rate": 2e-06,
+      "loss": 0.3548,
+      "step": 548
+    },
+    {
+      "epoch": 0.127363414917063,
+      "grad_norm": 10.728552291466901,
+      "learning_rate": 2e-06,
+      "loss": 0.2975,
+      "step": 549
+    },
+    {
+      "epoch": 0.12759540656536364,
+      "grad_norm": 12.197563885040825,
+      "learning_rate": 2e-06,
+      "loss": 0.3412,
+      "step": 550
+    },
+    {
+      "epoch": 0.12782739821366432,
+      "grad_norm": 11.460186845755205,
+      "learning_rate": 2e-06,
+      "loss": 0.3667,
+      "step": 551
+    },
+    {
+      "epoch": 0.12805938986196497,
+      "grad_norm": 24.191419011715134,
+      "learning_rate": 2e-06,
+      "loss": 0.4359,
+      "step": 552
+    },
+    {
+      "epoch": 0.12829138151026562,
+      "grad_norm": 15.072369818460848,
+      "learning_rate": 2e-06,
+      "loss": 0.3481,
+      "step": 553
+    },
+    {
+      "epoch": 0.1285233731585663,
+      "grad_norm": 15.493554128998317,
+      "learning_rate": 2e-06,
+      "loss": 0.3944,
+      "step": 554
+    },
+    {
+      "epoch": 0.12875536480686695,
+      "grad_norm": 10.299584834939997,
+      "learning_rate": 2e-06,
+      "loss": 0.2933,
+      "step": 555
+    },
+    {
+      "epoch": 0.12898735645516762,
+      "grad_norm": 19.667944578719855,
+      "learning_rate": 2e-06,
+      "loss": 0.3843,
+      "step": 556
+    },
+    {
+      "epoch": 0.12921934810346827,
+      "grad_norm": 13.020604199902435,
+      "learning_rate": 2e-06,
+      "loss": 0.3362,
+      "step": 557
+    },
+    {
+      "epoch": 0.12945133975176892,
+      "grad_norm": 18.494565081255427,
+      "learning_rate": 2e-06,
+      "loss": 0.2426,
+      "step": 558
+    },
+    {
+      "epoch": 0.1296833314000696,
+      "grad_norm": 11.481350296677014,
+      "learning_rate": 2e-06,
+      "loss": 0.3884,
+      "step": 559
+    },
+    {
+      "epoch": 0.12991532304837025,
+      "grad_norm": 14.37603161322684,
+      "learning_rate": 2e-06,
+      "loss": 0.3521,
+      "step": 560
+    },
+    {
+      "epoch": 0.13014731469667093,
+      "grad_norm": 17.945165746050844,
+      "learning_rate": 2e-06,
+      "loss": 0.4201,
+      "step": 561
+    },
+    {
+      "epoch": 0.13037930634497158,
+      "grad_norm": 10.206165197896242,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 562
+    },
+    {
+      "epoch": 0.13061129799327223,
+      "grad_norm": 11.366171051296087,
+      "learning_rate": 2e-06,
+      "loss": 0.2697,
+      "step": 563
+    },
+    {
+      "epoch": 0.1308432896415729,
+      "grad_norm": 17.908668657075456,
+      "learning_rate": 2e-06,
+      "loss": 0.3358,
+      "step": 564
+    },
+    {
+      "epoch": 0.13107528128987356,
+      "grad_norm": 15.699717818446853,
+      "learning_rate": 2e-06,
+      "loss": 0.379,
+      "step": 565
+    },
+    {
+      "epoch": 0.13130727293817424,
+      "grad_norm": 16.386214174399377,
+      "learning_rate": 2e-06,
+      "loss": 0.3298,
+      "step": 566
+    },
+    {
+      "epoch": 0.13153926458647489,
+      "grad_norm": 10.902641798900966,
+      "learning_rate": 2e-06,
+      "loss": 0.2858,
+      "step": 567
+    },
+    {
+      "epoch": 0.13177125623477554,
+      "grad_norm": 19.20552496348048,
+      "learning_rate": 2e-06,
+      "loss": 0.3325,
+      "step": 568
+    },
+    {
+      "epoch": 0.1320032478830762,
+      "grad_norm": 17.225944620309544,
+      "learning_rate": 2e-06,
+      "loss": 0.3848,
+      "step": 569
+    },
+    {
+      "epoch": 0.13223523953137686,
+      "grad_norm": 12.063065237924471,
+      "learning_rate": 2e-06,
+      "loss": 0.3181,
+      "step": 570
+    },
+    {
+      "epoch": 0.13246723117967754,
+      "grad_norm": 15.04183112468636,
+      "learning_rate": 2e-06,
+      "loss": 0.3095,
+      "step": 571
+    },
+    {
+      "epoch": 0.1326992228279782,
+      "grad_norm": 29.59285246867334,
+      "learning_rate": 2e-06,
+      "loss": 0.3663,
+      "step": 572
+    },
+    {
+      "epoch": 0.13293121447627884,
+      "grad_norm": 22.314922985673483,
+      "learning_rate": 2e-06,
+      "loss": 0.3433,
+      "step": 573
+    },
+    {
+      "epoch": 0.13316320612457952,
+      "grad_norm": 20.246625843146674,
+      "learning_rate": 2e-06,
+      "loss": 0.3287,
+      "step": 574
+    },
+    {
+      "epoch": 0.13339519777288017,
+      "grad_norm": 35.745096593143096,
+      "learning_rate": 2e-06,
+      "loss": 0.5195,
+      "step": 575
+    },
+    {
+      "epoch": 0.13362718942118085,
+      "grad_norm": 21.991019971295252,
+      "learning_rate": 2e-06,
+      "loss": 0.3267,
+      "step": 576
+    },
+    {
+      "epoch": 0.1338591810694815,
+      "grad_norm": 11.547842625012965,
+      "learning_rate": 2e-06,
+      "loss": 0.3013,
+      "step": 577
+    },
+    {
+      "epoch": 0.13409117271778215,
+      "grad_norm": 7.575492023294917,
+      "learning_rate": 2e-06,
+      "loss": 0.2593,
+      "step": 578
+    },
+    {
+      "epoch": 0.13432316436608283,
+      "grad_norm": 14.200955942542265,
+      "learning_rate": 2e-06,
+      "loss": 0.1945,
+      "step": 579
+    },
+    {
+      "epoch": 0.13455515601438348,
+      "grad_norm": 17.254167222368046,
+      "learning_rate": 2e-06,
+      "loss": 0.3549,
+      "step": 580
+    },
+    {
+      "epoch": 0.13478714766268415,
+      "grad_norm": 14.858177542842116,
+      "learning_rate": 2e-06,
+      "loss": 0.2782,
+      "step": 581
+    },
+    {
+      "epoch": 0.1350191393109848,
+      "grad_norm": 14.332877132531262,
+      "learning_rate": 2e-06,
+      "loss": 0.2892,
+      "step": 582
+    },
+    {
+      "epoch": 0.13525113095928545,
+      "grad_norm": 11.706174595821098,
+      "learning_rate": 2e-06,
+      "loss": 0.263,
+      "step": 583
+    },
+    {
+      "epoch": 0.13548312260758613,
+      "grad_norm": 13.445854658600583,
+      "learning_rate": 2e-06,
+      "loss": 0.3788,
+      "step": 584
+    },
+    {
+      "epoch": 0.13571511425588678,
+      "grad_norm": 11.033594620835709,
+      "learning_rate": 2e-06,
+      "loss": 0.3227,
+      "step": 585
+    },
+    {
+      "epoch": 0.13594710590418746,
+      "grad_norm": 19.704778968598895,
+      "learning_rate": 2e-06,
+      "loss": 0.3464,
+      "step": 586
+    },
+    {
+      "epoch": 0.1361790975524881,
+      "grad_norm": 23.207724263487698,
+      "learning_rate": 2e-06,
+      "loss": 0.4422,
+      "step": 587
+    },
+    {
+      "epoch": 0.13641108920078876,
+      "grad_norm": 17.446548566313393,
+      "learning_rate": 2e-06,
+      "loss": 0.3404,
+      "step": 588
+    },
+    {
+      "epoch": 0.13664308084908944,
+      "grad_norm": 12.890562334412161,
+      "learning_rate": 2e-06,
+      "loss": 0.3314,
+      "step": 589
+    },
+    {
+      "epoch": 0.1368750724973901,
+      "grad_norm": 9.320297975561534,
+      "learning_rate": 2e-06,
+      "loss": 0.2527,
+      "step": 590
+    },
+    {
+      "epoch": 0.13710706414569077,
+      "grad_norm": 13.408963579965981,
+      "learning_rate": 2e-06,
+      "loss": 0.3661,
+      "step": 591
+    },
+    {
+      "epoch": 0.13733905579399142,
+      "grad_norm": 17.926859871243074,
+      "learning_rate": 2e-06,
+      "loss": 0.3791,
+      "step": 592
+    },
+    {
+      "epoch": 0.13757104744229207,
+      "grad_norm": 6.07597891688596,
+      "learning_rate": 2e-06,
+      "loss": 0.2008,
+      "step": 593
+    },
+    {
+      "epoch": 0.13780303909059274,
+      "grad_norm": 19.555073576535285,
+      "learning_rate": 2e-06,
+      "loss": 0.331,
+      "step": 594
+    },
+    {
+      "epoch": 0.1380350307388934,
+      "grad_norm": 12.512008844378691,
+      "learning_rate": 2e-06,
+      "loss": 0.3331,
+      "step": 595
+    },
+    {
+      "epoch": 0.13826702238719407,
+      "grad_norm": 20.58792756025946,
+      "learning_rate": 2e-06,
+      "loss": 0.3247,
+      "step": 596
+    },
+    {
+      "epoch": 0.13849901403549472,
+      "grad_norm": 13.69854700104819,
+      "learning_rate": 2e-06,
+      "loss": 0.3676,
+      "step": 597
+    },
+    {
+      "epoch": 0.13873100568379537,
+      "grad_norm": 7.247642328163105,
+      "learning_rate": 2e-06,
+      "loss": 0.2787,
+      "step": 598
+    },
+    {
+      "epoch": 0.13896299733209605,
+      "grad_norm": 20.662562349843842,
+      "learning_rate": 2e-06,
+      "loss": 0.4137,
+      "step": 599
+    },
+    {
+      "epoch": 0.1391949889803967,
+      "grad_norm": 23.74211769605697,
+      "learning_rate": 2e-06,
+      "loss": 0.4117,
+      "step": 600
+    },
+    {
+      "epoch": 0.13942698062869738,
+      "grad_norm": 14.54527602752155,
+      "learning_rate": 2e-06,
+      "loss": 0.3589,
+      "step": 601
+    },
+    {
+      "epoch": 0.13965897227699803,
+      "grad_norm": 13.128633707615593,
+      "learning_rate": 2e-06,
+      "loss": 0.3302,
+      "step": 602
+    },
+    {
+      "epoch": 0.13989096392529868,
+      "grad_norm": 13.78506522209295,
+      "learning_rate": 2e-06,
+      "loss": 0.3327,
+      "step": 603
+    },
+    {
+      "epoch": 0.14012295557359936,
+      "grad_norm": 15.841352731539514,
+      "learning_rate": 2e-06,
+      "loss": 0.3152,
+      "step": 604
+    },
+    {
+      "epoch": 0.1403549472219,
+      "grad_norm": 9.421455470531463,
+      "learning_rate": 2e-06,
+      "loss": 0.3354,
+      "step": 605
+    },
+    {
+      "epoch": 0.14058693887020068,
+      "grad_norm": 10.08321808770723,
+      "learning_rate": 2e-06,
+      "loss": 0.2326,
+      "step": 606
+    },
+    {
+      "epoch": 0.14081893051850133,
+      "grad_norm": 18.689274942037507,
+      "learning_rate": 2e-06,
+      "loss": 0.4126,
+      "step": 607
+    },
+    {
+      "epoch": 0.14105092216680198,
+      "grad_norm": 12.358109062169618,
+      "learning_rate": 2e-06,
+      "loss": 0.3798,
+      "step": 608
+    },
+    {
+      "epoch": 0.14128291381510266,
+      "grad_norm": 21.92588737376762,
+      "learning_rate": 2e-06,
+      "loss": 0.4711,
+      "step": 609
+    },
+    {
+      "epoch": 0.1415149054634033,
+      "grad_norm": 19.45991230773491,
+      "learning_rate": 2e-06,
+      "loss": 0.4289,
+      "step": 610
+    },
+    {
+      "epoch": 0.141746897111704,
+      "grad_norm": 13.005478035817163,
+      "learning_rate": 2e-06,
+      "loss": 0.3862,
+      "step": 611
+    },
+    {
+      "epoch": 0.14197888876000464,
+      "grad_norm": 13.562324246271288,
+      "learning_rate": 2e-06,
+      "loss": 0.3999,
+      "step": 612
+    },
+    {
+      "epoch": 0.1422108804083053,
+      "grad_norm": 31.784848826892258,
+      "learning_rate": 2e-06,
+      "loss": 0.3502,
+      "step": 613
+    },
+    {
+      "epoch": 0.14244287205660597,
+      "grad_norm": 10.206030065161611,
+      "learning_rate": 2e-06,
+      "loss": 0.2569,
+      "step": 614
+    },
+    {
+      "epoch": 0.14267486370490662,
+      "grad_norm": 24.276173026549397,
+      "learning_rate": 2e-06,
+      "loss": 0.2952,
+      "step": 615
+    },
+    {
+      "epoch": 0.1429068553532073,
+      "grad_norm": 18.342906382940086,
+      "learning_rate": 2e-06,
+      "loss": 0.3026,
+      "step": 616
+    },
+    {
+      "epoch": 0.14313884700150795,
+      "grad_norm": 8.993574006033441,
+      "learning_rate": 2e-06,
+      "loss": 0.2586,
+      "step": 617
+    },
+    {
+      "epoch": 0.1433708386498086,
+      "grad_norm": 23.936719026871057,
+      "learning_rate": 2e-06,
+      "loss": 0.3789,
+      "step": 618
+    },
+    {
+      "epoch": 0.14360283029810927,
+      "grad_norm": 16.32194373617926,
+      "learning_rate": 2e-06,
+      "loss": 0.3839,
+      "step": 619
+    },
+    {
+      "epoch": 0.14383482194640992,
+      "grad_norm": 12.988145492420317,
+      "learning_rate": 2e-06,
+      "loss": 0.3824,
+      "step": 620
+    },
+    {
+      "epoch": 0.1440668135947106,
+      "grad_norm": 20.669035386443895,
+      "learning_rate": 2e-06,
+      "loss": 0.3758,
+      "step": 621
+    },
+    {
+      "epoch": 0.14429880524301125,
+      "grad_norm": 24.990471580039006,
+      "learning_rate": 2e-06,
+      "loss": 0.3434,
+      "step": 622
+    },
+    {
+      "epoch": 0.1445307968913119,
+      "grad_norm": 17.99722260635703,
+      "learning_rate": 2e-06,
+      "loss": 0.3738,
+      "step": 623
+    },
+    {
+      "epoch": 0.14476278853961258,
+      "grad_norm": 28.862374960031236,
+      "learning_rate": 2e-06,
+      "loss": 0.3462,
+      "step": 624
+    },
+    {
+      "epoch": 0.14499478018791323,
+      "grad_norm": 16.286544229911005,
+      "learning_rate": 2e-06,
+      "loss": 0.3677,
+      "step": 625
+    },
+    {
+      "epoch": 0.1452267718362139,
+      "grad_norm": 13.749182210734784,
+      "learning_rate": 2e-06,
+      "loss": 0.3409,
+      "step": 626
+    },
+    {
+      "epoch": 0.14545876348451456,
+      "grad_norm": 8.781576263071432,
+      "learning_rate": 2e-06,
+      "loss": 0.3408,
+      "step": 627
+    },
+    {
+      "epoch": 0.1456907551328152,
+      "grad_norm": 20.23008968468618,
+      "learning_rate": 2e-06,
+      "loss": 0.3334,
+      "step": 628
+    },
+    {
+      "epoch": 0.1459227467811159,
+      "grad_norm": 21.403961956791857,
+      "learning_rate": 2e-06,
+      "loss": 0.3627,
+      "step": 629
+    },
+    {
+      "epoch": 0.14615473842941654,
+      "grad_norm": 19.072138287773942,
+      "learning_rate": 2e-06,
+      "loss": 0.2983,
+      "step": 630
+    },
+    {
+      "epoch": 0.14638673007771721,
+      "grad_norm": 22.329610088782662,
+      "learning_rate": 2e-06,
+      "loss": 0.4157,
+      "step": 631
+    },
+    {
+      "epoch": 0.14661872172601786,
+      "grad_norm": 11.982533942082721,
+      "learning_rate": 2e-06,
+      "loss": 0.2553,
+      "step": 632
+    },
+    {
+      "epoch": 0.14685071337431851,
+      "grad_norm": 22.78573635381739,
+      "learning_rate": 2e-06,
+      "loss": 0.3195,
+      "step": 633
+    },
+    {
+      "epoch": 0.1470827050226192,
+      "grad_norm": 10.40653341397135,
+      "learning_rate": 2e-06,
+      "loss": 0.3246,
+      "step": 634
+    },
+    {
+      "epoch": 0.14731469667091984,
+      "grad_norm": 14.463244019994837,
+      "learning_rate": 2e-06,
+      "loss": 0.3692,
+      "step": 635
+    },
+    {
+      "epoch": 0.14754668831922052,
+      "grad_norm": 19.800480987537895,
+      "learning_rate": 2e-06,
+      "loss": 0.3389,
+      "step": 636
+    },
+    {
+      "epoch": 0.14777867996752117,
+      "grad_norm": 10.631478775111985,
+      "learning_rate": 2e-06,
+      "loss": 0.2379,
+      "step": 637
+    },
+    {
+      "epoch": 0.14801067161582182,
+      "grad_norm": 19.589467703354284,
+      "learning_rate": 2e-06,
+      "loss": 0.3806,
+      "step": 638
+    },
+    {
+      "epoch": 0.1482426632641225,
+      "grad_norm": 17.235561484760968,
+      "learning_rate": 2e-06,
+      "loss": 0.3159,
+      "step": 639
+    },
+    {
+      "epoch": 0.14847465491242315,
+      "grad_norm": 14.29645254007326,
+      "learning_rate": 2e-06,
+      "loss": 0.2914,
+      "step": 640
+    },
+    {
+      "epoch": 0.14870664656072383,
+      "grad_norm": 15.25682225167021,
+      "learning_rate": 2e-06,
+      "loss": 0.2814,
+      "step": 641
+    },
+    {
+      "epoch": 0.14893863820902448,
+      "grad_norm": 12.72336393966868,
+      "learning_rate": 2e-06,
+      "loss": 0.31,
+      "step": 642
+    },
+    {
+      "epoch": 0.14917062985732513,
+      "grad_norm": 34.56907190484323,
+      "learning_rate": 2e-06,
+      "loss": 0.394,
+      "step": 643
+    },
+    {
+      "epoch": 0.1494026215056258,
+      "grad_norm": 20.092572113695343,
+      "learning_rate": 2e-06,
+      "loss": 0.4555,
+      "step": 644
+    },
+    {
+      "epoch": 0.14963461315392645,
+      "grad_norm": 12.558275124737698,
+      "learning_rate": 2e-06,
+      "loss": 0.2319,
+      "step": 645
+    },
+    {
+      "epoch": 0.14986660480222713,
+      "grad_norm": 10.926979723682406,
+      "learning_rate": 2e-06,
+      "loss": 0.3174,
+      "step": 646
+    },
+    {
+      "epoch": 0.15009859645052778,
+      "grad_norm": 21.759994543709613,
+      "learning_rate": 2e-06,
+      "loss": 0.3082,
+      "step": 647
+    },
+    {
+      "epoch": 0.15033058809882843,
+      "grad_norm": 14.962828288114487,
+      "learning_rate": 2e-06,
+      "loss": 0.2977,
+      "step": 648
+    },
+    {
+      "epoch": 0.1505625797471291,
+      "grad_norm": 14.945826490956852,
+      "learning_rate": 2e-06,
+      "loss": 0.3518,
+      "step": 649
+    },
+    {
+      "epoch": 0.15079457139542976,
+      "grad_norm": 12.677037134321225,
+      "learning_rate": 2e-06,
+      "loss": 0.353,
+      "step": 650
+    },
+    {
+      "epoch": 0.15102656304373044,
+      "grad_norm": 16.329612153370707,
+      "learning_rate": 2e-06,
+      "loss": 0.3625,
+      "step": 651
+    },
+    {
+      "epoch": 0.1512585546920311,
+      "grad_norm": 25.316498260503856,
+      "learning_rate": 2e-06,
+      "loss": 0.3509,
+      "step": 652
+    },
+    {
+      "epoch": 0.15149054634033174,
+      "grad_norm": 9.505821526608154,
+      "learning_rate": 2e-06,
+      "loss": 0.3144,
+      "step": 653
+    },
+    {
+      "epoch": 0.15172253798863242,
+      "grad_norm": 23.54621096323739,
+      "learning_rate": 2e-06,
+      "loss": 0.3844,
+      "step": 654
+    },
+    {
+      "epoch": 0.15195452963693307,
+      "grad_norm": 11.882344502450662,
+      "learning_rate": 2e-06,
+      "loss": 0.3435,
+      "step": 655
+    },
+    {
+      "epoch": 0.15218652128523374,
+      "grad_norm": 21.91198574912016,
+      "learning_rate": 2e-06,
+      "loss": 0.3632,
+      "step": 656
+    },
+    {
+      "epoch": 0.1524185129335344,
+      "grad_norm": 12.393539964259297,
+      "learning_rate": 2e-06,
+      "loss": 0.278,
+      "step": 657
+    },
+    {
+      "epoch": 0.15265050458183504,
+      "grad_norm": 18.168511634243217,
+      "learning_rate": 2e-06,
+      "loss": 0.2467,
+      "step": 658
+    },
+    {
+      "epoch": 0.15288249623013572,
+      "grad_norm": 14.178757384068685,
+      "learning_rate": 2e-06,
+      "loss": 0.3257,
+      "step": 659
+    },
+    {
+      "epoch": 0.15311448787843637,
+      "grad_norm": 32.74130918364026,
+      "learning_rate": 2e-06,
+      "loss": 0.3114,
+      "step": 660
+    },
+    {
+      "epoch": 0.15334647952673705,
+      "grad_norm": 12.222002529077335,
+      "learning_rate": 2e-06,
+      "loss": 0.3026,
+      "step": 661
+    },
+    {
+      "epoch": 0.1535784711750377,
+      "grad_norm": 10.954915709678856,
+      "learning_rate": 2e-06,
+      "loss": 0.3513,
+      "step": 662
+    },
+    {
+      "epoch": 0.15381046282333835,
+      "grad_norm": 20.24850597550675,
+      "learning_rate": 2e-06,
+      "loss": 0.4142,
+      "step": 663
+    },
+    {
+      "epoch": 0.15404245447163903,
+      "grad_norm": 15.214951718141643,
+      "learning_rate": 2e-06,
+      "loss": 0.3569,
+      "step": 664
+    },
+    {
+      "epoch": 0.15427444611993968,
+      "grad_norm": 17.304343638378498,
+      "learning_rate": 2e-06,
+      "loss": 0.3982,
+      "step": 665
+    },
+    {
+      "epoch": 0.15450643776824036,
+      "grad_norm": 10.307606751687013,
+      "learning_rate": 2e-06,
+      "loss": 0.3254,
+      "step": 666
+    },
+    {
+      "epoch": 0.154738429416541,
+      "grad_norm": 15.932744708373177,
+      "learning_rate": 2e-06,
+      "loss": 0.3459,
+      "step": 667
+    },
+    {
+      "epoch": 0.15497042106484166,
+      "grad_norm": 22.80289035080992,
+      "learning_rate": 2e-06,
+      "loss": 0.3968,
+      "step": 668
+    },
+    {
+      "epoch": 0.15520241271314233,
+      "grad_norm": 23.14623782349953,
+      "learning_rate": 2e-06,
+      "loss": 0.2405,
+      "step": 669
+    },
+    {
+      "epoch": 0.15543440436144298,
+      "grad_norm": 18.0912009844887,
+      "learning_rate": 2e-06,
+      "loss": 0.338,
+      "step": 670
+    },
+    {
+      "epoch": 0.15566639600974366,
+      "grad_norm": 18.677691324496845,
+      "learning_rate": 2e-06,
+      "loss": 0.3047,
+      "step": 671
+    },
+    {
+      "epoch": 0.1558983876580443,
+      "grad_norm": 8.998825461022793,
+      "learning_rate": 2e-06,
+      "loss": 0.303,
+      "step": 672
+    },
+    {
+      "epoch": 0.15613037930634496,
+      "grad_norm": 18.045426173107654,
+      "learning_rate": 2e-06,
+      "loss": 0.3606,
+      "step": 673
+    },
+    {
+      "epoch": 0.15636237095464564,
+      "grad_norm": 11.154343481153669,
+      "learning_rate": 2e-06,
+      "loss": 0.3413,
+      "step": 674
+    },
+    {
+      "epoch": 0.1565943626029463,
+      "grad_norm": 13.332750429948154,
+      "learning_rate": 2e-06,
+      "loss": 0.3186,
+      "step": 675
+    },
+    {
+      "epoch": 0.15682635425124697,
+      "grad_norm": 15.74486222469269,
+      "learning_rate": 2e-06,
+      "loss": 0.3002,
+      "step": 676
+    },
+    {
+      "epoch": 0.15705834589954762,
+      "grad_norm": 19.667188570074135,
+      "learning_rate": 2e-06,
+      "loss": 0.4405,
+      "step": 677
+    },
+    {
+      "epoch": 0.15729033754784827,
+      "grad_norm": 21.031839435111806,
+      "learning_rate": 2e-06,
+      "loss": 0.3478,
+      "step": 678
+    },
+    {
+      "epoch": 0.15752232919614895,
+      "grad_norm": 10.39843349546204,
+      "learning_rate": 2e-06,
+      "loss": 0.266,
+      "step": 679
+    },
+    {
+      "epoch": 0.1577543208444496,
+      "grad_norm": 16.454956918604434,
+      "learning_rate": 2e-06,
+      "loss": 0.3867,
+      "step": 680
+    },
+    {
+      "epoch": 0.15798631249275025,
+      "grad_norm": 16.358326166838868,
+      "learning_rate": 2e-06,
+      "loss": 0.418,
+      "step": 681
+    },
+    {
+      "epoch": 0.15821830414105092,
+      "grad_norm": 25.612389596778975,
+      "learning_rate": 2e-06,
+      "loss": 0.3624,
+      "step": 682
+    },
+    {
+      "epoch": 0.15845029578935158,
+      "grad_norm": 10.449260622783445,
+      "learning_rate": 2e-06,
+      "loss": 0.2386,
+      "step": 683
+    },
+    {
+      "epoch": 0.15868228743765225,
+      "grad_norm": 11.170225689154774,
+      "learning_rate": 2e-06,
+      "loss": 0.2839,
+      "step": 684
+    },
+    {
+      "epoch": 0.1589142790859529,
+      "grad_norm": 20.379086082147165,
+      "learning_rate": 2e-06,
+      "loss": 0.362,
+      "step": 685
+    },
+    {
+      "epoch": 0.15914627073425355,
+      "grad_norm": 29.30564423884362,
+      "learning_rate": 2e-06,
+      "loss": 0.4104,
+      "step": 686
+    },
+    {
+      "epoch": 0.15937826238255423,
+      "grad_norm": 12.564718453938301,
+      "learning_rate": 2e-06,
+      "loss": 0.3009,
+      "step": 687
+    },
+    {
+      "epoch": 0.15961025403085488,
+      "grad_norm": 16.60002829086695,
+      "learning_rate": 2e-06,
+      "loss": 0.3076,
+      "step": 688
+    },
+    {
+      "epoch": 0.15984224567915556,
+      "grad_norm": 18.44420873851104,
+      "learning_rate": 2e-06,
+      "loss": 0.376,
+      "step": 689
+    },
+    {
+      "epoch": 0.1600742373274562,
+      "grad_norm": 14.716335866668372,
+      "learning_rate": 2e-06,
+      "loss": 0.3314,
+      "step": 690
+    },
+    {
+      "epoch": 0.16030622897575686,
+      "grad_norm": 11.136643251719736,
+      "learning_rate": 2e-06,
+      "loss": 0.3109,
+      "step": 691
+    },
+    {
+      "epoch": 0.16053822062405754,
+      "grad_norm": 15.005943488342586,
+      "learning_rate": 2e-06,
+      "loss": 0.3069,
+      "step": 692
+    },
+    {
+      "epoch": 0.1607702122723582,
+      "grad_norm": 24.658145570665926,
+      "learning_rate": 2e-06,
+      "loss": 0.3951,
+      "step": 693
+    },
+    {
+      "epoch": 0.16100220392065887,
+      "grad_norm": 10.62390063366631,
+      "learning_rate": 2e-06,
+      "loss": 0.2891,
+      "step": 694
+    },
+    {
+      "epoch": 0.16123419556895952,
+      "grad_norm": 14.095175484680707,
+      "learning_rate": 2e-06,
+      "loss": 0.3373,
+      "step": 695
+    },
+    {
+      "epoch": 0.16146618721726017,
+      "grad_norm": 12.851412497958066,
+      "learning_rate": 2e-06,
+      "loss": 0.3065,
+      "step": 696
+    },
+    {
+      "epoch": 0.16169817886556084,
+      "grad_norm": 8.31681767749006,
+      "learning_rate": 2e-06,
+      "loss": 0.3037,
+      "step": 697
+    },
+    {
+      "epoch": 0.1619301705138615,
+      "grad_norm": 15.879779774741346,
+      "learning_rate": 2e-06,
+      "loss": 0.2867,
+      "step": 698
+    },
+    {
+      "epoch": 0.16216216216216217,
+      "grad_norm": 15.024589339052074,
+      "learning_rate": 2e-06,
+      "loss": 0.3346,
+      "step": 699
+    },
+    {
+      "epoch": 0.16239415381046282,
+      "grad_norm": 25.728722695798243,
+      "learning_rate": 2e-06,
+      "loss": 0.3263,
+      "step": 700
+    },
+    {
+      "epoch": 0.16262614545876347,
+      "grad_norm": 13.962249772821057,
+      "learning_rate": 2e-06,
+      "loss": 0.4021,
+      "step": 701
+    },
+    {
+      "epoch": 0.16285813710706415,
+      "grad_norm": 16.6701984573303,
+      "learning_rate": 2e-06,
+      "loss": 0.2992,
+      "step": 702
+    },
+    {
+      "epoch": 0.1630901287553648,
+      "grad_norm": 20.731069653085676,
+      "learning_rate": 2e-06,
+      "loss": 0.2982,
+      "step": 703
+    },
+    {
+      "epoch": 0.16332212040366548,
+      "grad_norm": 22.497997523362585,
+      "learning_rate": 2e-06,
+      "loss": 0.334,
+      "step": 704
+    },
+    {
+      "epoch": 0.16355411205196613,
+      "grad_norm": 13.511717785171559,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 705
+    },
+    {
+      "epoch": 0.16378610370026678,
+      "grad_norm": 14.860067921359914,
+      "learning_rate": 2e-06,
+      "loss": 0.3137,
+      "step": 706
+    },
+    {
+      "epoch": 0.16401809534856746,
+      "grad_norm": 14.477755167374747,
+      "learning_rate": 2e-06,
+      "loss": 0.2181,
+      "step": 707
+    },
+    {
+      "epoch": 0.1642500869968681,
+      "grad_norm": 31.17589495214125,
+      "learning_rate": 2e-06,
+      "loss": 0.4513,
+      "step": 708
+    },
+    {
+      "epoch": 0.16448207864516878,
+      "grad_norm": 30.74207575332384,
+      "learning_rate": 2e-06,
+      "loss": 0.3825,
+      "step": 709
+    },
+    {
+      "epoch": 0.16471407029346943,
+      "grad_norm": 40.584780969108955,
+      "learning_rate": 2e-06,
+      "loss": 0.4184,
+      "step": 710
+    },
+    {
+      "epoch": 0.16494606194177008,
+      "grad_norm": 47.0752374589824,
+      "learning_rate": 2e-06,
+      "loss": 0.3444,
+      "step": 711
+    },
+    {
+      "epoch": 0.16517805359007076,
+      "grad_norm": 24.05198512196974,
+      "learning_rate": 2e-06,
+      "loss": 0.2746,
+      "step": 712
+    },
+    {
+      "epoch": 0.1654100452383714,
+      "grad_norm": 20.96193342212991,
+      "learning_rate": 2e-06,
+      "loss": 0.3844,
+      "step": 713
+    },
+    {
+      "epoch": 0.1656420368866721,
+      "grad_norm": 26.844108345388346,
+      "learning_rate": 2e-06,
+      "loss": 0.3303,
+      "step": 714
+    },
+    {
+      "epoch": 0.16587402853497274,
+      "grad_norm": 18.361559812521087,
+      "learning_rate": 2e-06,
+      "loss": 0.2599,
+      "step": 715
+    },
+    {
+      "epoch": 0.1661060201832734,
+      "grad_norm": 10.334556071049587,
+      "learning_rate": 2e-06,
+      "loss": 0.2415,
+      "step": 716
+    },
+    {
+      "epoch": 0.16633801183157407,
+      "grad_norm": 17.466059203458997,
+      "learning_rate": 2e-06,
+      "loss": 0.2934,
+      "step": 717
+    },
+    {
+      "epoch": 0.16657000347987472,
+      "grad_norm": 13.792496069247843,
+      "learning_rate": 2e-06,
+      "loss": 0.2913,
+      "step": 718
+    },
+    {
+      "epoch": 0.1668019951281754,
+      "grad_norm": 27.344398206578028,
+      "learning_rate": 2e-06,
+      "loss": 0.5,
+      "step": 719
+    },
+    {
+      "epoch": 0.16703398677647605,
+      "grad_norm": 16.570069610376084,
+      "learning_rate": 2e-06,
+      "loss": 0.3282,
+      "step": 720
+    },
+    {
+      "epoch": 0.1672659784247767,
+      "grad_norm": 12.517016004539395,
+      "learning_rate": 2e-06,
+      "loss": 0.316,
+      "step": 721
+    },
+    {
+      "epoch": 0.16749797007307737,
+      "grad_norm": 9.454686300920637,
+      "learning_rate": 2e-06,
+      "loss": 0.2546,
+      "step": 722
+    },
+    {
+      "epoch": 0.16772996172137802,
+      "grad_norm": 17.013538405587067,
+      "learning_rate": 2e-06,
+      "loss": 0.3796,
+      "step": 723
+    },
+    {
+      "epoch": 0.1679619533696787,
+      "grad_norm": 24.06500993181185,
+      "learning_rate": 2e-06,
+      "loss": 0.4782,
+      "step": 724
+    },
+    {
+      "epoch": 0.16819394501797935,
+      "grad_norm": 21.754943523584142,
+      "learning_rate": 2e-06,
+      "loss": 0.4722,
+      "step": 725
+    },
+    {
+      "epoch": 0.16842593666628,
+      "grad_norm": 19.33220164628588,
+      "learning_rate": 2e-06,
+      "loss": 0.3475,
+      "step": 726
+    },
+    {
+      "epoch": 0.16865792831458068,
+      "grad_norm": 14.11612192602318,
+      "learning_rate": 2e-06,
+      "loss": 0.3717,
+      "step": 727
+    },
+    {
+      "epoch": 0.16888991996288133,
+      "grad_norm": 20.07653914337054,
+      "learning_rate": 2e-06,
+      "loss": 0.3061,
+      "step": 728
+    },
+    {
+      "epoch": 0.169121911611182,
+      "grad_norm": 15.694126220394503,
+      "learning_rate": 2e-06,
+      "loss": 0.3679,
+      "step": 729
+    },
+    {
+      "epoch": 0.16935390325948266,
+      "grad_norm": 7.597244637446782,
+      "learning_rate": 2e-06,
+      "loss": 0.3023,
+      "step": 730
+    },
+    {
+      "epoch": 0.1695858949077833,
+      "grad_norm": 9.960347483549528,
+      "learning_rate": 2e-06,
+      "loss": 0.3413,
+      "step": 731
+    },
+    {
+      "epoch": 0.16981788655608399,
+      "grad_norm": 13.00232896611134,
+      "learning_rate": 2e-06,
+      "loss": 0.3,
+      "step": 732
+    },
+    {
+      "epoch": 0.17004987820438464,
+      "grad_norm": 18.64285936948962,
+      "learning_rate": 2e-06,
+      "loss": 0.3768,
+      "step": 733
+    },
+    {
+      "epoch": 0.1702818698526853,
+      "grad_norm": 19.56053782005133,
+      "learning_rate": 2e-06,
+      "loss": 0.3985,
+      "step": 734
+    },
+    {
+      "epoch": 0.17051386150098596,
+      "grad_norm": 10.622675208141215,
+      "learning_rate": 2e-06,
+      "loss": 0.3562,
+      "step": 735
+    },
+    {
+      "epoch": 0.1707458531492866,
+      "grad_norm": 15.804895543467035,
+      "learning_rate": 2e-06,
+      "loss": 0.3077,
+      "step": 736
+    },
+    {
+      "epoch": 0.1709778447975873,
+      "grad_norm": 18.447008280828154,
+      "learning_rate": 2e-06,
+      "loss": 0.3354,
+      "step": 737
+    },
+    {
+      "epoch": 0.17120983644588794,
+      "grad_norm": 14.621557018343863,
+      "learning_rate": 2e-06,
+      "loss": 0.459,
+      "step": 738
+    },
+    {
+      "epoch": 0.17144182809418862,
+      "grad_norm": 8.763421231576094,
+      "learning_rate": 2e-06,
+      "loss": 0.2067,
+      "step": 739
+    },
+    {
+      "epoch": 0.17167381974248927,
+      "grad_norm": 21.564285979581147,
+      "learning_rate": 2e-06,
+      "loss": 0.3686,
+      "step": 740
+    },
+    {
+      "epoch": 0.17190581139078992,
+      "grad_norm": 15.462860742192579,
+      "learning_rate": 2e-06,
+      "loss": 0.2838,
+      "step": 741
+    },
+    {
+      "epoch": 0.1721378030390906,
+      "grad_norm": 18.866472307797476,
+      "learning_rate": 2e-06,
+      "loss": 0.3532,
+      "step": 742
+    },
+    {
+      "epoch": 0.17236979468739125,
+      "grad_norm": 11.27689988835768,
+      "learning_rate": 2e-06,
+      "loss": 0.3262,
+      "step": 743
+    },
+    {
+      "epoch": 0.17260178633569193,
+      "grad_norm": 6.560862378907595,
+      "learning_rate": 2e-06,
+      "loss": 0.2374,
+      "step": 744
+    },
+    {
+      "epoch": 0.17283377798399258,
+      "grad_norm": 17.22733586413075,
+      "learning_rate": 2e-06,
+      "loss": 0.3586,
+      "step": 745
+    },
+    {
+      "epoch": 0.17306576963229323,
+      "grad_norm": 8.816541653445544,
+      "learning_rate": 2e-06,
+      "loss": 0.2539,
+      "step": 746
+    },
+    {
+      "epoch": 0.1732977612805939,
+      "grad_norm": 9.561143157495499,
+      "learning_rate": 2e-06,
+      "loss": 0.3186,
+      "step": 747
+    },
+    {
+      "epoch": 0.17352975292889455,
+      "grad_norm": 12.445921344090433,
+      "learning_rate": 2e-06,
+      "loss": 0.3107,
+      "step": 748
+    },
+    {
+      "epoch": 0.17376174457719523,
+      "grad_norm": 14.75541877744063,
+      "learning_rate": 2e-06,
+      "loss": 0.3665,
+      "step": 749
+    },
+    {
+      "epoch": 0.17399373622549588,
+      "grad_norm": 11.932681911379351,
+      "learning_rate": 2e-06,
+      "loss": 0.305,
+      "step": 750
+    },
+    {
+      "epoch": 0.17422572787379653,
+      "grad_norm": 13.778515110830838,
+      "learning_rate": 2e-06,
+      "loss": 0.2846,
+      "step": 751
+    },
+    {
+      "epoch": 0.1744577195220972,
+      "grad_norm": 17.98604309887815,
+      "learning_rate": 2e-06,
+      "loss": 0.3341,
+      "step": 752
+    },
+    {
+      "epoch": 0.17468971117039786,
+      "grad_norm": 17.334206140348677,
+      "learning_rate": 2e-06,
+      "loss": 0.2994,
+      "step": 753
+    },
+    {
+      "epoch": 0.17492170281869854,
+      "grad_norm": 17.67618987598563,
+      "learning_rate": 2e-06,
+      "loss": 0.4629,
+      "step": 754
+    },
+    {
+      "epoch": 0.1751536944669992,
+      "grad_norm": 13.136167347637274,
+      "learning_rate": 2e-06,
+      "loss": 0.3304,
+      "step": 755
+    },
+    {
+      "epoch": 0.17538568611529984,
+      "grad_norm": 8.952224830050373,
+      "learning_rate": 2e-06,
+      "loss": 0.3626,
+      "step": 756
+    },
+    {
+      "epoch": 0.17561767776360052,
+      "grad_norm": 11.193654263857521,
+      "learning_rate": 2e-06,
+      "loss": 0.2494,
+      "step": 757
+    },
+    {
+      "epoch": 0.17584966941190117,
+      "grad_norm": 22.320651074813423,
+      "learning_rate": 2e-06,
+      "loss": 0.3289,
+      "step": 758
+    },
+    {
+      "epoch": 0.17608166106020184,
+      "grad_norm": 17.535712646014808,
+      "learning_rate": 2e-06,
+      "loss": 0.4499,
+      "step": 759
+    },
+    {
+      "epoch": 0.1763136527085025,
+      "grad_norm": 12.733734517226111,
+      "learning_rate": 2e-06,
+      "loss": 0.3214,
+      "step": 760
+    },
+    {
+      "epoch": 0.17654564435680314,
+      "grad_norm": 13.353201137951201,
+      "learning_rate": 2e-06,
+      "loss": 0.3699,
+      "step": 761
+    },
+    {
+      "epoch": 0.17677763600510382,
+      "grad_norm": 9.201780284194042,
+      "learning_rate": 2e-06,
+      "loss": 0.3144,
+      "step": 762
+    },
+    {
+      "epoch": 0.17700962765340447,
+      "grad_norm": 9.586095433749264,
+      "learning_rate": 2e-06,
+      "loss": 0.2671,
+      "step": 763
+    },
+    {
+      "epoch": 0.17724161930170515,
+      "grad_norm": 18.8866400319037,
+      "learning_rate": 2e-06,
+      "loss": 0.3327,
+      "step": 764
+    },
+    {
+      "epoch": 0.1774736109500058,
+      "grad_norm": 16.97686477233965,
+      "learning_rate": 2e-06,
+      "loss": 0.3186,
+      "step": 765
+    },
+    {
+      "epoch": 0.17770560259830645,
+      "grad_norm": 23.109981923087553,
+      "learning_rate": 2e-06,
+      "loss": 0.3106,
+      "step": 766
+    },
+    {
+      "epoch": 0.17793759424660713,
+      "grad_norm": 16.775430605152785,
+      "learning_rate": 2e-06,
+      "loss": 0.3196,
+      "step": 767
+    },
+    {
+      "epoch": 0.17816958589490778,
+      "grad_norm": 11.87197374426931,
+      "learning_rate": 2e-06,
+      "loss": 0.2493,
+      "step": 768
+    },
+    {
+      "epoch": 0.17840157754320846,
+      "grad_norm": 18.34787640158497,
+      "learning_rate": 2e-06,
+      "loss": 0.4318,
+      "step": 769
+    },
+    {
+      "epoch": 0.1786335691915091,
+      "grad_norm": 12.462233834921067,
+      "learning_rate": 2e-06,
+      "loss": 0.3509,
+      "step": 770
+    },
+    {
+      "epoch": 0.17886556083980976,
+      "grad_norm": 9.523701984608357,
+      "learning_rate": 2e-06,
+      "loss": 0.2247,
+      "step": 771
+    },
+    {
+      "epoch": 0.17909755248811043,
+      "grad_norm": 20.77996685844607,
+      "learning_rate": 2e-06,
+      "loss": 0.3783,
+      "step": 772
+    },
+    {
+      "epoch": 0.17932954413641108,
+      "grad_norm": 14.594333033140241,
+      "learning_rate": 2e-06,
+      "loss": 0.2588,
+      "step": 773
+    },
+    {
+      "epoch": 0.17956153578471176,
+      "grad_norm": 11.82242238113082,
+      "learning_rate": 2e-06,
+      "loss": 0.2729,
+      "step": 774
+    },
+    {
+      "epoch": 0.1797935274330124,
+      "grad_norm": 13.30174298015603,
+      "learning_rate": 2e-06,
+      "loss": 0.3338,
+      "step": 775
+    },
+    {
+      "epoch": 0.18002551908131306,
+      "grad_norm": 21.963362480858713,
+      "learning_rate": 2e-06,
+      "loss": 0.4504,
+      "step": 776
+    },
+    {
+      "epoch": 0.18025751072961374,
+      "grad_norm": 16.089942933076735,
+      "learning_rate": 2e-06,
+      "loss": 0.2697,
+      "step": 777
+    },
+    {
+      "epoch": 0.1804895023779144,
+      "grad_norm": 22.34317850213985,
+      "learning_rate": 2e-06,
+      "loss": 0.3904,
+      "step": 778
+    },
+    {
+      "epoch": 0.18072149402621507,
+      "grad_norm": 23.94242658386868,
+      "learning_rate": 2e-06,
+      "loss": 0.38,
+      "step": 779
+    },
+    {
+      "epoch": 0.18095348567451572,
+      "grad_norm": 12.818302873757448,
+      "learning_rate": 2e-06,
+      "loss": 0.3119,
+      "step": 780
+    },
+    {
+      "epoch": 0.18118547732281637,
+      "grad_norm": 15.137628022957086,
+      "learning_rate": 2e-06,
+      "loss": 0.3781,
+      "step": 781
+    },
+    {
+      "epoch": 0.18141746897111705,
+      "grad_norm": 24.94966829807091,
+      "learning_rate": 2e-06,
+      "loss": 0.3605,
+      "step": 782
+    },
+    {
+      "epoch": 0.1816494606194177,
+      "grad_norm": 13.102831529880358,
+      "learning_rate": 2e-06,
+      "loss": 0.2806,
+      "step": 783
+    },
+    {
+      "epoch": 0.18188145226771837,
+      "grad_norm": 19.645424296337186,
+      "learning_rate": 2e-06,
+      "loss": 0.2957,
+      "step": 784
+    },
+    {
+      "epoch": 0.18211344391601902,
+      "grad_norm": 23.164457029140216,
+      "learning_rate": 2e-06,
+      "loss": 0.3374,
+      "step": 785
+    },
+    {
+      "epoch": 0.18234543556431967,
+      "grad_norm": 20.373224626818722,
+      "learning_rate": 2e-06,
+      "loss": 0.3663,
+      "step": 786
+    },
+    {
+      "epoch": 0.18257742721262035,
+      "grad_norm": 22.147470638665496,
+      "learning_rate": 2e-06,
+      "loss": 0.4535,
+      "step": 787
+    },
+    {
+      "epoch": 0.182809418860921,
+      "grad_norm": 15.781492413550604,
+      "learning_rate": 2e-06,
+      "loss": 0.3429,
+      "step": 788
+    },
+    {
+      "epoch": 0.18304141050922168,
+      "grad_norm": 14.581736302069048,
+      "learning_rate": 2e-06,
+      "loss": 0.3106,
+      "step": 789
+    },
+    {
+      "epoch": 0.18327340215752233,
+      "grad_norm": 20.75205079068296,
+      "learning_rate": 2e-06,
+      "loss": 0.2501,
+      "step": 790
+    },
+    {
+      "epoch": 0.18350539380582298,
+      "grad_norm": 11.99955208209882,
+      "learning_rate": 2e-06,
+      "loss": 0.3079,
+      "step": 791
+    },
+    {
+      "epoch": 0.18373738545412366,
+      "grad_norm": 23.105384860674214,
+      "learning_rate": 2e-06,
+      "loss": 0.3447,
+      "step": 792
+    },
+    {
+      "epoch": 0.1839693771024243,
+      "grad_norm": 15.03266460995024,
+      "learning_rate": 2e-06,
+      "loss": 0.3274,
+      "step": 793
+    },
+    {
+      "epoch": 0.184201368750725,
+      "grad_norm": 16.42107190060086,
+      "learning_rate": 2e-06,
+      "loss": 0.3123,
+      "step": 794
+    },
+    {
+      "epoch": 0.18443336039902564,
+      "grad_norm": 10.80224803928146,
+      "learning_rate": 2e-06,
+      "loss": 0.2789,
+      "step": 795
+    },
+    {
+      "epoch": 0.1846653520473263,
+      "grad_norm": 12.57701768453657,
+      "learning_rate": 2e-06,
+      "loss": 0.3366,
+      "step": 796
+    },
+    {
+      "epoch": 0.18489734369562696,
+      "grad_norm": 18.119585754482213,
+      "learning_rate": 2e-06,
+      "loss": 0.3537,
+      "step": 797
+    },
+    {
+      "epoch": 0.18512933534392761,
+      "grad_norm": 13.742042613570336,
+      "learning_rate": 2e-06,
+      "loss": 0.3081,
+      "step": 798
+    },
+    {
+      "epoch": 0.1853613269922283,
+      "grad_norm": 12.840783424332322,
+      "learning_rate": 2e-06,
+      "loss": 0.2905,
+      "step": 799
+    },
+    {
+      "epoch": 0.18559331864052894,
+      "grad_norm": 22.663177940728435,
+      "learning_rate": 2e-06,
+      "loss": 0.3788,
+      "step": 800
+    },
+    {
+      "epoch": 0.1858253102888296,
+      "grad_norm": 9.773808862135011,
+      "learning_rate": 2e-06,
+      "loss": 0.3019,
+      "step": 801
+    },
+    {
+      "epoch": 0.18605730193713027,
+      "grad_norm": 13.74093840790118,
+      "learning_rate": 2e-06,
+      "loss": 0.3438,
+      "step": 802
+    },
+    {
+      "epoch": 0.18628929358543092,
+      "grad_norm": 10.157630882906007,
+      "learning_rate": 2e-06,
+      "loss": 0.3606,
+      "step": 803
+    },
+    {
+      "epoch": 0.1865212852337316,
+      "grad_norm": 19.94680174631867,
+      "learning_rate": 2e-06,
+      "loss": 0.3932,
+      "step": 804
+    },
+    {
+      "epoch": 0.18675327688203225,
+      "grad_norm": 19.62427920903281,
+      "learning_rate": 2e-06,
+      "loss": 0.411,
+      "step": 805
+    },
+    {
+      "epoch": 0.1869852685303329,
+      "grad_norm": 20.749400909089836,
+      "learning_rate": 2e-06,
+      "loss": 0.3252,
+      "step": 806
+    },
+    {
+      "epoch": 0.18721726017863358,
+      "grad_norm": 16.912302806170956,
+      "learning_rate": 2e-06,
+      "loss": 0.3226,
+      "step": 807
+    },
+    {
+      "epoch": 0.18744925182693423,
+      "grad_norm": 7.836093031268887,
+      "learning_rate": 2e-06,
+      "loss": 0.3174,
+      "step": 808
+    },
+    {
+      "epoch": 0.1876812434752349,
+      "grad_norm": 10.36121215740947,
+      "learning_rate": 2e-06,
+      "loss": 0.2763,
+      "step": 809
+    },
+    {
+      "epoch": 0.18791323512353555,
+      "grad_norm": 11.951105686627765,
+      "learning_rate": 2e-06,
+      "loss": 0.2768,
+      "step": 810
+    },
+    {
+      "epoch": 0.1881452267718362,
+      "grad_norm": 14.9237046497379,
+      "learning_rate": 2e-06,
+      "loss": 0.3761,
+      "step": 811
+    },
+    {
+      "epoch": 0.18837721842013688,
+      "grad_norm": 14.974947243668653,
+      "learning_rate": 2e-06,
+      "loss": 0.3652,
+      "step": 812
+    },
+    {
+      "epoch": 0.18860921006843753,
+      "grad_norm": 21.727343631884633,
+      "learning_rate": 2e-06,
+      "loss": 0.288,
+      "step": 813
+    },
+    {
+      "epoch": 0.1888412017167382,
+      "grad_norm": 20.573896161218798,
+      "learning_rate": 2e-06,
+      "loss": 0.4486,
+      "step": 814
+    },
+    {
+      "epoch": 0.18907319336503886,
+      "grad_norm": 16.25349741628844,
+      "learning_rate": 2e-06,
+      "loss": 0.2948,
+      "step": 815
+    },
+    {
+      "epoch": 0.1893051850133395,
+      "grad_norm": 26.182205854500186,
+      "learning_rate": 2e-06,
+      "loss": 0.4274,
+      "step": 816
+    },
+    {
+      "epoch": 0.1895371766616402,
+      "grad_norm": 17.041520476227614,
+      "learning_rate": 2e-06,
+      "loss": 0.2824,
+      "step": 817
+    },
+    {
+      "epoch": 0.18976916830994084,
+      "grad_norm": 12.614284283681387,
+      "learning_rate": 2e-06,
+      "loss": 0.3327,
+      "step": 818
+    },
+    {
+      "epoch": 0.19000115995824152,
+      "grad_norm": 21.138920185559407,
+      "learning_rate": 2e-06,
+      "loss": 0.2783,
+      "step": 819
+    },
+    {
+      "epoch": 0.19023315160654217,
+      "grad_norm": 21.829436439600233,
+      "learning_rate": 2e-06,
+      "loss": 0.4277,
+      "step": 820
+    },
+    {
+      "epoch": 0.19046514325484282,
+      "grad_norm": 13.577464451034475,
+      "learning_rate": 2e-06,
+      "loss": 0.3058,
+      "step": 821
+    },
+    {
+      "epoch": 0.1906971349031435,
+      "grad_norm": 20.013660716589733,
+      "learning_rate": 2e-06,
+      "loss": 0.3382,
+      "step": 822
+    },
+    {
+      "epoch": 0.19092912655144414,
+      "grad_norm": 22.59776772609933,
+      "learning_rate": 2e-06,
+      "loss": 0.365,
+      "step": 823
+    },
+    {
+      "epoch": 0.19116111819974482,
+      "grad_norm": 14.284326073429842,
+      "learning_rate": 2e-06,
+      "loss": 0.2696,
+      "step": 824
+    },
+    {
+      "epoch": 0.19139310984804547,
+      "grad_norm": 21.38410672676771,
+      "learning_rate": 2e-06,
+      "loss": 0.4893,
+      "step": 825
+    },
+    {
+      "epoch": 0.19162510149634612,
+      "grad_norm": 15.799118469111308,
+      "learning_rate": 2e-06,
+      "loss": 0.3119,
+      "step": 826
+    },
+    {
+      "epoch": 0.1918570931446468,
+      "grad_norm": 8.29295418290862,
+      "learning_rate": 2e-06,
+      "loss": 0.353,
+      "step": 827
+    },
+    {
+      "epoch": 0.19208908479294745,
+      "grad_norm": 11.550631260812212,
+      "learning_rate": 2e-06,
+      "loss": 0.2806,
+      "step": 828
+    },
+    {
+      "epoch": 0.19232107644124813,
+      "grad_norm": 15.316290042899775,
+      "learning_rate": 2e-06,
+      "loss": 0.3393,
+      "step": 829
+    },
+    {
+      "epoch": 0.19255306808954878,
+      "grad_norm": 19.489646646108323,
+      "learning_rate": 2e-06,
+      "loss": 0.281,
+      "step": 830
+    },
+    {
+      "epoch": 0.19278505973784943,
+      "grad_norm": 9.929271193196863,
+      "learning_rate": 2e-06,
+      "loss": 0.2407,
+      "step": 831
+    },
+    {
+      "epoch": 0.1930170513861501,
+      "grad_norm": 14.739191818389179,
+      "learning_rate": 2e-06,
+      "loss": 0.4059,
+      "step": 832
+    },
+    {
+      "epoch": 0.19324904303445076,
+      "grad_norm": 15.977615302695332,
+      "learning_rate": 2e-06,
+      "loss": 0.3081,
+      "step": 833
+    },
+    {
+      "epoch": 0.1934810346827514,
+      "grad_norm": 17.21606409393626,
+      "learning_rate": 2e-06,
+      "loss": 0.2522,
+      "step": 834
+    },
+    {
+      "epoch": 0.19371302633105209,
+      "grad_norm": 10.464523778961931,
+      "learning_rate": 2e-06,
+      "loss": 0.3314,
+      "step": 835
+    },
+    {
+      "epoch": 0.19394501797935274,
+      "grad_norm": 21.76286939880773,
+      "learning_rate": 2e-06,
+      "loss": 0.2913,
+      "step": 836
+    },
+    {
+      "epoch": 0.1941770096276534,
+      "grad_norm": 19.049786854010637,
+      "learning_rate": 2e-06,
+      "loss": 0.3758,
+      "step": 837
+    },
+    {
+      "epoch": 0.19440900127595406,
+      "grad_norm": 11.044066209039185,
+      "learning_rate": 2e-06,
+      "loss": 0.2418,
+      "step": 838
+    },
+    {
+      "epoch": 0.1946409929242547,
+      "grad_norm": 8.866011503268515,
+      "learning_rate": 2e-06,
+      "loss": 0.3765,
+      "step": 839
+    },
+    {
+      "epoch": 0.1948729845725554,
+      "grad_norm": 13.395474628389369,
+      "learning_rate": 2e-06,
+      "loss": 0.2689,
+      "step": 840
+    },
+    {
+      "epoch": 0.19510497622085604,
+      "grad_norm": 23.036725052626185,
+      "learning_rate": 2e-06,
+      "loss": 0.32,
+      "step": 841
+    },
+    {
+      "epoch": 0.19533696786915672,
+      "grad_norm": 15.691738893928159,
+      "learning_rate": 2e-06,
+      "loss": 0.4299,
+      "step": 842
+    },
+    {
+      "epoch": 0.19556895951745737,
+      "grad_norm": 18.60561924491579,
+      "learning_rate": 2e-06,
+      "loss": 0.3365,
+      "step": 843
+    },
+    {
+      "epoch": 0.19580095116575802,
+      "grad_norm": 23.651757254556525,
+      "learning_rate": 2e-06,
+      "loss": 0.3196,
+      "step": 844
+    },
+    {
+      "epoch": 0.1960329428140587,
+      "grad_norm": 11.371947892209338,
+      "learning_rate": 2e-06,
+      "loss": 0.288,
+      "step": 845
+    },
+    {
+      "epoch": 0.19626493446235935,
+      "grad_norm": 10.090779614062537,
+      "learning_rate": 2e-06,
+      "loss": 0.3127,
+      "step": 846
+    },
+    {
+      "epoch": 0.19649692611066003,
+      "grad_norm": 17.306165471527102,
+      "learning_rate": 2e-06,
+      "loss": 0.3371,
+      "step": 847
+    },
+    {
+      "epoch": 0.19672891775896068,
+      "grad_norm": 17.408658612319552,
+      "learning_rate": 2e-06,
+      "loss": 0.3374,
+      "step": 848
+    },
+    {
+      "epoch": 0.19696090940726133,
+      "grad_norm": 31.10483274697328,
+      "learning_rate": 2e-06,
+      "loss": 0.419,
+      "step": 849
+    },
+    {
+      "epoch": 0.197192901055562,
+      "grad_norm": 13.739629867076612,
+      "learning_rate": 2e-06,
+      "loss": 0.4226,
+      "step": 850
+    },
+    {
+      "epoch": 0.19742489270386265,
+      "grad_norm": 18.10650904524103,
+      "learning_rate": 2e-06,
+      "loss": 0.4111,
+      "step": 851
+    },
+    {
+      "epoch": 0.19765688435216333,
+      "grad_norm": 9.724254659000593,
+      "learning_rate": 2e-06,
+      "loss": 0.2907,
+      "step": 852
+    },
+    {
+      "epoch": 0.19788887600046398,
+      "grad_norm": 11.70840979887669,
+      "learning_rate": 2e-06,
+      "loss": 0.3338,
+      "step": 853
+    },
+    {
+      "epoch": 0.19812086764876463,
+      "grad_norm": 18.223636670611263,
+      "learning_rate": 2e-06,
+      "loss": 0.3297,
+      "step": 854
+    },
+    {
+      "epoch": 0.1983528592970653,
+      "grad_norm": 19.969087791024695,
+      "learning_rate": 2e-06,
+      "loss": 0.3981,
+      "step": 855
+    },
+    {
+      "epoch": 0.19858485094536596,
+      "grad_norm": 24.668935395200606,
+      "learning_rate": 2e-06,
+      "loss": 0.4325,
+      "step": 856
+    },
+    {
+      "epoch": 0.19881684259366664,
+      "grad_norm": 8.569207714004078,
+      "learning_rate": 2e-06,
+      "loss": 0.2341,
+      "step": 857
+    },
+    {
+      "epoch": 0.1990488342419673,
+      "grad_norm": 18.141428991522456,
+      "learning_rate": 2e-06,
+      "loss": 0.3797,
+      "step": 858
+    },
+    {
+      "epoch": 0.19928082589026794,
+      "grad_norm": 13.83844732174976,
+      "learning_rate": 2e-06,
+      "loss": 0.3004,
+      "step": 859
+    },
+    {
+      "epoch": 0.19951281753856862,
+      "grad_norm": 7.570402229629171,
+      "learning_rate": 2e-06,
+      "loss": 0.2659,
+      "step": 860
+    },
+    {
+      "epoch": 0.19974480918686927,
+      "grad_norm": 19.151243048179083,
+      "learning_rate": 2e-06,
+      "loss": 0.3457,
+      "step": 861
+    },
+    {
+      "epoch": 0.19997680083516994,
+      "grad_norm": 9.023686026024441,
+      "learning_rate": 2e-06,
+      "loss": 0.2003,
+      "step": 862
+    },
+    {
+      "epoch": 0.2002087924834706,
+      "grad_norm": 10.652610316585804,
+      "learning_rate": 2e-06,
+      "loss": 0.2778,
+      "step": 863
+    },
+    {
+      "epoch": 0.20044078413177124,
+      "grad_norm": 21.104189125544075,
+      "learning_rate": 2e-06,
+      "loss": 0.3277,
+      "step": 864
+    },
+    {
+      "epoch": 0.20067277578007192,
+      "grad_norm": 17.871109485100916,
+      "learning_rate": 2e-06,
+      "loss": 0.3817,
+      "step": 865
+    },
+    {
+      "epoch": 0.20090476742837257,
+      "grad_norm": 10.595302363200277,
+      "learning_rate": 2e-06,
+      "loss": 0.3649,
+      "step": 866
+    },
+    {
+      "epoch": 0.20113675907667325,
+      "grad_norm": 18.58896905743985,
+      "learning_rate": 2e-06,
+      "loss": 0.2794,
+      "step": 867
+    },
+    {
+      "epoch": 0.2013687507249739,
+      "grad_norm": 8.56618420358794,
+      "learning_rate": 2e-06,
+      "loss": 0.332,
+      "step": 868
+    },
+    {
+      "epoch": 0.20160074237327455,
+      "grad_norm": 7.871954271104144,
+      "learning_rate": 2e-06,
+      "loss": 0.3256,
+      "step": 869
+    },
+    {
+      "epoch": 0.20183273402157523,
+      "grad_norm": 14.482295828257122,
+      "learning_rate": 2e-06,
+      "loss": 0.3513,
+      "step": 870
+    },
+    {
+      "epoch": 0.20206472566987588,
+      "grad_norm": 15.59436837133702,
+      "learning_rate": 2e-06,
+      "loss": 0.4059,
+      "step": 871
+    },
+    {
+      "epoch": 0.20229671731817656,
+      "grad_norm": 11.695894552841496,
+      "learning_rate": 2e-06,
+      "loss": 0.271,
+      "step": 872
+    },
+    {
+      "epoch": 0.2025287089664772,
+      "grad_norm": 9.279739517381302,
+      "learning_rate": 2e-06,
+      "loss": 0.3572,
+      "step": 873
+    },
+    {
+      "epoch": 0.20276070061477786,
+      "grad_norm": 9.436327949040212,
+      "learning_rate": 2e-06,
+      "loss": 0.3279,
+      "step": 874
+    },
+    {
+      "epoch": 0.20299269226307853,
+      "grad_norm": 14.37356404419197,
+      "learning_rate": 2e-06,
+      "loss": 0.3537,
+      "step": 875
+    },
+    {
+      "epoch": 0.20322468391137918,
+      "grad_norm": 10.460306167068557,
+      "learning_rate": 2e-06,
+      "loss": 0.3193,
+      "step": 876
+    },
+    {
+      "epoch": 0.20345667555967986,
+      "grad_norm": 9.957388619811127,
+      "learning_rate": 2e-06,
+      "loss": 0.2464,
+      "step": 877
+    },
+    {
+      "epoch": 0.2036886672079805,
+      "grad_norm": 14.043925791536124,
+      "learning_rate": 2e-06,
+      "loss": 0.3229,
+      "step": 878
+    },
+    {
+      "epoch": 0.20392065885628116,
+      "grad_norm": 10.916638475637072,
+      "learning_rate": 2e-06,
+      "loss": 0.3094,
+      "step": 879
+    },
+    {
+      "epoch": 0.20415265050458184,
+      "grad_norm": 14.207690648062343,
+      "learning_rate": 2e-06,
+      "loss": 0.2728,
+      "step": 880
+    },
+    {
+      "epoch": 0.2043846421528825,
+      "grad_norm": 29.774947857721138,
+      "learning_rate": 2e-06,
+      "loss": 0.4734,
+      "step": 881
+    },
+    {
+      "epoch": 0.20461663380118317,
+      "grad_norm": 12.886953112164875,
+      "learning_rate": 2e-06,
+      "loss": 0.2808,
+      "step": 882
+    },
+    {
+      "epoch": 0.20484862544948382,
+      "grad_norm": 54.83652013490153,
+      "learning_rate": 2e-06,
+      "loss": 0.4345,
+      "step": 883
+    },
+    {
+      "epoch": 0.20508061709778447,
+      "grad_norm": 13.836004473480973,
+      "learning_rate": 2e-06,
+      "loss": 0.2673,
+      "step": 884
+    },
+    {
+      "epoch": 0.20531260874608515,
+      "grad_norm": 16.236716597033208,
+      "learning_rate": 2e-06,
+      "loss": 0.3773,
+      "step": 885
+    },
+    {
+      "epoch": 0.2055446003943858,
+      "grad_norm": 9.584331646818612,
+      "learning_rate": 2e-06,
+      "loss": 0.2832,
+      "step": 886
+    },
+    {
+      "epoch": 0.20577659204268647,
+      "grad_norm": 14.986402451734643,
+      "learning_rate": 2e-06,
+      "loss": 0.3277,
+      "step": 887
+    },
+    {
+      "epoch": 0.20600858369098712,
+      "grad_norm": 17.0253000035854,
+      "learning_rate": 2e-06,
+      "loss": 0.3905,
+      "step": 888
+    },
+    {
+      "epoch": 0.20624057533928777,
+      "grad_norm": 16.450631095420082,
+      "learning_rate": 2e-06,
+      "loss": 0.4094,
+      "step": 889
+    },
+    {
+      "epoch": 0.20647256698758845,
+      "grad_norm": 18.750612286588492,
+      "learning_rate": 2e-06,
+      "loss": 0.3131,
+      "step": 890
+    },
+    {
+      "epoch": 0.2067045586358891,
+      "grad_norm": 18.73328993359991,
+      "learning_rate": 2e-06,
+      "loss": 0.3061,
+      "step": 891
+    },
+    {
+      "epoch": 0.20693655028418978,
+      "grad_norm": 12.70523808816334,
+      "learning_rate": 2e-06,
+      "loss": 0.282,
+      "step": 892
+    },
+    {
+      "epoch": 0.20716854193249043,
+      "grad_norm": 19.8824827414663,
+      "learning_rate": 2e-06,
+      "loss": 0.2741,
+      "step": 893
+    },
+    {
+      "epoch": 0.20740053358079108,
+      "grad_norm": 17.324392469861575,
+      "learning_rate": 2e-06,
+      "loss": 0.2701,
+      "step": 894
+    },
+    {
+      "epoch": 0.20763252522909176,
+      "grad_norm": 13.78937251217679,
+      "learning_rate": 2e-06,
+      "loss": 0.328,
+      "step": 895
+    },
+    {
+      "epoch": 0.2078645168773924,
+      "grad_norm": 16.027091968408882,
+      "learning_rate": 2e-06,
+      "loss": 0.3655,
+      "step": 896
+    },
+    {
+      "epoch": 0.20809650852569309,
+      "grad_norm": 7.60913786100639,
+      "learning_rate": 2e-06,
+      "loss": 0.1704,
+      "step": 897
+    },
+    {
+      "epoch": 0.20832850017399374,
+      "grad_norm": 28.677874376982054,
+      "learning_rate": 2e-06,
+      "loss": 0.4329,
+      "step": 898
+    },
+    {
+      "epoch": 0.20856049182229439,
+      "grad_norm": 8.221011139573621,
+      "learning_rate": 2e-06,
+      "loss": 0.2507,
+      "step": 899
+    },
+    {
+      "epoch": 0.20879248347059506,
+      "grad_norm": 11.4744661875519,
+      "learning_rate": 2e-06,
+      "loss": 0.3259,
+      "step": 900
+    },
+    {
+      "epoch": 0.20902447511889571,
+      "grad_norm": 15.294883355532624,
+      "learning_rate": 2e-06,
+      "loss": 0.3727,
+      "step": 901
+    },
+    {
+      "epoch": 0.2092564667671964,
+      "grad_norm": 19.055402797561403,
+      "learning_rate": 2e-06,
+      "loss": 0.303,
+      "step": 902
+    },
+    {
+      "epoch": 0.20948845841549704,
+      "grad_norm": 23.698316035534873,
+      "learning_rate": 2e-06,
+      "loss": 0.3667,
+      "step": 903
+    },
+    {
+      "epoch": 0.2097204500637977,
+      "grad_norm": 22.8054584811822,
+      "learning_rate": 2e-06,
+      "loss": 0.3585,
+      "step": 904
+    },
+    {
+      "epoch": 0.20995244171209837,
+      "grad_norm": 20.083038658083378,
+      "learning_rate": 2e-06,
+      "loss": 0.3192,
+      "step": 905
+    },
+    {
+      "epoch": 0.21018443336039902,
+      "grad_norm": 22.584586110449916,
+      "learning_rate": 2e-06,
+      "loss": 0.4078,
+      "step": 906
+    },
+    {
+      "epoch": 0.2104164250086997,
+      "grad_norm": 15.037422485453048,
+      "learning_rate": 2e-06,
+      "loss": 0.2889,
+      "step": 907
+    },
+    {
+      "epoch": 0.21064841665700035,
+      "grad_norm": 20.203868005835083,
+      "learning_rate": 2e-06,
+      "loss": 0.448,
+      "step": 908
+    },
+    {
+      "epoch": 0.210880408305301,
+      "grad_norm": 15.16465590647367,
+      "learning_rate": 2e-06,
+      "loss": 0.3637,
+      "step": 909
+    },
+    {
+      "epoch": 0.21111239995360168,
+      "grad_norm": 5.2078995129577805,
+      "learning_rate": 2e-06,
+      "loss": 0.1984,
+      "step": 910
+    },
+    {
+      "epoch": 0.21134439160190233,
+      "grad_norm": 17.776999304081468,
+      "learning_rate": 2e-06,
+      "loss": 0.4419,
+      "step": 911
+    },
+    {
+      "epoch": 0.211576383250203,
+      "grad_norm": 20.801335368698737,
+      "learning_rate": 2e-06,
+      "loss": 0.3527,
+      "step": 912
+    },
+    {
+      "epoch": 0.21180837489850365,
+      "grad_norm": 12.89318139470534,
+      "learning_rate": 2e-06,
+      "loss": 0.2925,
+      "step": 913
+    },
+    {
+      "epoch": 0.2120403665468043,
+      "grad_norm": 13.279124604676984,
+      "learning_rate": 2e-06,
+      "loss": 0.2692,
+      "step": 914
+    },
+    {
+      "epoch": 0.21227235819510498,
+      "grad_norm": 31.361102208165416,
+      "learning_rate": 2e-06,
+      "loss": 0.4164,
+      "step": 915
+    },
+    {
+      "epoch": 0.21250434984340563,
+      "grad_norm": 14.926166059180384,
+      "learning_rate": 2e-06,
+      "loss": 0.2819,
+      "step": 916
+    },
+    {
+      "epoch": 0.2127363414917063,
+      "grad_norm": 9.505395111229008,
+      "learning_rate": 2e-06,
+      "loss": 0.3,
+      "step": 917
+    },
+    {
+      "epoch": 0.21296833314000696,
+      "grad_norm": 18.19667325838775,
+      "learning_rate": 2e-06,
+      "loss": 0.2777,
+      "step": 918
+    },
+    {
+      "epoch": 0.2132003247883076,
+      "grad_norm": 34.244764055740525,
+      "learning_rate": 2e-06,
+      "loss": 0.2731,
+      "step": 919
+    },
+    {
+      "epoch": 0.2134323164366083,
+      "grad_norm": 13.827735412000134,
+      "learning_rate": 2e-06,
+      "loss": 0.312,
+      "step": 920
+    },
+    {
+      "epoch": 0.21366430808490894,
+      "grad_norm": 20.705363084487367,
+      "learning_rate": 2e-06,
+      "loss": 0.3044,
+      "step": 921
+    },
+    {
+      "epoch": 0.21389629973320962,
+      "grad_norm": 23.492053633413406,
+      "learning_rate": 2e-06,
+      "loss": 0.3717,
+      "step": 922
+    },
+    {
+      "epoch": 0.21412829138151027,
+      "grad_norm": 17.796209396317078,
+      "learning_rate": 2e-06,
+      "loss": 0.3501,
+      "step": 923
+    },
+    {
+      "epoch": 0.21436028302981092,
+      "grad_norm": 12.432070649047894,
+      "learning_rate": 2e-06,
+      "loss": 0.2967,
+      "step": 924
+    },
+    {
+      "epoch": 0.2145922746781116,
+      "grad_norm": 7.377036669491922,
+      "learning_rate": 2e-06,
+      "loss": 0.2317,
+      "step": 925
+    },
+    {
+      "epoch": 0.21482426632641224,
+      "grad_norm": 9.763829157513072,
+      "learning_rate": 2e-06,
+      "loss": 0.3069,
+      "step": 926
+    },
+    {
+      "epoch": 0.21505625797471292,
+      "grad_norm": 14.067087869685949,
+      "learning_rate": 2e-06,
+      "loss": 0.3761,
+      "step": 927
+    },
+    {
+      "epoch": 0.21528824962301357,
+      "grad_norm": 18.626123583576774,
+      "learning_rate": 2e-06,
+      "loss": 0.2611,
+      "step": 928
+    },
+    {
+      "epoch": 0.21552024127131422,
+      "grad_norm": 13.482003556824175,
+      "learning_rate": 2e-06,
+      "loss": 0.3761,
+      "step": 929
+    },
+    {
+      "epoch": 0.2157522329196149,
+      "grad_norm": 13.154857034380507,
+      "learning_rate": 2e-06,
+      "loss": 0.2722,
+      "step": 930
+    },
+    {
+      "epoch": 0.21598422456791555,
+      "grad_norm": 31.45359846464405,
+      "learning_rate": 2e-06,
+      "loss": 0.3893,
+      "step": 931
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 17.454676536343737,
+      "learning_rate": 2e-06,
+      "loss": 0.3429,
+      "step": 932
+    },
+    {
+      "epoch": 0.21644820786451688,
+      "grad_norm": 13.781514708400756,
+      "learning_rate": 2e-06,
+      "loss": 0.3244,
+      "step": 933
+    },
+    {
+      "epoch": 0.21668019951281753,
+      "grad_norm": 17.4126190084326,
+      "learning_rate": 2e-06,
+      "loss": 0.3813,
+      "step": 934
+    },
+    {
+      "epoch": 0.2169121911611182,
+      "grad_norm": 18.19944774510841,
+      "learning_rate": 2e-06,
+      "loss": 0.3342,
+      "step": 935
+    },
+    {
+      "epoch": 0.21714418280941886,
+      "grad_norm": 18.326170459536367,
+      "learning_rate": 2e-06,
+      "loss": 0.3103,
+      "step": 936
+    },
+    {
+      "epoch": 0.21737617445771953,
+      "grad_norm": 10.250425954162694,
+      "learning_rate": 2e-06,
+      "loss": 0.3748,
+      "step": 937
+    },
+    {
+      "epoch": 0.21760816610602018,
+      "grad_norm": 16.066598951701888,
+      "learning_rate": 2e-06,
+      "loss": 0.4068,
+      "step": 938
+    },
+    {
+      "epoch": 0.21784015775432083,
+      "grad_norm": 10.189922232130474,
+      "learning_rate": 2e-06,
+      "loss": 0.3328,
+      "step": 939
+    },
+    {
+      "epoch": 0.2180721494026215,
+      "grad_norm": 11.083488151842678,
+      "learning_rate": 2e-06,
+      "loss": 0.3137,
+      "step": 940
+    },
+    {
+      "epoch": 0.21830414105092216,
+      "grad_norm": 15.80352256233468,
+      "learning_rate": 2e-06,
+      "loss": 0.417,
+      "step": 941
+    },
+    {
+      "epoch": 0.21853613269922284,
+      "grad_norm": 10.397398946874906,
+      "learning_rate": 2e-06,
+      "loss": 0.258,
+      "step": 942
+    },
+    {
+      "epoch": 0.2187681243475235,
+      "grad_norm": 6.2699124866115605,
+      "learning_rate": 2e-06,
+      "loss": 0.2945,
+      "step": 943
+    },
+    {
+      "epoch": 0.21900011599582414,
+      "grad_norm": 23.96778950909715,
+      "learning_rate": 2e-06,
+      "loss": 0.3493,
+      "step": 944
+    },
+    {
+      "epoch": 0.21923210764412482,
+      "grad_norm": 12.741819491680975,
+      "learning_rate": 2e-06,
+      "loss": 0.3068,
+      "step": 945
+    },
+    {
+      "epoch": 0.21946409929242547,
+      "grad_norm": 22.214441345100916,
+      "learning_rate": 2e-06,
+      "loss": 0.306,
+      "step": 946
+    },
+    {
+      "epoch": 0.21969609094072615,
+      "grad_norm": 12.91661021107639,
+      "learning_rate": 2e-06,
+      "loss": 0.2215,
+      "step": 947
+    },
+    {
+      "epoch": 0.2199280825890268,
+      "grad_norm": 15.890385515844484,
+      "learning_rate": 2e-06,
+      "loss": 0.3135,
+      "step": 948
+    },
+    {
+      "epoch": 0.22016007423732745,
+      "grad_norm": 17.44548027799827,
+      "learning_rate": 2e-06,
+      "loss": 0.4428,
+      "step": 949
+    },
+    {
+      "epoch": 0.22039206588562812,
+      "grad_norm": 29.612385060726385,
+      "learning_rate": 2e-06,
+      "loss": 0.4595,
+      "step": 950
+    },
+    {
+      "epoch": 0.22062405753392877,
+      "grad_norm": 14.192682314656473,
+      "learning_rate": 2e-06,
+      "loss": 0.3099,
+      "step": 951
+    },
+    {
+      "epoch": 0.22085604918222945,
+      "grad_norm": 22.44367441783778,
+      "learning_rate": 2e-06,
+      "loss": 0.4008,
+      "step": 952
+    },
+    {
+      "epoch": 0.2210880408305301,
+      "grad_norm": 16.51988527879832,
+      "learning_rate": 2e-06,
+      "loss": 0.3694,
+      "step": 953
+    },
+    {
+      "epoch": 0.22132003247883075,
+      "grad_norm": 13.145552437440195,
+      "learning_rate": 2e-06,
+      "loss": 0.2829,
+      "step": 954
+    },
+    {
+      "epoch": 0.22155202412713143,
+      "grad_norm": 17.045976637894437,
+      "learning_rate": 2e-06,
+      "loss": 0.3363,
+      "step": 955
+    },
+    {
+      "epoch": 0.22178401577543208,
+      "grad_norm": 7.152755972453679,
+      "learning_rate": 2e-06,
+      "loss": 0.2048,
+      "step": 956
+    },
+    {
+      "epoch": 0.22201600742373276,
+      "grad_norm": 14.476364773777387,
+      "learning_rate": 2e-06,
+      "loss": 0.3139,
+      "step": 957
+    },
+    {
+      "epoch": 0.2222479990720334,
+      "grad_norm": 20.17448656913104,
+      "learning_rate": 2e-06,
+      "loss": 0.4056,
+      "step": 958
+    },
+    {
+      "epoch": 0.22247999072033406,
+      "grad_norm": 17.447954517156088,
+      "learning_rate": 2e-06,
+      "loss": 0.3249,
+      "step": 959
+    },
+    {
+      "epoch": 0.22271198236863474,
+      "grad_norm": 16.47574513287585,
+      "learning_rate": 2e-06,
+      "loss": 0.3578,
+      "step": 960
+    },
+    {
+      "epoch": 0.2229439740169354,
+      "grad_norm": 18.833569055388487,
+      "learning_rate": 2e-06,
+      "loss": 0.2698,
+      "step": 961
+    },
+    {
+      "epoch": 0.22317596566523606,
+      "grad_norm": 12.878810264947404,
+      "learning_rate": 2e-06,
+      "loss": 0.322,
+      "step": 962
+    },
+    {
+      "epoch": 0.22340795731353671,
+      "grad_norm": 9.754848029157317,
+      "learning_rate": 2e-06,
+      "loss": 0.3028,
+      "step": 963
+    },
+    {
+      "epoch": 0.22363994896183736,
+      "grad_norm": 10.458699501656957,
+      "learning_rate": 2e-06,
+      "loss": 0.2659,
+      "step": 964
+    },
+    {
+      "epoch": 0.22387194061013804,
+      "grad_norm": 9.213002126694942,
+      "learning_rate": 2e-06,
+      "loss": 0.2413,
+      "step": 965
+    },
+    {
+      "epoch": 0.2241039322584387,
+      "grad_norm": 22.77852040491952,
+      "learning_rate": 2e-06,
+      "loss": 0.3402,
+      "step": 966
+    },
+    {
+      "epoch": 0.22433592390673937,
+      "grad_norm": 19.97429495238275,
+      "learning_rate": 2e-06,
+      "loss": 0.3356,
+      "step": 967
+    },
+    {
+      "epoch": 0.22456791555504002,
+      "grad_norm": 22.725310459439754,
+      "learning_rate": 2e-06,
+      "loss": 0.3688,
+      "step": 968
+    },
+    {
+      "epoch": 0.22479990720334067,
+      "grad_norm": 15.880456669293403,
+      "learning_rate": 2e-06,
+      "loss": 0.2583,
+      "step": 969
+    },
+    {
+      "epoch": 0.22503189885164135,
+      "grad_norm": 12.562977799213726,
+      "learning_rate": 2e-06,
+      "loss": 0.3211,
+      "step": 970
+    },
+    {
+      "epoch": 0.225263890499942,
+      "grad_norm": 38.52385320084602,
+      "learning_rate": 2e-06,
+      "loss": 0.4099,
+      "step": 971
+    },
+    {
+      "epoch": 0.22549588214824268,
+      "grad_norm": 19.568782203246148,
+      "learning_rate": 2e-06,
+      "loss": 0.3392,
+      "step": 972
+    },
+    {
+      "epoch": 0.22572787379654333,
+      "grad_norm": 22.75231297936888,
+      "learning_rate": 2e-06,
+      "loss": 0.3981,
+      "step": 973
+    },
+    {
+      "epoch": 0.22595986544484398,
+      "grad_norm": 23.093165664624546,
+      "learning_rate": 2e-06,
+      "loss": 0.4342,
+      "step": 974
+    },
+    {
+      "epoch": 0.22619185709314465,
+      "grad_norm": 21.85862109103549,
+      "learning_rate": 2e-06,
+      "loss": 0.3366,
+      "step": 975
+    },
+    {
+      "epoch": 0.2264238487414453,
+      "grad_norm": 15.783198583306087,
+      "learning_rate": 2e-06,
+      "loss": 0.3545,
+      "step": 976
+    },
+    {
+      "epoch": 0.22665584038974598,
+      "grad_norm": 20.00115524370116,
+      "learning_rate": 2e-06,
+      "loss": 0.419,
+      "step": 977
+    },
+    {
+      "epoch": 0.22688783203804663,
+      "grad_norm": 17.13575397262277,
+      "learning_rate": 2e-06,
+      "loss": 0.4164,
+      "step": 978
+    },
+    {
+      "epoch": 0.22711982368634728,
+      "grad_norm": 20.66236442506612,
+      "learning_rate": 2e-06,
+      "loss": 0.3921,
+      "step": 979
+    },
+    {
+      "epoch": 0.22735181533464796,
+      "grad_norm": 26.425130152598022,
+      "learning_rate": 2e-06,
+      "loss": 0.4785,
+      "step": 980
+    },
+    {
+      "epoch": 0.2275838069829486,
+      "grad_norm": 18.382125908922546,
+      "learning_rate": 2e-06,
+      "loss": 0.3235,
+      "step": 981
+    },
+    {
+      "epoch": 0.2278157986312493,
+      "grad_norm": 22.287573999128128,
+      "learning_rate": 2e-06,
+      "loss": 0.3582,
+      "step": 982
+    },
+    {
+      "epoch": 0.22804779027954994,
+      "grad_norm": 15.318127091663474,
+      "learning_rate": 2e-06,
+      "loss": 0.3105,
+      "step": 983
+    },
+    {
+      "epoch": 0.2282797819278506,
+      "grad_norm": 15.2128289752655,
+      "learning_rate": 2e-06,
+      "loss": 0.2805,
+      "step": 984
+    },
+    {
+      "epoch": 0.22851177357615127,
+      "grad_norm": 13.6362169257321,
+      "learning_rate": 2e-06,
+      "loss": 0.3603,
+      "step": 985
+    },
+    {
+      "epoch": 0.22874376522445192,
+      "grad_norm": 19.427830286160788,
+      "learning_rate": 2e-06,
+      "loss": 0.3149,
+      "step": 986
+    },
+    {
+      "epoch": 0.22897575687275257,
+      "grad_norm": 10.236059122445331,
+      "learning_rate": 2e-06,
+      "loss": 0.341,
+      "step": 987
+    },
+    {
+      "epoch": 0.22920774852105325,
+      "grad_norm": 20.821455151022416,
+      "learning_rate": 2e-06,
+      "loss": 0.4091,
+      "step": 988
+    },
+    {
+      "epoch": 0.2294397401693539,
+      "grad_norm": 45.94263705822538,
+      "learning_rate": 2e-06,
+      "loss": 0.2858,
+      "step": 989
+    },
+    {
+      "epoch": 0.22967173181765457,
+      "grad_norm": 10.338890421609332,
+      "learning_rate": 2e-06,
+      "loss": 0.378,
+      "step": 990
+    },
+    {
+      "epoch": 0.22990372346595522,
+      "grad_norm": 16.299094568182728,
+      "learning_rate": 2e-06,
+      "loss": 0.3556,
+      "step": 991
+    },
+    {
+      "epoch": 0.23013571511425587,
+      "grad_norm": 26.428391344147883,
+      "learning_rate": 2e-06,
+      "loss": 0.4697,
+      "step": 992
+    },
+    {
+      "epoch": 0.23036770676255655,
+      "grad_norm": 13.437299905823535,
+      "learning_rate": 2e-06,
+      "loss": 0.3016,
+      "step": 993
+    },
+    {
+      "epoch": 0.2305996984108572,
+      "grad_norm": 12.436162205821939,
+      "learning_rate": 2e-06,
+      "loss": 0.3067,
+      "step": 994
+    },
+    {
+      "epoch": 0.23083169005915788,
+      "grad_norm": 9.291303336441867,
+      "learning_rate": 2e-06,
+      "loss": 0.2107,
+      "step": 995
+    },
+    {
+      "epoch": 0.23106368170745853,
+      "grad_norm": 18.667844626673965,
+      "learning_rate": 2e-06,
+      "loss": 0.277,
+      "step": 996
+    },
+    {
+      "epoch": 0.23129567335575918,
+      "grad_norm": 10.092912796644061,
+      "learning_rate": 2e-06,
+      "loss": 0.3074,
+      "step": 997
+    },
+    {
+      "epoch": 0.23152766500405986,
+      "grad_norm": 9.506478549775954,
+      "learning_rate": 2e-06,
+      "loss": 0.213,
+      "step": 998
+    },
+    {
+      "epoch": 0.2317596566523605,
+      "grad_norm": 9.16061150676762,
+      "learning_rate": 2e-06,
+      "loss": 0.2959,
+      "step": 999
+    },
+    {
+      "epoch": 0.23199164830066119,
+      "grad_norm": 13.513819503869733,
+      "learning_rate": 2e-06,
+      "loss": 0.3028,
+      "step": 1000
+    },
+    {
+      "epoch": 0.23222363994896184,
+      "grad_norm": 12.741939938596376,
+      "learning_rate": 2e-06,
+      "loss": 0.3136,
+      "step": 1001
+    },
+    {
+      "epoch": 0.23245563159726249,
+      "grad_norm": 19.698673117365807,
+      "learning_rate": 2e-06,
+      "loss": 0.3715,
+      "step": 1002
+    },
+    {
+      "epoch": 0.23268762324556316,
+      "grad_norm": 19.48596566945166,
+      "learning_rate": 2e-06,
+      "loss": 0.3778,
+      "step": 1003
+    },
+    {
+      "epoch": 0.2329196148938638,
+      "grad_norm": 13.556562374482986,
+      "learning_rate": 2e-06,
+      "loss": 0.3665,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2331516065421645,
+      "grad_norm": 9.87402518242158,
+      "learning_rate": 2e-06,
+      "loss": 0.3285,
+      "step": 1005
+    },
+    {
+      "epoch": 0.23338359819046514,
+      "grad_norm": 13.973925681529213,
+      "learning_rate": 2e-06,
+      "loss": 0.3313,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2336155898387658,
+      "grad_norm": 12.270413839177959,
+      "learning_rate": 2e-06,
+      "loss": 0.2936,
+      "step": 1007
+    },
+    {
+      "epoch": 0.23384758148706647,
+      "grad_norm": 10.414829485394955,
+      "learning_rate": 2e-06,
+      "loss": 0.3295,
+      "step": 1008
+    },
+    {
+      "epoch": 0.23407957313536712,
+      "grad_norm": 8.359953878679415,
+      "learning_rate": 2e-06,
+      "loss": 0.2915,
+      "step": 1009
+    },
+    {
+      "epoch": 0.2343115647836678,
+      "grad_norm": 21.553315281917737,
+      "learning_rate": 2e-06,
+      "loss": 0.244,
+      "step": 1010
+    },
+    {
+      "epoch": 0.23454355643196845,
+      "grad_norm": 15.282341825830445,
+      "learning_rate": 2e-06,
+      "loss": 0.364,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2347755480802691,
+      "grad_norm": 11.105284562316186,
+      "learning_rate": 2e-06,
+      "loss": 0.3318,
+      "step": 1012
+    },
+    {
+      "epoch": 0.23500753972856978,
+      "grad_norm": 26.70366182684123,
+      "learning_rate": 2e-06,
+      "loss": 0.4745,
+      "step": 1013
+    },
+    {
+      "epoch": 0.23523953137687043,
+      "grad_norm": 13.969126715794628,
+      "learning_rate": 2e-06,
+      "loss": 0.3891,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2354715230251711,
+      "grad_norm": 7.149208235386196,
+      "learning_rate": 2e-06,
+      "loss": 0.2255,
+      "step": 1015
+    },
+    {
+      "epoch": 0.23570351467347175,
+      "grad_norm": 9.282816481833471,
+      "learning_rate": 2e-06,
+      "loss": 0.2991,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2359355063217724,
+      "grad_norm": 11.997722779384231,
+      "learning_rate": 2e-06,
+      "loss": 0.3355,
+      "step": 1017
+    },
+    {
+      "epoch": 0.23616749797007308,
+      "grad_norm": 7.261743927913489,
+      "learning_rate": 2e-06,
+      "loss": 0.2568,
+      "step": 1018
+    },
+    {
+      "epoch": 0.23639948961837373,
+      "grad_norm": 9.285412136288146,
+      "learning_rate": 2e-06,
+      "loss": 0.2552,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2366314812666744,
+      "grad_norm": 14.550212842325665,
+      "learning_rate": 2e-06,
+      "loss": 0.2656,
+      "step": 1020
+    },
+    {
+      "epoch": 0.23686347291497506,
+      "grad_norm": 13.672680558038127,
+      "learning_rate": 2e-06,
+      "loss": 0.3041,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2370954645632757,
+      "grad_norm": 13.175353642289746,
+      "learning_rate": 2e-06,
+      "loss": 0.2793,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2373274562115764,
+      "grad_norm": 14.857712932971321,
+      "learning_rate": 2e-06,
+      "loss": 0.2785,
+      "step": 1023
+    },
+    {
+      "epoch": 0.23755944785987704,
+      "grad_norm": 14.940824776770855,
+      "learning_rate": 2e-06,
+      "loss": 0.2517,
+      "step": 1024
+    },
+    {
+      "epoch": 0.23779143950817772,
+      "grad_norm": 27.551149012916042,
+      "learning_rate": 2e-06,
+      "loss": 0.412,
+      "step": 1025
+    },
+    {
+      "epoch": 0.23802343115647837,
+      "grad_norm": 13.21079485337451,
+      "learning_rate": 2e-06,
+      "loss": 0.2841,
+      "step": 1026
+    },
+    {
+      "epoch": 0.23825542280477902,
+      "grad_norm": 8.073937246007285,
+      "learning_rate": 2e-06,
+      "loss": 0.4117,
+      "step": 1027
+    },
+    {
+      "epoch": 0.2384874144530797,
+      "grad_norm": 15.211564630308953,
+      "learning_rate": 2e-06,
+      "loss": 0.3509,
+      "step": 1028
+    },
+    {
+      "epoch": 0.23871940610138034,
+      "grad_norm": 6.753994429637856,
+      "learning_rate": 2e-06,
+      "loss": 0.2745,
+      "step": 1029
+    },
+    {
+      "epoch": 0.23895139774968102,
+      "grad_norm": 16.579544298722894,
+      "learning_rate": 2e-06,
+      "loss": 0.3769,
+      "step": 1030
+    },
+    {
+      "epoch": 0.23918338939798167,
+      "grad_norm": 12.312689236209662,
+      "learning_rate": 2e-06,
+      "loss": 0.2979,
+      "step": 1031
+    },
+    {
+      "epoch": 0.23941538104628232,
+      "grad_norm": 22.961333721387835,
+      "learning_rate": 2e-06,
+      "loss": 0.2667,
+      "step": 1032
+    },
+    {
+      "epoch": 0.239647372694583,
+      "grad_norm": 26.38181638368504,
+      "learning_rate": 2e-06,
+      "loss": 0.2407,
+      "step": 1033
+    },
+    {
+      "epoch": 0.23987936434288365,
+      "grad_norm": 21.122533742687466,
+      "learning_rate": 2e-06,
+      "loss": 0.324,
+      "step": 1034
+    },
+    {
+      "epoch": 0.24011135599118433,
+      "grad_norm": 9.385543716224886,
+      "learning_rate": 2e-06,
+      "loss": 0.2322,
+      "step": 1035
+    },
+    {
+      "epoch": 0.24034334763948498,
+      "grad_norm": 22.411151234064008,
+      "learning_rate": 2e-06,
+      "loss": 0.3298,
+      "step": 1036
+    },
+    {
+      "epoch": 0.24057533928778563,
+      "grad_norm": 8.712554848424903,
+      "learning_rate": 2e-06,
+      "loss": 0.2662,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2408073309360863,
+      "grad_norm": 23.670062322774925,
+      "learning_rate": 2e-06,
+      "loss": 0.3704,
+      "step": 1038
+    },
+    {
+      "epoch": 0.24103932258438696,
+      "grad_norm": 14.416551486323577,
+      "learning_rate": 2e-06,
+      "loss": 0.339,
+      "step": 1039
+    },
+    {
+      "epoch": 0.24127131423268763,
+      "grad_norm": 11.777740161029039,
+      "learning_rate": 2e-06,
+      "loss": 0.307,
+      "step": 1040
+    },
+    {
+      "epoch": 0.24150330588098828,
+      "grad_norm": 16.333864517062604,
+      "learning_rate": 2e-06,
+      "loss": 0.3238,
+      "step": 1041
+    },
+    {
+      "epoch": 0.24173529752928893,
+      "grad_norm": 16.200709138501136,
+      "learning_rate": 2e-06,
+      "loss": 0.3817,
+      "step": 1042
+    },
+    {
+      "epoch": 0.2419672891775896,
+      "grad_norm": 16.75392028228238,
+      "learning_rate": 2e-06,
+      "loss": 0.3657,
+      "step": 1043
+    },
+    {
+      "epoch": 0.24219928082589026,
+      "grad_norm": 16.7005437223624,
+      "learning_rate": 2e-06,
+      "loss": 0.2499,
+      "step": 1044
+    },
+    {
+      "epoch": 0.24243127247419094,
+      "grad_norm": 8.676317977311099,
+      "learning_rate": 2e-06,
+      "loss": 0.3238,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2426632641224916,
+      "grad_norm": 25.531733213211815,
+      "learning_rate": 2e-06,
+      "loss": 0.3677,
+      "step": 1046
+    },
+    {
+      "epoch": 0.24289525577079224,
+      "grad_norm": 15.53633794877907,
+      "learning_rate": 2e-06,
+      "loss": 0.3253,
+      "step": 1047
+    },
+    {
+      "epoch": 0.24312724741909292,
+      "grad_norm": 19.81381137188102,
+      "learning_rate": 2e-06,
+      "loss": 0.3648,
+      "step": 1048
+    },
+    {
+      "epoch": 0.24335923906739357,
+      "grad_norm": 12.25975324260386,
+      "learning_rate": 2e-06,
+      "loss": 0.1973,
+      "step": 1049
+    },
+    {
+      "epoch": 0.24359123071569425,
+      "grad_norm": 14.00865679850547,
+      "learning_rate": 2e-06,
+      "loss": 0.3152,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2438232223639949,
+      "grad_norm": 12.392977697215564,
+      "learning_rate": 2e-06,
+      "loss": 0.3584,
+      "step": 1051
+    },
+    {
+      "epoch": 0.24405521401229555,
+      "grad_norm": 13.265639482781935,
+      "learning_rate": 2e-06,
+      "loss": 0.3614,
+      "step": 1052
+    },
+    {
+      "epoch": 0.24428720566059622,
+      "grad_norm": 12.65724022381264,
+      "learning_rate": 2e-06,
+      "loss": 0.2673,
+      "step": 1053
+    },
+    {
+      "epoch": 0.24451919730889687,
+      "grad_norm": 19.42837132743988,
+      "learning_rate": 2e-06,
+      "loss": 0.3473,
+      "step": 1054
+    },
+    {
+      "epoch": 0.24475118895719755,
+      "grad_norm": 14.077878346698146,
+      "learning_rate": 2e-06,
+      "loss": 0.3495,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2449831806054982,
+      "grad_norm": 14.184080625576248,
+      "learning_rate": 2e-06,
+      "loss": 0.2904,
+      "step": 1056
+    },
+    {
+      "epoch": 0.24521517225379885,
+      "grad_norm": 14.88015498904286,
+      "learning_rate": 2e-06,
+      "loss": 0.2611,
+      "step": 1057
+    },
+    {
+      "epoch": 0.24544716390209953,
+      "grad_norm": 18.393845134928288,
+      "learning_rate": 2e-06,
+      "loss": 0.3404,
+      "step": 1058
+    },
+    {
+      "epoch": 0.24567915555040018,
+      "grad_norm": 12.814313086491294,
+      "learning_rate": 2e-06,
+      "loss": 0.3263,
+      "step": 1059
+    },
+    {
+      "epoch": 0.24591114719870086,
+      "grad_norm": 10.02474449009755,
+      "learning_rate": 2e-06,
+      "loss": 0.1984,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2461431388470015,
+      "grad_norm": 22.079614637832556,
+      "learning_rate": 2e-06,
+      "loss": 0.405,
+      "step": 1061
+    },
+    {
+      "epoch": 0.24637513049530216,
+      "grad_norm": 20.259832978927108,
+      "learning_rate": 2e-06,
+      "loss": 0.3395,
+      "step": 1062
+    },
+    {
+      "epoch": 0.24660712214360284,
+      "grad_norm": 15.223930734276674,
+      "learning_rate": 2e-06,
+      "loss": 0.3289,
+      "step": 1063
+    },
+    {
+      "epoch": 0.24683911379190349,
+      "grad_norm": 15.643931293073702,
+      "learning_rate": 2e-06,
+      "loss": 0.2943,
+      "step": 1064
+    },
+    {
+      "epoch": 0.24707110544020416,
+      "grad_norm": 10.985354961154403,
+      "learning_rate": 2e-06,
+      "loss": 0.2698,
+      "step": 1065
+    },
+    {
+      "epoch": 0.24730309708850481,
+      "grad_norm": 14.823172015971181,
+      "learning_rate": 2e-06,
+      "loss": 0.2952,
+      "step": 1066
+    },
+    {
+      "epoch": 0.24753508873680546,
+      "grad_norm": 23.05124526115744,
+      "learning_rate": 2e-06,
+      "loss": 0.3497,
+      "step": 1067
+    },
+    {
+      "epoch": 0.24776708038510614,
+      "grad_norm": 20.657372065022596,
+      "learning_rate": 2e-06,
+      "loss": 0.3718,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2479990720334068,
+      "grad_norm": 8.63115021917154,
+      "learning_rate": 2e-06,
+      "loss": 0.2653,
+      "step": 1069
+    },
+    {
+      "epoch": 0.24823106368170747,
+      "grad_norm": 9.151095904091301,
+      "learning_rate": 2e-06,
+      "loss": 0.2463,
+      "step": 1070
+    },
+    {
+      "epoch": 0.24846305533000812,
+      "grad_norm": 17.992692596229116,
+      "learning_rate": 2e-06,
+      "loss": 0.2568,
+      "step": 1071
+    },
+    {
+      "epoch": 0.24869504697830877,
+      "grad_norm": 14.4339784894775,
+      "learning_rate": 2e-06,
+      "loss": 0.298,
+      "step": 1072
+    },
+    {
+      "epoch": 0.24892703862660945,
+      "grad_norm": 12.390189004009237,
+      "learning_rate": 2e-06,
+      "loss": 0.2571,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2491590302749101,
+      "grad_norm": 12.565613467230603,
+      "learning_rate": 2e-06,
+      "loss": 0.2154,
+      "step": 1074
+    },
+    {
+      "epoch": 0.24939102192321078,
+      "grad_norm": 23.2959019541224,
+      "learning_rate": 2e-06,
+      "loss": 0.3911,
+      "step": 1075
+    },
+    {
+      "epoch": 0.24962301357151143,
+      "grad_norm": 14.039259928255964,
+      "learning_rate": 2e-06,
+      "loss": 0.2285,
+      "step": 1076
+    },
+    {
+      "epoch": 0.24985500521981208,
+      "grad_norm": 19.598970977130815,
+      "learning_rate": 2e-06,
+      "loss": 0.4487,
+      "step": 1077
+    },
+    {
+      "epoch": 0.25008699686811275,
+      "grad_norm": 12.73183841773306,
+      "learning_rate": 2e-06,
+      "loss": 0.3772,
+      "step": 1078
+    },
+    {
+      "epoch": 0.25031898851641343,
+      "grad_norm": 17.74810849129701,
+      "learning_rate": 2e-06,
+      "loss": 0.3162,
+      "step": 1079
+    },
+    {
+      "epoch": 0.25055098016471405,
+      "grad_norm": 23.191357988387217,
+      "learning_rate": 2e-06,
+      "loss": 0.4153,
+      "step": 1080
+    },
+    {
+      "epoch": 0.25078297181301473,
+      "grad_norm": 20.19849388943989,
+      "learning_rate": 2e-06,
+      "loss": 0.4059,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2510149634613154,
+      "grad_norm": 11.232999446060697,
+      "learning_rate": 2e-06,
+      "loss": 0.3919,
+      "step": 1082
+    },
+    {
+      "epoch": 0.25124695510961603,
+      "grad_norm": 16.45997289285503,
+      "learning_rate": 2e-06,
+      "loss": 0.3498,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2514789467579167,
+      "grad_norm": 15.723585351231751,
+      "learning_rate": 2e-06,
+      "loss": 0.2632,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2517109384062174,
+      "grad_norm": 7.330028584689176,
+      "learning_rate": 2e-06,
+      "loss": 0.2012,
+      "step": 1085
+    },
+    {
+      "epoch": 0.251942930054518,
+      "grad_norm": 13.88113627267473,
+      "learning_rate": 2e-06,
+      "loss": 0.2686,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2521749217028187,
+      "grad_norm": 19.509094614015467,
+      "learning_rate": 2e-06,
+      "loss": 0.3515,
+      "step": 1087
+    },
+    {
+      "epoch": 0.25240691335111937,
+      "grad_norm": 20.16682881221863,
+      "learning_rate": 2e-06,
+      "loss": 0.4119,
+      "step": 1088
+    },
+    {
+      "epoch": 0.25263890499942004,
+      "grad_norm": 11.957155629814546,
+      "learning_rate": 2e-06,
+      "loss": 0.3324,
+      "step": 1089
+    },
+    {
+      "epoch": 0.25287089664772067,
+      "grad_norm": 21.806967508834347,
+      "learning_rate": 2e-06,
+      "loss": 0.4624,
+      "step": 1090
+    },
+    {
+      "epoch": 0.25310288829602134,
+      "grad_norm": 13.749685076401235,
+      "learning_rate": 2e-06,
+      "loss": 0.3831,
+      "step": 1091
+    },
+    {
+      "epoch": 0.253334879944322,
+      "grad_norm": 16.656854478247116,
+      "learning_rate": 2e-06,
+      "loss": 0.2794,
+      "step": 1092
+    },
+    {
+      "epoch": 0.25356687159262264,
+      "grad_norm": 17.627596146798037,
+      "learning_rate": 2e-06,
+      "loss": 0.3494,
+      "step": 1093
+    },
+    {
+      "epoch": 0.2537988632409233,
+      "grad_norm": 19.745119628195066,
+      "learning_rate": 2e-06,
+      "loss": 0.4636,
+      "step": 1094
+    },
+    {
+      "epoch": 0.254030854889224,
+      "grad_norm": 17.18830926296359,
+      "learning_rate": 2e-06,
+      "loss": 0.3584,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2542628465375246,
+      "grad_norm": 9.4163939741239,
+      "learning_rate": 2e-06,
+      "loss": 0.3597,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2544948381858253,
+      "grad_norm": 16.018761774541133,
+      "learning_rate": 2e-06,
+      "loss": 0.3183,
+      "step": 1097
+    },
+    {
+      "epoch": 0.254726829834126,
+      "grad_norm": 17.99094305205517,
+      "learning_rate": 2e-06,
+      "loss": 0.3835,
+      "step": 1098
+    },
+    {
+      "epoch": 0.25495882148242666,
+      "grad_norm": 19.22211629507197,
+      "learning_rate": 2e-06,
+      "loss": 0.2728,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2551908131307273,
+      "grad_norm": 11.092634757643538,
+      "learning_rate": 2e-06,
+      "loss": 0.2726,
+      "step": 1100
+    },
+    {
+      "epoch": 0.25542280477902796,
+      "grad_norm": 27.83008647407575,
+      "learning_rate": 2e-06,
+      "loss": 0.4326,
+      "step": 1101
+    },
+    {
+      "epoch": 0.25565479642732863,
+      "grad_norm": 15.001540429809646,
+      "learning_rate": 2e-06,
+      "loss": 0.3416,
+      "step": 1102
+    },
+    {
+      "epoch": 0.25588678807562926,
+      "grad_norm": 15.144143645466546,
+      "learning_rate": 2e-06,
+      "loss": 0.3476,
+      "step": 1103
+    },
+    {
+      "epoch": 0.25611877972392993,
+      "grad_norm": 15.645113127825065,
+      "learning_rate": 2e-06,
+      "loss": 0.3551,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2563507713722306,
+      "grad_norm": 10.928220909805345,
+      "learning_rate": 2e-06,
+      "loss": 0.2243,
+      "step": 1105
+    },
+    {
+      "epoch": 0.25658276302053123,
+      "grad_norm": 11.445144928478546,
+      "learning_rate": 2e-06,
+      "loss": 0.2368,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2568147546688319,
+      "grad_norm": 7.806367904222486,
+      "learning_rate": 2e-06,
+      "loss": 0.211,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2570467463171326,
+      "grad_norm": 22.770218588703784,
+      "learning_rate": 2e-06,
+      "loss": 0.4157,
+      "step": 1108
+    },
+    {
+      "epoch": 0.25727873796543327,
+      "grad_norm": 24.118256132266875,
+      "learning_rate": 2e-06,
+      "loss": 0.3311,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2575107296137339,
+      "grad_norm": 14.418260243799653,
+      "learning_rate": 2e-06,
+      "loss": 0.2665,
+      "step": 1110
+    },
+    {
+      "epoch": 0.25774272126203457,
+      "grad_norm": 17.95557610022089,
+      "learning_rate": 2e-06,
+      "loss": 0.4597,
+      "step": 1111
+    },
+    {
+      "epoch": 0.25797471291033525,
+      "grad_norm": 9.998203620016008,
+      "learning_rate": 2e-06,
+      "loss": 0.264,
+      "step": 1112
+    },
+    {
+      "epoch": 0.25820670455863587,
+      "grad_norm": 15.98410256214507,
+      "learning_rate": 2e-06,
+      "loss": 0.3399,
+      "step": 1113
+    },
+    {
+      "epoch": 0.25843869620693655,
+      "grad_norm": 14.65274151237759,
+      "learning_rate": 2e-06,
+      "loss": 0.3357,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2586706878552372,
+      "grad_norm": 15.099128606884662,
+      "learning_rate": 2e-06,
+      "loss": 0.254,
+      "step": 1115
+    },
+    {
+      "epoch": 0.25890267950353785,
+      "grad_norm": 24.82947299465006,
+      "learning_rate": 2e-06,
+      "loss": 0.3305,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2591346711518385,
+      "grad_norm": 13.261084761295198,
+      "learning_rate": 2e-06,
+      "loss": 0.3377,
+      "step": 1117
+    },
+    {
+      "epoch": 0.2593666628001392,
+      "grad_norm": 22.691482458679683,
+      "learning_rate": 2e-06,
+      "loss": 0.3698,
+      "step": 1118
+    },
+    {
+      "epoch": 0.2595986544484399,
+      "grad_norm": 14.41997073811382,
+      "learning_rate": 2e-06,
+      "loss": 0.4644,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2598306460967405,
+      "grad_norm": 15.695126032511784,
+      "learning_rate": 2e-06,
+      "loss": 0.3084,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2600626377450412,
+      "grad_norm": 20.452165103407637,
+      "learning_rate": 2e-06,
+      "loss": 0.3405,
+      "step": 1121
+    },
+    {
+      "epoch": 0.26029462939334186,
+      "grad_norm": 19.152384814044723,
+      "learning_rate": 2e-06,
+      "loss": 0.3941,
+      "step": 1122
+    },
+    {
+      "epoch": 0.2605266210416425,
+      "grad_norm": 10.929796244886692,
+      "learning_rate": 2e-06,
+      "loss": 0.3085,
+      "step": 1123
+    },
+    {
+      "epoch": 0.26075861268994316,
+      "grad_norm": 17.796433072113317,
+      "learning_rate": 2e-06,
+      "loss": 0.4109,
+      "step": 1124
+    },
+    {
+      "epoch": 0.26099060433824384,
+      "grad_norm": 10.425466920495225,
+      "learning_rate": 2e-06,
+      "loss": 0.3517,
+      "step": 1125
+    },
+    {
+      "epoch": 0.26122259598654446,
+      "grad_norm": 6.015552175860235,
+      "learning_rate": 2e-06,
+      "loss": 0.2988,
+      "step": 1126
+    },
+    {
+      "epoch": 0.26145458763484514,
+      "grad_norm": 6.126586375825967,
+      "learning_rate": 2e-06,
+      "loss": 0.2415,
+      "step": 1127
+    },
+    {
+      "epoch": 0.2616865792831458,
+      "grad_norm": 11.863623548309842,
+      "learning_rate": 2e-06,
+      "loss": 0.2739,
+      "step": 1128
+    },
+    {
+      "epoch": 0.2619185709314465,
+      "grad_norm": 14.909953393587152,
+      "learning_rate": 2e-06,
+      "loss": 0.3223,
+      "step": 1129
+    },
+    {
+      "epoch": 0.2621505625797471,
+      "grad_norm": 14.646054099158548,
+      "learning_rate": 2e-06,
+      "loss": 0.3307,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2623825542280478,
+      "grad_norm": 13.629312091258925,
+      "learning_rate": 2e-06,
+      "loss": 0.3004,
+      "step": 1131
+    },
+    {
+      "epoch": 0.26261454587634847,
+      "grad_norm": 24.391131892080814,
+      "learning_rate": 2e-06,
+      "loss": 0.3546,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2628465375246491,
+      "grad_norm": 9.685129053473396,
+      "learning_rate": 2e-06,
+      "loss": 0.3143,
+      "step": 1133
+    },
+    {
+      "epoch": 0.26307852917294977,
+      "grad_norm": 19.920212953026695,
+      "learning_rate": 2e-06,
+      "loss": 0.3734,
+      "step": 1134
+    },
+    {
+      "epoch": 0.26331052082125045,
+      "grad_norm": 18.873355014808933,
+      "learning_rate": 2e-06,
+      "loss": 0.2707,
+      "step": 1135
+    },
+    {
+      "epoch": 0.26354251246955107,
+      "grad_norm": 15.679849289037026,
+      "learning_rate": 2e-06,
+      "loss": 0.338,
+      "step": 1136
+    },
+    {
+      "epoch": 0.26377450411785175,
+      "grad_norm": 13.53924087047476,
+      "learning_rate": 2e-06,
+      "loss": 0.3066,
+      "step": 1137
+    },
+    {
+      "epoch": 0.2640064957661524,
+      "grad_norm": 11.745611010016228,
+      "learning_rate": 2e-06,
+      "loss": 0.3399,
+      "step": 1138
+    },
+    {
+      "epoch": 0.2642384874144531,
+      "grad_norm": 15.225621934010864,
+      "learning_rate": 2e-06,
+      "loss": 0.244,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2644704790627537,
+      "grad_norm": 10.40783059783653,
+      "learning_rate": 2e-06,
+      "loss": 0.2386,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2647024707110544,
+      "grad_norm": 15.851959230703185,
+      "learning_rate": 2e-06,
+      "loss": 0.3335,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2649344623593551,
+      "grad_norm": 22.66210347750043,
+      "learning_rate": 2e-06,
+      "loss": 0.3033,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2651664540076557,
+      "grad_norm": 26.031860194956312,
+      "learning_rate": 2e-06,
+      "loss": 0.3799,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2653984456559564,
+      "grad_norm": 18.188321934019086,
+      "learning_rate": 2e-06,
+      "loss": 0.3478,
+      "step": 1144
+    },
+    {
+      "epoch": 0.26563043730425706,
+      "grad_norm": 9.418783234825963,
+      "learning_rate": 2e-06,
+      "loss": 0.2319,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2658624289525577,
+      "grad_norm": 12.05426823213287,
+      "learning_rate": 2e-06,
+      "loss": 0.4227,
+      "step": 1146
+    },
+    {
+      "epoch": 0.26609442060085836,
+      "grad_norm": 26.72325329812725,
+      "learning_rate": 2e-06,
+      "loss": 0.3139,
+      "step": 1147
+    },
+    {
+      "epoch": 0.26632641224915904,
+      "grad_norm": 16.733829455551838,
+      "learning_rate": 2e-06,
+      "loss": 0.3918,
+      "step": 1148
+    },
+    {
+      "epoch": 0.2665584038974597,
+      "grad_norm": 14.249457828238011,
+      "learning_rate": 2e-06,
+      "loss": 0.242,
+      "step": 1149
+    },
+    {
+      "epoch": 0.26679039554576034,
+      "grad_norm": 19.86506070787913,
+      "learning_rate": 2e-06,
+      "loss": 0.3798,
+      "step": 1150
+    },
+    {
+      "epoch": 0.267022387194061,
+      "grad_norm": 13.956094717202392,
+      "learning_rate": 2e-06,
+      "loss": 0.2447,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2672543788423617,
+      "grad_norm": 10.456531626943017,
+      "learning_rate": 2e-06,
+      "loss": 0.3394,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2674863704906623,
+      "grad_norm": 11.825638890958947,
+      "learning_rate": 2e-06,
+      "loss": 0.2495,
+      "step": 1153
+    },
+    {
+      "epoch": 0.267718362138963,
+      "grad_norm": 16.889888110765746,
+      "learning_rate": 2e-06,
+      "loss": 0.3971,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2679503537872637,
+      "grad_norm": 14.904612459119354,
+      "learning_rate": 2e-06,
+      "loss": 0.2268,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2681823454355643,
+      "grad_norm": 15.751958535325018,
+      "learning_rate": 2e-06,
+      "loss": 0.2576,
+      "step": 1156
+    },
+    {
+      "epoch": 0.268414337083865,
+      "grad_norm": 11.142765466493485,
+      "learning_rate": 2e-06,
+      "loss": 0.3464,
+      "step": 1157
+    },
+    {
+      "epoch": 0.26864632873216565,
+      "grad_norm": 11.463689658847697,
+      "learning_rate": 2e-06,
+      "loss": 0.2759,
+      "step": 1158
+    },
+    {
+      "epoch": 0.26887832038046633,
+      "grad_norm": 15.396711358653583,
+      "learning_rate": 2e-06,
+      "loss": 0.2698,
+      "step": 1159
+    },
+    {
+      "epoch": 0.26911031202876695,
+      "grad_norm": 20.355215766836594,
+      "learning_rate": 2e-06,
+      "loss": 0.2896,
+      "step": 1160
+    },
+    {
+      "epoch": 0.26934230367706763,
+      "grad_norm": 22.16248015841271,
+      "learning_rate": 2e-06,
+      "loss": 0.3936,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2695742953253683,
+      "grad_norm": 25.343187131455633,
+      "learning_rate": 2e-06,
+      "loss": 0.3586,
+      "step": 1162
+    },
+    {
+      "epoch": 0.26980628697366893,
+      "grad_norm": 7.171393164606418,
+      "learning_rate": 2e-06,
+      "loss": 0.1983,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2700382786219696,
+      "grad_norm": 19.05142625020754,
+      "learning_rate": 2e-06,
+      "loss": 0.3657,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2702702702702703,
+      "grad_norm": 8.95382335460153,
+      "learning_rate": 2e-06,
+      "loss": 0.3213,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2705022619185709,
+      "grad_norm": 12.81735944131886,
+      "learning_rate": 2e-06,
+      "loss": 0.3069,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2707342535668716,
+      "grad_norm": 17.805420961529475,
+      "learning_rate": 2e-06,
+      "loss": 0.2557,
+      "step": 1167
+    },
+    {
+      "epoch": 0.27096624521517226,
+      "grad_norm": 16.874564931891637,
+      "learning_rate": 2e-06,
+      "loss": 0.3154,
+      "step": 1168
+    },
+    {
+      "epoch": 0.27119823686347294,
+      "grad_norm": 11.799433595683006,
+      "learning_rate": 2e-06,
+      "loss": 0.2859,
+      "step": 1169
+    },
+    {
+      "epoch": 0.27143022851177356,
+      "grad_norm": 9.39700626037502,
+      "learning_rate": 2e-06,
+      "loss": 0.2184,
+      "step": 1170
+    },
+    {
+      "epoch": 0.27166222016007424,
+      "grad_norm": 17.073536459242636,
+      "learning_rate": 2e-06,
+      "loss": 0.2849,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2718942118083749,
+      "grad_norm": 21.22630275472056,
+      "learning_rate": 2e-06,
+      "loss": 0.3663,
+      "step": 1172
+    },
+    {
+      "epoch": 0.27212620345667554,
+      "grad_norm": 12.59115651190941,
+      "learning_rate": 2e-06,
+      "loss": 0.3009,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2723581951049762,
+      "grad_norm": 16.968572807837223,
+      "learning_rate": 2e-06,
+      "loss": 0.4332,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2725901867532769,
+      "grad_norm": 9.655637385536188,
+      "learning_rate": 2e-06,
+      "loss": 0.2609,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2728221784015775,
+      "grad_norm": 13.890112671278628,
+      "learning_rate": 2e-06,
+      "loss": 0.2625,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2730541700498782,
+      "grad_norm": 12.606139803898643,
+      "learning_rate": 2e-06,
+      "loss": 0.3785,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2732861616981789,
+      "grad_norm": 13.28056042143577,
+      "learning_rate": 2e-06,
+      "loss": 0.2906,
+      "step": 1178
+    },
+    {
+      "epoch": 0.27351815334647955,
+      "grad_norm": 30.27887213841802,
+      "learning_rate": 2e-06,
+      "loss": 0.4022,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2737501449947802,
+      "grad_norm": 7.930895909161131,
+      "learning_rate": 2e-06,
+      "loss": 0.2247,
+      "step": 1180
+    },
+    {
+      "epoch": 0.27398213664308085,
+      "grad_norm": 13.533083064817841,
+      "learning_rate": 2e-06,
+      "loss": 0.2595,
+      "step": 1181
+    },
+    {
+      "epoch": 0.27421412829138153,
+      "grad_norm": 10.212297449937365,
+      "learning_rate": 2e-06,
+      "loss": 0.275,
+      "step": 1182
+    },
+    {
+      "epoch": 0.27444611993968215,
+      "grad_norm": 10.988417124060847,
+      "learning_rate": 2e-06,
+      "loss": 0.213,
+      "step": 1183
+    },
+    {
+      "epoch": 0.27467811158798283,
+      "grad_norm": 24.17803575510525,
+      "learning_rate": 2e-06,
+      "loss": 0.4077,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2749101032362835,
+      "grad_norm": 9.680825814303265,
+      "learning_rate": 2e-06,
+      "loss": 0.2429,
+      "step": 1185
+    },
+    {
+      "epoch": 0.27514209488458413,
+      "grad_norm": 6.049044301937683,
+      "learning_rate": 2e-06,
+      "loss": 0.2024,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2753740865328848,
+      "grad_norm": 18.00295476464138,
+      "learning_rate": 2e-06,
+      "loss": 0.2559,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2756060781811855,
+      "grad_norm": 15.30270566627656,
+      "learning_rate": 2e-06,
+      "loss": 0.3257,
+      "step": 1188
+    },
+    {
+      "epoch": 0.27583806982948617,
+      "grad_norm": 21.694395810281325,
+      "learning_rate": 2e-06,
+      "loss": 0.3079,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2760700614777868,
+      "grad_norm": 8.331967188153596,
+      "learning_rate": 2e-06,
+      "loss": 0.1824,
+      "step": 1190
+    },
+    {
+      "epoch": 0.27630205312608747,
+      "grad_norm": 15.415548630144007,
+      "learning_rate": 2e-06,
+      "loss": 0.3489,
+      "step": 1191
+    },
+    {
+      "epoch": 0.27653404477438814,
+      "grad_norm": 14.233338810505089,
+      "learning_rate": 2e-06,
+      "loss": 0.1883,
+      "step": 1192
+    },
+    {
+      "epoch": 0.27676603642268877,
+      "grad_norm": 13.344673584990591,
+      "learning_rate": 2e-06,
+      "loss": 0.2581,
+      "step": 1193
+    },
+    {
+      "epoch": 0.27699802807098944,
+      "grad_norm": 8.892289670198297,
+      "learning_rate": 2e-06,
+      "loss": 0.3472,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2772300197192901,
+      "grad_norm": 7.283026978981145,
+      "learning_rate": 2e-06,
+      "loss": 0.2081,
+      "step": 1195
+    },
+    {
+      "epoch": 0.27746201136759074,
+      "grad_norm": 20.765644495960995,
+      "learning_rate": 2e-06,
+      "loss": 0.3821,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2776940030158914,
+      "grad_norm": 11.812462517867834,
+      "learning_rate": 2e-06,
+      "loss": 0.2431,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2779259946641921,
+      "grad_norm": 21.853687986115002,
+      "learning_rate": 2e-06,
+      "loss": 0.4368,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2781579863124928,
+      "grad_norm": 29.536970035347192,
+      "learning_rate": 2e-06,
+      "loss": 0.2561,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2783899779607934,
+      "grad_norm": 15.492748748291122,
+      "learning_rate": 2e-06,
+      "loss": 0.2616,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2786219696090941,
+      "grad_norm": 11.006289448991424,
+      "learning_rate": 2e-06,
+      "loss": 0.2992,
+      "step": 1201
+    },
+    {
+      "epoch": 0.27885396125739476,
+      "grad_norm": 16.026435056899846,
+      "learning_rate": 2e-06,
+      "loss": 0.2834,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2790859529056954,
+      "grad_norm": 14.183111450349523,
+      "learning_rate": 2e-06,
+      "loss": 0.3305,
+      "step": 1203
+    },
+    {
+      "epoch": 0.27931794455399606,
+      "grad_norm": 12.028101213734947,
+      "learning_rate": 2e-06,
+      "loss": 0.4311,
+      "step": 1204
+    },
+    {
+      "epoch": 0.27954993620229673,
+      "grad_norm": 15.728963078237921,
+      "learning_rate": 2e-06,
+      "loss": 0.2617,
+      "step": 1205
+    },
+    {
+      "epoch": 0.27978192785059736,
+      "grad_norm": 18.74674073643408,
+      "learning_rate": 2e-06,
+      "loss": 0.2332,
+      "step": 1206
+    },
+    {
+      "epoch": 0.28001391949889803,
+      "grad_norm": 7.8767526680463735,
+      "learning_rate": 2e-06,
+      "loss": 0.2635,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2802459111471987,
+      "grad_norm": 19.589274869050506,
+      "learning_rate": 2e-06,
+      "loss": 0.3261,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2804779027954994,
+      "grad_norm": 11.569333600691113,
+      "learning_rate": 2e-06,
+      "loss": 0.2491,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2807098944438,
+      "grad_norm": 20.27838346921668,
+      "learning_rate": 2e-06,
+      "loss": 0.365,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2809418860921007,
+      "grad_norm": 16.174668249143103,
+      "learning_rate": 2e-06,
+      "loss": 0.2677,
+      "step": 1211
+    },
+    {
+      "epoch": 0.28117387774040137,
+      "grad_norm": 9.592205036160628,
+      "learning_rate": 2e-06,
+      "loss": 0.4643,
+      "step": 1212
+    },
+    {
+      "epoch": 0.281405869388702,
+      "grad_norm": 11.243429081806028,
+      "learning_rate": 2e-06,
+      "loss": 0.2655,
+      "step": 1213
+    },
+    {
+      "epoch": 0.28163786103700267,
+      "grad_norm": 22.27363746188085,
+      "learning_rate": 2e-06,
+      "loss": 0.4154,
+      "step": 1214
+    },
+    {
+      "epoch": 0.28186985268530335,
+      "grad_norm": 16.73996394226011,
+      "learning_rate": 2e-06,
+      "loss": 0.3184,
+      "step": 1215
+    },
+    {
+      "epoch": 0.28210184433360397,
+      "grad_norm": 15.017671158539848,
+      "learning_rate": 2e-06,
+      "loss": 0.184,
+      "step": 1216
+    },
+    {
+      "epoch": 0.28233383598190465,
+      "grad_norm": 10.718819280489564,
+      "learning_rate": 2e-06,
+      "loss": 0.226,
+      "step": 1217
+    },
+    {
+      "epoch": 0.2825658276302053,
+      "grad_norm": 16.07308944496508,
+      "learning_rate": 2e-06,
+      "loss": 0.4901,
+      "step": 1218
+    },
+    {
+      "epoch": 0.28279781927850595,
+      "grad_norm": 21.358455166015435,
+      "learning_rate": 2e-06,
+      "loss": 0.3208,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2830298109268066,
+      "grad_norm": 11.131587047899961,
+      "learning_rate": 2e-06,
+      "loss": 0.4496,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2832618025751073,
+      "grad_norm": 13.583352811882996,
+      "learning_rate": 2e-06,
+      "loss": 0.2877,
+      "step": 1221
+    },
+    {
+      "epoch": 0.283493794223408,
+      "grad_norm": 18.275221017841407,
+      "learning_rate": 2e-06,
+      "loss": 0.346,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2837257858717086,
+      "grad_norm": 13.559004500486491,
+      "learning_rate": 2e-06,
+      "loss": 0.305,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2839577775200093,
+      "grad_norm": 12.193009244722145,
+      "learning_rate": 2e-06,
+      "loss": 0.3228,
+      "step": 1224
+    },
+    {
+      "epoch": 0.28418976916830996,
+      "grad_norm": 16.417106957183268,
+      "learning_rate": 2e-06,
+      "loss": 0.3148,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2844217608166106,
+      "grad_norm": 6.873063427629634,
+      "learning_rate": 2e-06,
+      "loss": 0.2427,
+      "step": 1226
+    },
+    {
+      "epoch": 0.28465375246491126,
+      "grad_norm": 14.000907125813223,
+      "learning_rate": 2e-06,
+      "loss": 0.266,
+      "step": 1227
+    },
+    {
+      "epoch": 0.28488574411321194,
+      "grad_norm": 6.750238251953412,
+      "learning_rate": 2e-06,
+      "loss": 0.2787,
+      "step": 1228
+    },
+    {
+      "epoch": 0.28511773576151256,
+      "grad_norm": 9.187628279617533,
+      "learning_rate": 2e-06,
+      "loss": 0.3106,
+      "step": 1229
+    },
+    {
+      "epoch": 0.28534972740981324,
+      "grad_norm": 26.616245578904888,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2855817190581139,
+      "grad_norm": 14.429684634728005,
+      "learning_rate": 2e-06,
+      "loss": 0.3717,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2858137107064146,
+      "grad_norm": 15.696007478073717,
+      "learning_rate": 2e-06,
+      "loss": 0.3244,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2860457023547152,
+      "grad_norm": 21.349186599737536,
+      "learning_rate": 2e-06,
+      "loss": 0.3987,
+      "step": 1233
+    },
+    {
+      "epoch": 0.2862776940030159,
+      "grad_norm": 8.12123750761895,
+      "learning_rate": 2e-06,
+      "loss": 0.2641,
+      "step": 1234
+    },
+    {
+      "epoch": 0.28650968565131657,
+      "grad_norm": 15.21984604745377,
+      "learning_rate": 2e-06,
+      "loss": 0.2997,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2867416772996172,
+      "grad_norm": 12.726342178155084,
+      "learning_rate": 2e-06,
+      "loss": 0.3333,
+      "step": 1236
+    },
+    {
+      "epoch": 0.28697366894791787,
+      "grad_norm": 16.90525595728918,
+      "learning_rate": 2e-06,
+      "loss": 0.3284,
+      "step": 1237
+    },
+    {
+      "epoch": 0.28720566059621855,
+      "grad_norm": 17.579056592317873,
+      "learning_rate": 2e-06,
+      "loss": 0.3973,
+      "step": 1238
+    },
+    {
+      "epoch": 0.28743765224451917,
+      "grad_norm": 15.35796271778437,
+      "learning_rate": 2e-06,
+      "loss": 0.3152,
+      "step": 1239
+    },
+    {
+      "epoch": 0.28766964389281985,
+      "grad_norm": 21.02104399086055,
+      "learning_rate": 2e-06,
+      "loss": 0.3467,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2879016355411205,
+      "grad_norm": 5.394065636271639,
+      "learning_rate": 2e-06,
+      "loss": 0.217,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2881336271894212,
+      "grad_norm": 18.591886314525397,
+      "learning_rate": 2e-06,
+      "loss": 0.3213,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2883656188377218,
+      "grad_norm": 22.04197488485568,
+      "learning_rate": 2e-06,
+      "loss": 0.303,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2885976104860225,
+      "grad_norm": 12.995736617500656,
+      "learning_rate": 2e-06,
+      "loss": 0.2668,
+      "step": 1244
+    },
+    {
+      "epoch": 0.2888296021343232,
+      "grad_norm": 23.646397067968046,
+      "learning_rate": 2e-06,
+      "loss": 0.3176,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2890615937826238,
+      "grad_norm": 11.409118689634937,
+      "learning_rate": 2e-06,
+      "loss": 0.2784,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2892935854309245,
+      "grad_norm": 11.742027071172174,
+      "learning_rate": 2e-06,
+      "loss": 0.4533,
+      "step": 1247
+    },
+    {
+      "epoch": 0.28952557707922516,
+      "grad_norm": 8.003574151157743,
+      "learning_rate": 2e-06,
+      "loss": 0.2324,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2897575687275258,
+      "grad_norm": 18.4898821926245,
+      "learning_rate": 2e-06,
+      "loss": 0.3342,
+      "step": 1249
+    },
+    {
+      "epoch": 0.28998956037582646,
+      "grad_norm": 19.712215280172444,
+      "learning_rate": 2e-06,
+      "loss": 0.3294,
+      "step": 1250
+    },
+    {
+      "epoch": 0.29022155202412714,
+      "grad_norm": 12.006452342747606,
+      "learning_rate": 2e-06,
+      "loss": 0.2672,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2904535436724278,
+      "grad_norm": 9.992023893613998,
+      "learning_rate": 2e-06,
+      "loss": 0.2019,
+      "step": 1252
+    },
+    {
+      "epoch": 0.29068553532072844,
+      "grad_norm": 13.6861127169211,
+      "learning_rate": 2e-06,
+      "loss": 0.255,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2909175269690291,
+      "grad_norm": 15.02484608301213,
+      "learning_rate": 2e-06,
+      "loss": 0.3086,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2911495186173298,
+      "grad_norm": 9.942533776328188,
+      "learning_rate": 2e-06,
+      "loss": 0.2822,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2913815102656304,
+      "grad_norm": 15.070113001242195,
+      "learning_rate": 2e-06,
+      "loss": 0.2637,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2916135019139311,
+      "grad_norm": 20.127966390681635,
+      "learning_rate": 2e-06,
+      "loss": 0.3592,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2918454935622318,
+      "grad_norm": 16.783537665196086,
+      "learning_rate": 2e-06,
+      "loss": 0.4248,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2920774852105324,
+      "grad_norm": 14.996063668291244,
+      "learning_rate": 2e-06,
+      "loss": 0.3356,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2923094768588331,
+      "grad_norm": 13.280194937801802,
+      "learning_rate": 2e-06,
+      "loss": 0.2595,
+      "step": 1260
+    },
+    {
+      "epoch": 0.29254146850713375,
+      "grad_norm": 12.639332668504503,
+      "learning_rate": 2e-06,
+      "loss": 0.3405,
+      "step": 1261
+    },
+    {
+      "epoch": 0.29277346015543443,
+      "grad_norm": 11.504927643928248,
+      "learning_rate": 2e-06,
+      "loss": 0.2966,
+      "step": 1262
+    },
+    {
+      "epoch": 0.29300545180373505,
+      "grad_norm": 20.752244836231803,
+      "learning_rate": 2e-06,
+      "loss": 0.4051,
+      "step": 1263
+    },
+    {
+      "epoch": 0.29323744345203573,
+      "grad_norm": 11.657319471913775,
+      "learning_rate": 2e-06,
+      "loss": 0.2778,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2934694351003364,
+      "grad_norm": 16.861970554754627,
+      "learning_rate": 2e-06,
+      "loss": 0.3725,
+      "step": 1265
+    },
+    {
+      "epoch": 0.29370142674863703,
+      "grad_norm": 14.885281184009598,
+      "learning_rate": 2e-06,
+      "loss": 0.2854,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2939334183969377,
+      "grad_norm": 9.448485631920036,
+      "learning_rate": 2e-06,
+      "loss": 0.3432,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2941654100452384,
+      "grad_norm": 14.217616728146512,
+      "learning_rate": 2e-06,
+      "loss": 0.3417,
+      "step": 1268
+    },
+    {
+      "epoch": 0.294397401693539,
+      "grad_norm": 9.39865709086706,
+      "learning_rate": 2e-06,
+      "loss": 0.2979,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2946293933418397,
+      "grad_norm": 19.94441583861392,
+      "learning_rate": 2e-06,
+      "loss": 0.3479,
+      "step": 1270
+    },
+    {
+      "epoch": 0.29486138499014036,
+      "grad_norm": 11.9760265364321,
+      "learning_rate": 2e-06,
+      "loss": 0.3599,
+      "step": 1271
+    },
+    {
+      "epoch": 0.29509337663844104,
+      "grad_norm": 11.245365111905581,
+      "learning_rate": 2e-06,
+      "loss": 0.2648,
+      "step": 1272
+    },
+    {
+      "epoch": 0.29532536828674166,
+      "grad_norm": 14.520006307802863,
+      "learning_rate": 2e-06,
+      "loss": 0.2594,
+      "step": 1273
+    },
+    {
+      "epoch": 0.29555735993504234,
+      "grad_norm": 18.75988516752072,
+      "learning_rate": 2e-06,
+      "loss": 0.2644,
+      "step": 1274
+    },
+    {
+      "epoch": 0.295789351583343,
+      "grad_norm": 15.494061198744825,
+      "learning_rate": 2e-06,
+      "loss": 0.3706,
+      "step": 1275
+    },
+    {
+      "epoch": 0.29602134323164364,
+      "grad_norm": 18.800626470937203,
+      "learning_rate": 2e-06,
+      "loss": 0.3949,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2962533348799443,
+      "grad_norm": 21.137356291700847,
+      "learning_rate": 2e-06,
+      "loss": 0.4411,
+      "step": 1277
+    },
+    {
+      "epoch": 0.296485326528245,
+      "grad_norm": 14.69093229850745,
+      "learning_rate": 2e-06,
+      "loss": 0.3186,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2967173181765456,
+      "grad_norm": 19.660974625480087,
+      "learning_rate": 2e-06,
+      "loss": 0.3263,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2969493098248463,
+      "grad_norm": 11.241549838681076,
+      "learning_rate": 2e-06,
+      "loss": 0.2473,
+      "step": 1280
+    },
+    {
+      "epoch": 0.297181301473147,
+      "grad_norm": 13.986806951706525,
+      "learning_rate": 2e-06,
+      "loss": 0.2402,
+      "step": 1281
+    },
+    {
+      "epoch": 0.29741329312144765,
+      "grad_norm": 14.285571512151392,
+      "learning_rate": 2e-06,
+      "loss": 0.374,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2976452847697483,
+      "grad_norm": 10.958162154047578,
+      "learning_rate": 2e-06,
+      "loss": 0.2306,
+      "step": 1283
+    },
+    {
+      "epoch": 0.29787727641804895,
+      "grad_norm": 18.532356058362055,
+      "learning_rate": 2e-06,
+      "loss": 0.3142,
+      "step": 1284
+    },
+    {
+      "epoch": 0.29810926806634963,
+      "grad_norm": 12.891377637696152,
+      "learning_rate": 2e-06,
+      "loss": 0.2602,
+      "step": 1285
+    },
+    {
+      "epoch": 0.29834125971465025,
+      "grad_norm": 16.11181170520358,
+      "learning_rate": 2e-06,
+      "loss": 0.3242,
+      "step": 1286
+    },
+    {
+      "epoch": 0.29857325136295093,
+      "grad_norm": 9.615362986016594,
+      "learning_rate": 2e-06,
+      "loss": 0.2893,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2988052430112516,
+      "grad_norm": 17.945990919432717,
+      "learning_rate": 2e-06,
+      "loss": 0.3861,
+      "step": 1288
+    },
+    {
+      "epoch": 0.29903723465955223,
+      "grad_norm": 17.99362450227029,
+      "learning_rate": 2e-06,
+      "loss": 0.3724,
+      "step": 1289
+    },
+    {
+      "epoch": 0.2992692263078529,
+      "grad_norm": 17.107598158507745,
+      "learning_rate": 2e-06,
+      "loss": 0.3102,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2995012179561536,
+      "grad_norm": 10.520699209722633,
+      "learning_rate": 2e-06,
+      "loss": 0.2517,
+      "step": 1291
+    },
+    {
+      "epoch": 0.29973320960445426,
+      "grad_norm": 22.305669883971103,
+      "learning_rate": 2e-06,
+      "loss": 0.3745,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2999652012527549,
+      "grad_norm": 23.331585901418507,
+      "learning_rate": 2e-06,
+      "loss": 0.297,
+      "step": 1293
+    },
+    {
+      "epoch": 0.30019719290105557,
+      "grad_norm": 9.13037425685583,
+      "learning_rate": 2e-06,
+      "loss": 0.2517,
+      "step": 1294
+    },
+    {
+      "epoch": 0.30042918454935624,
+      "grad_norm": 25.03598967873586,
+      "learning_rate": 2e-06,
+      "loss": 0.2891,
+      "step": 1295
+    },
+    {
+      "epoch": 0.30066117619765687,
+      "grad_norm": 6.227195868422087,
+      "learning_rate": 2e-06,
+      "loss": 0.2703,
+      "step": 1296
+    },
+    {
+      "epoch": 0.30089316784595754,
+      "grad_norm": 10.524631044435568,
+      "learning_rate": 2e-06,
+      "loss": 0.2402,
+      "step": 1297
+    },
+    {
+      "epoch": 0.3011251594942582,
+      "grad_norm": 14.452260998510923,
+      "learning_rate": 2e-06,
+      "loss": 0.2147,
+      "step": 1298
+    },
+    {
+      "epoch": 0.30135715114255884,
+      "grad_norm": 16.572736184416854,
+      "learning_rate": 2e-06,
+      "loss": 0.3713,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3015891427908595,
+      "grad_norm": 12.092715650876997,
+      "learning_rate": 2e-06,
+      "loss": 0.4643,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3018211344391602,
+      "grad_norm": 11.763924015078992,
+      "learning_rate": 2e-06,
+      "loss": 0.249,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3020531260874609,
+      "grad_norm": 11.297415978022148,
+      "learning_rate": 2e-06,
+      "loss": 0.3153,
+      "step": 1302
+    },
+    {
+      "epoch": 0.3022851177357615,
+      "grad_norm": 22.418021514922543,
+      "learning_rate": 2e-06,
+      "loss": 0.3138,
+      "step": 1303
+    },
+    {
+      "epoch": 0.3025171093840622,
+      "grad_norm": 20.535356510555992,
+      "learning_rate": 2e-06,
+      "loss": 0.2667,
+      "step": 1304
+    },
+    {
+      "epoch": 0.30274910103236286,
+      "grad_norm": 15.480983737195427,
+      "learning_rate": 2e-06,
+      "loss": 0.3294,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3029810926806635,
+      "grad_norm": 18.091773010729597,
+      "learning_rate": 2e-06,
+      "loss": 0.3949,
+      "step": 1306
+    },
+    {
+      "epoch": 0.30321308432896416,
+      "grad_norm": 11.142072308484599,
+      "learning_rate": 2e-06,
+      "loss": 0.299,
+      "step": 1307
+    },
+    {
+      "epoch": 0.30344507597726483,
+      "grad_norm": 10.747760068034689,
+      "learning_rate": 2e-06,
+      "loss": 0.2483,
+      "step": 1308
+    },
+    {
+      "epoch": 0.30367706762556546,
+      "grad_norm": 16.711980069760436,
+      "learning_rate": 2e-06,
+      "loss": 0.3531,
+      "step": 1309
+    },
+    {
+      "epoch": 0.30390905927386613,
+      "grad_norm": 17.991188041443273,
+      "learning_rate": 2e-06,
+      "loss": 0.2774,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3041410509221668,
+      "grad_norm": 11.576762335855799,
+      "learning_rate": 2e-06,
+      "loss": 0.395,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3043730425704675,
+      "grad_norm": 10.466971442475415,
+      "learning_rate": 2e-06,
+      "loss": 0.3074,
+      "step": 1312
+    },
+    {
+      "epoch": 0.3046050342187681,
+      "grad_norm": 18.42622648855536,
+      "learning_rate": 2e-06,
+      "loss": 0.2287,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3048370258670688,
+      "grad_norm": 15.540120049867527,
+      "learning_rate": 2e-06,
+      "loss": 0.3262,
+      "step": 1314
+    },
+    {
+      "epoch": 0.30506901751536947,
+      "grad_norm": 17.587207558740374,
+      "learning_rate": 2e-06,
+      "loss": 0.3041,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3053010091636701,
+      "grad_norm": 17.44985967664141,
+      "learning_rate": 2e-06,
+      "loss": 0.2743,
+      "step": 1316
+    },
+    {
+      "epoch": 0.30553300081197077,
+      "grad_norm": 8.54965869254267,
+      "learning_rate": 2e-06,
+      "loss": 0.2416,
+      "step": 1317
+    },
+    {
+      "epoch": 0.30576499246027145,
+      "grad_norm": 8.979997168292295,
+      "learning_rate": 2e-06,
+      "loss": 0.348,
+      "step": 1318
+    },
+    {
+      "epoch": 0.30599698410857207,
+      "grad_norm": 28.59983116861295,
+      "learning_rate": 2e-06,
+      "loss": 0.4125,
+      "step": 1319
+    },
+    {
+      "epoch": 0.30622897575687275,
+      "grad_norm": 8.07380333347449,
+      "learning_rate": 2e-06,
+      "loss": 0.2392,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3064609674051734,
+      "grad_norm": 10.676150094033074,
+      "learning_rate": 2e-06,
+      "loss": 0.2199,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3066929590534741,
+      "grad_norm": 17.299675012306995,
+      "learning_rate": 2e-06,
+      "loss": 0.2338,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3069249507017747,
+      "grad_norm": 11.668388317530653,
+      "learning_rate": 2e-06,
+      "loss": 0.3085,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3071569423500754,
+      "grad_norm": 13.351683945674617,
+      "learning_rate": 2e-06,
+      "loss": 0.3607,
+      "step": 1324
+    },
+    {
+      "epoch": 0.3073889339983761,
+      "grad_norm": 13.197158579360199,
+      "learning_rate": 2e-06,
+      "loss": 0.3151,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3076209256466767,
+      "grad_norm": 7.283939134417037,
+      "learning_rate": 2e-06,
+      "loss": 0.1872,
+      "step": 1326
+    },
+    {
+      "epoch": 0.3078529172949774,
+      "grad_norm": 10.299996868078303,
+      "learning_rate": 2e-06,
+      "loss": 0.3248,
+      "step": 1327
+    },
+    {
+      "epoch": 0.30808490894327806,
+      "grad_norm": 24.67342231320794,
+      "learning_rate": 2e-06,
+      "loss": 0.3031,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3083169005915787,
+      "grad_norm": 9.223137640402232,
+      "learning_rate": 2e-06,
+      "loss": 0.3476,
+      "step": 1329
+    },
+    {
+      "epoch": 0.30854889223987936,
+      "grad_norm": 13.917374776095183,
+      "learning_rate": 2e-06,
+      "loss": 0.2151,
+      "step": 1330
+    },
+    {
+      "epoch": 0.30878088388818004,
+      "grad_norm": 22.274647106743906,
+      "learning_rate": 2e-06,
+      "loss": 0.3928,
+      "step": 1331
+    },
+    {
+      "epoch": 0.3090128755364807,
+      "grad_norm": 14.705277230122983,
+      "learning_rate": 2e-06,
+      "loss": 0.3337,
+      "step": 1332
+    },
+    {
+      "epoch": 0.30924486718478134,
+      "grad_norm": 16.660797472432396,
+      "learning_rate": 2e-06,
+      "loss": 0.3067,
+      "step": 1333
+    },
+    {
+      "epoch": 0.309476858833082,
+      "grad_norm": 10.20320184386022,
+      "learning_rate": 2e-06,
+      "loss": 0.2807,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3097088504813827,
+      "grad_norm": 14.164750183745458,
+      "learning_rate": 2e-06,
+      "loss": 0.3187,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3099408421296833,
+      "grad_norm": 17.276330857778227,
+      "learning_rate": 2e-06,
+      "loss": 0.4243,
+      "step": 1336
+    },
+    {
+      "epoch": 0.310172833777984,
+      "grad_norm": 15.575398625815712,
+      "learning_rate": 2e-06,
+      "loss": 0.3158,
+      "step": 1337
+    },
+    {
+      "epoch": 0.31040482542628467,
+      "grad_norm": 18.940988228755096,
+      "learning_rate": 2e-06,
+      "loss": 0.2767,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3106368170745853,
+      "grad_norm": 15.728923178327832,
+      "learning_rate": 2e-06,
+      "loss": 0.2889,
+      "step": 1339
+    },
+    {
+      "epoch": 0.31086880872288597,
+      "grad_norm": 12.861312651073032,
+      "learning_rate": 2e-06,
+      "loss": 0.3422,
+      "step": 1340
+    },
+    {
+      "epoch": 0.31110080037118665,
+      "grad_norm": 7.844519890976591,
+      "learning_rate": 2e-06,
+      "loss": 0.1895,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3113327920194873,
+      "grad_norm": 16.205539259080346,
+      "learning_rate": 2e-06,
+      "loss": 0.319,
+      "step": 1342
+    },
+    {
+      "epoch": 0.31156478366778795,
+      "grad_norm": 13.541084983572475,
+      "learning_rate": 2e-06,
+      "loss": 0.3174,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3117967753160886,
+      "grad_norm": 15.138498429814257,
+      "learning_rate": 2e-06,
+      "loss": 0.2069,
+      "step": 1344
+    },
+    {
+      "epoch": 0.3120287669643893,
+      "grad_norm": 17.90538146115018,
+      "learning_rate": 2e-06,
+      "loss": 0.307,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3122607586126899,
+      "grad_norm": 16.52864730654945,
+      "learning_rate": 2e-06,
+      "loss": 0.2979,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3124927502609906,
+      "grad_norm": 10.476892047549619,
+      "learning_rate": 2e-06,
+      "loss": 0.29,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3127247419092913,
+      "grad_norm": 22.11804081976467,
+      "learning_rate": 2e-06,
+      "loss": 0.2963,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3129567335575919,
+      "grad_norm": 14.362786675309335,
+      "learning_rate": 2e-06,
+      "loss": 0.2286,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3131887252058926,
+      "grad_norm": 7.571507878073395,
+      "learning_rate": 2e-06,
+      "loss": 0.1847,
+      "step": 1350
+    },
+    {
+      "epoch": 0.31342071685419326,
+      "grad_norm": 18.40040614466039,
+      "learning_rate": 2e-06,
+      "loss": 0.3439,
+      "step": 1351
+    },
+    {
+      "epoch": 0.31365270850249394,
+      "grad_norm": 29.00454963048086,
+      "learning_rate": 2e-06,
+      "loss": 0.2897,
+      "step": 1352
+    },
+    {
+      "epoch": 0.31388470015079456,
+      "grad_norm": 12.1743071611877,
+      "learning_rate": 2e-06,
+      "loss": 0.3104,
+      "step": 1353
+    },
+    {
+      "epoch": 0.31411669179909524,
+      "grad_norm": 13.407021819241502,
+      "learning_rate": 2e-06,
+      "loss": 0.2899,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3143486834473959,
+      "grad_norm": 13.353584074233229,
+      "learning_rate": 2e-06,
+      "loss": 0.1983,
+      "step": 1355
+    },
+    {
+      "epoch": 0.31458067509569654,
+      "grad_norm": 17.248859424682657,
+      "learning_rate": 2e-06,
+      "loss": 0.3753,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3148126667439972,
+      "grad_norm": 10.014448704532423,
+      "learning_rate": 2e-06,
+      "loss": 0.1834,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3150446583922979,
+      "grad_norm": 21.763953176545574,
+      "learning_rate": 2e-06,
+      "loss": 0.3596,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3152766500405985,
+      "grad_norm": 19.940820725887995,
+      "learning_rate": 2e-06,
+      "loss": 0.2957,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3155086416888992,
+      "grad_norm": 18.912097481709182,
+      "learning_rate": 2e-06,
+      "loss": 0.3685,
+      "step": 1360
+    },
+    {
+      "epoch": 0.31574063333719987,
+      "grad_norm": 11.346471266770104,
+      "learning_rate": 2e-06,
+      "loss": 0.2816,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3159726249855005,
+      "grad_norm": 28.193240828763855,
+      "learning_rate": 2e-06,
+      "loss": 0.4203,
+      "step": 1362
+    },
+    {
+      "epoch": 0.31620461663380117,
+      "grad_norm": 16.73102518384111,
+      "learning_rate": 2e-06,
+      "loss": 0.3564,
+      "step": 1363
+    },
+    {
+      "epoch": 0.31643660828210185,
+      "grad_norm": 9.763506994333493,
+      "learning_rate": 2e-06,
+      "loss": 0.2947,
+      "step": 1364
+    },
+    {
+      "epoch": 0.31666859993040253,
+      "grad_norm": 16.116070672609183,
+      "learning_rate": 2e-06,
+      "loss": 0.2653,
+      "step": 1365
+    },
+    {
+      "epoch": 0.31690059157870315,
+      "grad_norm": 16.30539281629766,
+      "learning_rate": 2e-06,
+      "loss": 0.2291,
+      "step": 1366
+    },
+    {
+      "epoch": 0.31713258322700383,
+      "grad_norm": 19.019320340396867,
+      "learning_rate": 2e-06,
+      "loss": 0.3052,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3173645748753045,
+      "grad_norm": 18.228935063012884,
+      "learning_rate": 2e-06,
+      "loss": 0.4026,
+      "step": 1368
+    },
+    {
+      "epoch": 0.31759656652360513,
+      "grad_norm": 18.242836131999766,
+      "learning_rate": 2e-06,
+      "loss": 0.2933,
+      "step": 1369
+    },
+    {
+      "epoch": 0.3178285581719058,
+      "grad_norm": 11.764367942970903,
+      "learning_rate": 2e-06,
+      "loss": 0.2875,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3180605498202065,
+      "grad_norm": 16.85260895830024,
+      "learning_rate": 2e-06,
+      "loss": 0.3813,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3182925414685071,
+      "grad_norm": 15.517837361759963,
+      "learning_rate": 2e-06,
+      "loss": 0.3062,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3185245331168078,
+      "grad_norm": 16.41448843521921,
+      "learning_rate": 2e-06,
+      "loss": 0.449,
+      "step": 1373
+    },
+    {
+      "epoch": 0.31875652476510846,
+      "grad_norm": 21.721345612017643,
+      "learning_rate": 2e-06,
+      "loss": 0.3781,
+      "step": 1374
+    },
+    {
+      "epoch": 0.31898851641340914,
+      "grad_norm": 19.833120601711492,
+      "learning_rate": 2e-06,
+      "loss": 0.3795,
+      "step": 1375
+    },
+    {
+      "epoch": 0.31922050806170976,
+      "grad_norm": 7.1809944456557,
+      "learning_rate": 2e-06,
+      "loss": 0.2324,
+      "step": 1376
+    },
+    {
+      "epoch": 0.31945249971001044,
+      "grad_norm": 27.35532202666317,
+      "learning_rate": 2e-06,
+      "loss": 0.4693,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3196844913583111,
+      "grad_norm": 14.076587114302116,
+      "learning_rate": 2e-06,
+      "loss": 0.2823,
+      "step": 1378
+    },
+    {
+      "epoch": 0.31991648300661174,
+      "grad_norm": 8.780774638050996,
+      "learning_rate": 2e-06,
+      "loss": 0.2473,
+      "step": 1379
+    },
+    {
+      "epoch": 0.3201484746549124,
+      "grad_norm": 11.116062441609776,
+      "learning_rate": 2e-06,
+      "loss": 0.2186,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3203804663032131,
+      "grad_norm": 8.693893969917667,
+      "learning_rate": 2e-06,
+      "loss": 0.2627,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3206124579515137,
+      "grad_norm": 10.604765368834062,
+      "learning_rate": 2e-06,
+      "loss": 0.1967,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3208444495998144,
+      "grad_norm": 17.142486434058437,
+      "learning_rate": 2e-06,
+      "loss": 0.3608,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3210764412481151,
+      "grad_norm": 12.196660226224022,
+      "learning_rate": 2e-06,
+      "loss": 0.2737,
+      "step": 1384
+    },
+    {
+      "epoch": 0.32130843289641575,
+      "grad_norm": 7.812040078901983,
+      "learning_rate": 2e-06,
+      "loss": 0.2015,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3215404245447164,
+      "grad_norm": 12.31031716391756,
+      "learning_rate": 2e-06,
+      "loss": 0.3417,
+      "step": 1386
+    },
+    {
+      "epoch": 0.32177241619301705,
+      "grad_norm": 5.6895948115371455,
+      "learning_rate": 2e-06,
+      "loss": 0.2746,
+      "step": 1387
+    },
+    {
+      "epoch": 0.32200440784131773,
+      "grad_norm": 26.500274825195785,
+      "learning_rate": 2e-06,
+      "loss": 0.3603,
+      "step": 1388
+    },
+    {
+      "epoch": 0.32223639948961835,
+      "grad_norm": 11.59197120918137,
+      "learning_rate": 2e-06,
+      "loss": 0.2567,
+      "step": 1389
+    },
+    {
+      "epoch": 0.32246839113791903,
+      "grad_norm": 21.22159104373813,
+      "learning_rate": 2e-06,
+      "loss": 0.2878,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3227003827862197,
+      "grad_norm": 7.653899247960632,
+      "learning_rate": 2e-06,
+      "loss": 0.1782,
+      "step": 1391
+    },
+    {
+      "epoch": 0.32293237443452033,
+      "grad_norm": 24.21228884650294,
+      "learning_rate": 2e-06,
+      "loss": 0.4119,
+      "step": 1392
+    },
+    {
+      "epoch": 0.323164366082821,
+      "grad_norm": 19.528609223344635,
+      "learning_rate": 2e-06,
+      "loss": 0.369,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3233963577311217,
+      "grad_norm": 30.983200470204057,
+      "learning_rate": 2e-06,
+      "loss": 0.5261,
+      "step": 1394
+    },
+    {
+      "epoch": 0.32362834937942236,
+      "grad_norm": 16.414562055891793,
+      "learning_rate": 2e-06,
+      "loss": 0.2037,
+      "step": 1395
+    },
+    {
+      "epoch": 0.323860341027723,
+      "grad_norm": 13.884976117418837,
+      "learning_rate": 2e-06,
+      "loss": 0.3318,
+      "step": 1396
+    },
+    {
+      "epoch": 0.32409233267602366,
+      "grad_norm": 17.823166997510224,
+      "learning_rate": 2e-06,
+      "loss": 0.2978,
+      "step": 1397
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 16.094709140807357,
+      "learning_rate": 2e-06,
+      "loss": 0.3587,
+      "step": 1398
+    },
+    {
+      "epoch": 0.32455631597262496,
+      "grad_norm": 21.996263140110763,
+      "learning_rate": 2e-06,
+      "loss": 0.4132,
+      "step": 1399
+    },
+    {
+      "epoch": 0.32478830762092564,
+      "grad_norm": 15.17287807041555,
+      "learning_rate": 2e-06,
+      "loss": 0.1584,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3250202992692263,
+      "grad_norm": 9.076859822738712,
+      "learning_rate": 2e-06,
+      "loss": 0.1804,
+      "step": 1401
+    },
+    {
+      "epoch": 0.32525229091752694,
+      "grad_norm": 12.833433028804583,
+      "learning_rate": 2e-06,
+      "loss": 0.3237,
+      "step": 1402
+    },
+    {
+      "epoch": 0.3254842825658276,
+      "grad_norm": 21.081007557262744,
+      "learning_rate": 2e-06,
+      "loss": 0.3414,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3257162742141283,
+      "grad_norm": 24.44170266561206,
+      "learning_rate": 2e-06,
+      "loss": 0.4184,
+      "step": 1404
+    },
+    {
+      "epoch": 0.325948265862429,
+      "grad_norm": 15.761112175024815,
+      "learning_rate": 2e-06,
+      "loss": 0.3424,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3261802575107296,
+      "grad_norm": 26.682059157151674,
+      "learning_rate": 2e-06,
+      "loss": 0.3506,
+      "step": 1406
+    },
+    {
+      "epoch": 0.3264122491590303,
+      "grad_norm": 26.625870041802795,
+      "learning_rate": 2e-06,
+      "loss": 0.4061,
+      "step": 1407
+    },
+    {
+      "epoch": 0.32664424080733095,
+      "grad_norm": 8.524660616405798,
+      "learning_rate": 2e-06,
+      "loss": 0.3291,
+      "step": 1408
+    },
+    {
+      "epoch": 0.3268762324556316,
+      "grad_norm": 13.785566655746681,
+      "learning_rate": 2e-06,
+      "loss": 0.2278,
+      "step": 1409
+    },
+    {
+      "epoch": 0.32710822410393225,
+      "grad_norm": 10.510689880378692,
+      "learning_rate": 2e-06,
+      "loss": 0.2374,
+      "step": 1410
+    },
+    {
+      "epoch": 0.32734021575223293,
+      "grad_norm": 13.282208345671345,
+      "learning_rate": 2e-06,
+      "loss": 0.3365,
+      "step": 1411
+    },
+    {
+      "epoch": 0.32757220740053355,
+      "grad_norm": 7.200604791903009,
+      "learning_rate": 2e-06,
+      "loss": 0.3043,
+      "step": 1412
+    },
+    {
+      "epoch": 0.32780419904883423,
+      "grad_norm": 16.380763960725297,
+      "learning_rate": 2e-06,
+      "loss": 0.336,
+      "step": 1413
+    },
+    {
+      "epoch": 0.3280361906971349,
+      "grad_norm": 7.862667605788348,
+      "learning_rate": 2e-06,
+      "loss": 0.2364,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3282681823454356,
+      "grad_norm": 13.792537373401919,
+      "learning_rate": 2e-06,
+      "loss": 0.3635,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3285001739937362,
+      "grad_norm": 21.034934320250013,
+      "learning_rate": 2e-06,
+      "loss": 0.4046,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3287321656420369,
+      "grad_norm": 22.086878012364966,
+      "learning_rate": 2e-06,
+      "loss": 0.2852,
+      "step": 1417
+    },
+    {
+      "epoch": 0.32896415729033757,
+      "grad_norm": 14.784229950459304,
+      "learning_rate": 2e-06,
+      "loss": 0.2574,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3291961489386382,
+      "grad_norm": 12.373195748036627,
+      "learning_rate": 2e-06,
+      "loss": 0.2822,
+      "step": 1419
+    },
+    {
+      "epoch": 0.32942814058693887,
+      "grad_norm": 38.01272152718663,
+      "learning_rate": 2e-06,
+      "loss": 0.3616,
+      "step": 1420
+    },
+    {
+      "epoch": 0.32966013223523954,
+      "grad_norm": 11.991749136874276,
+      "learning_rate": 2e-06,
+      "loss": 0.353,
+      "step": 1421
+    },
+    {
+      "epoch": 0.32989212388354017,
+      "grad_norm": 26.97583446550163,
+      "learning_rate": 2e-06,
+      "loss": 0.3066,
+      "step": 1422
+    },
+    {
+      "epoch": 0.33012411553184084,
+      "grad_norm": 15.834084651170842,
+      "learning_rate": 2e-06,
+      "loss": 0.298,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3303561071801415,
+      "grad_norm": 18.92977879289124,
+      "learning_rate": 2e-06,
+      "loss": 0.392,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3305880988284422,
+      "grad_norm": 13.530259540842376,
+      "learning_rate": 2e-06,
+      "loss": 0.3095,
+      "step": 1425
+    },
+    {
+      "epoch": 0.3308200904767428,
+      "grad_norm": 10.827914794149137,
+      "learning_rate": 2e-06,
+      "loss": 0.3398,
+      "step": 1426
+    },
+    {
+      "epoch": 0.3310520821250435,
+      "grad_norm": 9.71939624119191,
+      "learning_rate": 2e-06,
+      "loss": 0.3422,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3312840737733442,
+      "grad_norm": 8.993853404795685,
+      "learning_rate": 2e-06,
+      "loss": 0.2793,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3315160654216448,
+      "grad_norm": 11.037296435326578,
+      "learning_rate": 2e-06,
+      "loss": 0.3753,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3317480570699455,
+      "grad_norm": 15.519371017718381,
+      "learning_rate": 2e-06,
+      "loss": 0.3152,
+      "step": 1430
+    },
+    {
+      "epoch": 0.33198004871824616,
+      "grad_norm": 12.947799922606766,
+      "learning_rate": 2e-06,
+      "loss": 0.3816,
+      "step": 1431
+    },
+    {
+      "epoch": 0.3322120403665468,
+      "grad_norm": 43.44503329570567,
+      "learning_rate": 2e-06,
+      "loss": 0.4163,
+      "step": 1432
+    },
+    {
+      "epoch": 0.33244403201484746,
+      "grad_norm": 10.85914909093172,
+      "learning_rate": 2e-06,
+      "loss": 0.3539,
+      "step": 1433
+    },
+    {
+      "epoch": 0.33267602366314813,
+      "grad_norm": 16.45369684618797,
+      "learning_rate": 2e-06,
+      "loss": 0.223,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3329080153114488,
+      "grad_norm": 11.856791453456106,
+      "learning_rate": 2e-06,
+      "loss": 0.2651,
+      "step": 1435
+    },
+    {
+      "epoch": 0.33314000695974944,
+      "grad_norm": 28.227771959540448,
+      "learning_rate": 2e-06,
+      "loss": 0.356,
+      "step": 1436
+    },
+    {
+      "epoch": 0.3333719986080501,
+      "grad_norm": 16.11516272567042,
+      "learning_rate": 2e-06,
+      "loss": 0.3581,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3336039902563508,
+      "grad_norm": 17.205293488598823,
+      "learning_rate": 2e-06,
+      "loss": 0.305,
+      "step": 1438
+    },
+    {
+      "epoch": 0.3338359819046514,
+      "grad_norm": 18.59915835527598,
+      "learning_rate": 2e-06,
+      "loss": 0.2897,
+      "step": 1439
+    },
+    {
+      "epoch": 0.3340679735529521,
+      "grad_norm": 9.665209145110666,
+      "learning_rate": 2e-06,
+      "loss": 0.3308,
+      "step": 1440
+    },
+    {
+      "epoch": 0.33429996520125277,
+      "grad_norm": 9.68040316568584,
+      "learning_rate": 2e-06,
+      "loss": 0.1829,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3345319568495534,
+      "grad_norm": 13.189970236007511,
+      "learning_rate": 2e-06,
+      "loss": 0.3357,
+      "step": 1442
+    },
+    {
+      "epoch": 0.33476394849785407,
+      "grad_norm": 19.650408705244487,
+      "learning_rate": 2e-06,
+      "loss": 0.3785,
+      "step": 1443
+    },
+    {
+      "epoch": 0.33499594014615475,
+      "grad_norm": 28.296732624494975,
+      "learning_rate": 2e-06,
+      "loss": 0.335,
+      "step": 1444
+    },
+    {
+      "epoch": 0.3352279317944554,
+      "grad_norm": 18.467670187497443,
+      "learning_rate": 2e-06,
+      "loss": 0.3028,
+      "step": 1445
+    },
+    {
+      "epoch": 0.33545992344275605,
+      "grad_norm": 18.270227526717235,
+      "learning_rate": 2e-06,
+      "loss": 0.2988,
+      "step": 1446
+    },
+    {
+      "epoch": 0.3356919150910567,
+      "grad_norm": 11.52991657081852,
+      "learning_rate": 2e-06,
+      "loss": 0.2774,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3359239067393574,
+      "grad_norm": 23.744389862640357,
+      "learning_rate": 2e-06,
+      "loss": 0.548,
+      "step": 1448
+    },
+    {
+      "epoch": 0.336155898387658,
+      "grad_norm": 18.275621368601392,
+      "learning_rate": 2e-06,
+      "loss": 0.4032,
+      "step": 1449
+    },
+    {
+      "epoch": 0.3363878900359587,
+      "grad_norm": 14.829177620147869,
+      "learning_rate": 2e-06,
+      "loss": 0.3983,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3366198816842594,
+      "grad_norm": 13.640819060025391,
+      "learning_rate": 2e-06,
+      "loss": 0.3402,
+      "step": 1451
+    },
+    {
+      "epoch": 0.33685187333256,
+      "grad_norm": 20.325335878357638,
+      "learning_rate": 2e-06,
+      "loss": 0.2938,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3370838649808607,
+      "grad_norm": 10.1654491529205,
+      "learning_rate": 2e-06,
+      "loss": 0.3499,
+      "step": 1453
+    },
+    {
+      "epoch": 0.33731585662916136,
+      "grad_norm": 18.11067937334146,
+      "learning_rate": 2e-06,
+      "loss": 0.3405,
+      "step": 1454
+    },
+    {
+      "epoch": 0.33754784827746204,
+      "grad_norm": 7.213728305081726,
+      "learning_rate": 2e-06,
+      "loss": 0.3285,
+      "step": 1455
+    },
+    {
+      "epoch": 0.33777983992576266,
+      "grad_norm": 8.111878159324867,
+      "learning_rate": 2e-06,
+      "loss": 0.2244,
+      "step": 1456
+    },
+    {
+      "epoch": 0.33801183157406334,
+      "grad_norm": 16.26670030048152,
+      "learning_rate": 2e-06,
+      "loss": 0.3037,
+      "step": 1457
+    },
+    {
+      "epoch": 0.338243823222364,
+      "grad_norm": 24.565372997992675,
+      "learning_rate": 2e-06,
+      "loss": 0.3137,
+      "step": 1458
+    },
+    {
+      "epoch": 0.33847581487066464,
+      "grad_norm": 22.886787785735034,
+      "learning_rate": 2e-06,
+      "loss": 0.3186,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3387078065189653,
+      "grad_norm": 21.436646107450535,
+      "learning_rate": 2e-06,
+      "loss": 0.2531,
+      "step": 1460
+    },
+    {
+      "epoch": 0.338939798167266,
+      "grad_norm": 12.328393696710915,
+      "learning_rate": 2e-06,
+      "loss": 0.3254,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3391717898155666,
+      "grad_norm": 15.206807262565173,
+      "learning_rate": 2e-06,
+      "loss": 0.3559,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3394037814638673,
+      "grad_norm": 14.31015543426358,
+      "learning_rate": 2e-06,
+      "loss": 0.2817,
+      "step": 1463
+    },
+    {
+      "epoch": 0.33963577311216797,
+      "grad_norm": 21.67014327512649,
+      "learning_rate": 2e-06,
+      "loss": 0.3082,
+      "step": 1464
+    },
+    {
+      "epoch": 0.33986776476046865,
+      "grad_norm": 7.896970994527335,
+      "learning_rate": 2e-06,
+      "loss": 0.2985,
+      "step": 1465
+    },
+    {
+      "epoch": 0.34009975640876927,
+      "grad_norm": 10.908642187522922,
+      "learning_rate": 2e-06,
+      "loss": 0.4137,
+      "step": 1466
+    },
+    {
+      "epoch": 0.34033174805706995,
+      "grad_norm": 14.30417006499471,
+      "learning_rate": 2e-06,
+      "loss": 0.3977,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3405637397053706,
+      "grad_norm": 12.724543568104627,
+      "learning_rate": 2e-06,
+      "loss": 0.2614,
+      "step": 1468
+    },
+    {
+      "epoch": 0.34079573135367125,
+      "grad_norm": 13.860781856965406,
+      "learning_rate": 2e-06,
+      "loss": 0.3516,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3410277230019719,
+      "grad_norm": 8.11891853375226,
+      "learning_rate": 2e-06,
+      "loss": 0.2225,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3412597146502726,
+      "grad_norm": 18.084721461864714,
+      "learning_rate": 2e-06,
+      "loss": 0.2586,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3414917062985732,
+      "grad_norm": 7.181534455328431,
+      "learning_rate": 2e-06,
+      "loss": 0.2283,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3417236979468739,
+      "grad_norm": 10.249481209532288,
+      "learning_rate": 2e-06,
+      "loss": 0.2586,
+      "step": 1473
+    },
+    {
+      "epoch": 0.3419556895951746,
+      "grad_norm": 20.235711545193375,
+      "learning_rate": 2e-06,
+      "loss": 0.2793,
+      "step": 1474
+    },
+    {
+      "epoch": 0.34218768124347526,
+      "grad_norm": 12.615681034876618,
+      "learning_rate": 2e-06,
+      "loss": 0.2957,
+      "step": 1475
+    },
+    {
+      "epoch": 0.3424196728917759,
+      "grad_norm": 20.373845008527187,
+      "learning_rate": 2e-06,
+      "loss": 0.5045,
+      "step": 1476
+    },
+    {
+      "epoch": 0.34265166454007656,
+      "grad_norm": 23.646655211696096,
+      "learning_rate": 2e-06,
+      "loss": 0.3433,
+      "step": 1477
+    },
+    {
+      "epoch": 0.34288365618837724,
+      "grad_norm": 10.779640440881222,
+      "learning_rate": 2e-06,
+      "loss": 0.26,
+      "step": 1478
+    },
+    {
+      "epoch": 0.34311564783667786,
+      "grad_norm": 14.955799866697927,
+      "learning_rate": 2e-06,
+      "loss": 0.3575,
+      "step": 1479
+    },
+    {
+      "epoch": 0.34334763948497854,
+      "grad_norm": 23.34240373648745,
+      "learning_rate": 2e-06,
+      "loss": 0.3543,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3435796311332792,
+      "grad_norm": 27.198043313140015,
+      "learning_rate": 2e-06,
+      "loss": 0.3609,
+      "step": 1481
+    },
+    {
+      "epoch": 0.34381162278157984,
+      "grad_norm": 4.740917917219977,
+      "learning_rate": 2e-06,
+      "loss": 0.2064,
+      "step": 1482
+    },
+    {
+      "epoch": 0.3440436144298805,
+      "grad_norm": 17.595434056508047,
+      "learning_rate": 2e-06,
+      "loss": 0.289,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3442756060781812,
+      "grad_norm": 9.572257970860454,
+      "learning_rate": 2e-06,
+      "loss": 0.3235,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3445075977264819,
+      "grad_norm": 21.185559442915206,
+      "learning_rate": 2e-06,
+      "loss": 0.3622,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3447395893747825,
+      "grad_norm": 19.125381912342608,
+      "learning_rate": 2e-06,
+      "loss": 0.3033,
+      "step": 1486
+    },
+    {
+      "epoch": 0.3449715810230832,
+      "grad_norm": 14.398206313708432,
+      "learning_rate": 2e-06,
+      "loss": 0.3011,
+      "step": 1487
+    },
+    {
+      "epoch": 0.34520357267138385,
+      "grad_norm": 20.4757543531186,
+      "learning_rate": 2e-06,
+      "loss": 0.3481,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3454355643196845,
+      "grad_norm": 13.521107412007973,
+      "learning_rate": 2e-06,
+      "loss": 0.3284,
+      "step": 1489
+    },
+    {
+      "epoch": 0.34566755596798515,
+      "grad_norm": 12.82873921884128,
+      "learning_rate": 2e-06,
+      "loss": 0.333,
+      "step": 1490
+    },
+    {
+      "epoch": 0.34589954761628583,
+      "grad_norm": 16.50102131285493,
+      "learning_rate": 2e-06,
+      "loss": 0.3682,
+      "step": 1491
+    },
+    {
+      "epoch": 0.34613153926458645,
+      "grad_norm": 15.968468514516362,
+      "learning_rate": 2e-06,
+      "loss": 0.3645,
+      "step": 1492
+    },
+    {
+      "epoch": 0.34636353091288713,
+      "grad_norm": 15.411384633795938,
+      "learning_rate": 2e-06,
+      "loss": 0.2859,
+      "step": 1493
+    },
+    {
+      "epoch": 0.3465955225611878,
+      "grad_norm": 10.528121007020545,
+      "learning_rate": 2e-06,
+      "loss": 0.3602,
+      "step": 1494
+    },
+    {
+      "epoch": 0.3468275142094885,
+      "grad_norm": 12.50351889191618,
+      "learning_rate": 2e-06,
+      "loss": 0.2303,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3470595058577891,
+      "grad_norm": 22.53265667953538,
+      "learning_rate": 2e-06,
+      "loss": 0.3537,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3472914975060898,
+      "grad_norm": 13.161514617319222,
+      "learning_rate": 2e-06,
+      "loss": 0.3082,
+      "step": 1497
+    },
+    {
+      "epoch": 0.34752348915439046,
+      "grad_norm": 12.514909309899709,
+      "learning_rate": 2e-06,
+      "loss": 0.3135,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3477554808026911,
+      "grad_norm": 27.200429723788545,
+      "learning_rate": 2e-06,
+      "loss": 0.3255,
+      "step": 1499
+    },
+    {
+      "epoch": 0.34798747245099176,
+      "grad_norm": 14.382791715885393,
+      "learning_rate": 2e-06,
+      "loss": 0.2845,
+      "step": 1500
+    },
+    {
+      "epoch": 0.34821946409929244,
+      "grad_norm": 27.770489340950228,
+      "learning_rate": 2e-06,
+      "loss": 0.3061,
+      "step": 1501
+    },
+    {
+      "epoch": 0.34845145574759306,
+      "grad_norm": 13.103877107565271,
+      "learning_rate": 2e-06,
+      "loss": 0.2417,
+      "step": 1502
+    },
+    {
+      "epoch": 0.34868344739589374,
+      "grad_norm": 22.095341754977923,
+      "learning_rate": 2e-06,
+      "loss": 0.3161,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3489154390441944,
+      "grad_norm": 17.66831432896543,
+      "learning_rate": 2e-06,
+      "loss": 0.3733,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3491474306924951,
+      "grad_norm": 23.243882502598925,
+      "learning_rate": 2e-06,
+      "loss": 0.2508,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3493794223407957,
+      "grad_norm": 14.198012691546005,
+      "learning_rate": 2e-06,
+      "loss": 0.2499,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3496114139890964,
+      "grad_norm": 16.38842201576127,
+      "learning_rate": 2e-06,
+      "loss": 0.3034,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3498434056373971,
+      "grad_norm": 12.030502731601553,
+      "learning_rate": 2e-06,
+      "loss": 0.2351,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3500753972856977,
+      "grad_norm": 12.003382984655511,
+      "learning_rate": 2e-06,
+      "loss": 0.2391,
+      "step": 1509
+    },
+    {
+      "epoch": 0.3503073889339984,
+      "grad_norm": 16.653026981636675,
+      "learning_rate": 2e-06,
+      "loss": 0.267,
+      "step": 1510
+    },
+    {
+      "epoch": 0.35053938058229905,
+      "grad_norm": 26.30925965099654,
+      "learning_rate": 2e-06,
+      "loss": 0.2996,
+      "step": 1511
+    },
+    {
+      "epoch": 0.3507713722305997,
+      "grad_norm": 8.268051727806895,
+      "learning_rate": 2e-06,
+      "loss": 0.3707,
+      "step": 1512
+    },
+    {
+      "epoch": 0.35100336387890035,
+      "grad_norm": 16.492831306112947,
+      "learning_rate": 2e-06,
+      "loss": 0.3764,
+      "step": 1513
+    },
+    {
+      "epoch": 0.35123535552720103,
+      "grad_norm": 12.953009967608708,
+      "learning_rate": 2e-06,
+      "loss": 0.3184,
+      "step": 1514
+    },
+    {
+      "epoch": 0.35146734717550165,
+      "grad_norm": 13.868907702767919,
+      "learning_rate": 2e-06,
+      "loss": 0.275,
+      "step": 1515
+    },
+    {
+      "epoch": 0.35169933882380233,
+      "grad_norm": 14.707904973561801,
+      "learning_rate": 2e-06,
+      "loss": 0.3275,
+      "step": 1516
+    },
+    {
+      "epoch": 0.351931330472103,
+      "grad_norm": 18.927232123905426,
+      "learning_rate": 2e-06,
+      "loss": 0.4729,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3521633221204037,
+      "grad_norm": 26.683198978700762,
+      "learning_rate": 2e-06,
+      "loss": 0.3891,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3523953137687043,
+      "grad_norm": 23.200118728090455,
+      "learning_rate": 2e-06,
+      "loss": 0.3123,
+      "step": 1519
+    },
+    {
+      "epoch": 0.352627305417005,
+      "grad_norm": 12.772671965285351,
+      "learning_rate": 2e-06,
+      "loss": 0.2973,
+      "step": 1520
+    },
+    {
+      "epoch": 0.35285929706530567,
+      "grad_norm": 19.03558091042126,
+      "learning_rate": 2e-06,
+      "loss": 0.3802,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3530912887136063,
+      "grad_norm": 18.97245741876137,
+      "learning_rate": 2e-06,
+      "loss": 0.3188,
+      "step": 1522
+    },
+    {
+      "epoch": 0.35332328036190697,
+      "grad_norm": 14.336059096837559,
+      "learning_rate": 2e-06,
+      "loss": 0.3388,
+      "step": 1523
+    },
+    {
+      "epoch": 0.35355527201020764,
+      "grad_norm": 11.160440481969392,
+      "learning_rate": 2e-06,
+      "loss": 0.2824,
+      "step": 1524
+    },
+    {
+      "epoch": 0.35378726365850827,
+      "grad_norm": 23.929660422992608,
+      "learning_rate": 2e-06,
+      "loss": 0.311,
+      "step": 1525
+    },
+    {
+      "epoch": 0.35401925530680894,
+      "grad_norm": 23.916293169824055,
+      "learning_rate": 2e-06,
+      "loss": 0.4163,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3542512469551096,
+      "grad_norm": 20.043119923355555,
+      "learning_rate": 2e-06,
+      "loss": 0.329,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3544832386034103,
+      "grad_norm": 12.250980750726233,
+      "learning_rate": 2e-06,
+      "loss": 0.2934,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3547152302517109,
+      "grad_norm": 14.48483674657129,
+      "learning_rate": 2e-06,
+      "loss": 0.3425,
+      "step": 1529
+    },
+    {
+      "epoch": 0.3549472219000116,
+      "grad_norm": 11.553582803477648,
+      "learning_rate": 2e-06,
+      "loss": 0.3147,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3551792135483123,
+      "grad_norm": 11.560989065944293,
+      "learning_rate": 2e-06,
+      "loss": 0.2615,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3554112051966129,
+      "grad_norm": 11.17415333216845,
+      "learning_rate": 2e-06,
+      "loss": 0.3311,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3556431968449136,
+      "grad_norm": 12.185737847018876,
+      "learning_rate": 2e-06,
+      "loss": 0.2516,
+      "step": 1533
+    },
+    {
+      "epoch": 0.35587518849321426,
+      "grad_norm": 8.191288874419087,
+      "learning_rate": 2e-06,
+      "loss": 0.1495,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3561071801415149,
+      "grad_norm": 10.949054148249836,
+      "learning_rate": 2e-06,
+      "loss": 0.343,
+      "step": 1535
+    },
+    {
+      "epoch": 0.35633917178981556,
+      "grad_norm": 12.278173030546192,
+      "learning_rate": 2e-06,
+      "loss": 0.3576,
+      "step": 1536
+    },
+    {
+      "epoch": 0.35657116343811623,
+      "grad_norm": 13.977798272440518,
+      "learning_rate": 2e-06,
+      "loss": 0.2952,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3568031550864169,
+      "grad_norm": 11.83754990738849,
+      "learning_rate": 2e-06,
+      "loss": 0.2399,
+      "step": 1538
+    },
+    {
+      "epoch": 0.35703514673471753,
+      "grad_norm": 16.49812571960583,
+      "learning_rate": 2e-06,
+      "loss": 0.3304,
+      "step": 1539
+    },
+    {
+      "epoch": 0.3572671383830182,
+      "grad_norm": 17.13588560605484,
+      "learning_rate": 2e-06,
+      "loss": 0.3186,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3574991300313189,
+      "grad_norm": 18.403814271161163,
+      "learning_rate": 2e-06,
+      "loss": 0.2804,
+      "step": 1541
+    },
+    {
+      "epoch": 0.3577311216796195,
+      "grad_norm": 12.827052629703486,
+      "learning_rate": 2e-06,
+      "loss": 0.2458,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3579631133279202,
+      "grad_norm": 12.760937396611903,
+      "learning_rate": 2e-06,
+      "loss": 0.2407,
+      "step": 1543
+    },
+    {
+      "epoch": 0.35819510497622087,
+      "grad_norm": 16.539137967377687,
+      "learning_rate": 2e-06,
+      "loss": 0.383,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3584270966245215,
+      "grad_norm": 20.039107817355806,
+      "learning_rate": 2e-06,
+      "loss": 0.3159,
+      "step": 1545
+    },
+    {
+      "epoch": 0.35865908827282217,
+      "grad_norm": 13.64428431793954,
+      "learning_rate": 2e-06,
+      "loss": 0.2665,
+      "step": 1546
+    },
+    {
+      "epoch": 0.35889107992112285,
+      "grad_norm": 12.42057044851482,
+      "learning_rate": 2e-06,
+      "loss": 0.3102,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3591230715694235,
+      "grad_norm": 22.523044251326006,
+      "learning_rate": 2e-06,
+      "loss": 0.2732,
+      "step": 1548
+    },
+    {
+      "epoch": 0.35935506321772415,
+      "grad_norm": 16.928365338505053,
+      "learning_rate": 2e-06,
+      "loss": 0.3164,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3595870548660248,
+      "grad_norm": 18.232182331487515,
+      "learning_rate": 2e-06,
+      "loss": 0.2678,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3598190465143255,
+      "grad_norm": 21.978434417091222,
+      "learning_rate": 2e-06,
+      "loss": 0.3644,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3600510381626261,
+      "grad_norm": 7.471359625072732,
+      "learning_rate": 2e-06,
+      "loss": 0.273,
+      "step": 1552
+    },
+    {
+      "epoch": 0.3602830298109268,
+      "grad_norm": 19.050119083208457,
+      "learning_rate": 2e-06,
+      "loss": 0.2834,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3605150214592275,
+      "grad_norm": 8.666454101947402,
+      "learning_rate": 2e-06,
+      "loss": 0.2591,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3607470131075281,
+      "grad_norm": 14.71395010544609,
+      "learning_rate": 2e-06,
+      "loss": 0.329,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3609790047558288,
+      "grad_norm": 15.222939559224029,
+      "learning_rate": 2e-06,
+      "loss": 0.2663,
+      "step": 1556
+    },
+    {
+      "epoch": 0.36121099640412946,
+      "grad_norm": 16.19223554787744,
+      "learning_rate": 2e-06,
+      "loss": 0.4093,
+      "step": 1557
+    },
+    {
+      "epoch": 0.36144298805243014,
+      "grad_norm": 12.171363940214102,
+      "learning_rate": 2e-06,
+      "loss": 0.2952,
+      "step": 1558
+    },
+    {
+      "epoch": 0.36167497970073076,
+      "grad_norm": 15.245125862878577,
+      "learning_rate": 2e-06,
+      "loss": 0.4228,
+      "step": 1559
+    },
+    {
+      "epoch": 0.36190697134903144,
+      "grad_norm": 19.04913330454029,
+      "learning_rate": 2e-06,
+      "loss": 0.3924,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3621389629973321,
+      "grad_norm": 15.994627326757682,
+      "learning_rate": 2e-06,
+      "loss": 0.3107,
+      "step": 1561
+    },
+    {
+      "epoch": 0.36237095464563274,
+      "grad_norm": 21.858725247811957,
+      "learning_rate": 2e-06,
+      "loss": 0.4131,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3626029462939334,
+      "grad_norm": 12.86548452333493,
+      "learning_rate": 2e-06,
+      "loss": 0.3945,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3628349379422341,
+      "grad_norm": 11.10614066767929,
+      "learning_rate": 2e-06,
+      "loss": 0.2737,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3630669295905347,
+      "grad_norm": 7.711392585957257,
+      "learning_rate": 2e-06,
+      "loss": 0.2551,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3632989212388354,
+      "grad_norm": 13.634176534870537,
+      "learning_rate": 2e-06,
+      "loss": 0.2395,
+      "step": 1566
+    },
+    {
+      "epoch": 0.36353091288713607,
+      "grad_norm": 10.852789631053897,
+      "learning_rate": 2e-06,
+      "loss": 0.2562,
+      "step": 1567
+    },
+    {
+      "epoch": 0.36376290453543675,
+      "grad_norm": 10.476346649167414,
+      "learning_rate": 2e-06,
+      "loss": 0.2577,
+      "step": 1568
+    },
+    {
+      "epoch": 0.36399489618373737,
+      "grad_norm": 15.507037512786859,
+      "learning_rate": 2e-06,
+      "loss": 0.2622,
+      "step": 1569
+    },
+    {
+      "epoch": 0.36422688783203805,
+      "grad_norm": 19.351470038814583,
+      "learning_rate": 2e-06,
+      "loss": 0.4442,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3644588794803387,
+      "grad_norm": 27.313074395348565,
+      "learning_rate": 2e-06,
+      "loss": 0.4793,
+      "step": 1571
+    },
+    {
+      "epoch": 0.36469087112863935,
+      "grad_norm": 16.532665789399566,
+      "learning_rate": 2e-06,
+      "loss": 0.3345,
+      "step": 1572
+    },
+    {
+      "epoch": 0.36492286277694,
+      "grad_norm": 13.405927824510398,
+      "learning_rate": 2e-06,
+      "loss": 0.3608,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3651548544252407,
+      "grad_norm": 10.875804546859849,
+      "learning_rate": 2e-06,
+      "loss": 0.3128,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3653868460735413,
+      "grad_norm": 18.823706038628313,
+      "learning_rate": 2e-06,
+      "loss": 0.3562,
+      "step": 1575
+    },
+    {
+      "epoch": 0.365618837721842,
+      "grad_norm": 18.61416232947668,
+      "learning_rate": 2e-06,
+      "loss": 0.2618,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3658508293701427,
+      "grad_norm": 24.02212692496374,
+      "learning_rate": 2e-06,
+      "loss": 0.3201,
+      "step": 1577
+    },
+    {
+      "epoch": 0.36608282101844336,
+      "grad_norm": 10.800817628224939,
+      "learning_rate": 2e-06,
+      "loss": 0.2754,
+      "step": 1578
+    },
+    {
+      "epoch": 0.366314812666744,
+      "grad_norm": 9.361010948025115,
+      "learning_rate": 2e-06,
+      "loss": 0.2917,
+      "step": 1579
+    },
+    {
+      "epoch": 0.36654680431504466,
+      "grad_norm": 24.186205696257296,
+      "learning_rate": 2e-06,
+      "loss": 0.3642,
+      "step": 1580
+    },
+    {
+      "epoch": 0.36677879596334534,
+      "grad_norm": 18.65446560349194,
+      "learning_rate": 2e-06,
+      "loss": 0.3754,
+      "step": 1581
+    },
+    {
+      "epoch": 0.36701078761164596,
+      "grad_norm": 24.572827718997146,
+      "learning_rate": 2e-06,
+      "loss": 0.3356,
+      "step": 1582
+    },
+    {
+      "epoch": 0.36724277925994664,
+      "grad_norm": 25.30961911705546,
+      "learning_rate": 2e-06,
+      "loss": 0.3706,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3674747709082473,
+      "grad_norm": 19.183888435579004,
+      "learning_rate": 2e-06,
+      "loss": 0.391,
+      "step": 1584
+    },
+    {
+      "epoch": 0.36770676255654794,
+      "grad_norm": 16.32160541185497,
+      "learning_rate": 2e-06,
+      "loss": 0.2556,
+      "step": 1585
+    },
+    {
+      "epoch": 0.3679387542048486,
+      "grad_norm": 24.10960800511522,
+      "learning_rate": 2e-06,
+      "loss": 0.3104,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3681707458531493,
+      "grad_norm": 15.480985134797434,
+      "learning_rate": 2e-06,
+      "loss": 0.3862,
+      "step": 1587
+    },
+    {
+      "epoch": 0.36840273750145,
+      "grad_norm": 12.142741471902808,
+      "learning_rate": 2e-06,
+      "loss": 0.3515,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3686347291497506,
+      "grad_norm": 11.702276399480478,
+      "learning_rate": 2e-06,
+      "loss": 0.311,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3688667207980513,
+      "grad_norm": 16.70719405848102,
+      "learning_rate": 2e-06,
+      "loss": 0.3269,
+      "step": 1590
+    },
+    {
+      "epoch": 0.36909871244635195,
+      "grad_norm": 12.12225751120716,
+      "learning_rate": 2e-06,
+      "loss": 0.3404,
+      "step": 1591
+    },
+    {
+      "epoch": 0.3693307040946526,
+      "grad_norm": 11.512749205669706,
+      "learning_rate": 2e-06,
+      "loss": 0.2705,
+      "step": 1592
+    },
+    {
+      "epoch": 0.36956269574295325,
+      "grad_norm": 7.127614749150153,
+      "learning_rate": 2e-06,
+      "loss": 0.2072,
+      "step": 1593
+    },
+    {
+      "epoch": 0.36979468739125393,
+      "grad_norm": 23.513456031877478,
+      "learning_rate": 2e-06,
+      "loss": 0.3113,
+      "step": 1594
+    },
+    {
+      "epoch": 0.37002667903955455,
+      "grad_norm": 8.289546448002808,
+      "learning_rate": 2e-06,
+      "loss": 0.3195,
+      "step": 1595
+    },
+    {
+      "epoch": 0.37025867068785523,
+      "grad_norm": 15.104930483599123,
+      "learning_rate": 2e-06,
+      "loss": 0.2434,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3704906623361559,
+      "grad_norm": 17.009568657067756,
+      "learning_rate": 2e-06,
+      "loss": 0.2678,
+      "step": 1597
+    },
+    {
+      "epoch": 0.3707226539844566,
+      "grad_norm": 17.809610643285815,
+      "learning_rate": 2e-06,
+      "loss": 0.387,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3709546456327572,
+      "grad_norm": 20.129437927118126,
+      "learning_rate": 2e-06,
+      "loss": 0.3004,
+      "step": 1599
+    },
+    {
+      "epoch": 0.3711866372810579,
+      "grad_norm": 15.245701649509975,
+      "learning_rate": 2e-06,
+      "loss": 0.3453,
+      "step": 1600
+    },
+    {
+      "epoch": 0.37141862892935856,
+      "grad_norm": 12.497912883042815,
+      "learning_rate": 2e-06,
+      "loss": 0.2064,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3716506205776592,
+      "grad_norm": 12.763218076540506,
+      "learning_rate": 2e-06,
+      "loss": 0.3223,
+      "step": 1602
+    },
+    {
+      "epoch": 0.37188261222595986,
+      "grad_norm": 14.76310202808672,
+      "learning_rate": 2e-06,
+      "loss": 0.4025,
+      "step": 1603
+    },
+    {
+      "epoch": 0.37211460387426054,
+      "grad_norm": 10.854093941092742,
+      "learning_rate": 2e-06,
+      "loss": 0.2103,
+      "step": 1604
+    },
+    {
+      "epoch": 0.37234659552256116,
+      "grad_norm": 20.010957585816257,
+      "learning_rate": 2e-06,
+      "loss": 0.502,
+      "step": 1605
+    },
+    {
+      "epoch": 0.37257858717086184,
+      "grad_norm": 20.461827465123676,
+      "learning_rate": 2e-06,
+      "loss": 0.3678,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3728105788191625,
+      "grad_norm": 12.35407993340569,
+      "learning_rate": 2e-06,
+      "loss": 0.2872,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3730425704674632,
+      "grad_norm": 16.69775007977456,
+      "learning_rate": 2e-06,
+      "loss": 0.3112,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3732745621157638,
+      "grad_norm": 16.24608996111506,
+      "learning_rate": 2e-06,
+      "loss": 0.3128,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3735065537640645,
+      "grad_norm": 21.652530385792378,
+      "learning_rate": 2e-06,
+      "loss": 0.31,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3737385454123652,
+      "grad_norm": 12.622235889749975,
+      "learning_rate": 2e-06,
+      "loss": 0.3509,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3739705370606658,
+      "grad_norm": 17.157816059926006,
+      "learning_rate": 2e-06,
+      "loss": 0.3221,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3742025287089665,
+      "grad_norm": 9.928656920548773,
+      "learning_rate": 2e-06,
+      "loss": 0.2486,
+      "step": 1613
+    },
+    {
+      "epoch": 0.37443452035726715,
+      "grad_norm": 7.878328610402713,
+      "learning_rate": 2e-06,
+      "loss": 0.2231,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3746665120055678,
+      "grad_norm": 21.179210126353382,
+      "learning_rate": 2e-06,
+      "loss": 0.3815,
+      "step": 1615
+    },
+    {
+      "epoch": 0.37489850365386845,
+      "grad_norm": 19.483475307452156,
+      "learning_rate": 2e-06,
+      "loss": 0.2831,
+      "step": 1616
+    },
+    {
+      "epoch": 0.37513049530216913,
+      "grad_norm": 18.918324334501875,
+      "learning_rate": 2e-06,
+      "loss": 0.3439,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3753624869504698,
+      "grad_norm": 13.974426869543807,
+      "learning_rate": 2e-06,
+      "loss": 0.3879,
+      "step": 1618
+    },
+    {
+      "epoch": 0.37559447859877043,
+      "grad_norm": 26.12250512557313,
+      "learning_rate": 2e-06,
+      "loss": 0.4059,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3758264702470711,
+      "grad_norm": 16.61178037941796,
+      "learning_rate": 2e-06,
+      "loss": 0.4071,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3760584618953718,
+      "grad_norm": 11.416733585531539,
+      "learning_rate": 2e-06,
+      "loss": 0.2167,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3762904535436724,
+      "grad_norm": 13.319214185114532,
+      "learning_rate": 2e-06,
+      "loss": 0.3348,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3765224451919731,
+      "grad_norm": 13.655422759429916,
+      "learning_rate": 2e-06,
+      "loss": 0.389,
+      "step": 1623
+    },
+    {
+      "epoch": 0.37675443684027377,
+      "grad_norm": 12.053182330123681,
+      "learning_rate": 2e-06,
+      "loss": 0.2747,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3769864284885744,
+      "grad_norm": 18.83496077772956,
+      "learning_rate": 2e-06,
+      "loss": 0.3359,
+      "step": 1625
+    },
+    {
+      "epoch": 0.37721842013687507,
+      "grad_norm": 17.905370457467995,
+      "learning_rate": 2e-06,
+      "loss": 0.3528,
+      "step": 1626
+    },
+    {
+      "epoch": 0.37745041178517574,
+      "grad_norm": 17.468656650631704,
+      "learning_rate": 2e-06,
+      "loss": 0.2534,
+      "step": 1627
+    },
+    {
+      "epoch": 0.3776824034334764,
+      "grad_norm": 23.713256640121145,
+      "learning_rate": 2e-06,
+      "loss": 0.2424,
+      "step": 1628
+    },
+    {
+      "epoch": 0.37791439508177704,
+      "grad_norm": 11.843935984726398,
+      "learning_rate": 2e-06,
+      "loss": 0.3334,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3781463867300777,
+      "grad_norm": 16.972846583250554,
+      "learning_rate": 2e-06,
+      "loss": 0.3979,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3783783783783784,
+      "grad_norm": 16.37480579398604,
+      "learning_rate": 2e-06,
+      "loss": 0.3796,
+      "step": 1631
+    },
+    {
+      "epoch": 0.378610370026679,
+      "grad_norm": 19.360782354701076,
+      "learning_rate": 2e-06,
+      "loss": 0.3681,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3788423616749797,
+      "grad_norm": 10.154342633825937,
+      "learning_rate": 2e-06,
+      "loss": 0.2564,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3790743533232804,
+      "grad_norm": 22.689267404415812,
+      "learning_rate": 2e-06,
+      "loss": 0.3216,
+      "step": 1634
+    },
+    {
+      "epoch": 0.379306344971581,
+      "grad_norm": 22.7895374079743,
+      "learning_rate": 2e-06,
+      "loss": 0.3923,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3795383366198817,
+      "grad_norm": 23.763833374699864,
+      "learning_rate": 2e-06,
+      "loss": 0.3587,
+      "step": 1636
+    },
+    {
+      "epoch": 0.37977032826818236,
+      "grad_norm": 9.367475197108027,
+      "learning_rate": 2e-06,
+      "loss": 0.2917,
+      "step": 1637
+    },
+    {
+      "epoch": 0.38000231991648303,
+      "grad_norm": 13.235342399140125,
+      "learning_rate": 2e-06,
+      "loss": 0.242,
+      "step": 1638
+    },
+    {
+      "epoch": 0.38023431156478366,
+      "grad_norm": 16.83143948537525,
+      "learning_rate": 2e-06,
+      "loss": 0.3111,
+      "step": 1639
+    },
+    {
+      "epoch": 0.38046630321308433,
+      "grad_norm": 13.590884945633038,
+      "learning_rate": 2e-06,
+      "loss": 0.3164,
+      "step": 1640
+    },
+    {
+      "epoch": 0.380698294861385,
+      "grad_norm": 9.172619097320478,
+      "learning_rate": 2e-06,
+      "loss": 0.2266,
+      "step": 1641
+    },
+    {
+      "epoch": 0.38093028650968563,
+      "grad_norm": 35.29786602689293,
+      "learning_rate": 2e-06,
+      "loss": 0.405,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3811622781579863,
+      "grad_norm": 21.620132837757843,
+      "learning_rate": 2e-06,
+      "loss": 0.3185,
+      "step": 1643
+    },
+    {
+      "epoch": 0.381394269806287,
+      "grad_norm": 15.970781077578724,
+      "learning_rate": 2e-06,
+      "loss": 0.2512,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3816262614545876,
+      "grad_norm": 14.70064658110598,
+      "learning_rate": 2e-06,
+      "loss": 0.2931,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3818582531028883,
+      "grad_norm": 13.887474908567368,
+      "learning_rate": 2e-06,
+      "loss": 0.2872,
+      "step": 1646
+    },
+    {
+      "epoch": 0.38209024475118897,
+      "grad_norm": 19.884616625123037,
+      "learning_rate": 2e-06,
+      "loss": 0.4227,
+      "step": 1647
+    },
+    {
+      "epoch": 0.38232223639948965,
+      "grad_norm": 15.302936516623225,
+      "learning_rate": 2e-06,
+      "loss": 0.3547,
+      "step": 1648
+    },
+    {
+      "epoch": 0.38255422804779027,
+      "grad_norm": 14.32883875962087,
+      "learning_rate": 2e-06,
+      "loss": 0.2966,
+      "step": 1649
+    },
+    {
+      "epoch": 0.38278621969609095,
+      "grad_norm": 11.768538296099742,
+      "learning_rate": 2e-06,
+      "loss": 0.2556,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3830182113443916,
+      "grad_norm": 22.138118983802702,
+      "learning_rate": 2e-06,
+      "loss": 0.3406,
+      "step": 1651
+    },
+    {
+      "epoch": 0.38325020299269225,
+      "grad_norm": 13.11426253348709,
+      "learning_rate": 2e-06,
+      "loss": 0.2848,
+      "step": 1652
+    },
+    {
+      "epoch": 0.3834821946409929,
+      "grad_norm": 21.608807534604892,
+      "learning_rate": 2e-06,
+      "loss": 0.2958,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3837141862892936,
+      "grad_norm": 9.369420646181656,
+      "learning_rate": 2e-06,
+      "loss": 0.3086,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3839461779375942,
+      "grad_norm": 21.073996636387196,
+      "learning_rate": 2e-06,
+      "loss": 0.2979,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3841781695858949,
+      "grad_norm": 20.865881418922612,
+      "learning_rate": 2e-06,
+      "loss": 0.2507,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3844101612341956,
+      "grad_norm": 18.164322650255222,
+      "learning_rate": 2e-06,
+      "loss": 0.2604,
+      "step": 1657
+    },
+    {
+      "epoch": 0.38464215288249626,
+      "grad_norm": 10.166821633179827,
+      "learning_rate": 2e-06,
+      "loss": 0.2436,
+      "step": 1658
+    },
+    {
+      "epoch": 0.3848741445307969,
+      "grad_norm": 20.213626738684884,
+      "learning_rate": 2e-06,
+      "loss": 0.3739,
+      "step": 1659
+    },
+    {
+      "epoch": 0.38510613617909756,
+      "grad_norm": 21.43928377483437,
+      "learning_rate": 2e-06,
+      "loss": 0.3391,
+      "step": 1660
+    },
+    {
+      "epoch": 0.38533812782739824,
+      "grad_norm": 12.13247627608106,
+      "learning_rate": 2e-06,
+      "loss": 0.3681,
+      "step": 1661
+    },
+    {
+      "epoch": 0.38557011947569886,
+      "grad_norm": 15.29816035374616,
+      "learning_rate": 2e-06,
+      "loss": 0.2788,
+      "step": 1662
+    },
+    {
+      "epoch": 0.38580211112399954,
+      "grad_norm": 31.623005766886344,
+      "learning_rate": 2e-06,
+      "loss": 0.2836,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3860341027723002,
+      "grad_norm": 11.67738748278718,
+      "learning_rate": 2e-06,
+      "loss": 0.2403,
+      "step": 1664
+    },
+    {
+      "epoch": 0.38626609442060084,
+      "grad_norm": 13.416501155518167,
+      "learning_rate": 2e-06,
+      "loss": 0.2573,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3864980860689015,
+      "grad_norm": 11.920384086621059,
+      "learning_rate": 2e-06,
+      "loss": 0.2362,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3867300777172022,
+      "grad_norm": 18.72931889505314,
+      "learning_rate": 2e-06,
+      "loss": 0.3919,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3869620693655028,
+      "grad_norm": 8.137213985487316,
+      "learning_rate": 2e-06,
+      "loss": 0.2741,
+      "step": 1668
+    },
+    {
+      "epoch": 0.3871940610138035,
+      "grad_norm": 18.030404210623193,
+      "learning_rate": 2e-06,
+      "loss": 0.2867,
+      "step": 1669
+    },
+    {
+      "epoch": 0.38742605266210417,
+      "grad_norm": 23.792439026114092,
+      "learning_rate": 2e-06,
+      "loss": 0.3914,
+      "step": 1670
+    },
+    {
+      "epoch": 0.38765804431040485,
+      "grad_norm": 16.554147682332868,
+      "learning_rate": 2e-06,
+      "loss": 0.398,
+      "step": 1671
+    },
+    {
+      "epoch": 0.38789003595870547,
+      "grad_norm": 18.15107420377198,
+      "learning_rate": 2e-06,
+      "loss": 0.3045,
+      "step": 1672
+    },
+    {
+      "epoch": 0.38812202760700615,
+      "grad_norm": 16.492763041860865,
+      "learning_rate": 2e-06,
+      "loss": 0.3261,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3883540192553068,
+      "grad_norm": 19.950535800151258,
+      "learning_rate": 2e-06,
+      "loss": 0.3572,
+      "step": 1674
+    },
+    {
+      "epoch": 0.38858601090360745,
+      "grad_norm": 12.332895478452894,
+      "learning_rate": 2e-06,
+      "loss": 0.2484,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3888180025519081,
+      "grad_norm": 18.816090795558736,
+      "learning_rate": 2e-06,
+      "loss": 0.4445,
+      "step": 1676
+    },
+    {
+      "epoch": 0.3890499942002088,
+      "grad_norm": 11.172173233194435,
+      "learning_rate": 2e-06,
+      "loss": 0.2278,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3892819858485094,
+      "grad_norm": 22.84534502367574,
+      "learning_rate": 2e-06,
+      "loss": 0.4344,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3895139774968101,
+      "grad_norm": 9.327583213833936,
+      "learning_rate": 2e-06,
+      "loss": 0.1992,
+      "step": 1679
+    },
+    {
+      "epoch": 0.3897459691451108,
+      "grad_norm": 11.737114391340048,
+      "learning_rate": 2e-06,
+      "loss": 0.2423,
+      "step": 1680
+    },
+    {
+      "epoch": 0.38997796079341146,
+      "grad_norm": 12.417614516132264,
+      "learning_rate": 2e-06,
+      "loss": 0.3601,
+      "step": 1681
+    },
+    {
+      "epoch": 0.3902099524417121,
+      "grad_norm": 20.0655001781096,
+      "learning_rate": 2e-06,
+      "loss": 0.3668,
+      "step": 1682
+    },
+    {
+      "epoch": 0.39044194409001276,
+      "grad_norm": 11.796345412021118,
+      "learning_rate": 2e-06,
+      "loss": 0.319,
+      "step": 1683
+    },
+    {
+      "epoch": 0.39067393573831344,
+      "grad_norm": 17.12049151202,
+      "learning_rate": 2e-06,
+      "loss": 0.3473,
+      "step": 1684
+    },
+    {
+      "epoch": 0.39090592738661406,
+      "grad_norm": 13.345906672000439,
+      "learning_rate": 2e-06,
+      "loss": 0.2233,
+      "step": 1685
+    },
+    {
+      "epoch": 0.39113791903491474,
+      "grad_norm": 13.240562612933,
+      "learning_rate": 2e-06,
+      "loss": 0.3004,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3913699106832154,
+      "grad_norm": 14.035831267305534,
+      "learning_rate": 2e-06,
+      "loss": 0.3141,
+      "step": 1687
+    },
+    {
+      "epoch": 0.39160190233151604,
+      "grad_norm": 13.460145672208174,
+      "learning_rate": 2e-06,
+      "loss": 0.2406,
+      "step": 1688
+    },
+    {
+      "epoch": 0.3918338939798167,
+      "grad_norm": 13.097576569454555,
+      "learning_rate": 2e-06,
+      "loss": 0.2964,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3920658856281174,
+      "grad_norm": 7.234995154351904,
+      "learning_rate": 2e-06,
+      "loss": 0.3318,
+      "step": 1690
+    },
+    {
+      "epoch": 0.39229787727641807,
+      "grad_norm": 21.42498802164779,
+      "learning_rate": 2e-06,
+      "loss": 0.3629,
+      "step": 1691
+    },
+    {
+      "epoch": 0.3925298689247187,
+      "grad_norm": 29.42585971810702,
+      "learning_rate": 2e-06,
+      "loss": 0.398,
+      "step": 1692
+    },
+    {
+      "epoch": 0.3927618605730194,
+      "grad_norm": 14.67197715062144,
+      "learning_rate": 2e-06,
+      "loss": 0.2969,
+      "step": 1693
+    },
+    {
+      "epoch": 0.39299385222132005,
+      "grad_norm": 26.94553900507181,
+      "learning_rate": 2e-06,
+      "loss": 0.4015,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3932258438696207,
+      "grad_norm": 14.144027051794579,
+      "learning_rate": 2e-06,
+      "loss": 0.3245,
+      "step": 1695
+    },
+    {
+      "epoch": 0.39345783551792135,
+      "grad_norm": 13.121990072480953,
+      "learning_rate": 2e-06,
+      "loss": 0.3125,
+      "step": 1696
+    },
+    {
+      "epoch": 0.39368982716622203,
+      "grad_norm": 14.27997838244899,
+      "learning_rate": 2e-06,
+      "loss": 0.34,
+      "step": 1697
+    },
+    {
+      "epoch": 0.39392181881452265,
+      "grad_norm": 13.623194592867334,
+      "learning_rate": 2e-06,
+      "loss": 0.2952,
+      "step": 1698
+    },
+    {
+      "epoch": 0.39415381046282333,
+      "grad_norm": 11.2125716282906,
+      "learning_rate": 2e-06,
+      "loss": 0.2761,
+      "step": 1699
+    },
+    {
+      "epoch": 0.394385802111124,
+      "grad_norm": 12.100281502825506,
+      "learning_rate": 2e-06,
+      "loss": 0.3967,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3946177937594247,
+      "grad_norm": 16.621454147354402,
+      "learning_rate": 2e-06,
+      "loss": 0.3108,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3948497854077253,
+      "grad_norm": 10.874602148419658,
+      "learning_rate": 2e-06,
+      "loss": 0.288,
+      "step": 1702
+    },
+    {
+      "epoch": 0.395081777056026,
+      "grad_norm": 9.863341650478107,
+      "learning_rate": 2e-06,
+      "loss": 0.2274,
+      "step": 1703
+    },
+    {
+      "epoch": 0.39531376870432666,
+      "grad_norm": 12.068471550080465,
+      "learning_rate": 2e-06,
+      "loss": 0.2886,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3955457603526273,
+      "grad_norm": 12.144258760523668,
+      "learning_rate": 2e-06,
+      "loss": 0.2762,
+      "step": 1705
+    },
+    {
+      "epoch": 0.39577775200092796,
+      "grad_norm": 14.432399080933754,
+      "learning_rate": 2e-06,
+      "loss": 0.378,
+      "step": 1706
+    },
+    {
+      "epoch": 0.39600974364922864,
+      "grad_norm": 21.47189822882105,
+      "learning_rate": 2e-06,
+      "loss": 0.379,
+      "step": 1707
+    },
+    {
+      "epoch": 0.39624173529752926,
+      "grad_norm": 14.139723328501157,
+      "learning_rate": 2e-06,
+      "loss": 0.2892,
+      "step": 1708
+    },
+    {
+      "epoch": 0.39647372694582994,
+      "grad_norm": 12.874557329749697,
+      "learning_rate": 2e-06,
+      "loss": 0.25,
+      "step": 1709
+    },
+    {
+      "epoch": 0.3967057185941306,
+      "grad_norm": 19.74261744906016,
+      "learning_rate": 2e-06,
+      "loss": 0.2884,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3969377102424313,
+      "grad_norm": 13.526740825328071,
+      "learning_rate": 2e-06,
+      "loss": 0.2469,
+      "step": 1711
+    },
+    {
+      "epoch": 0.3971697018907319,
+      "grad_norm": 17.29573806918465,
+      "learning_rate": 2e-06,
+      "loss": 0.3666,
+      "step": 1712
+    },
+    {
+      "epoch": 0.3974016935390326,
+      "grad_norm": 22.00074367090561,
+      "learning_rate": 2e-06,
+      "loss": 0.3315,
+      "step": 1713
+    },
+    {
+      "epoch": 0.3976336851873333,
+      "grad_norm": 12.216104847810138,
+      "learning_rate": 2e-06,
+      "loss": 0.3578,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3978656768356339,
+      "grad_norm": 33.55844174882175,
+      "learning_rate": 2e-06,
+      "loss": 0.4602,
+      "step": 1715
+    },
+    {
+      "epoch": 0.3980976684839346,
+      "grad_norm": 6.798634249805131,
+      "learning_rate": 2e-06,
+      "loss": 0.178,
+      "step": 1716
+    },
+    {
+      "epoch": 0.39832966013223525,
+      "grad_norm": 17.31017228664811,
+      "learning_rate": 2e-06,
+      "loss": 0.202,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3985616517805359,
+      "grad_norm": 15.660089205257279,
+      "learning_rate": 2e-06,
+      "loss": 0.3996,
+      "step": 1718
+    },
+    {
+      "epoch": 0.39879364342883655,
+      "grad_norm": 21.208381895451023,
+      "learning_rate": 2e-06,
+      "loss": 0.3099,
+      "step": 1719
+    },
+    {
+      "epoch": 0.39902563507713723,
+      "grad_norm": 14.960757327235939,
+      "learning_rate": 2e-06,
+      "loss": 0.2789,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3992576267254379,
+      "grad_norm": 30.451286730187253,
+      "learning_rate": 2e-06,
+      "loss": 0.3973,
+      "step": 1721
+    },
+    {
+      "epoch": 0.39948961837373853,
+      "grad_norm": 29.620181933264263,
+      "learning_rate": 2e-06,
+      "loss": 0.3298,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3997216100220392,
+      "grad_norm": 13.07179496764708,
+      "learning_rate": 2e-06,
+      "loss": 0.3204,
+      "step": 1723
+    },
+    {
+      "epoch": 0.3999536016703399,
+      "grad_norm": 31.938037556895573,
+      "learning_rate": 2e-06,
+      "loss": 0.4647,
+      "step": 1724
+    },
+    {
+      "epoch": 0.4001855933186405,
+      "grad_norm": 11.521405989449157,
+      "learning_rate": 2e-06,
+      "loss": 0.3678,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4004175849669412,
+      "grad_norm": 13.366109839116676,
+      "learning_rate": 2e-06,
+      "loss": 0.4173,
+      "step": 1726
+    },
+    {
+      "epoch": 0.40064957661524186,
+      "grad_norm": 17.831226552760167,
+      "learning_rate": 2e-06,
+      "loss": 0.3475,
+      "step": 1727
+    },
+    {
+      "epoch": 0.4008815682635425,
+      "grad_norm": 16.687838976166226,
+      "learning_rate": 2e-06,
+      "loss": 0.1913,
+      "step": 1728
+    },
+    {
+      "epoch": 0.40111355991184316,
+      "grad_norm": 8.855862056448904,
+      "learning_rate": 2e-06,
+      "loss": 0.2669,
+      "step": 1729
+    },
+    {
+      "epoch": 0.40134555156014384,
+      "grad_norm": 20.294045592533415,
+      "learning_rate": 2e-06,
+      "loss": 0.4138,
+      "step": 1730
+    },
+    {
+      "epoch": 0.4015775432084445,
+      "grad_norm": 20.282307181816307,
+      "learning_rate": 2e-06,
+      "loss": 0.3926,
+      "step": 1731
+    },
+    {
+      "epoch": 0.40180953485674514,
+      "grad_norm": 17.150362589886832,
+      "learning_rate": 2e-06,
+      "loss": 0.1616,
+      "step": 1732
+    },
+    {
+      "epoch": 0.4020415265050458,
+      "grad_norm": 16.07804665946086,
+      "learning_rate": 2e-06,
+      "loss": 0.2544,
+      "step": 1733
+    },
+    {
+      "epoch": 0.4022735181533465,
+      "grad_norm": 15.330586830512058,
+      "learning_rate": 2e-06,
+      "loss": 0.3177,
+      "step": 1734
+    },
+    {
+      "epoch": 0.4025055098016471,
+      "grad_norm": 12.730645638537759,
+      "learning_rate": 2e-06,
+      "loss": 0.303,
+      "step": 1735
+    },
+    {
+      "epoch": 0.4027375014499478,
+      "grad_norm": 16.595958589744118,
+      "learning_rate": 2e-06,
+      "loss": 0.3585,
+      "step": 1736
+    },
+    {
+      "epoch": 0.4029694930982485,
+      "grad_norm": 8.706544673603664,
+      "learning_rate": 2e-06,
+      "loss": 0.311,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4032014847465491,
+      "grad_norm": 16.090059723709984,
+      "learning_rate": 2e-06,
+      "loss": 0.2455,
+      "step": 1738
+    },
+    {
+      "epoch": 0.4034334763948498,
+      "grad_norm": 8.675772602038379,
+      "learning_rate": 2e-06,
+      "loss": 0.1731,
+      "step": 1739
+    },
+    {
+      "epoch": 0.40366546804315045,
+      "grad_norm": 18.296364172981882,
+      "learning_rate": 2e-06,
+      "loss": 0.3908,
+      "step": 1740
+    },
+    {
+      "epoch": 0.40389745969145113,
+      "grad_norm": 16.115322502756978,
+      "learning_rate": 2e-06,
+      "loss": 0.2754,
+      "step": 1741
+    },
+    {
+      "epoch": 0.40412945133975176,
+      "grad_norm": 20.310432071130496,
+      "learning_rate": 2e-06,
+      "loss": 0.3214,
+      "step": 1742
+    },
+    {
+      "epoch": 0.40436144298805243,
+      "grad_norm": 15.865946919832218,
+      "learning_rate": 2e-06,
+      "loss": 0.3414,
+      "step": 1743
+    },
+    {
+      "epoch": 0.4045934346363531,
+      "grad_norm": 14.325433367552268,
+      "learning_rate": 2e-06,
+      "loss": 0.3477,
+      "step": 1744
+    },
+    {
+      "epoch": 0.40482542628465373,
+      "grad_norm": 16.91522871754405,
+      "learning_rate": 2e-06,
+      "loss": 0.2881,
+      "step": 1745
+    },
+    {
+      "epoch": 0.4050574179329544,
+      "grad_norm": 12.156285411915164,
+      "learning_rate": 2e-06,
+      "loss": 0.2757,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4052894095812551,
+      "grad_norm": 15.700298066784649,
+      "learning_rate": 2e-06,
+      "loss": 0.2911,
+      "step": 1747
+    },
+    {
+      "epoch": 0.4055214012295557,
+      "grad_norm": 14.070074963571189,
+      "learning_rate": 2e-06,
+      "loss": 0.2735,
+      "step": 1748
+    },
+    {
+      "epoch": 0.4057533928778564,
+      "grad_norm": 20.567228387763485,
+      "learning_rate": 2e-06,
+      "loss": 0.3966,
+      "step": 1749
+    },
+    {
+      "epoch": 0.40598538452615707,
+      "grad_norm": 21.568521644903555,
+      "learning_rate": 2e-06,
+      "loss": 0.4069,
+      "step": 1750
+    },
+    {
+      "epoch": 0.40621737617445774,
+      "grad_norm": 24.98305458957807,
+      "learning_rate": 2e-06,
+      "loss": 0.3,
+      "step": 1751
+    },
+    {
+      "epoch": 0.40644936782275837,
+      "grad_norm": 11.182159916834218,
+      "learning_rate": 2e-06,
+      "loss": 0.2476,
+      "step": 1752
+    },
+    {
+      "epoch": 0.40668135947105905,
+      "grad_norm": 17.16773776017756,
+      "learning_rate": 2e-06,
+      "loss": 0.2769,
+      "step": 1753
+    },
+    {
+      "epoch": 0.4069133511193597,
+      "grad_norm": 8.087817209372279,
+      "learning_rate": 2e-06,
+      "loss": 0.2333,
+      "step": 1754
+    },
+    {
+      "epoch": 0.40714534276766035,
+      "grad_norm": 9.737889285495338,
+      "learning_rate": 2e-06,
+      "loss": 0.3203,
+      "step": 1755
+    },
+    {
+      "epoch": 0.407377334415961,
+      "grad_norm": 13.276234387033147,
+      "learning_rate": 2e-06,
+      "loss": 0.3975,
+      "step": 1756
+    },
+    {
+      "epoch": 0.4076093260642617,
+      "grad_norm": 12.833120524941462,
+      "learning_rate": 2e-06,
+      "loss": 0.3393,
+      "step": 1757
+    },
+    {
+      "epoch": 0.4078413177125623,
+      "grad_norm": 10.864122378055994,
+      "learning_rate": 2e-06,
+      "loss": 0.213,
+      "step": 1758
+    },
+    {
+      "epoch": 0.408073309360863,
+      "grad_norm": 17.43651215931162,
+      "learning_rate": 2e-06,
+      "loss": 0.2571,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4083053010091637,
+      "grad_norm": 9.462835101636259,
+      "learning_rate": 2e-06,
+      "loss": 0.2363,
+      "step": 1760
+    },
+    {
+      "epoch": 0.40853729265746436,
+      "grad_norm": 14.60179515190759,
+      "learning_rate": 2e-06,
+      "loss": 0.358,
+      "step": 1761
+    },
+    {
+      "epoch": 0.408769284305765,
+      "grad_norm": 9.35250829968875,
+      "learning_rate": 2e-06,
+      "loss": 0.2229,
+      "step": 1762
+    },
+    {
+      "epoch": 0.40900127595406566,
+      "grad_norm": 16.893224625393078,
+      "learning_rate": 2e-06,
+      "loss": 0.2516,
+      "step": 1763
+    },
+    {
+      "epoch": 0.40923326760236634,
+      "grad_norm": 18.305246064820754,
+      "learning_rate": 2e-06,
+      "loss": 0.4163,
+      "step": 1764
+    },
+    {
+      "epoch": 0.40946525925066696,
+      "grad_norm": 14.924968231349286,
+      "learning_rate": 2e-06,
+      "loss": 0.4418,
+      "step": 1765
+    },
+    {
+      "epoch": 0.40969725089896764,
+      "grad_norm": 19.839080541303428,
+      "learning_rate": 2e-06,
+      "loss": 0.298,
+      "step": 1766
+    },
+    {
+      "epoch": 0.4099292425472683,
+      "grad_norm": 11.601214327833912,
+      "learning_rate": 2e-06,
+      "loss": 0.2756,
+      "step": 1767
+    },
+    {
+      "epoch": 0.41016123419556894,
+      "grad_norm": 10.968524968551193,
+      "learning_rate": 2e-06,
+      "loss": 0.276,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4103932258438696,
+      "grad_norm": 6.826904440619662,
+      "learning_rate": 2e-06,
+      "loss": 0.255,
+      "step": 1769
+    },
+    {
+      "epoch": 0.4106252174921703,
+      "grad_norm": 17.323631423776984,
+      "learning_rate": 2e-06,
+      "loss": 0.3229,
+      "step": 1770
+    },
+    {
+      "epoch": 0.41085720914047097,
+      "grad_norm": 13.28354425099362,
+      "learning_rate": 2e-06,
+      "loss": 0.3321,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4110892007887716,
+      "grad_norm": 10.03252119573095,
+      "learning_rate": 2e-06,
+      "loss": 0.308,
+      "step": 1772
+    },
+    {
+      "epoch": 0.41132119243707227,
+      "grad_norm": 23.554278728776346,
+      "learning_rate": 2e-06,
+      "loss": 0.3286,
+      "step": 1773
+    },
+    {
+      "epoch": 0.41155318408537295,
+      "grad_norm": 13.998688298441905,
+      "learning_rate": 2e-06,
+      "loss": 0.2936,
+      "step": 1774
+    },
+    {
+      "epoch": 0.41178517573367357,
+      "grad_norm": 22.02491607538067,
+      "learning_rate": 2e-06,
+      "loss": 0.3387,
+      "step": 1775
+    },
+    {
+      "epoch": 0.41201716738197425,
+      "grad_norm": 13.99636863261507,
+      "learning_rate": 2e-06,
+      "loss": 0.2389,
+      "step": 1776
+    },
+    {
+      "epoch": 0.4122491590302749,
+      "grad_norm": 14.508439431491093,
+      "learning_rate": 2e-06,
+      "loss": 0.2818,
+      "step": 1777
+    },
+    {
+      "epoch": 0.41248115067857555,
+      "grad_norm": 22.92519999043949,
+      "learning_rate": 2e-06,
+      "loss": 0.4278,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4127131423268762,
+      "grad_norm": 14.599442491978975,
+      "learning_rate": 2e-06,
+      "loss": 0.4189,
+      "step": 1779
+    },
+    {
+      "epoch": 0.4129451339751769,
+      "grad_norm": 20.83978644941764,
+      "learning_rate": 2e-06,
+      "loss": 0.2988,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4131771256234776,
+      "grad_norm": 16.8812419501217,
+      "learning_rate": 2e-06,
+      "loss": 0.2913,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4134091172717782,
+      "grad_norm": 22.735272830579238,
+      "learning_rate": 2e-06,
+      "loss": 0.3513,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4136411089200789,
+      "grad_norm": 9.301251987696757,
+      "learning_rate": 2e-06,
+      "loss": 0.305,
+      "step": 1783
+    },
+    {
+      "epoch": 0.41387310056837956,
+      "grad_norm": 18.72825758685982,
+      "learning_rate": 2e-06,
+      "loss": 0.3613,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4141050922166802,
+      "grad_norm": 33.62199677330242,
+      "learning_rate": 2e-06,
+      "loss": 0.3825,
+      "step": 1785
+    },
+    {
+      "epoch": 0.41433708386498086,
+      "grad_norm": 9.097995766326937,
+      "learning_rate": 2e-06,
+      "loss": 0.3069,
+      "step": 1786
+    },
+    {
+      "epoch": 0.41456907551328154,
+      "grad_norm": 5.898798150314532,
+      "learning_rate": 2e-06,
+      "loss": 0.2112,
+      "step": 1787
+    },
+    {
+      "epoch": 0.41480106716158216,
+      "grad_norm": 16.7077409628407,
+      "learning_rate": 2e-06,
+      "loss": 0.2519,
+      "step": 1788
+    },
+    {
+      "epoch": 0.41503305880988284,
+      "grad_norm": 14.29526419345486,
+      "learning_rate": 2e-06,
+      "loss": 0.2502,
+      "step": 1789
+    },
+    {
+      "epoch": 0.4152650504581835,
+      "grad_norm": 18.288221743113432,
+      "learning_rate": 2e-06,
+      "loss": 0.3396,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4154970421064842,
+      "grad_norm": 8.845939616997287,
+      "learning_rate": 2e-06,
+      "loss": 0.2663,
+      "step": 1791
+    },
+    {
+      "epoch": 0.4157290337547848,
+      "grad_norm": 12.965756776065295,
+      "learning_rate": 2e-06,
+      "loss": 0.3467,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4159610254030855,
+      "grad_norm": 15.914929213419684,
+      "learning_rate": 2e-06,
+      "loss": 0.3533,
+      "step": 1793
+    },
+    {
+      "epoch": 0.41619301705138617,
+      "grad_norm": 14.67947860185235,
+      "learning_rate": 2e-06,
+      "loss": 0.337,
+      "step": 1794
+    },
+    {
+      "epoch": 0.4164250086996868,
+      "grad_norm": 19.295061646529685,
+      "learning_rate": 2e-06,
+      "loss": 0.3472,
+      "step": 1795
+    },
+    {
+      "epoch": 0.41665700034798747,
+      "grad_norm": 8.408293380915646,
+      "learning_rate": 2e-06,
+      "loss": 0.2147,
+      "step": 1796
+    },
+    {
+      "epoch": 0.41688899199628815,
+      "grad_norm": 10.062098294402563,
+      "learning_rate": 2e-06,
+      "loss": 0.2418,
+      "step": 1797
+    },
+    {
+      "epoch": 0.41712098364458877,
+      "grad_norm": 11.463553835251005,
+      "learning_rate": 2e-06,
+      "loss": 0.2798,
+      "step": 1798
+    },
+    {
+      "epoch": 0.41735297529288945,
+      "grad_norm": 18.860146153362425,
+      "learning_rate": 2e-06,
+      "loss": 0.3719,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4175849669411901,
+      "grad_norm": 13.389079107989657,
+      "learning_rate": 2e-06,
+      "loss": 0.2659,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4178169585894908,
+      "grad_norm": 19.569806795295246,
+      "learning_rate": 2e-06,
+      "loss": 0.3267,
+      "step": 1801
+    },
+    {
+      "epoch": 0.41804895023779143,
+      "grad_norm": 12.072125881404723,
+      "learning_rate": 2e-06,
+      "loss": 0.2558,
+      "step": 1802
+    },
+    {
+      "epoch": 0.4182809418860921,
+      "grad_norm": 20.466583975370828,
+      "learning_rate": 2e-06,
+      "loss": 0.3409,
+      "step": 1803
+    },
+    {
+      "epoch": 0.4185129335343928,
+      "grad_norm": 16.68337800654914,
+      "learning_rate": 2e-06,
+      "loss": 0.258,
+      "step": 1804
+    },
+    {
+      "epoch": 0.4187449251826934,
+      "grad_norm": 13.062693644167947,
+      "learning_rate": 2e-06,
+      "loss": 0.2986,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4189769168309941,
+      "grad_norm": 14.091819481275975,
+      "learning_rate": 2e-06,
+      "loss": 0.2461,
+      "step": 1806
+    },
+    {
+      "epoch": 0.41920890847929476,
+      "grad_norm": 18.249361092614027,
+      "learning_rate": 2e-06,
+      "loss": 0.3251,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4194409001275954,
+      "grad_norm": 18.720542222479533,
+      "learning_rate": 2e-06,
+      "loss": 0.2487,
+      "step": 1808
+    },
+    {
+      "epoch": 0.41967289177589606,
+      "grad_norm": 6.835853742556412,
+      "learning_rate": 2e-06,
+      "loss": 0.1909,
+      "step": 1809
+    },
+    {
+      "epoch": 0.41990488342419674,
+      "grad_norm": 21.313152980029592,
+      "learning_rate": 2e-06,
+      "loss": 0.2814,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4201368750724974,
+      "grad_norm": 19.15105976427836,
+      "learning_rate": 2e-06,
+      "loss": 0.383,
+      "step": 1811
+    },
+    {
+      "epoch": 0.42036886672079804,
+      "grad_norm": 11.295461296397983,
+      "learning_rate": 2e-06,
+      "loss": 0.2857,
+      "step": 1812
+    },
+    {
+      "epoch": 0.4206008583690987,
+      "grad_norm": 7.541556412734428,
+      "learning_rate": 2e-06,
+      "loss": 0.2447,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4208328500173994,
+      "grad_norm": 18.291856914356643,
+      "learning_rate": 2e-06,
+      "loss": 0.4423,
+      "step": 1814
+    },
+    {
+      "epoch": 0.4210648416657,
+      "grad_norm": 8.321430630856858,
+      "learning_rate": 2e-06,
+      "loss": 0.2377,
+      "step": 1815
+    },
+    {
+      "epoch": 0.4212968333140007,
+      "grad_norm": 8.615068388386895,
+      "learning_rate": 2e-06,
+      "loss": 0.2333,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4215288249623014,
+      "grad_norm": 18.587239540548637,
+      "learning_rate": 2e-06,
+      "loss": 0.4264,
+      "step": 1817
+    },
+    {
+      "epoch": 0.421760816610602,
+      "grad_norm": 19.258406635069775,
+      "learning_rate": 2e-06,
+      "loss": 0.3658,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4219928082589027,
+      "grad_norm": 4.718414240134861,
+      "learning_rate": 2e-06,
+      "loss": 0.1482,
+      "step": 1819
+    },
+    {
+      "epoch": 0.42222479990720335,
+      "grad_norm": 19.28316315364071,
+      "learning_rate": 2e-06,
+      "loss": 0.2905,
+      "step": 1820
+    },
+    {
+      "epoch": 0.422456791555504,
+      "grad_norm": 12.782558855185856,
+      "learning_rate": 2e-06,
+      "loss": 0.3345,
+      "step": 1821
+    },
+    {
+      "epoch": 0.42268878320380465,
+      "grad_norm": 21.96367145052797,
+      "learning_rate": 2e-06,
+      "loss": 0.339,
+      "step": 1822
+    },
+    {
+      "epoch": 0.42292077485210533,
+      "grad_norm": 15.16140918689698,
+      "learning_rate": 2e-06,
+      "loss": 0.3114,
+      "step": 1823
+    },
+    {
+      "epoch": 0.423152766500406,
+      "grad_norm": 10.192502648774663,
+      "learning_rate": 2e-06,
+      "loss": 0.1998,
+      "step": 1824
+    },
+    {
+      "epoch": 0.42338475814870663,
+      "grad_norm": 16.432028981570053,
+      "learning_rate": 2e-06,
+      "loss": 0.2548,
+      "step": 1825
+    },
+    {
+      "epoch": 0.4236167497970073,
+      "grad_norm": 12.767243541706549,
+      "learning_rate": 2e-06,
+      "loss": 0.2338,
+      "step": 1826
+    },
+    {
+      "epoch": 0.423848741445308,
+      "grad_norm": 23.907500035971236,
+      "learning_rate": 2e-06,
+      "loss": 0.3951,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4240807330936086,
+      "grad_norm": 20.369900827476446,
+      "learning_rate": 2e-06,
+      "loss": 0.3829,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4243127247419093,
+      "grad_norm": 12.49686598283754,
+      "learning_rate": 2e-06,
+      "loss": 0.2322,
+      "step": 1829
+    },
+    {
+      "epoch": 0.42454471639020996,
+      "grad_norm": 17.01708522916691,
+      "learning_rate": 2e-06,
+      "loss": 0.291,
+      "step": 1830
+    },
+    {
+      "epoch": 0.4247767080385106,
+      "grad_norm": 8.448959698676287,
+      "learning_rate": 2e-06,
+      "loss": 0.2344,
+      "step": 1831
+    },
+    {
+      "epoch": 0.42500869968681126,
+      "grad_norm": 13.391004267278555,
+      "learning_rate": 2e-06,
+      "loss": 0.2412,
+      "step": 1832
+    },
+    {
+      "epoch": 0.42524069133511194,
+      "grad_norm": 32.59089985580288,
+      "learning_rate": 2e-06,
+      "loss": 0.4774,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4254726829834126,
+      "grad_norm": 14.407359323922236,
+      "learning_rate": 2e-06,
+      "loss": 0.3374,
+      "step": 1834
+    },
+    {
+      "epoch": 0.42570467463171324,
+      "grad_norm": 12.317732601842495,
+      "learning_rate": 2e-06,
+      "loss": 0.2778,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4259366662800139,
+      "grad_norm": 19.8003091897629,
+      "learning_rate": 2e-06,
+      "loss": 0.418,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4261686579283146,
+      "grad_norm": 11.484951169175416,
+      "learning_rate": 2e-06,
+      "loss": 0.217,
+      "step": 1837
+    },
+    {
+      "epoch": 0.4264006495766152,
+      "grad_norm": 13.45733271265659,
+      "learning_rate": 2e-06,
+      "loss": 0.31,
+      "step": 1838
+    },
+    {
+      "epoch": 0.4266326412249159,
+      "grad_norm": 15.472653333707216,
+      "learning_rate": 2e-06,
+      "loss": 0.2737,
+      "step": 1839
+    },
+    {
+      "epoch": 0.4268646328732166,
+      "grad_norm": 24.25670999804905,
+      "learning_rate": 2e-06,
+      "loss": 0.3363,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4270966245215172,
+      "grad_norm": 19.94460399795094,
+      "learning_rate": 2e-06,
+      "loss": 0.3361,
+      "step": 1841
+    },
+    {
+      "epoch": 0.4273286161698179,
+      "grad_norm": 13.063517439435312,
+      "learning_rate": 2e-06,
+      "loss": 0.3786,
+      "step": 1842
+    },
+    {
+      "epoch": 0.42756060781811855,
+      "grad_norm": 10.281433902687104,
+      "learning_rate": 2e-06,
+      "loss": 0.3698,
+      "step": 1843
+    },
+    {
+      "epoch": 0.42779259946641923,
+      "grad_norm": 14.490270130967572,
+      "learning_rate": 2e-06,
+      "loss": 0.3,
+      "step": 1844
+    },
+    {
+      "epoch": 0.42802459111471985,
+      "grad_norm": 12.412525094210439,
+      "learning_rate": 2e-06,
+      "loss": 0.3253,
+      "step": 1845
+    },
+    {
+      "epoch": 0.42825658276302053,
+      "grad_norm": 20.377294425853872,
+      "learning_rate": 2e-06,
+      "loss": 0.3974,
+      "step": 1846
+    },
+    {
+      "epoch": 0.4284885744113212,
+      "grad_norm": 18.736659411511127,
+      "learning_rate": 2e-06,
+      "loss": 0.4384,
+      "step": 1847
+    },
+    {
+      "epoch": 0.42872056605962183,
+      "grad_norm": 12.415806850795308,
+      "learning_rate": 2e-06,
+      "loss": 0.3279,
+      "step": 1848
+    },
+    {
+      "epoch": 0.4289525577079225,
+      "grad_norm": 23.342252148205862,
+      "learning_rate": 2e-06,
+      "loss": 0.2838,
+      "step": 1849
+    },
+    {
+      "epoch": 0.4291845493562232,
+      "grad_norm": 13.276574562564289,
+      "learning_rate": 2e-06,
+      "loss": 0.2406,
+      "step": 1850
+    },
+    {
+      "epoch": 0.4294165410045238,
+      "grad_norm": 10.567426494535969,
+      "learning_rate": 2e-06,
+      "loss": 0.2654,
+      "step": 1851
+    },
+    {
+      "epoch": 0.4296485326528245,
+      "grad_norm": 8.000618948096822,
+      "learning_rate": 2e-06,
+      "loss": 0.2258,
+      "step": 1852
+    },
+    {
+      "epoch": 0.42988052430112517,
+      "grad_norm": 8.852477043978167,
+      "learning_rate": 2e-06,
+      "loss": 0.292,
+      "step": 1853
+    },
+    {
+      "epoch": 0.43011251594942584,
+      "grad_norm": 8.604882799928584,
+      "learning_rate": 2e-06,
+      "loss": 0.2531,
+      "step": 1854
+    },
+    {
+      "epoch": 0.43034450759772647,
+      "grad_norm": 15.729433265309597,
+      "learning_rate": 2e-06,
+      "loss": 0.3744,
+      "step": 1855
+    },
+    {
+      "epoch": 0.43057649924602714,
+      "grad_norm": 15.22527467684575,
+      "learning_rate": 2e-06,
+      "loss": 0.3011,
+      "step": 1856
+    },
+    {
+      "epoch": 0.4308084908943278,
+      "grad_norm": 14.485139169410346,
+      "learning_rate": 2e-06,
+      "loss": 0.2902,
+      "step": 1857
+    },
+    {
+      "epoch": 0.43104048254262844,
+      "grad_norm": 28.39814798556948,
+      "learning_rate": 2e-06,
+      "loss": 0.3888,
+      "step": 1858
+    },
+    {
+      "epoch": 0.4312724741909291,
+      "grad_norm": 11.340683696031968,
+      "learning_rate": 2e-06,
+      "loss": 0.2898,
+      "step": 1859
+    },
+    {
+      "epoch": 0.4315044658392298,
+      "grad_norm": 9.388704159058799,
+      "learning_rate": 2e-06,
+      "loss": 0.2385,
+      "step": 1860
+    },
+    {
+      "epoch": 0.4317364574875304,
+      "grad_norm": 12.78526627182259,
+      "learning_rate": 2e-06,
+      "loss": 0.2516,
+      "step": 1861
+    },
+    {
+      "epoch": 0.4319684491358311,
+      "grad_norm": 11.221452479221814,
+      "learning_rate": 2e-06,
+      "loss": 0.2505,
+      "step": 1862
+    },
+    {
+      "epoch": 0.4322004407841318,
+      "grad_norm": 8.805058760349361,
+      "learning_rate": 2e-06,
+      "loss": 0.3661,
+      "step": 1863
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 20.94913580794226,
+      "learning_rate": 2e-06,
+      "loss": 0.4231,
+      "step": 1864
+    },
+    {
+      "epoch": 0.4326644240807331,
+      "grad_norm": 18.669289662216194,
+      "learning_rate": 2e-06,
+      "loss": 0.3046,
+      "step": 1865
+    },
+    {
+      "epoch": 0.43289641572903376,
+      "grad_norm": 12.36670710729202,
+      "learning_rate": 2e-06,
+      "loss": 0.2493,
+      "step": 1866
+    },
+    {
+      "epoch": 0.43312840737733443,
+      "grad_norm": 11.374741286569185,
+      "learning_rate": 2e-06,
+      "loss": 0.3335,
+      "step": 1867
+    },
+    {
+      "epoch": 0.43336039902563506,
+      "grad_norm": 33.88185912182654,
+      "learning_rate": 2e-06,
+      "loss": 0.3517,
+      "step": 1868
+    },
+    {
+      "epoch": 0.43359239067393573,
+      "grad_norm": 22.20508173062101,
+      "learning_rate": 2e-06,
+      "loss": 0.2955,
+      "step": 1869
+    },
+    {
+      "epoch": 0.4338243823222364,
+      "grad_norm": 7.340339839461144,
+      "learning_rate": 2e-06,
+      "loss": 0.1892,
+      "step": 1870
+    },
+    {
+      "epoch": 0.43405637397053703,
+      "grad_norm": 20.551325961108073,
+      "learning_rate": 2e-06,
+      "loss": 0.3139,
+      "step": 1871
+    },
+    {
+      "epoch": 0.4342883656188377,
+      "grad_norm": 14.707610072766458,
+      "learning_rate": 2e-06,
+      "loss": 0.3444,
+      "step": 1872
+    },
+    {
+      "epoch": 0.4345203572671384,
+      "grad_norm": 15.411962906532079,
+      "learning_rate": 2e-06,
+      "loss": 0.3017,
+      "step": 1873
+    },
+    {
+      "epoch": 0.43475234891543907,
+      "grad_norm": 13.893129257683169,
+      "learning_rate": 2e-06,
+      "loss": 0.3638,
+      "step": 1874
+    },
+    {
+      "epoch": 0.4349843405637397,
+      "grad_norm": 16.280513594402603,
+      "learning_rate": 2e-06,
+      "loss": 0.2568,
+      "step": 1875
+    },
+    {
+      "epoch": 0.43521633221204037,
+      "grad_norm": 18.16785510645439,
+      "learning_rate": 2e-06,
+      "loss": 0.3422,
+      "step": 1876
+    },
+    {
+      "epoch": 0.43544832386034105,
+      "grad_norm": 9.750925911740046,
+      "learning_rate": 2e-06,
+      "loss": 0.2206,
+      "step": 1877
+    },
+    {
+      "epoch": 0.43568031550864167,
+      "grad_norm": 13.192130318022839,
+      "learning_rate": 2e-06,
+      "loss": 0.3196,
+      "step": 1878
+    },
+    {
+      "epoch": 0.43591230715694235,
+      "grad_norm": 12.223261978383052,
+      "learning_rate": 2e-06,
+      "loss": 0.2715,
+      "step": 1879
+    },
+    {
+      "epoch": 0.436144298805243,
+      "grad_norm": 11.059180241388743,
+      "learning_rate": 2e-06,
+      "loss": 0.2609,
+      "step": 1880
+    },
+    {
+      "epoch": 0.43637629045354365,
+      "grad_norm": 15.662230117945928,
+      "learning_rate": 2e-06,
+      "loss": 0.2097,
+      "step": 1881
+    },
+    {
+      "epoch": 0.4366082821018443,
+      "grad_norm": 7.331387077331177,
+      "learning_rate": 2e-06,
+      "loss": 0.2691,
+      "step": 1882
+    },
+    {
+      "epoch": 0.436840273750145,
+      "grad_norm": 12.61809767481381,
+      "learning_rate": 2e-06,
+      "loss": 0.2374,
+      "step": 1883
+    },
+    {
+      "epoch": 0.4370722653984457,
+      "grad_norm": 12.78740040484834,
+      "learning_rate": 2e-06,
+      "loss": 0.2293,
+      "step": 1884
+    },
+    {
+      "epoch": 0.4373042570467463,
+      "grad_norm": 14.424100789930472,
+      "learning_rate": 2e-06,
+      "loss": 0.3617,
+      "step": 1885
+    },
+    {
+      "epoch": 0.437536248695047,
+      "grad_norm": 80.96212131101774,
+      "learning_rate": 2e-06,
+      "loss": 0.4131,
+      "step": 1886
+    },
+    {
+      "epoch": 0.43776824034334766,
+      "grad_norm": 15.011266652601709,
+      "learning_rate": 2e-06,
+      "loss": 0.3044,
+      "step": 1887
+    },
+    {
+      "epoch": 0.4380002319916483,
+      "grad_norm": 11.718612332506902,
+      "learning_rate": 2e-06,
+      "loss": 0.2712,
+      "step": 1888
+    },
+    {
+      "epoch": 0.43823222363994896,
+      "grad_norm": 9.69556958468024,
+      "learning_rate": 2e-06,
+      "loss": 0.2875,
+      "step": 1889
+    },
+    {
+      "epoch": 0.43846421528824964,
+      "grad_norm": 13.37107182380002,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 1890
+    },
+    {
+      "epoch": 0.43869620693655026,
+      "grad_norm": 25.52509004317602,
+      "learning_rate": 2e-06,
+      "loss": 0.2865,
+      "step": 1891
+    },
+    {
+      "epoch": 0.43892819858485094,
+      "grad_norm": 25.557432614648956,
+      "learning_rate": 2e-06,
+      "loss": 0.2095,
+      "step": 1892
+    },
+    {
+      "epoch": 0.4391601902331516,
+      "grad_norm": 10.227165672394735,
+      "learning_rate": 2e-06,
+      "loss": 0.1997,
+      "step": 1893
+    },
+    {
+      "epoch": 0.4393921818814523,
+      "grad_norm": 10.833971024024654,
+      "learning_rate": 2e-06,
+      "loss": 0.3021,
+      "step": 1894
+    },
+    {
+      "epoch": 0.4396241735297529,
+      "grad_norm": 6.8477809683100395,
+      "learning_rate": 2e-06,
+      "loss": 0.2074,
+      "step": 1895
+    },
+    {
+      "epoch": 0.4398561651780536,
+      "grad_norm": 12.612097699683517,
+      "learning_rate": 2e-06,
+      "loss": 0.268,
+      "step": 1896
+    },
+    {
+      "epoch": 0.44008815682635427,
+      "grad_norm": 14.93894889485441,
+      "learning_rate": 2e-06,
+      "loss": 0.295,
+      "step": 1897
+    },
+    {
+      "epoch": 0.4403201484746549,
+      "grad_norm": 12.675335108037462,
+      "learning_rate": 2e-06,
+      "loss": 0.2563,
+      "step": 1898
+    },
+    {
+      "epoch": 0.44055214012295557,
+      "grad_norm": 14.079067493345839,
+      "learning_rate": 2e-06,
+      "loss": 0.252,
+      "step": 1899
+    },
+    {
+      "epoch": 0.44078413177125625,
+      "grad_norm": 22.713016972391877,
+      "learning_rate": 2e-06,
+      "loss": 0.3575,
+      "step": 1900
+    },
+    {
+      "epoch": 0.44101612341955687,
+      "grad_norm": 16.909479818805984,
+      "learning_rate": 2e-06,
+      "loss": 0.2408,
+      "step": 1901
+    },
+    {
+      "epoch": 0.44124811506785755,
+      "grad_norm": 26.53615495358387,
+      "learning_rate": 2e-06,
+      "loss": 0.3782,
+      "step": 1902
+    },
+    {
+      "epoch": 0.4414801067161582,
+      "grad_norm": 11.637938103347429,
+      "learning_rate": 2e-06,
+      "loss": 0.2748,
+      "step": 1903
+    },
+    {
+      "epoch": 0.4417120983644589,
+      "grad_norm": 13.790174253848365,
+      "learning_rate": 2e-06,
+      "loss": 0.3713,
+      "step": 1904
+    },
+    {
+      "epoch": 0.4419440900127595,
+      "grad_norm": 12.622131028191134,
+      "learning_rate": 2e-06,
+      "loss": 0.2575,
+      "step": 1905
+    },
+    {
+      "epoch": 0.4421760816610602,
+      "grad_norm": 11.39969737402925,
+      "learning_rate": 2e-06,
+      "loss": 0.2648,
+      "step": 1906
+    },
+    {
+      "epoch": 0.4424080733093609,
+      "grad_norm": 16.675642271360555,
+      "learning_rate": 2e-06,
+      "loss": 0.4741,
+      "step": 1907
+    },
+    {
+      "epoch": 0.4426400649576615,
+      "grad_norm": 10.78164946540168,
+      "learning_rate": 2e-06,
+      "loss": 0.2563,
+      "step": 1908
+    },
+    {
+      "epoch": 0.4428720566059622,
+      "grad_norm": 19.68038839531077,
+      "learning_rate": 2e-06,
+      "loss": 0.4349,
+      "step": 1909
+    },
+    {
+      "epoch": 0.44310404825426286,
+      "grad_norm": 14.168099510681532,
+      "learning_rate": 2e-06,
+      "loss": 0.3591,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4433360399025635,
+      "grad_norm": 13.257912219731839,
+      "learning_rate": 2e-06,
+      "loss": 0.2278,
+      "step": 1911
+    },
+    {
+      "epoch": 0.44356803155086416,
+      "grad_norm": 28.111517131517505,
+      "learning_rate": 2e-06,
+      "loss": 0.2837,
+      "step": 1912
+    },
+    {
+      "epoch": 0.44380002319916484,
+      "grad_norm": 25.10965649267154,
+      "learning_rate": 2e-06,
+      "loss": 0.3274,
+      "step": 1913
+    },
+    {
+      "epoch": 0.4440320148474655,
+      "grad_norm": 14.93557483973632,
+      "learning_rate": 2e-06,
+      "loss": 0.2515,
+      "step": 1914
+    },
+    {
+      "epoch": 0.44426400649576614,
+      "grad_norm": 16.642468175359035,
+      "learning_rate": 2e-06,
+      "loss": 0.3378,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4444959981440668,
+      "grad_norm": 6.84438606505179,
+      "learning_rate": 2e-06,
+      "loss": 0.3409,
+      "step": 1916
+    },
+    {
+      "epoch": 0.4447279897923675,
+      "grad_norm": 8.477797064840939,
+      "learning_rate": 2e-06,
+      "loss": 0.3412,
+      "step": 1917
+    },
+    {
+      "epoch": 0.4449599814406681,
+      "grad_norm": 12.99853283907464,
+      "learning_rate": 2e-06,
+      "loss": 0.2441,
+      "step": 1918
+    },
+    {
+      "epoch": 0.4451919730889688,
+      "grad_norm": 10.977481652548592,
+      "learning_rate": 2e-06,
+      "loss": 0.3244,
+      "step": 1919
+    },
+    {
+      "epoch": 0.4454239647372695,
+      "grad_norm": 10.570234822426068,
+      "learning_rate": 2e-06,
+      "loss": 0.2877,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4456559563855701,
+      "grad_norm": 6.860732361741255,
+      "learning_rate": 2e-06,
+      "loss": 0.2442,
+      "step": 1921
+    },
+    {
+      "epoch": 0.4458879480338708,
+      "grad_norm": 19.302410055233295,
+      "learning_rate": 2e-06,
+      "loss": 0.4067,
+      "step": 1922
+    },
+    {
+      "epoch": 0.44611993968217145,
+      "grad_norm": 18.20793820249486,
+      "learning_rate": 2e-06,
+      "loss": 0.3678,
+      "step": 1923
+    },
+    {
+      "epoch": 0.44635193133047213,
+      "grad_norm": 8.97289273170762,
+      "learning_rate": 2e-06,
+      "loss": 0.2128,
+      "step": 1924
+    },
+    {
+      "epoch": 0.44658392297877275,
+      "grad_norm": 6.3471643353963,
+      "learning_rate": 2e-06,
+      "loss": 0.1974,
+      "step": 1925
+    },
+    {
+      "epoch": 0.44681591462707343,
+      "grad_norm": 12.638237055112246,
+      "learning_rate": 2e-06,
+      "loss": 0.3043,
+      "step": 1926
+    },
+    {
+      "epoch": 0.4470479062753741,
+      "grad_norm": 16.3087905268165,
+      "learning_rate": 2e-06,
+      "loss": 0.3217,
+      "step": 1927
+    },
+    {
+      "epoch": 0.44727989792367473,
+      "grad_norm": 6.312405813391441,
+      "learning_rate": 2e-06,
+      "loss": 0.2761,
+      "step": 1928
+    },
+    {
+      "epoch": 0.4475118895719754,
+      "grad_norm": 9.42320757238294,
+      "learning_rate": 2e-06,
+      "loss": 0.178,
+      "step": 1929
+    },
+    {
+      "epoch": 0.4477438812202761,
+      "grad_norm": 14.571508418966776,
+      "learning_rate": 2e-06,
+      "loss": 0.3671,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4479758728685767,
+      "grad_norm": 11.19866672252569,
+      "learning_rate": 2e-06,
+      "loss": 0.4218,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4482078645168774,
+      "grad_norm": 18.094999969369557,
+      "learning_rate": 2e-06,
+      "loss": 0.2873,
+      "step": 1932
+    },
+    {
+      "epoch": 0.44843985616517806,
+      "grad_norm": 17.776408186188583,
+      "learning_rate": 2e-06,
+      "loss": 0.212,
+      "step": 1933
+    },
+    {
+      "epoch": 0.44867184781347874,
+      "grad_norm": 13.319460218034406,
+      "learning_rate": 2e-06,
+      "loss": 0.4009,
+      "step": 1934
+    },
+    {
+      "epoch": 0.44890383946177936,
+      "grad_norm": 20.49299261920701,
+      "learning_rate": 2e-06,
+      "loss": 0.3528,
+      "step": 1935
+    },
+    {
+      "epoch": 0.44913583111008004,
+      "grad_norm": 12.207768363429762,
+      "learning_rate": 2e-06,
+      "loss": 0.2797,
+      "step": 1936
+    },
+    {
+      "epoch": 0.4493678227583807,
+      "grad_norm": 13.241157529714169,
+      "learning_rate": 2e-06,
+      "loss": 0.2463,
+      "step": 1937
+    },
+    {
+      "epoch": 0.44959981440668134,
+      "grad_norm": 11.964061797665712,
+      "learning_rate": 2e-06,
+      "loss": 0.3486,
+      "step": 1938
+    },
+    {
+      "epoch": 0.449831806054982,
+      "grad_norm": 17.760405028259854,
+      "learning_rate": 2e-06,
+      "loss": 0.2826,
+      "step": 1939
+    },
+    {
+      "epoch": 0.4500637977032827,
+      "grad_norm": 18.835286779046086,
+      "learning_rate": 2e-06,
+      "loss": 0.2983,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4502957893515833,
+      "grad_norm": 10.061056247616934,
+      "learning_rate": 2e-06,
+      "loss": 0.2066,
+      "step": 1941
+    },
+    {
+      "epoch": 0.450527780999884,
+      "grad_norm": 15.190416553972092,
+      "learning_rate": 2e-06,
+      "loss": 0.2897,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4507597726481847,
+      "grad_norm": 18.100146573334026,
+      "learning_rate": 2e-06,
+      "loss": 0.2557,
+      "step": 1943
+    },
+    {
+      "epoch": 0.45099176429648535,
+      "grad_norm": 12.145254242892683,
+      "learning_rate": 2e-06,
+      "loss": 0.2617,
+      "step": 1944
+    },
+    {
+      "epoch": 0.451223755944786,
+      "grad_norm": 8.675074169684049,
+      "learning_rate": 2e-06,
+      "loss": 0.2641,
+      "step": 1945
+    },
+    {
+      "epoch": 0.45145574759308665,
+      "grad_norm": 14.379004029242951,
+      "learning_rate": 2e-06,
+      "loss": 0.3171,
+      "step": 1946
+    },
+    {
+      "epoch": 0.45168773924138733,
+      "grad_norm": 16.35450216910894,
+      "learning_rate": 2e-06,
+      "loss": 0.2428,
+      "step": 1947
+    },
+    {
+      "epoch": 0.45191973088968795,
+      "grad_norm": 15.501856211045736,
+      "learning_rate": 2e-06,
+      "loss": 0.2962,
+      "step": 1948
+    },
+    {
+      "epoch": 0.45215172253798863,
+      "grad_norm": 13.519337257061718,
+      "learning_rate": 2e-06,
+      "loss": 0.2869,
+      "step": 1949
+    },
+    {
+      "epoch": 0.4523837141862893,
+      "grad_norm": 22.08178341082319,
+      "learning_rate": 2e-06,
+      "loss": 0.3634,
+      "step": 1950
+    },
+    {
+      "epoch": 0.45261570583458993,
+      "grad_norm": 19.244466854707706,
+      "learning_rate": 2e-06,
+      "loss": 0.4142,
+      "step": 1951
+    },
+    {
+      "epoch": 0.4528476974828906,
+      "grad_norm": 13.45744682479427,
+      "learning_rate": 2e-06,
+      "loss": 0.3247,
+      "step": 1952
+    },
+    {
+      "epoch": 0.4530796891311913,
+      "grad_norm": 19.971997627098528,
+      "learning_rate": 2e-06,
+      "loss": 0.2832,
+      "step": 1953
+    },
+    {
+      "epoch": 0.45331168077949197,
+      "grad_norm": 7.784071208122583,
+      "learning_rate": 2e-06,
+      "loss": 0.1705,
+      "step": 1954
+    },
+    {
+      "epoch": 0.4535436724277926,
+      "grad_norm": 26.175747345649796,
+      "learning_rate": 2e-06,
+      "loss": 0.2603,
+      "step": 1955
+    },
+    {
+      "epoch": 0.45377566407609327,
+      "grad_norm": 9.977944745048129,
+      "learning_rate": 2e-06,
+      "loss": 0.3606,
+      "step": 1956
+    },
+    {
+      "epoch": 0.45400765572439394,
+      "grad_norm": 9.241863633870448,
+      "learning_rate": 2e-06,
+      "loss": 0.1683,
+      "step": 1957
+    },
+    {
+      "epoch": 0.45423964737269457,
+      "grad_norm": 7.854276987996447,
+      "learning_rate": 2e-06,
+      "loss": 0.2218,
+      "step": 1958
+    },
+    {
+      "epoch": 0.45447163902099524,
+      "grad_norm": 17.508549301128976,
+      "learning_rate": 2e-06,
+      "loss": 0.265,
+      "step": 1959
+    },
+    {
+      "epoch": 0.4547036306692959,
+      "grad_norm": 14.495906592291902,
+      "learning_rate": 2e-06,
+      "loss": 0.2092,
+      "step": 1960
+    },
+    {
+      "epoch": 0.45493562231759654,
+      "grad_norm": 11.396050739221131,
+      "learning_rate": 2e-06,
+      "loss": 0.2218,
+      "step": 1961
+    },
+    {
+      "epoch": 0.4551676139658972,
+      "grad_norm": 17.21341794034456,
+      "learning_rate": 2e-06,
+      "loss": 0.4191,
+      "step": 1962
+    },
+    {
+      "epoch": 0.4553996056141979,
+      "grad_norm": 39.71999834351291,
+      "learning_rate": 2e-06,
+      "loss": 0.4364,
+      "step": 1963
+    },
+    {
+      "epoch": 0.4556315972624986,
+      "grad_norm": 14.344933276382463,
+      "learning_rate": 2e-06,
+      "loss": 0.2449,
+      "step": 1964
+    },
+    {
+      "epoch": 0.4558635889107992,
+      "grad_norm": 15.083856831023695,
+      "learning_rate": 2e-06,
+      "loss": 0.2171,
+      "step": 1965
+    },
+    {
+      "epoch": 0.4560955805590999,
+      "grad_norm": 30.1833620615398,
+      "learning_rate": 2e-06,
+      "loss": 0.3772,
+      "step": 1966
+    },
+    {
+      "epoch": 0.45632757220740056,
+      "grad_norm": 5.694038865160228,
+      "learning_rate": 2e-06,
+      "loss": 0.1896,
+      "step": 1967
+    },
+    {
+      "epoch": 0.4565595638557012,
+      "grad_norm": 7.238584483897142,
+      "learning_rate": 2e-06,
+      "loss": 0.2347,
+      "step": 1968
+    },
+    {
+      "epoch": 0.45679155550400186,
+      "grad_norm": 17.40421564491547,
+      "learning_rate": 2e-06,
+      "loss": 0.3307,
+      "step": 1969
+    },
+    {
+      "epoch": 0.45702354715230253,
+      "grad_norm": 11.75504415856113,
+      "learning_rate": 2e-06,
+      "loss": 0.2683,
+      "step": 1970
+    },
+    {
+      "epoch": 0.45725553880060316,
+      "grad_norm": 12.08785488011168,
+      "learning_rate": 2e-06,
+      "loss": 0.2822,
+      "step": 1971
+    },
+    {
+      "epoch": 0.45748753044890383,
+      "grad_norm": 5.746202672849507,
+      "learning_rate": 2e-06,
+      "loss": 0.2149,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4577195220972045,
+      "grad_norm": 10.556481746351935,
+      "learning_rate": 2e-06,
+      "loss": 0.2994,
+      "step": 1973
+    },
+    {
+      "epoch": 0.45795151374550513,
+      "grad_norm": 13.933486628952526,
+      "learning_rate": 2e-06,
+      "loss": 0.2687,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4581835053938058,
+      "grad_norm": 19.368610008099733,
+      "learning_rate": 2e-06,
+      "loss": 0.2651,
+      "step": 1975
+    },
+    {
+      "epoch": 0.4584154970421065,
+      "grad_norm": 7.269647480537984,
+      "learning_rate": 2e-06,
+      "loss": 0.2359,
+      "step": 1976
+    },
+    {
+      "epoch": 0.45864748869040717,
+      "grad_norm": 14.717656516693904,
+      "learning_rate": 2e-06,
+      "loss": 0.3163,
+      "step": 1977
+    },
+    {
+      "epoch": 0.4588794803387078,
+      "grad_norm": 13.554029836416927,
+      "learning_rate": 2e-06,
+      "loss": 0.3661,
+      "step": 1978
+    },
+    {
+      "epoch": 0.45911147198700847,
+      "grad_norm": 14.72655286380002,
+      "learning_rate": 2e-06,
+      "loss": 0.2868,
+      "step": 1979
+    },
+    {
+      "epoch": 0.45934346363530915,
+      "grad_norm": 19.754084297636542,
+      "learning_rate": 2e-06,
+      "loss": 0.4112,
+      "step": 1980
+    },
+    {
+      "epoch": 0.45957545528360977,
+      "grad_norm": 18.040682007952025,
+      "learning_rate": 2e-06,
+      "loss": 0.3363,
+      "step": 1981
+    },
+    {
+      "epoch": 0.45980744693191045,
+      "grad_norm": 23.563968224340655,
+      "learning_rate": 2e-06,
+      "loss": 0.2206,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4600394385802111,
+      "grad_norm": 14.553620166412069,
+      "learning_rate": 2e-06,
+      "loss": 0.2806,
+      "step": 1983
+    },
+    {
+      "epoch": 0.46027143022851175,
+      "grad_norm": 7.424935719456725,
+      "learning_rate": 2e-06,
+      "loss": 0.2433,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4605034218768124,
+      "grad_norm": 12.02707486472308,
+      "learning_rate": 2e-06,
+      "loss": 0.1909,
+      "step": 1985
+    },
+    {
+      "epoch": 0.4607354135251131,
+      "grad_norm": 26.349289928963117,
+      "learning_rate": 2e-06,
+      "loss": 0.288,
+      "step": 1986
+    },
+    {
+      "epoch": 0.4609674051734138,
+      "grad_norm": 11.77000173436258,
+      "learning_rate": 2e-06,
+      "loss": 0.2468,
+      "step": 1987
+    },
+    {
+      "epoch": 0.4611993968217144,
+      "grad_norm": 14.221580423288055,
+      "learning_rate": 2e-06,
+      "loss": 0.324,
+      "step": 1988
+    },
+    {
+      "epoch": 0.4614313884700151,
+      "grad_norm": 19.491108377638728,
+      "learning_rate": 2e-06,
+      "loss": 0.413,
+      "step": 1989
+    },
+    {
+      "epoch": 0.46166338011831576,
+      "grad_norm": 16.717819048175322,
+      "learning_rate": 2e-06,
+      "loss": 0.4165,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4618953717666164,
+      "grad_norm": 16.506916137038182,
+      "learning_rate": 2e-06,
+      "loss": 0.371,
+      "step": 1991
+    },
+    {
+      "epoch": 0.46212736341491706,
+      "grad_norm": 12.255157261779837,
+      "learning_rate": 2e-06,
+      "loss": 0.3191,
+      "step": 1992
+    },
+    {
+      "epoch": 0.46235935506321774,
+      "grad_norm": 10.690498300962876,
+      "learning_rate": 2e-06,
+      "loss": 0.2414,
+      "step": 1993
+    },
+    {
+      "epoch": 0.46259134671151836,
+      "grad_norm": 9.877054825920622,
+      "learning_rate": 2e-06,
+      "loss": 0.3728,
+      "step": 1994
+    },
+    {
+      "epoch": 0.46282333835981904,
+      "grad_norm": 15.340735025865651,
+      "learning_rate": 2e-06,
+      "loss": 0.4747,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4630553300081197,
+      "grad_norm": 15.341221735588649,
+      "learning_rate": 2e-06,
+      "loss": 0.3361,
+      "step": 1996
+    },
+    {
+      "epoch": 0.4632873216564204,
+      "grad_norm": 10.945005995090467,
+      "learning_rate": 2e-06,
+      "loss": 0.2797,
+      "step": 1997
+    },
+    {
+      "epoch": 0.463519313304721,
+      "grad_norm": 5.284853062707283,
+      "learning_rate": 2e-06,
+      "loss": 0.1841,
+      "step": 1998
+    },
+    {
+      "epoch": 0.4637513049530217,
+      "grad_norm": 11.633188950114901,
+      "learning_rate": 2e-06,
+      "loss": 0.2856,
+      "step": 1999
+    },
+    {
+      "epoch": 0.46398329660132237,
+      "grad_norm": 6.859718365049897,
+      "learning_rate": 2e-06,
+      "loss": 0.3048,
+      "step": 2000
+    },
+    {
+      "epoch": 0.464215288249623,
+      "grad_norm": 16.589305902090143,
+      "learning_rate": 2e-06,
+      "loss": 0.4069,
+      "step": 2001
+    },
+    {
+      "epoch": 0.46444727989792367,
+      "grad_norm": 10.342457937117073,
+      "learning_rate": 2e-06,
+      "loss": 0.3057,
+      "step": 2002
+    },
+    {
+      "epoch": 0.46467927154622435,
+      "grad_norm": 12.333270314902068,
+      "learning_rate": 2e-06,
+      "loss": 0.2834,
+      "step": 2003
+    },
+    {
+      "epoch": 0.46491126319452497,
+      "grad_norm": 22.423928853702186,
+      "learning_rate": 2e-06,
+      "loss": 0.3894,
+      "step": 2004
+    },
+    {
+      "epoch": 0.46514325484282565,
+      "grad_norm": 13.766616278689142,
+      "learning_rate": 2e-06,
+      "loss": 0.375,
+      "step": 2005
+    },
+    {
+      "epoch": 0.4653752464911263,
+      "grad_norm": 16.62716850114615,
+      "learning_rate": 2e-06,
+      "loss": 0.3018,
+      "step": 2006
+    },
+    {
+      "epoch": 0.465607238139427,
+      "grad_norm": 18.43425031492219,
+      "learning_rate": 2e-06,
+      "loss": 0.3473,
+      "step": 2007
+    },
+    {
+      "epoch": 0.4658392297877276,
+      "grad_norm": 8.281568359570956,
+      "learning_rate": 2e-06,
+      "loss": 0.2622,
+      "step": 2008
+    },
+    {
+      "epoch": 0.4660712214360283,
+      "grad_norm": 14.635641990408606,
+      "learning_rate": 2e-06,
+      "loss": 0.2376,
+      "step": 2009
+    },
+    {
+      "epoch": 0.466303213084329,
+      "grad_norm": 16.20241517427514,
+      "learning_rate": 2e-06,
+      "loss": 0.3306,
+      "step": 2010
+    },
+    {
+      "epoch": 0.4665352047326296,
+      "grad_norm": 11.51127206202647,
+      "learning_rate": 2e-06,
+      "loss": 0.2747,
+      "step": 2011
+    },
+    {
+      "epoch": 0.4667671963809303,
+      "grad_norm": 58.30128405919469,
+      "learning_rate": 2e-06,
+      "loss": 0.3029,
+      "step": 2012
+    },
+    {
+      "epoch": 0.46699918802923096,
+      "grad_norm": 13.9317037243102,
+      "learning_rate": 2e-06,
+      "loss": 0.3309,
+      "step": 2013
+    },
+    {
+      "epoch": 0.4672311796775316,
+      "grad_norm": 20.31824848079081,
+      "learning_rate": 2e-06,
+      "loss": 0.3761,
+      "step": 2014
+    },
+    {
+      "epoch": 0.46746317132583226,
+      "grad_norm": 9.457890163970369,
+      "learning_rate": 2e-06,
+      "loss": 0.3379,
+      "step": 2015
+    },
+    {
+      "epoch": 0.46769516297413294,
+      "grad_norm": 18.053400087811408,
+      "learning_rate": 2e-06,
+      "loss": 0.3155,
+      "step": 2016
+    },
+    {
+      "epoch": 0.4679271546224336,
+      "grad_norm": 11.634119001512087,
+      "learning_rate": 2e-06,
+      "loss": 0.3429,
+      "step": 2017
+    },
+    {
+      "epoch": 0.46815914627073424,
+      "grad_norm": 13.942355437046755,
+      "learning_rate": 2e-06,
+      "loss": 0.3103,
+      "step": 2018
+    },
+    {
+      "epoch": 0.4683911379190349,
+      "grad_norm": 14.711577692597876,
+      "learning_rate": 2e-06,
+      "loss": 0.3242,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4686231295673356,
+      "grad_norm": 12.822970373558434,
+      "learning_rate": 2e-06,
+      "loss": 0.2824,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4688551212156362,
+      "grad_norm": 7.922564466961149,
+      "learning_rate": 2e-06,
+      "loss": 0.3574,
+      "step": 2021
+    },
+    {
+      "epoch": 0.4690871128639369,
+      "grad_norm": 16.645591544402976,
+      "learning_rate": 2e-06,
+      "loss": 0.26,
+      "step": 2022
+    },
+    {
+      "epoch": 0.4693191045122376,
+      "grad_norm": 17.150991755358596,
+      "learning_rate": 2e-06,
+      "loss": 0.3091,
+      "step": 2023
+    },
+    {
+      "epoch": 0.4695510961605382,
+      "grad_norm": 19.355662866561232,
+      "learning_rate": 2e-06,
+      "loss": 0.3248,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4697830878088389,
+      "grad_norm": 8.98763782114109,
+      "learning_rate": 2e-06,
+      "loss": 0.2328,
+      "step": 2025
+    },
+    {
+      "epoch": 0.47001507945713955,
+      "grad_norm": 10.750289888646794,
+      "learning_rate": 2e-06,
+      "loss": 0.2208,
+      "step": 2026
+    },
+    {
+      "epoch": 0.47024707110544023,
+      "grad_norm": 15.590133390676085,
+      "learning_rate": 2e-06,
+      "loss": 0.2389,
+      "step": 2027
+    },
+    {
+      "epoch": 0.47047906275374085,
+      "grad_norm": 16.63610849979652,
+      "learning_rate": 2e-06,
+      "loss": 0.3112,
+      "step": 2028
+    },
+    {
+      "epoch": 0.47071105440204153,
+      "grad_norm": 10.792337381759852,
+      "learning_rate": 2e-06,
+      "loss": 0.3559,
+      "step": 2029
+    },
+    {
+      "epoch": 0.4709430460503422,
+      "grad_norm": 14.816149884219211,
+      "learning_rate": 2e-06,
+      "loss": 0.3003,
+      "step": 2030
+    },
+    {
+      "epoch": 0.47117503769864283,
+      "grad_norm": 10.341262243880099,
+      "learning_rate": 2e-06,
+      "loss": 0.3463,
+      "step": 2031
+    },
+    {
+      "epoch": 0.4714070293469435,
+      "grad_norm": 17.72305433274037,
+      "learning_rate": 2e-06,
+      "loss": 0.4462,
+      "step": 2032
+    },
+    {
+      "epoch": 0.4716390209952442,
+      "grad_norm": 16.83950046901211,
+      "learning_rate": 2e-06,
+      "loss": 0.2417,
+      "step": 2033
+    },
+    {
+      "epoch": 0.4718710126435448,
+      "grad_norm": 18.10327281457699,
+      "learning_rate": 2e-06,
+      "loss": 0.3096,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4721030042918455,
+      "grad_norm": 14.254882642959036,
+      "learning_rate": 2e-06,
+      "loss": 0.3366,
+      "step": 2035
+    },
+    {
+      "epoch": 0.47233499594014616,
+      "grad_norm": 21.676286798359246,
+      "learning_rate": 2e-06,
+      "loss": 0.3546,
+      "step": 2036
+    },
+    {
+      "epoch": 0.47256698758844684,
+      "grad_norm": 12.720790909738758,
+      "learning_rate": 2e-06,
+      "loss": 0.3007,
+      "step": 2037
+    },
+    {
+      "epoch": 0.47279897923674746,
+      "grad_norm": 11.498119426389403,
+      "learning_rate": 2e-06,
+      "loss": 0.2476,
+      "step": 2038
+    },
+    {
+      "epoch": 0.47303097088504814,
+      "grad_norm": 20.816542225879648,
+      "learning_rate": 2e-06,
+      "loss": 0.4009,
+      "step": 2039
+    },
+    {
+      "epoch": 0.4732629625333488,
+      "grad_norm": 16.82688865312617,
+      "learning_rate": 2e-06,
+      "loss": 0.2652,
+      "step": 2040
+    },
+    {
+      "epoch": 0.47349495418164944,
+      "grad_norm": 9.70477771542857,
+      "learning_rate": 2e-06,
+      "loss": 0.2974,
+      "step": 2041
+    },
+    {
+      "epoch": 0.4737269458299501,
+      "grad_norm": 13.296534871793732,
+      "learning_rate": 2e-06,
+      "loss": 0.2746,
+      "step": 2042
+    },
+    {
+      "epoch": 0.4739589374782508,
+      "grad_norm": 14.065427392790259,
+      "learning_rate": 2e-06,
+      "loss": 0.3149,
+      "step": 2043
+    },
+    {
+      "epoch": 0.4741909291265514,
+      "grad_norm": 22.38889426658264,
+      "learning_rate": 2e-06,
+      "loss": 0.433,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4744229207748521,
+      "grad_norm": 14.376475918423807,
+      "learning_rate": 2e-06,
+      "loss": 0.3366,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4746549124231528,
+      "grad_norm": 9.277440540725683,
+      "learning_rate": 2e-06,
+      "loss": 0.337,
+      "step": 2046
+    },
+    {
+      "epoch": 0.47488690407145345,
+      "grad_norm": 17.340785257332662,
+      "learning_rate": 2e-06,
+      "loss": 0.2725,
+      "step": 2047
+    },
+    {
+      "epoch": 0.4751188957197541,
+      "grad_norm": 9.985084330908725,
+      "learning_rate": 2e-06,
+      "loss": 0.2141,
+      "step": 2048
+    },
+    {
+      "epoch": 0.47535088736805475,
+      "grad_norm": 10.633352809709951,
+      "learning_rate": 2e-06,
+      "loss": 0.2789,
+      "step": 2049
+    },
+    {
+      "epoch": 0.47558287901635543,
+      "grad_norm": 23.121629632798786,
+      "learning_rate": 2e-06,
+      "loss": 0.2321,
+      "step": 2050
+    },
+    {
+      "epoch": 0.47581487066465605,
+      "grad_norm": 10.891572671577103,
+      "learning_rate": 2e-06,
+      "loss": 0.3372,
+      "step": 2051
+    },
+    {
+      "epoch": 0.47604686231295673,
+      "grad_norm": 23.9600752713284,
+      "learning_rate": 2e-06,
+      "loss": 0.3471,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4762788539612574,
+      "grad_norm": 15.048695431676713,
+      "learning_rate": 2e-06,
+      "loss": 0.4033,
+      "step": 2053
+    },
+    {
+      "epoch": 0.47651084560955803,
+      "grad_norm": 19.555167465786873,
+      "learning_rate": 2e-06,
+      "loss": 0.3338,
+      "step": 2054
+    },
+    {
+      "epoch": 0.4767428372578587,
+      "grad_norm": 27.78200124896194,
+      "learning_rate": 2e-06,
+      "loss": 0.4486,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4769748289061594,
+      "grad_norm": 12.272255303881282,
+      "learning_rate": 2e-06,
+      "loss": 0.3018,
+      "step": 2056
+    },
+    {
+      "epoch": 0.47720682055446006,
+      "grad_norm": 15.556624908312893,
+      "learning_rate": 2e-06,
+      "loss": 0.5194,
+      "step": 2057
+    },
+    {
+      "epoch": 0.4774388122027607,
+      "grad_norm": 16.338581145118493,
+      "learning_rate": 2e-06,
+      "loss": 0.3547,
+      "step": 2058
+    },
+    {
+      "epoch": 0.47767080385106137,
+      "grad_norm": 9.074037753398741,
+      "learning_rate": 2e-06,
+      "loss": 0.2408,
+      "step": 2059
+    },
+    {
+      "epoch": 0.47790279549936204,
+      "grad_norm": 10.66378411033167,
+      "learning_rate": 2e-06,
+      "loss": 0.3439,
+      "step": 2060
+    },
+    {
+      "epoch": 0.47813478714766267,
+      "grad_norm": 15.797434529468434,
+      "learning_rate": 2e-06,
+      "loss": 0.2944,
+      "step": 2061
+    },
+    {
+      "epoch": 0.47836677879596334,
+      "grad_norm": 13.227433292179827,
+      "learning_rate": 2e-06,
+      "loss": 0.339,
+      "step": 2062
+    },
+    {
+      "epoch": 0.478598770444264,
+      "grad_norm": 12.03833396979913,
+      "learning_rate": 2e-06,
+      "loss": 0.3403,
+      "step": 2063
+    },
+    {
+      "epoch": 0.47883076209256464,
+      "grad_norm": 24.553130563738854,
+      "learning_rate": 2e-06,
+      "loss": 0.3004,
+      "step": 2064
+    },
+    {
+      "epoch": 0.4790627537408653,
+      "grad_norm": 20.737155578901096,
+      "learning_rate": 2e-06,
+      "loss": 0.4112,
+      "step": 2065
+    },
+    {
+      "epoch": 0.479294745389166,
+      "grad_norm": 27.884478252430842,
+      "learning_rate": 2e-06,
+      "loss": 0.3297,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4795267370374667,
+      "grad_norm": 13.103378328133143,
+      "learning_rate": 2e-06,
+      "loss": 0.3001,
+      "step": 2067
+    },
+    {
+      "epoch": 0.4797587286857673,
+      "grad_norm": 17.80927820776487,
+      "learning_rate": 2e-06,
+      "loss": 0.4377,
+      "step": 2068
+    },
+    {
+      "epoch": 0.479990720334068,
+      "grad_norm": 20.661695218080254,
+      "learning_rate": 2e-06,
+      "loss": 0.3473,
+      "step": 2069
+    },
+    {
+      "epoch": 0.48022271198236866,
+      "grad_norm": 28.217870160429282,
+      "learning_rate": 2e-06,
+      "loss": 0.3783,
+      "step": 2070
+    },
+    {
+      "epoch": 0.4804547036306693,
+      "grad_norm": 17.477461536493266,
+      "learning_rate": 2e-06,
+      "loss": 0.3352,
+      "step": 2071
+    },
+    {
+      "epoch": 0.48068669527896996,
+      "grad_norm": 23.889925987937147,
+      "learning_rate": 2e-06,
+      "loss": 0.3111,
+      "step": 2072
+    },
+    {
+      "epoch": 0.48091868692727063,
+      "grad_norm": 7.9886444116729205,
+      "learning_rate": 2e-06,
+      "loss": 0.3253,
+      "step": 2073
+    },
+    {
+      "epoch": 0.48115067857557126,
+      "grad_norm": 19.73259296094963,
+      "learning_rate": 2e-06,
+      "loss": 0.3367,
+      "step": 2074
+    },
+    {
+      "epoch": 0.48138267022387193,
+      "grad_norm": 9.559685773496563,
+      "learning_rate": 2e-06,
+      "loss": 0.234,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4816146618721726,
+      "grad_norm": 9.743531669983685,
+      "learning_rate": 2e-06,
+      "loss": 0.2975,
+      "step": 2076
+    },
+    {
+      "epoch": 0.4818466535204733,
+      "grad_norm": 16.81635360695747,
+      "learning_rate": 2e-06,
+      "loss": 0.2438,
+      "step": 2077
+    },
+    {
+      "epoch": 0.4820786451687739,
+      "grad_norm": 11.790522180861018,
+      "learning_rate": 2e-06,
+      "loss": 0.2424,
+      "step": 2078
+    },
+    {
+      "epoch": 0.4823106368170746,
+      "grad_norm": 14.632744832313552,
+      "learning_rate": 2e-06,
+      "loss": 0.3295,
+      "step": 2079
+    },
+    {
+      "epoch": 0.48254262846537527,
+      "grad_norm": 11.717947162424096,
+      "learning_rate": 2e-06,
+      "loss": 0.3074,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4827746201136759,
+      "grad_norm": 8.98069286186794,
+      "learning_rate": 2e-06,
+      "loss": 0.2319,
+      "step": 2081
+    },
+    {
+      "epoch": 0.48300661176197657,
+      "grad_norm": 10.350196437648169,
+      "learning_rate": 2e-06,
+      "loss": 0.3483,
+      "step": 2082
+    },
+    {
+      "epoch": 0.48323860341027725,
+      "grad_norm": 18.872461989879593,
+      "learning_rate": 2e-06,
+      "loss": 0.3595,
+      "step": 2083
+    },
+    {
+      "epoch": 0.48347059505857787,
+      "grad_norm": 18.968088081083174,
+      "learning_rate": 2e-06,
+      "loss": 0.3471,
+      "step": 2084
+    },
+    {
+      "epoch": 0.48370258670687855,
+      "grad_norm": 14.399667825466722,
+      "learning_rate": 2e-06,
+      "loss": 0.3329,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4839345783551792,
+      "grad_norm": 13.474511688092948,
+      "learning_rate": 2e-06,
+      "loss": 0.3144,
+      "step": 2086
+    },
+    {
+      "epoch": 0.4841665700034799,
+      "grad_norm": 11.216722839581688,
+      "learning_rate": 2e-06,
+      "loss": 0.2157,
+      "step": 2087
+    },
+    {
+      "epoch": 0.4843985616517805,
+      "grad_norm": 14.86844187814411,
+      "learning_rate": 2e-06,
+      "loss": 0.4169,
+      "step": 2088
+    },
+    {
+      "epoch": 0.4846305533000812,
+      "grad_norm": 7.8447621775911545,
+      "learning_rate": 2e-06,
+      "loss": 0.3091,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4848625449483819,
+      "grad_norm": 10.254262575831111,
+      "learning_rate": 2e-06,
+      "loss": 0.359,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4850945365966825,
+      "grad_norm": 11.876234611995246,
+      "learning_rate": 2e-06,
+      "loss": 0.269,
+      "step": 2091
+    },
+    {
+      "epoch": 0.4853265282449832,
+      "grad_norm": 14.067135704743569,
+      "learning_rate": 2e-06,
+      "loss": 0.3312,
+      "step": 2092
+    },
+    {
+      "epoch": 0.48555851989328386,
+      "grad_norm": 7.128865969322003,
+      "learning_rate": 2e-06,
+      "loss": 0.235,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4857905115415845,
+      "grad_norm": 26.351599690207127,
+      "learning_rate": 2e-06,
+      "loss": 0.3259,
+      "step": 2094
+    },
+    {
+      "epoch": 0.48602250318988516,
+      "grad_norm": 16.241201805268105,
+      "learning_rate": 2e-06,
+      "loss": 0.2984,
+      "step": 2095
+    },
+    {
+      "epoch": 0.48625449483818584,
+      "grad_norm": 14.767969302463698,
+      "learning_rate": 2e-06,
+      "loss": 0.3041,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4864864864864865,
+      "grad_norm": 15.7565988823627,
+      "learning_rate": 2e-06,
+      "loss": 0.329,
+      "step": 2097
+    },
+    {
+      "epoch": 0.48671847813478714,
+      "grad_norm": 16.203714255793574,
+      "learning_rate": 2e-06,
+      "loss": 0.2604,
+      "step": 2098
+    },
+    {
+      "epoch": 0.4869504697830878,
+      "grad_norm": 10.42518604548034,
+      "learning_rate": 2e-06,
+      "loss": 0.3288,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4871824614313885,
+      "grad_norm": 22.287666620106165,
+      "learning_rate": 2e-06,
+      "loss": 0.3003,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4874144530796891,
+      "grad_norm": 15.199158506152191,
+      "learning_rate": 2e-06,
+      "loss": 0.2466,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4876464447279898,
+      "grad_norm": 17.558767455624956,
+      "learning_rate": 2e-06,
+      "loss": 0.3145,
+      "step": 2102
+    },
+    {
+      "epoch": 0.48787843637629047,
+      "grad_norm": 18.285953331339602,
+      "learning_rate": 2e-06,
+      "loss": 0.4843,
+      "step": 2103
+    },
+    {
+      "epoch": 0.4881104280245911,
+      "grad_norm": 15.432844223638583,
+      "learning_rate": 2e-06,
+      "loss": 0.2451,
+      "step": 2104
+    },
+    {
+      "epoch": 0.48834241967289177,
+      "grad_norm": 11.767138467996935,
+      "learning_rate": 2e-06,
+      "loss": 0.2904,
+      "step": 2105
+    },
+    {
+      "epoch": 0.48857441132119245,
+      "grad_norm": 14.707804934793018,
+      "learning_rate": 2e-06,
+      "loss": 0.3309,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4888064029694931,
+      "grad_norm": 12.873460858804348,
+      "learning_rate": 2e-06,
+      "loss": 0.2538,
+      "step": 2107
+    },
+    {
+      "epoch": 0.48903839461779375,
+      "grad_norm": 15.551365719566405,
+      "learning_rate": 2e-06,
+      "loss": 0.3376,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4892703862660944,
+      "grad_norm": 22.354249058364896,
+      "learning_rate": 2e-06,
+      "loss": 0.3482,
+      "step": 2109
+    },
+    {
+      "epoch": 0.4895023779143951,
+      "grad_norm": 12.13426092309433,
+      "learning_rate": 2e-06,
+      "loss": 0.2767,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4897343695626957,
+      "grad_norm": 6.378416323954354,
+      "learning_rate": 2e-06,
+      "loss": 0.1642,
+      "step": 2111
+    },
+    {
+      "epoch": 0.4899663612109964,
+      "grad_norm": 12.172120073384693,
+      "learning_rate": 2e-06,
+      "loss": 0.2751,
+      "step": 2112
+    },
+    {
+      "epoch": 0.4901983528592971,
+      "grad_norm": 8.006798240440455,
+      "learning_rate": 2e-06,
+      "loss": 0.2359,
+      "step": 2113
+    },
+    {
+      "epoch": 0.4904303445075977,
+      "grad_norm": 15.544069438253967,
+      "learning_rate": 2e-06,
+      "loss": 0.3287,
+      "step": 2114
+    },
+    {
+      "epoch": 0.4906623361558984,
+      "grad_norm": 7.0112959850423735,
+      "learning_rate": 2e-06,
+      "loss": 0.1766,
+      "step": 2115
+    },
+    {
+      "epoch": 0.49089432780419906,
+      "grad_norm": 15.549511728081491,
+      "learning_rate": 2e-06,
+      "loss": 0.3081,
+      "step": 2116
+    },
+    {
+      "epoch": 0.49112631945249974,
+      "grad_norm": 11.256547217001314,
+      "learning_rate": 2e-06,
+      "loss": 0.3567,
+      "step": 2117
+    },
+    {
+      "epoch": 0.49135831110080036,
+      "grad_norm": 14.704389837873286,
+      "learning_rate": 2e-06,
+      "loss": 0.2985,
+      "step": 2118
+    },
+    {
+      "epoch": 0.49159030274910104,
+      "grad_norm": 17.727161833868298,
+      "learning_rate": 2e-06,
+      "loss": 0.3379,
+      "step": 2119
+    },
+    {
+      "epoch": 0.4918222943974017,
+      "grad_norm": 13.301220933708462,
+      "learning_rate": 2e-06,
+      "loss": 0.2558,
+      "step": 2120
+    },
+    {
+      "epoch": 0.49205428604570234,
+      "grad_norm": 6.108889993757156,
+      "learning_rate": 2e-06,
+      "loss": 0.2027,
+      "step": 2121
+    },
+    {
+      "epoch": 0.492286277694003,
+      "grad_norm": 10.868425584914485,
+      "learning_rate": 2e-06,
+      "loss": 0.2438,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4925182693423037,
+      "grad_norm": 17.635453142218786,
+      "learning_rate": 2e-06,
+      "loss": 0.4297,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4927502609906043,
+      "grad_norm": 17.445941519621602,
+      "learning_rate": 2e-06,
+      "loss": 0.436,
+      "step": 2124
+    },
+    {
+      "epoch": 0.492982252638905,
+      "grad_norm": 16.86257932055074,
+      "learning_rate": 2e-06,
+      "loss": 0.334,
+      "step": 2125
+    },
+    {
+      "epoch": 0.49321424428720567,
+      "grad_norm": 14.557790324443094,
+      "learning_rate": 2e-06,
+      "loss": 0.3895,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4934462359355063,
+      "grad_norm": 15.365709948457615,
+      "learning_rate": 2e-06,
+      "loss": 0.2738,
+      "step": 2127
+    },
+    {
+      "epoch": 0.49367822758380697,
+      "grad_norm": 11.727258104871728,
+      "learning_rate": 2e-06,
+      "loss": 0.1987,
+      "step": 2128
+    },
+    {
+      "epoch": 0.49391021923210765,
+      "grad_norm": 13.763772152669365,
+      "learning_rate": 2e-06,
+      "loss": 0.2981,
+      "step": 2129
+    },
+    {
+      "epoch": 0.49414221088040833,
+      "grad_norm": 12.891289421629347,
+      "learning_rate": 2e-06,
+      "loss": 0.241,
+      "step": 2130
+    },
+    {
+      "epoch": 0.49437420252870895,
+      "grad_norm": 15.264107197755068,
+      "learning_rate": 2e-06,
+      "loss": 0.319,
+      "step": 2131
+    },
+    {
+      "epoch": 0.49460619417700963,
+      "grad_norm": 19.049486690093694,
+      "learning_rate": 2e-06,
+      "loss": 0.2844,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4948381858253103,
+      "grad_norm": 20.289091498324936,
+      "learning_rate": 2e-06,
+      "loss": 0.3354,
+      "step": 2133
+    },
+    {
+      "epoch": 0.49507017747361093,
+      "grad_norm": 8.846363557878725,
+      "learning_rate": 2e-06,
+      "loss": 0.2455,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4953021691219116,
+      "grad_norm": 19.68036789620874,
+      "learning_rate": 2e-06,
+      "loss": 0.3254,
+      "step": 2135
+    },
+    {
+      "epoch": 0.4955341607702123,
+      "grad_norm": 19.01167373777457,
+      "learning_rate": 2e-06,
+      "loss": 0.3668,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4957661524185129,
+      "grad_norm": 20.271394808053454,
+      "learning_rate": 2e-06,
+      "loss": 0.3329,
+      "step": 2137
+    },
+    {
+      "epoch": 0.4959981440668136,
+      "grad_norm": 13.713450589308374,
+      "learning_rate": 2e-06,
+      "loss": 0.3172,
+      "step": 2138
+    },
+    {
+      "epoch": 0.49623013571511426,
+      "grad_norm": 15.681288599458853,
+      "learning_rate": 2e-06,
+      "loss": 0.2464,
+      "step": 2139
+    },
+    {
+      "epoch": 0.49646212736341494,
+      "grad_norm": 22.489577644012517,
+      "learning_rate": 2e-06,
+      "loss": 0.466,
+      "step": 2140
+    },
+    {
+      "epoch": 0.49669411901171556,
+      "grad_norm": 11.929339914832337,
+      "learning_rate": 2e-06,
+      "loss": 0.2438,
+      "step": 2141
+    },
+    {
+      "epoch": 0.49692611066001624,
+      "grad_norm": 6.908041019643261,
+      "learning_rate": 2e-06,
+      "loss": 0.1916,
+      "step": 2142
+    },
+    {
+      "epoch": 0.4971581023083169,
+      "grad_norm": 16.87423395285013,
+      "learning_rate": 2e-06,
+      "loss": 0.3377,
+      "step": 2143
+    },
+    {
+      "epoch": 0.49739009395661754,
+      "grad_norm": 21.484609766510122,
+      "learning_rate": 2e-06,
+      "loss": 0.3187,
+      "step": 2144
+    },
+    {
+      "epoch": 0.4976220856049182,
+      "grad_norm": 10.005933510006772,
+      "learning_rate": 2e-06,
+      "loss": 0.2997,
+      "step": 2145
+    },
+    {
+      "epoch": 0.4978540772532189,
+      "grad_norm": 20.390856830943978,
+      "learning_rate": 2e-06,
+      "loss": 0.2964,
+      "step": 2146
+    },
+    {
+      "epoch": 0.4980860689015195,
+      "grad_norm": 20.54252638356709,
+      "learning_rate": 2e-06,
+      "loss": 0.3739,
+      "step": 2147
+    },
+    {
+      "epoch": 0.4983180605498202,
+      "grad_norm": 6.440030667693158,
+      "learning_rate": 2e-06,
+      "loss": 0.213,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4985500521981209,
+      "grad_norm": 13.909759042917415,
+      "learning_rate": 2e-06,
+      "loss": 0.1993,
+      "step": 2149
+    },
+    {
+      "epoch": 0.49878204384642155,
+      "grad_norm": 13.442729423292715,
+      "learning_rate": 2e-06,
+      "loss": 0.1918,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4990140354947222,
+      "grad_norm": 9.777832167761424,
+      "learning_rate": 2e-06,
+      "loss": 0.1691,
+      "step": 2151
+    },
+    {
+      "epoch": 0.49924602714302285,
+      "grad_norm": 7.528993896065364,
+      "learning_rate": 2e-06,
+      "loss": 0.2213,
+      "step": 2152
+    },
+    {
+      "epoch": 0.49947801879132353,
+      "grad_norm": 25.537326149047804,
+      "learning_rate": 2e-06,
+      "loss": 0.4174,
+      "step": 2153
+    },
+    {
+      "epoch": 0.49971001043962415,
+      "grad_norm": 13.598357491155793,
+      "learning_rate": 2e-06,
+      "loss": 0.3848,
+      "step": 2154
+    },
+    {
+      "epoch": 0.49994200208792483,
+      "grad_norm": 22.388810531659942,
+      "learning_rate": 2e-06,
+      "loss": 0.3423,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5001739937362255,
+      "grad_norm": 13.93373343416442,
+      "learning_rate": 2e-06,
+      "loss": 0.2189,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5004059853845262,
+      "grad_norm": 14.328986795854332,
+      "learning_rate": 2e-06,
+      "loss": 0.3242,
+      "step": 2157
+    },
+    {
+      "epoch": 0.5006379770328269,
+      "grad_norm": 8.904671756440253,
+      "learning_rate": 2e-06,
+      "loss": 0.2041,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5008699686811274,
+      "grad_norm": 23.515968807666702,
+      "learning_rate": 2e-06,
+      "loss": 0.4471,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5011019603294281,
+      "grad_norm": 21.827464695996884,
+      "learning_rate": 2e-06,
+      "loss": 0.3769,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5013339519777288,
+      "grad_norm": 18.861402954271224,
+      "learning_rate": 2e-06,
+      "loss": 0.372,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5015659436260295,
+      "grad_norm": 16.535871849830876,
+      "learning_rate": 2e-06,
+      "loss": 0.2858,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5017979352743301,
+      "grad_norm": 10.303277065169574,
+      "learning_rate": 2e-06,
+      "loss": 0.1672,
+      "step": 2163
+    },
+    {
+      "epoch": 0.5020299269226308,
+      "grad_norm": 15.957852373614422,
+      "learning_rate": 2e-06,
+      "loss": 0.2696,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5022619185709315,
+      "grad_norm": 26.862222528082594,
+      "learning_rate": 2e-06,
+      "loss": 0.3697,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5024939102192321,
+      "grad_norm": 13.53436358948289,
+      "learning_rate": 2e-06,
+      "loss": 0.3213,
+      "step": 2166
+    },
+    {
+      "epoch": 0.5027259018675327,
+      "grad_norm": 29.75890831496249,
+      "learning_rate": 2e-06,
+      "loss": 0.3747,
+      "step": 2167
+    },
+    {
+      "epoch": 0.5029578935158334,
+      "grad_norm": 17.201119948928586,
+      "learning_rate": 2e-06,
+      "loss": 0.4157,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5031898851641341,
+      "grad_norm": 20.322957627453476,
+      "learning_rate": 2e-06,
+      "loss": 0.3425,
+      "step": 2169
+    },
+    {
+      "epoch": 0.5034218768124348,
+      "grad_norm": 7.242148992234226,
+      "learning_rate": 2e-06,
+      "loss": 0.2237,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5036538684607355,
+      "grad_norm": 11.3995466559775,
+      "learning_rate": 2e-06,
+      "loss": 0.2041,
+      "step": 2171
+    },
+    {
+      "epoch": 0.503885860109036,
+      "grad_norm": 13.692661532868897,
+      "learning_rate": 2e-06,
+      "loss": 0.2766,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5041178517573367,
+      "grad_norm": 13.045830152901408,
+      "learning_rate": 2e-06,
+      "loss": 0.2802,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5043498434056374,
+      "grad_norm": 13.19375863975848,
+      "learning_rate": 2e-06,
+      "loss": 0.3715,
+      "step": 2174
+    },
+    {
+      "epoch": 0.504581835053938,
+      "grad_norm": 12.549473738148507,
+      "learning_rate": 2e-06,
+      "loss": 0.2536,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5048138267022387,
+      "grad_norm": 18.270592559002413,
+      "learning_rate": 2e-06,
+      "loss": 0.2249,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5050458183505394,
+      "grad_norm": 22.483451955144304,
+      "learning_rate": 2e-06,
+      "loss": 0.3166,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5052778099988401,
+      "grad_norm": 17.316063660340365,
+      "learning_rate": 2e-06,
+      "loss": 0.2714,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5055098016471407,
+      "grad_norm": 10.947393037556699,
+      "learning_rate": 2e-06,
+      "loss": 0.3,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5057417932954413,
+      "grad_norm": 22.039873971526152,
+      "learning_rate": 2e-06,
+      "loss": 0.4411,
+      "step": 2180
+    },
+    {
+      "epoch": 0.505973784943742,
+      "grad_norm": 10.035311774289827,
+      "learning_rate": 2e-06,
+      "loss": 0.2457,
+      "step": 2181
+    },
+    {
+      "epoch": 0.5062057765920427,
+      "grad_norm": 20.400488855842067,
+      "learning_rate": 2e-06,
+      "loss": 0.3309,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5064377682403434,
+      "grad_norm": 11.139661608372032,
+      "learning_rate": 2e-06,
+      "loss": 0.2716,
+      "step": 2183
+    },
+    {
+      "epoch": 0.506669759888644,
+      "grad_norm": 49.22644449344963,
+      "learning_rate": 2e-06,
+      "loss": 0.413,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5069017515369447,
+      "grad_norm": 19.282249953371107,
+      "learning_rate": 2e-06,
+      "loss": 0.2559,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5071337431852453,
+      "grad_norm": 9.96155712975433,
+      "learning_rate": 2e-06,
+      "loss": 0.2553,
+      "step": 2186
+    },
+    {
+      "epoch": 0.507365734833546,
+      "grad_norm": 11.389788449201504,
+      "learning_rate": 2e-06,
+      "loss": 0.2749,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5075977264818466,
+      "grad_norm": 9.573367203354213,
+      "learning_rate": 2e-06,
+      "loss": 0.325,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5078297181301473,
+      "grad_norm": 17.638556783160844,
+      "learning_rate": 2e-06,
+      "loss": 0.3245,
+      "step": 2189
+    },
+    {
+      "epoch": 0.508061709778448,
+      "grad_norm": 25.24343689451894,
+      "learning_rate": 2e-06,
+      "loss": 0.4962,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5082937014267487,
+      "grad_norm": 15.432898553377086,
+      "learning_rate": 2e-06,
+      "loss": 0.2843,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5085256930750492,
+      "grad_norm": 18.526829636362876,
+      "learning_rate": 2e-06,
+      "loss": 0.3193,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5087576847233499,
+      "grad_norm": 20.829626240692,
+      "learning_rate": 2e-06,
+      "loss": 0.2898,
+      "step": 2193
+    },
+    {
+      "epoch": 0.5089896763716506,
+      "grad_norm": 12.080813626427616,
+      "learning_rate": 2e-06,
+      "loss": 0.2743,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5092216680199513,
+      "grad_norm": 8.755550213350691,
+      "learning_rate": 2e-06,
+      "loss": 0.2772,
+      "step": 2195
+    },
+    {
+      "epoch": 0.509453659668252,
+      "grad_norm": 17.579316445335433,
+      "learning_rate": 2e-06,
+      "loss": 0.3266,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5096856513165526,
+      "grad_norm": 17.325365704374104,
+      "learning_rate": 2e-06,
+      "loss": 0.2805,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5099176429648533,
+      "grad_norm": 14.492615581490334,
+      "learning_rate": 2e-06,
+      "loss": 0.3094,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5101496346131539,
+      "grad_norm": 14.234190346875504,
+      "learning_rate": 2e-06,
+      "loss": 0.2348,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5103816262614546,
+      "grad_norm": 7.5398495123313465,
+      "learning_rate": 2e-06,
+      "loss": 0.1718,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5106136179097552,
+      "grad_norm": 6.97541242627694,
+      "learning_rate": 2e-06,
+      "loss": 0.3266,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5108456095580559,
+      "grad_norm": 12.721195042383458,
+      "learning_rate": 2e-06,
+      "loss": 0.253,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5110776012063566,
+      "grad_norm": 7.105303434910561,
+      "learning_rate": 2e-06,
+      "loss": 0.2085,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5113095928546573,
+      "grad_norm": 16.65842865369534,
+      "learning_rate": 2e-06,
+      "loss": 0.326,
+      "step": 2204
+    },
+    {
+      "epoch": 0.511541584502958,
+      "grad_norm": 24.447691272915243,
+      "learning_rate": 2e-06,
+      "loss": 0.3282,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5117735761512585,
+      "grad_norm": 7.472587814339079,
+      "learning_rate": 2e-06,
+      "loss": 0.1978,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5120055677995592,
+      "grad_norm": 27.452162551642193,
+      "learning_rate": 2e-06,
+      "loss": 0.2873,
+      "step": 2207
+    },
+    {
+      "epoch": 0.5122375594478599,
+      "grad_norm": 12.531644069360812,
+      "learning_rate": 2e-06,
+      "loss": 0.2842,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5124695510961605,
+      "grad_norm": 13.022291960747255,
+      "learning_rate": 2e-06,
+      "loss": 0.3008,
+      "step": 2209
+    },
+    {
+      "epoch": 0.5127015427444612,
+      "grad_norm": 11.436705149111399,
+      "learning_rate": 2e-06,
+      "loss": 0.2965,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5129335343927619,
+      "grad_norm": 9.325082731789037,
+      "learning_rate": 2e-06,
+      "loss": 0.2812,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5131655260410625,
+      "grad_norm": 15.090660618989261,
+      "learning_rate": 2e-06,
+      "loss": 0.2271,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5133975176893631,
+      "grad_norm": 12.092657076627773,
+      "learning_rate": 2e-06,
+      "loss": 0.2794,
+      "step": 2213
+    },
+    {
+      "epoch": 0.5136295093376638,
+      "grad_norm": 16.494372122641952,
+      "learning_rate": 2e-06,
+      "loss": 0.3281,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5138615009859645,
+      "grad_norm": 12.449203842828988,
+      "learning_rate": 2e-06,
+      "loss": 0.2594,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5140934926342652,
+      "grad_norm": 15.52348629468538,
+      "learning_rate": 2e-06,
+      "loss": 0.3694,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5143254842825659,
+      "grad_norm": 17.681642491435536,
+      "learning_rate": 2e-06,
+      "loss": 0.3573,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5145574759308665,
+      "grad_norm": 6.604458338708483,
+      "learning_rate": 2e-06,
+      "loss": 0.2758,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5147894675791671,
+      "grad_norm": 13.732407906659214,
+      "learning_rate": 2e-06,
+      "loss": 0.304,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5150214592274678,
+      "grad_norm": 19.435164902984656,
+      "learning_rate": 2e-06,
+      "loss": 0.2704,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5152534508757685,
+      "grad_norm": 17.545281421409236,
+      "learning_rate": 2e-06,
+      "loss": 0.3262,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5154854425240691,
+      "grad_norm": 9.242074142541009,
+      "learning_rate": 2e-06,
+      "loss": 0.2567,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5157174341723698,
+      "grad_norm": 15.786352449218043,
+      "learning_rate": 2e-06,
+      "loss": 0.3375,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5159494258206705,
+      "grad_norm": 5.905837709522254,
+      "learning_rate": 2e-06,
+      "loss": 0.2207,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5161814174689712,
+      "grad_norm": 14.735766964453777,
+      "learning_rate": 2e-06,
+      "loss": 0.3925,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5164134091172717,
+      "grad_norm": 15.461827147246586,
+      "learning_rate": 2e-06,
+      "loss": 0.3894,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5166454007655724,
+      "grad_norm": 10.058630282691166,
+      "learning_rate": 2e-06,
+      "loss": 0.2264,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5168773924138731,
+      "grad_norm": 6.832754603678109,
+      "learning_rate": 2e-06,
+      "loss": 0.281,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5171093840621738,
+      "grad_norm": 16.511843498691217,
+      "learning_rate": 2e-06,
+      "loss": 0.3108,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5173413757104744,
+      "grad_norm": 19.925078422841,
+      "learning_rate": 2e-06,
+      "loss": 0.3226,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5175733673587751,
+      "grad_norm": 23.279579301825684,
+      "learning_rate": 2e-06,
+      "loss": 0.3833,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5178053590070757,
+      "grad_norm": 10.204446780260694,
+      "learning_rate": 2e-06,
+      "loss": 0.2996,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5180373506553764,
+      "grad_norm": 12.123237316873128,
+      "learning_rate": 2e-06,
+      "loss": 0.2906,
+      "step": 2233
+    },
+    {
+      "epoch": 0.518269342303677,
+      "grad_norm": 11.329122471507537,
+      "learning_rate": 2e-06,
+      "loss": 0.3399,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5185013339519777,
+      "grad_norm": 24.975527522311257,
+      "learning_rate": 2e-06,
+      "loss": 0.4095,
+      "step": 2235
+    },
+    {
+      "epoch": 0.5187333256002784,
+      "grad_norm": 15.940577202895128,
+      "learning_rate": 2e-06,
+      "loss": 0.3415,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5189653172485791,
+      "grad_norm": 20.14661131715556,
+      "learning_rate": 2e-06,
+      "loss": 0.3974,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5191973088968798,
+      "grad_norm": 25.747100784967884,
+      "learning_rate": 2e-06,
+      "loss": 0.5037,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5194293005451803,
+      "grad_norm": 14.208535073812218,
+      "learning_rate": 2e-06,
+      "loss": 0.2438,
+      "step": 2239
+    },
+    {
+      "epoch": 0.519661292193481,
+      "grad_norm": 17.03550596170188,
+      "learning_rate": 2e-06,
+      "loss": 0.3653,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5198932838417817,
+      "grad_norm": 20.93553667412997,
+      "learning_rate": 2e-06,
+      "loss": 0.2732,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5201252754900824,
+      "grad_norm": 10.008111793167586,
+      "learning_rate": 2e-06,
+      "loss": 0.2699,
+      "step": 2242
+    },
+    {
+      "epoch": 0.520357267138383,
+      "grad_norm": 20.26414491317248,
+      "learning_rate": 2e-06,
+      "loss": 0.3475,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5205892587866837,
+      "grad_norm": 24.230542480582827,
+      "learning_rate": 2e-06,
+      "loss": 0.3996,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5208212504349844,
+      "grad_norm": 20.463428916265258,
+      "learning_rate": 2e-06,
+      "loss": 0.2895,
+      "step": 2245
+    },
+    {
+      "epoch": 0.521053242083285,
+      "grad_norm": 12.477672191195685,
+      "learning_rate": 2e-06,
+      "loss": 0.2641,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5212852337315856,
+      "grad_norm": 100.35080063444947,
+      "learning_rate": 2e-06,
+      "loss": 0.2666,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5215172253798863,
+      "grad_norm": 24.628890125223037,
+      "learning_rate": 2e-06,
+      "loss": 0.3212,
+      "step": 2248
+    },
+    {
+      "epoch": 0.521749217028187,
+      "grad_norm": 9.204998287585962,
+      "learning_rate": 2e-06,
+      "loss": 0.2658,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5219812086764877,
+      "grad_norm": 18.863388131065673,
+      "learning_rate": 2e-06,
+      "loss": 0.3096,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5222132003247884,
+      "grad_norm": 15.814258917614072,
+      "learning_rate": 2e-06,
+      "loss": 0.2506,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5224451919730889,
+      "grad_norm": 12.02163164782188,
+      "learning_rate": 2e-06,
+      "loss": 0.2721,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5226771836213896,
+      "grad_norm": 13.910898698958496,
+      "learning_rate": 2e-06,
+      "loss": 0.2853,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5229091752696903,
+      "grad_norm": 17.072635763484605,
+      "learning_rate": 2e-06,
+      "loss": 0.2623,
+      "step": 2254
+    },
+    {
+      "epoch": 0.523141166917991,
+      "grad_norm": 14.665093774873432,
+      "learning_rate": 2e-06,
+      "loss": 0.3196,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5233731585662916,
+      "grad_norm": 17.038607429232925,
+      "learning_rate": 2e-06,
+      "loss": 0.3338,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5236051502145923,
+      "grad_norm": 8.32717226665899,
+      "learning_rate": 2e-06,
+      "loss": 0.2404,
+      "step": 2257
+    },
+    {
+      "epoch": 0.523837141862893,
+      "grad_norm": 16.883098831842254,
+      "learning_rate": 2e-06,
+      "loss": 0.2216,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5240691335111936,
+      "grad_norm": 20.377352713674426,
+      "learning_rate": 2e-06,
+      "loss": 0.3238,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5243011251594942,
+      "grad_norm": 12.458863951225556,
+      "learning_rate": 2e-06,
+      "loss": 0.3009,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5245331168077949,
+      "grad_norm": 11.239890354739732,
+      "learning_rate": 2e-06,
+      "loss": 0.3214,
+      "step": 2261
+    },
+    {
+      "epoch": 0.5247651084560956,
+      "grad_norm": 18.620507924392726,
+      "learning_rate": 2e-06,
+      "loss": 0.3723,
+      "step": 2262
+    },
+    {
+      "epoch": 0.5249971001043963,
+      "grad_norm": 13.254486794790614,
+      "learning_rate": 2e-06,
+      "loss": 0.2997,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5252290917526969,
+      "grad_norm": 13.22770783264009,
+      "learning_rate": 2e-06,
+      "loss": 0.3,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5254610834009976,
+      "grad_norm": 17.009437904290014,
+      "learning_rate": 2e-06,
+      "loss": 0.3899,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5256930750492982,
+      "grad_norm": 20.946786476780414,
+      "learning_rate": 2e-06,
+      "loss": 0.3711,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5259250666975989,
+      "grad_norm": 13.43035616771182,
+      "learning_rate": 2e-06,
+      "loss": 0.2795,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5261570583458995,
+      "grad_norm": 15.701373292627125,
+      "learning_rate": 2e-06,
+      "loss": 0.34,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5263890499942002,
+      "grad_norm": 11.788398579381099,
+      "learning_rate": 2e-06,
+      "loss": 0.2095,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5266210416425009,
+      "grad_norm": 21.228215030045135,
+      "learning_rate": 2e-06,
+      "loss": 0.3226,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5268530332908016,
+      "grad_norm": 13.830399196968724,
+      "learning_rate": 2e-06,
+      "loss": 0.2479,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5270850249391021,
+      "grad_norm": 19.86319616283615,
+      "learning_rate": 2e-06,
+      "loss": 0.3438,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5273170165874028,
+      "grad_norm": 11.956746996575461,
+      "learning_rate": 2e-06,
+      "loss": 0.2824,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5275490082357035,
+      "grad_norm": 11.180178543415238,
+      "learning_rate": 2e-06,
+      "loss": 0.2802,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5277809998840042,
+      "grad_norm": 9.880631130726139,
+      "learning_rate": 2e-06,
+      "loss": 0.3052,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5280129915323049,
+      "grad_norm": 9.766750451583386,
+      "learning_rate": 2e-06,
+      "loss": 0.2437,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5282449831806055,
+      "grad_norm": 18.584827551294715,
+      "learning_rate": 2e-06,
+      "loss": 0.3032,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5284769748289062,
+      "grad_norm": 13.19289878347082,
+      "learning_rate": 2e-06,
+      "loss": 0.2042,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5287089664772068,
+      "grad_norm": 8.841165204840243,
+      "learning_rate": 2e-06,
+      "loss": 0.1651,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5289409581255075,
+      "grad_norm": 12.913411705213157,
+      "learning_rate": 2e-06,
+      "loss": 0.2826,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5291729497738081,
+      "grad_norm": 16.76037343633817,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5294049414221088,
+      "grad_norm": 15.153622033605663,
+      "learning_rate": 2e-06,
+      "loss": 0.2516,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5296369330704095,
+      "grad_norm": 9.680270585277633,
+      "learning_rate": 2e-06,
+      "loss": 0.2359,
+      "step": 2283
+    },
+    {
+      "epoch": 0.5298689247187102,
+      "grad_norm": 22.396002284150175,
+      "learning_rate": 2e-06,
+      "loss": 0.3231,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5301009163670107,
+      "grad_norm": 16.939145166831942,
+      "learning_rate": 2e-06,
+      "loss": 0.4175,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5303329080153114,
+      "grad_norm": 13.878641540351467,
+      "learning_rate": 2e-06,
+      "loss": 0.2965,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5305648996636121,
+      "grad_norm": 14.818456238172777,
+      "learning_rate": 2e-06,
+      "loss": 0.2961,
+      "step": 2287
+    },
+    {
+      "epoch": 0.5307968913119128,
+      "grad_norm": 12.767136730539981,
+      "learning_rate": 2e-06,
+      "loss": 0.258,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5310288829602134,
+      "grad_norm": 10.171811680648132,
+      "learning_rate": 2e-06,
+      "loss": 0.2368,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5312608746085141,
+      "grad_norm": 16.018420281641706,
+      "learning_rate": 2e-06,
+      "loss": 0.3154,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5314928662568148,
+      "grad_norm": 13.07454574100996,
+      "learning_rate": 2e-06,
+      "loss": 0.2373,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5317248579051154,
+      "grad_norm": 13.353026905275584,
+      "learning_rate": 2e-06,
+      "loss": 0.4516,
+      "step": 2292
+    },
+    {
+      "epoch": 0.531956849553416,
+      "grad_norm": 12.39102054984825,
+      "learning_rate": 2e-06,
+      "loss": 0.2753,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5321888412017167,
+      "grad_norm": 17.34864838714827,
+      "learning_rate": 2e-06,
+      "loss": 0.3422,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5324208328500174,
+      "grad_norm": 16.64109635116588,
+      "learning_rate": 2e-06,
+      "loss": 0.1882,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5326528244983181,
+      "grad_norm": 14.168599271666439,
+      "learning_rate": 2e-06,
+      "loss": 0.4041,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5328848161466188,
+      "grad_norm": 16.024561602681622,
+      "learning_rate": 2e-06,
+      "loss": 0.2647,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5331168077949194,
+      "grad_norm": 15.266324319466408,
+      "learning_rate": 2e-06,
+      "loss": 0.3007,
+      "step": 2298
+    },
+    {
+      "epoch": 0.53334879944322,
+      "grad_norm": 14.756981269977882,
+      "learning_rate": 2e-06,
+      "loss": 0.2312,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5335807910915207,
+      "grad_norm": 20.247224134109825,
+      "learning_rate": 2e-06,
+      "loss": 0.302,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5338127827398214,
+      "grad_norm": 18.21501807524069,
+      "learning_rate": 2e-06,
+      "loss": 0.2802,
+      "step": 2301
+    },
+    {
+      "epoch": 0.534044774388122,
+      "grad_norm": 9.884233780654693,
+      "learning_rate": 2e-06,
+      "loss": 0.2565,
+      "step": 2302
+    },
+    {
+      "epoch": 0.5342767660364227,
+      "grad_norm": 17.423012073490213,
+      "learning_rate": 2e-06,
+      "loss": 0.3299,
+      "step": 2303
+    },
+    {
+      "epoch": 0.5345087576847234,
+      "grad_norm": 18.25976551054708,
+      "learning_rate": 2e-06,
+      "loss": 0.4491,
+      "step": 2304
+    },
+    {
+      "epoch": 0.534740749333024,
+      "grad_norm": 11.64769360043185,
+      "learning_rate": 2e-06,
+      "loss": 0.2243,
+      "step": 2305
+    },
+    {
+      "epoch": 0.5349727409813246,
+      "grad_norm": 14.89009167595676,
+      "learning_rate": 2e-06,
+      "loss": 0.2694,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5352047326296253,
+      "grad_norm": 17.75693597667897,
+      "learning_rate": 2e-06,
+      "loss": 0.4001,
+      "step": 2307
+    },
+    {
+      "epoch": 0.535436724277926,
+      "grad_norm": 17.159122883056842,
+      "learning_rate": 2e-06,
+      "loss": 0.3234,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5356687159262267,
+      "grad_norm": 13.326925118558393,
+      "learning_rate": 2e-06,
+      "loss": 0.2136,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5359007075745273,
+      "grad_norm": 24.72149462549094,
+      "learning_rate": 2e-06,
+      "loss": 0.3685,
+      "step": 2310
+    },
+    {
+      "epoch": 0.536132699222828,
+      "grad_norm": 8.387532478488598,
+      "learning_rate": 2e-06,
+      "loss": 0.2289,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5363646908711286,
+      "grad_norm": 18.357082381108455,
+      "learning_rate": 2e-06,
+      "loss": 0.3841,
+      "step": 2312
+    },
+    {
+      "epoch": 0.5365966825194293,
+      "grad_norm": 12.95410078049241,
+      "learning_rate": 2e-06,
+      "loss": 0.3858,
+      "step": 2313
+    },
+    {
+      "epoch": 0.53682867416773,
+      "grad_norm": 13.089473379625638,
+      "learning_rate": 2e-06,
+      "loss": 0.2664,
+      "step": 2314
+    },
+    {
+      "epoch": 0.5370606658160306,
+      "grad_norm": 19.433935514249765,
+      "learning_rate": 2e-06,
+      "loss": 0.2824,
+      "step": 2315
+    },
+    {
+      "epoch": 0.5372926574643313,
+      "grad_norm": 10.02339829476107,
+      "learning_rate": 2e-06,
+      "loss": 0.2437,
+      "step": 2316
+    },
+    {
+      "epoch": 0.537524649112632,
+      "grad_norm": 12.552941976553708,
+      "learning_rate": 2e-06,
+      "loss": 0.3056,
+      "step": 2317
+    },
+    {
+      "epoch": 0.5377566407609327,
+      "grad_norm": 7.083892332328677,
+      "learning_rate": 2e-06,
+      "loss": 0.1957,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5379886324092332,
+      "grad_norm": 20.42922099874839,
+      "learning_rate": 2e-06,
+      "loss": 0.4354,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5382206240575339,
+      "grad_norm": 20.31467380319267,
+      "learning_rate": 2e-06,
+      "loss": 0.3671,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5384526157058346,
+      "grad_norm": 14.633241874519927,
+      "learning_rate": 2e-06,
+      "loss": 0.2382,
+      "step": 2321
+    },
+    {
+      "epoch": 0.5386846073541353,
+      "grad_norm": 12.98039438440369,
+      "learning_rate": 2e-06,
+      "loss": 0.2664,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5389165990024359,
+      "grad_norm": 16.565699553079266,
+      "learning_rate": 2e-06,
+      "loss": 0.2974,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5391485906507366,
+      "grad_norm": 15.238803129232114,
+      "learning_rate": 2e-06,
+      "loss": 0.343,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5393805822990372,
+      "grad_norm": 6.829254929370642,
+      "learning_rate": 2e-06,
+      "loss": 0.1979,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5396125739473379,
+      "grad_norm": 9.858620413535052,
+      "learning_rate": 2e-06,
+      "loss": 0.2957,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5398445655956385,
+      "grad_norm": 6.932270398065999,
+      "learning_rate": 2e-06,
+      "loss": 0.2907,
+      "step": 2327
+    },
+    {
+      "epoch": 0.5400765572439392,
+      "grad_norm": 12.710146071255126,
+      "learning_rate": 2e-06,
+      "loss": 0.2605,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5403085488922399,
+      "grad_norm": 7.804004164747687,
+      "learning_rate": 2e-06,
+      "loss": 0.1733,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 21.81106908709857,
+      "learning_rate": 2e-06,
+      "loss": 0.3851,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5407725321888412,
+      "grad_norm": 5.8748588911926145,
+      "learning_rate": 2e-06,
+      "loss": 0.2951,
+      "step": 2331
+    },
+    {
+      "epoch": 0.5410045238371418,
+      "grad_norm": 16.2691670015431,
+      "learning_rate": 2e-06,
+      "loss": 0.2875,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5412365154854425,
+      "grad_norm": 17.236361159889725,
+      "learning_rate": 2e-06,
+      "loss": 0.2956,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5414685071337432,
+      "grad_norm": 10.273390136146526,
+      "learning_rate": 2e-06,
+      "loss": 0.3004,
+      "step": 2334
+    },
+    {
+      "epoch": 0.5417004987820438,
+      "grad_norm": 20.87612590795775,
+      "learning_rate": 2e-06,
+      "loss": 0.3434,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5419324904303445,
+      "grad_norm": 27.08238944500166,
+      "learning_rate": 2e-06,
+      "loss": 0.5375,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5421644820786452,
+      "grad_norm": 6.215206436525487,
+      "learning_rate": 2e-06,
+      "loss": 0.1769,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5423964737269459,
+      "grad_norm": 15.724186262250118,
+      "learning_rate": 2e-06,
+      "loss": 0.3055,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5426284653752464,
+      "grad_norm": 11.368510674204973,
+      "learning_rate": 2e-06,
+      "loss": 0.2766,
+      "step": 2339
+    },
+    {
+      "epoch": 0.5428604570235471,
+      "grad_norm": 18.357210746849432,
+      "learning_rate": 2e-06,
+      "loss": 0.3577,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5430924486718478,
+      "grad_norm": 12.542863291662933,
+      "learning_rate": 2e-06,
+      "loss": 0.2229,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5433244403201485,
+      "grad_norm": 12.147225659096481,
+      "learning_rate": 2e-06,
+      "loss": 0.3056,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5435564319684492,
+      "grad_norm": 12.837094753441027,
+      "learning_rate": 2e-06,
+      "loss": 0.3689,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5437884236167498,
+      "grad_norm": 14.999446526361522,
+      "learning_rate": 2e-06,
+      "loss": 0.2976,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5440204152650504,
+      "grad_norm": 26.954049953958133,
+      "learning_rate": 2e-06,
+      "loss": 0.4802,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5442524069133511,
+      "grad_norm": 15.525353341388612,
+      "learning_rate": 2e-06,
+      "loss": 0.3723,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5444843985616518,
+      "grad_norm": 18.84480370516095,
+      "learning_rate": 2e-06,
+      "loss": 0.4144,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5447163902099524,
+      "grad_norm": 9.673118895122698,
+      "learning_rate": 2e-06,
+      "loss": 0.1825,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5449483818582531,
+      "grad_norm": 6.578330514551692,
+      "learning_rate": 2e-06,
+      "loss": 0.228,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5451803735065538,
+      "grad_norm": 21.222832338767525,
+      "learning_rate": 2e-06,
+      "loss": 0.3208,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5454123651548545,
+      "grad_norm": 19.578665813588188,
+      "learning_rate": 2e-06,
+      "loss": 0.3288,
+      "step": 2351
+    },
+    {
+      "epoch": 0.545644356803155,
+      "grad_norm": 17.952013710643097,
+      "learning_rate": 2e-06,
+      "loss": 0.2957,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5458763484514557,
+      "grad_norm": 16.066508436001325,
+      "learning_rate": 2e-06,
+      "loss": 0.3123,
+      "step": 2353
+    },
+    {
+      "epoch": 0.5461083400997564,
+      "grad_norm": 8.080325358675177,
+      "learning_rate": 2e-06,
+      "loss": 0.267,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5463403317480571,
+      "grad_norm": 16.236401419663363,
+      "learning_rate": 2e-06,
+      "loss": 0.3246,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5465723233963578,
+      "grad_norm": 23.272151572351987,
+      "learning_rate": 2e-06,
+      "loss": 0.353,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5468043150446584,
+      "grad_norm": 13.538462553357316,
+      "learning_rate": 2e-06,
+      "loss": 0.2789,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5470363066929591,
+      "grad_norm": 16.407723112546826,
+      "learning_rate": 2e-06,
+      "loss": 0.2905,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5472682983412597,
+      "grad_norm": 8.124259163275745,
+      "learning_rate": 2e-06,
+      "loss": 0.2108,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5475002899895604,
+      "grad_norm": 20.691277150417193,
+      "learning_rate": 2e-06,
+      "loss": 0.3986,
+      "step": 2360
+    },
+    {
+      "epoch": 0.547732281637861,
+      "grad_norm": 24.34822370897118,
+      "learning_rate": 2e-06,
+      "loss": 0.3407,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5479642732861617,
+      "grad_norm": 14.216553246002624,
+      "learning_rate": 2e-06,
+      "loss": 0.3413,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5481962649344624,
+      "grad_norm": 14.793477667414159,
+      "learning_rate": 2e-06,
+      "loss": 0.3285,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5484282565827631,
+      "grad_norm": 12.641379310724114,
+      "learning_rate": 2e-06,
+      "loss": 0.2678,
+      "step": 2364
+    },
+    {
+      "epoch": 0.5486602482310636,
+      "grad_norm": 13.317895196477977,
+      "learning_rate": 2e-06,
+      "loss": 0.325,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5488922398793643,
+      "grad_norm": 16.161930551381584,
+      "learning_rate": 2e-06,
+      "loss": 0.3113,
+      "step": 2366
+    },
+    {
+      "epoch": 0.549124231527665,
+      "grad_norm": 7.385801035359576,
+      "learning_rate": 2e-06,
+      "loss": 0.2321,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5493562231759657,
+      "grad_norm": 11.702660325886935,
+      "learning_rate": 2e-06,
+      "loss": 0.3183,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5495882148242663,
+      "grad_norm": 14.062392473354258,
+      "learning_rate": 2e-06,
+      "loss": 0.254,
+      "step": 2369
+    },
+    {
+      "epoch": 0.549820206472567,
+      "grad_norm": 18.748060159797518,
+      "learning_rate": 2e-06,
+      "loss": 0.3368,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5500521981208677,
+      "grad_norm": 14.300497850364994,
+      "learning_rate": 2e-06,
+      "loss": 0.3072,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5502841897691683,
+      "grad_norm": 10.418969804626075,
+      "learning_rate": 2e-06,
+      "loss": 0.2342,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5505161814174689,
+      "grad_norm": 20.44987186445462,
+      "learning_rate": 2e-06,
+      "loss": 0.3585,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5507481730657696,
+      "grad_norm": 20.05953857845853,
+      "learning_rate": 2e-06,
+      "loss": 0.2701,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5509801647140703,
+      "grad_norm": 16.11962443192865,
+      "learning_rate": 2e-06,
+      "loss": 0.2922,
+      "step": 2375
+    },
+    {
+      "epoch": 0.551212156362371,
+      "grad_norm": 10.646811964940431,
+      "learning_rate": 2e-06,
+      "loss": 0.2887,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5514441480106717,
+      "grad_norm": 18.017398046246253,
+      "learning_rate": 2e-06,
+      "loss": 0.3291,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5516761396589723,
+      "grad_norm": 9.988883639517411,
+      "learning_rate": 2e-06,
+      "loss": 0.2606,
+      "step": 2378
+    },
+    {
+      "epoch": 0.5519081313072729,
+      "grad_norm": 12.064044391011736,
+      "learning_rate": 2e-06,
+      "loss": 0.3694,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5521401229555736,
+      "grad_norm": 11.821170069247135,
+      "learning_rate": 2e-06,
+      "loss": 0.1949,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5523721146038743,
+      "grad_norm": 9.066163899738006,
+      "learning_rate": 2e-06,
+      "loss": 0.1487,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5526041062521749,
+      "grad_norm": 11.322561241704996,
+      "learning_rate": 2e-06,
+      "loss": 0.3068,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5528360979004756,
+      "grad_norm": 7.512021724930349,
+      "learning_rate": 2e-06,
+      "loss": 0.2746,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5530680895487763,
+      "grad_norm": 9.300021551053998,
+      "learning_rate": 2e-06,
+      "loss": 0.2483,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5533000811970769,
+      "grad_norm": 18.989662991902502,
+      "learning_rate": 2e-06,
+      "loss": 0.3323,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5535320728453775,
+      "grad_norm": 10.986524529354641,
+      "learning_rate": 2e-06,
+      "loss": 0.1893,
+      "step": 2386
+    },
+    {
+      "epoch": 0.5537640644936782,
+      "grad_norm": 10.924830004619258,
+      "learning_rate": 2e-06,
+      "loss": 0.311,
+      "step": 2387
+    },
+    {
+      "epoch": 0.5539960561419789,
+      "grad_norm": 14.54093682002165,
+      "learning_rate": 2e-06,
+      "loss": 0.2469,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5542280477902796,
+      "grad_norm": 23.32331970457638,
+      "learning_rate": 2e-06,
+      "loss": 0.3154,
+      "step": 2389
+    },
+    {
+      "epoch": 0.5544600394385802,
+      "grad_norm": 14.801663169850901,
+      "learning_rate": 2e-06,
+      "loss": 0.2924,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5546920310868809,
+      "grad_norm": 12.918355697187497,
+      "learning_rate": 2e-06,
+      "loss": 0.2827,
+      "step": 2391
+    },
+    {
+      "epoch": 0.5549240227351815,
+      "grad_norm": 17.089038419122126,
+      "learning_rate": 2e-06,
+      "loss": 0.321,
+      "step": 2392
+    },
+    {
+      "epoch": 0.5551560143834822,
+      "grad_norm": 9.67915384286139,
+      "learning_rate": 2e-06,
+      "loss": 0.2727,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5553880060317828,
+      "grad_norm": 7.766141239959042,
+      "learning_rate": 2e-06,
+      "loss": 0.191,
+      "step": 2394
+    },
+    {
+      "epoch": 0.5556199976800835,
+      "grad_norm": 18.090735119556687,
+      "learning_rate": 2e-06,
+      "loss": 0.3184,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5558519893283842,
+      "grad_norm": 8.505508622938232,
+      "learning_rate": 2e-06,
+      "loss": 0.2925,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5560839809766849,
+      "grad_norm": 14.387996447013048,
+      "learning_rate": 2e-06,
+      "loss": 0.3296,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5563159726249856,
+      "grad_norm": 16.06393041702412,
+      "learning_rate": 2e-06,
+      "loss": 0.3835,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5565479642732861,
+      "grad_norm": 13.847206072754162,
+      "learning_rate": 2e-06,
+      "loss": 0.2664,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5567799559215868,
+      "grad_norm": 6.8657066601555625,
+      "learning_rate": 2e-06,
+      "loss": 0.2501,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5570119475698875,
+      "grad_norm": 19.52261317249476,
+      "learning_rate": 2e-06,
+      "loss": 0.3469,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5572439392181882,
+      "grad_norm": 8.778377305253947,
+      "learning_rate": 2e-06,
+      "loss": 0.3689,
+      "step": 2402
+    },
+    {
+      "epoch": 0.5574759308664888,
+      "grad_norm": 16.378735216573,
+      "learning_rate": 2e-06,
+      "loss": 0.3538,
+      "step": 2403
+    },
+    {
+      "epoch": 0.5577079225147895,
+      "grad_norm": 11.787808238531637,
+      "learning_rate": 2e-06,
+      "loss": 0.2358,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5579399141630901,
+      "grad_norm": 9.171065652456917,
+      "learning_rate": 2e-06,
+      "loss": 0.2658,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5581719058113908,
+      "grad_norm": 10.920307577783632,
+      "learning_rate": 2e-06,
+      "loss": 0.3377,
+      "step": 2406
+    },
+    {
+      "epoch": 0.5584038974596914,
+      "grad_norm": 16.384068560006575,
+      "learning_rate": 2e-06,
+      "loss": 0.245,
+      "step": 2407
+    },
+    {
+      "epoch": 0.5586358891079921,
+      "grad_norm": 15.407554795589347,
+      "learning_rate": 2e-06,
+      "loss": 0.316,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5588678807562928,
+      "grad_norm": 8.255536852752579,
+      "learning_rate": 2e-06,
+      "loss": 0.3018,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5590998724045935,
+      "grad_norm": 15.595870272913924,
+      "learning_rate": 2e-06,
+      "loss": 0.2373,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5593318640528941,
+      "grad_norm": 10.573620346477986,
+      "learning_rate": 2e-06,
+      "loss": 0.2634,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5595638557011947,
+      "grad_norm": 13.44498052742146,
+      "learning_rate": 2e-06,
+      "loss": 0.2921,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5597958473494954,
+      "grad_norm": 18.226161572351923,
+      "learning_rate": 2e-06,
+      "loss": 0.3855,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5600278389977961,
+      "grad_norm": 26.04602255991392,
+      "learning_rate": 2e-06,
+      "loss": 0.3388,
+      "step": 2414
+    },
+    {
+      "epoch": 0.5602598306460967,
+      "grad_norm": 11.40825468917101,
+      "learning_rate": 2e-06,
+      "loss": 0.2476,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5604918222943974,
+      "grad_norm": 8.047637255378097,
+      "learning_rate": 2e-06,
+      "loss": 0.2142,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5607238139426981,
+      "grad_norm": 9.763552278815421,
+      "learning_rate": 2e-06,
+      "loss": 0.2166,
+      "step": 2417
+    },
+    {
+      "epoch": 0.5609558055909988,
+      "grad_norm": 15.426847438028386,
+      "learning_rate": 2e-06,
+      "loss": 0.3242,
+      "step": 2418
+    },
+    {
+      "epoch": 0.5611877972392993,
+      "grad_norm": 10.019695584582262,
+      "learning_rate": 2e-06,
+      "loss": 0.2315,
+      "step": 2419
+    },
+    {
+      "epoch": 0.5614197888876,
+      "grad_norm": 11.950391616471022,
+      "learning_rate": 2e-06,
+      "loss": 0.3032,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5616517805359007,
+      "grad_norm": 9.821360624161986,
+      "learning_rate": 2e-06,
+      "loss": 0.2753,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5618837721842014,
+      "grad_norm": 13.779153309516882,
+      "learning_rate": 2e-06,
+      "loss": 0.2562,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5621157638325021,
+      "grad_norm": 20.55957101204331,
+      "learning_rate": 2e-06,
+      "loss": 0.348,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5623477554808027,
+      "grad_norm": 10.717575945681206,
+      "learning_rate": 2e-06,
+      "loss": 0.3057,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5625797471291033,
+      "grad_norm": 9.5656433019114,
+      "learning_rate": 2e-06,
+      "loss": 0.2073,
+      "step": 2425
+    },
+    {
+      "epoch": 0.562811738777404,
+      "grad_norm": 14.943321776155047,
+      "learning_rate": 2e-06,
+      "loss": 0.2825,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5630437304257047,
+      "grad_norm": 14.854031581718427,
+      "learning_rate": 2e-06,
+      "loss": 0.4032,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5632757220740053,
+      "grad_norm": 19.945968462228393,
+      "learning_rate": 2e-06,
+      "loss": 0.2607,
+      "step": 2428
+    },
+    {
+      "epoch": 0.563507713722306,
+      "grad_norm": 21.348685688736698,
+      "learning_rate": 2e-06,
+      "loss": 0.2983,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5637397053706067,
+      "grad_norm": 8.603273140615295,
+      "learning_rate": 2e-06,
+      "loss": 0.296,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5639716970189074,
+      "grad_norm": 18.347215498971828,
+      "learning_rate": 2e-06,
+      "loss": 0.3555,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5642036886672079,
+      "grad_norm": 11.246191700500239,
+      "learning_rate": 2e-06,
+      "loss": 0.2004,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5644356803155086,
+      "grad_norm": 14.682817929588682,
+      "learning_rate": 2e-06,
+      "loss": 0.3376,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5646676719638093,
+      "grad_norm": 20.22286878096074,
+      "learning_rate": 2e-06,
+      "loss": 0.2333,
+      "step": 2434
+    },
+    {
+      "epoch": 0.56489966361211,
+      "grad_norm": 16.49444735633445,
+      "learning_rate": 2e-06,
+      "loss": 0.2472,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5651316552604106,
+      "grad_norm": 6.68551153484024,
+      "learning_rate": 2e-06,
+      "loss": 0.225,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5653636469087113,
+      "grad_norm": 9.686076248914613,
+      "learning_rate": 2e-06,
+      "loss": 0.3447,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5655956385570119,
+      "grad_norm": 17.76599719791514,
+      "learning_rate": 2e-06,
+      "loss": 0.2478,
+      "step": 2438
+    },
+    {
+      "epoch": 0.5658276302053126,
+      "grad_norm": 21.389768804288273,
+      "learning_rate": 2e-06,
+      "loss": 0.3174,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5660596218536132,
+      "grad_norm": 14.049780815178298,
+      "learning_rate": 2e-06,
+      "loss": 0.3321,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5662916135019139,
+      "grad_norm": 14.033113374604143,
+      "learning_rate": 2e-06,
+      "loss": 0.3,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5665236051502146,
+      "grad_norm": 22.448007786902185,
+      "learning_rate": 2e-06,
+      "loss": 0.259,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5667555967985153,
+      "grad_norm": 14.619950092423217,
+      "learning_rate": 2e-06,
+      "loss": 0.3202,
+      "step": 2443
+    },
+    {
+      "epoch": 0.566987588446816,
+      "grad_norm": 19.221172213265596,
+      "learning_rate": 2e-06,
+      "loss": 0.4239,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5672195800951165,
+      "grad_norm": 10.266028401673237,
+      "learning_rate": 2e-06,
+      "loss": 0.2292,
+      "step": 2445
+    },
+    {
+      "epoch": 0.5674515717434172,
+      "grad_norm": 16.665276447968612,
+      "learning_rate": 2e-06,
+      "loss": 0.2707,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5676835633917179,
+      "grad_norm": 10.6274177866923,
+      "learning_rate": 2e-06,
+      "loss": 0.3535,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5679155550400186,
+      "grad_norm": 11.8099997640705,
+      "learning_rate": 2e-06,
+      "loss": 0.366,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5681475466883192,
+      "grad_norm": 12.555980498186655,
+      "learning_rate": 2e-06,
+      "loss": 0.3625,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5683795383366199,
+      "grad_norm": 19.410923811816254,
+      "learning_rate": 2e-06,
+      "loss": 0.2926,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5686115299849206,
+      "grad_norm": 2.8525262146093486,
+      "learning_rate": 2e-06,
+      "loss": 0.1367,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5688435216332212,
+      "grad_norm": 10.782898016379185,
+      "learning_rate": 2e-06,
+      "loss": 0.3504,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5690755132815218,
+      "grad_norm": 16.295736831717836,
+      "learning_rate": 2e-06,
+      "loss": 0.3744,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5693075049298225,
+      "grad_norm": 12.536042109056044,
+      "learning_rate": 2e-06,
+      "loss": 0.2732,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5695394965781232,
+      "grad_norm": 14.381777859735644,
+      "learning_rate": 2e-06,
+      "loss": 0.3513,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5697714882264239,
+      "grad_norm": 18.1258827344621,
+      "learning_rate": 2e-06,
+      "loss": 0.2716,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5700034798747246,
+      "grad_norm": 10.674561459760259,
+      "learning_rate": 2e-06,
+      "loss": 0.314,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5702354715230251,
+      "grad_norm": 16.690016743500063,
+      "learning_rate": 2e-06,
+      "loss": 0.3153,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5704674631713258,
+      "grad_norm": 17.419336891896343,
+      "learning_rate": 2e-06,
+      "loss": 0.2838,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5706994548196265,
+      "grad_norm": 9.199259539316124,
+      "learning_rate": 2e-06,
+      "loss": 0.315,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5709314464679272,
+      "grad_norm": 11.35509051682607,
+      "learning_rate": 2e-06,
+      "loss": 0.3286,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5711634381162278,
+      "grad_norm": 13.829657807396838,
+      "learning_rate": 2e-06,
+      "loss": 0.2636,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5713954297645285,
+      "grad_norm": 13.03723236578468,
+      "learning_rate": 2e-06,
+      "loss": 0.3142,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5716274214128292,
+      "grad_norm": 14.110551716191281,
+      "learning_rate": 2e-06,
+      "loss": 0.323,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5718594130611298,
+      "grad_norm": 9.653588565786267,
+      "learning_rate": 2e-06,
+      "loss": 0.2312,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5720914047094304,
+      "grad_norm": 9.622618204933014,
+      "learning_rate": 2e-06,
+      "loss": 0.3371,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5723233963577311,
+      "grad_norm": 26.944013997098477,
+      "learning_rate": 2e-06,
+      "loss": 0.3133,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5725553880060318,
+      "grad_norm": 22.606898378576506,
+      "learning_rate": 2e-06,
+      "loss": 0.4094,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5727873796543325,
+      "grad_norm": 14.59816805685614,
+      "learning_rate": 2e-06,
+      "loss": 0.4009,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5730193713026331,
+      "grad_norm": 16.673847044715497,
+      "learning_rate": 2e-06,
+      "loss": 0.3406,
+      "step": 2470
+    },
+    {
+      "epoch": 0.5732513629509338,
+      "grad_norm": 14.847720782363652,
+      "learning_rate": 2e-06,
+      "loss": 0.2871,
+      "step": 2471
+    },
+    {
+      "epoch": 0.5734833545992344,
+      "grad_norm": 18.607494659772954,
+      "learning_rate": 2e-06,
+      "loss": 0.4925,
+      "step": 2472
+    },
+    {
+      "epoch": 0.5737153462475351,
+      "grad_norm": 14.499260801159384,
+      "learning_rate": 2e-06,
+      "loss": 0.3471,
+      "step": 2473
+    },
+    {
+      "epoch": 0.5739473378958357,
+      "grad_norm": 10.426182616781508,
+      "learning_rate": 2e-06,
+      "loss": 0.2871,
+      "step": 2474
+    },
+    {
+      "epoch": 0.5741793295441364,
+      "grad_norm": 11.674093773224836,
+      "learning_rate": 2e-06,
+      "loss": 0.2291,
+      "step": 2475
+    },
+    {
+      "epoch": 0.5744113211924371,
+      "grad_norm": 17.47014738833898,
+      "learning_rate": 2e-06,
+      "loss": 0.2749,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5746433128407378,
+      "grad_norm": 11.770554342989735,
+      "learning_rate": 2e-06,
+      "loss": 0.264,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5748753044890383,
+      "grad_norm": 20.94092375824433,
+      "learning_rate": 2e-06,
+      "loss": 0.3868,
+      "step": 2478
+    },
+    {
+      "epoch": 0.575107296137339,
+      "grad_norm": 13.262166143545379,
+      "learning_rate": 2e-06,
+      "loss": 0.2915,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5753392877856397,
+      "grad_norm": 8.214286704099807,
+      "learning_rate": 2e-06,
+      "loss": 0.3086,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5755712794339404,
+      "grad_norm": 8.507786708511688,
+      "learning_rate": 2e-06,
+      "loss": 0.2798,
+      "step": 2481
+    },
+    {
+      "epoch": 0.575803271082241,
+      "grad_norm": 18.642647129706745,
+      "learning_rate": 2e-06,
+      "loss": 0.3305,
+      "step": 2482
+    },
+    {
+      "epoch": 0.5760352627305417,
+      "grad_norm": 15.689231004790807,
+      "learning_rate": 2e-06,
+      "loss": 0.2448,
+      "step": 2483
+    },
+    {
+      "epoch": 0.5762672543788424,
+      "grad_norm": 12.418802098698727,
+      "learning_rate": 2e-06,
+      "loss": 0.2637,
+      "step": 2484
+    },
+    {
+      "epoch": 0.576499246027143,
+      "grad_norm": 11.158407856020904,
+      "learning_rate": 2e-06,
+      "loss": 0.3159,
+      "step": 2485
+    },
+    {
+      "epoch": 0.5767312376754437,
+      "grad_norm": 19.61483635744511,
+      "learning_rate": 2e-06,
+      "loss": 0.294,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5769632293237443,
+      "grad_norm": 14.865542445369957,
+      "learning_rate": 2e-06,
+      "loss": 0.2789,
+      "step": 2487
+    },
+    {
+      "epoch": 0.577195220972045,
+      "grad_norm": 15.83240384681972,
+      "learning_rate": 2e-06,
+      "loss": 0.3269,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5774272126203457,
+      "grad_norm": 15.78410172720427,
+      "learning_rate": 2e-06,
+      "loss": 0.3453,
+      "step": 2489
+    },
+    {
+      "epoch": 0.5776592042686464,
+      "grad_norm": 8.634452953530715,
+      "learning_rate": 2e-06,
+      "loss": 0.2287,
+      "step": 2490
+    },
+    {
+      "epoch": 0.577891195916947,
+      "grad_norm": 12.732603792922383,
+      "learning_rate": 2e-06,
+      "loss": 0.2653,
+      "step": 2491
+    },
+    {
+      "epoch": 0.5781231875652476,
+      "grad_norm": 16.61088777946793,
+      "learning_rate": 2e-06,
+      "loss": 0.2511,
+      "step": 2492
+    },
+    {
+      "epoch": 0.5783551792135483,
+      "grad_norm": 15.248936766635833,
+      "learning_rate": 2e-06,
+      "loss": 0.2938,
+      "step": 2493
+    },
+    {
+      "epoch": 0.578587170861849,
+      "grad_norm": 22.75939149932287,
+      "learning_rate": 2e-06,
+      "loss": 0.4973,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5788191625101496,
+      "grad_norm": 13.169262362496072,
+      "learning_rate": 2e-06,
+      "loss": 0.3417,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5790511541584503,
+      "grad_norm": 14.869556808260299,
+      "learning_rate": 2e-06,
+      "loss": 0.3089,
+      "step": 2496
+    },
+    {
+      "epoch": 0.579283145806751,
+      "grad_norm": 7.219403236004867,
+      "learning_rate": 2e-06,
+      "loss": 0.1704,
+      "step": 2497
+    },
+    {
+      "epoch": 0.5795151374550516,
+      "grad_norm": 20.08537920234524,
+      "learning_rate": 2e-06,
+      "loss": 0.2664,
+      "step": 2498
+    },
+    {
+      "epoch": 0.5797471291033522,
+      "grad_norm": 6.917849407953844,
+      "learning_rate": 2e-06,
+      "loss": 0.2856,
+      "step": 2499
+    },
+    {
+      "epoch": 0.5799791207516529,
+      "grad_norm": 17.85469764329847,
+      "learning_rate": 2e-06,
+      "loss": 0.3051,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5802111123999536,
+      "grad_norm": 9.285602566523718,
+      "learning_rate": 2e-06,
+      "loss": 0.1979,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5804431040482543,
+      "grad_norm": 12.654586179547152,
+      "learning_rate": 2e-06,
+      "loss": 0.2821,
+      "step": 2502
+    },
+    {
+      "epoch": 0.580675095696555,
+      "grad_norm": 20.484107836541714,
+      "learning_rate": 2e-06,
+      "loss": 0.4086,
+      "step": 2503
+    },
+    {
+      "epoch": 0.5809070873448556,
+      "grad_norm": 30.810539020797318,
+      "learning_rate": 2e-06,
+      "loss": 0.4463,
+      "step": 2504
+    },
+    {
+      "epoch": 0.5811390789931562,
+      "grad_norm": 8.239820096389755,
+      "learning_rate": 2e-06,
+      "loss": 0.1982,
+      "step": 2505
+    },
+    {
+      "epoch": 0.5813710706414569,
+      "grad_norm": 6.987799616443373,
+      "learning_rate": 2e-06,
+      "loss": 0.2571,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5816030622897576,
+      "grad_norm": 9.824023287155521,
+      "learning_rate": 2e-06,
+      "loss": 0.2848,
+      "step": 2507
+    },
+    {
+      "epoch": 0.5818350539380582,
+      "grad_norm": 19.77791150461645,
+      "learning_rate": 2e-06,
+      "loss": 0.3736,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5820670455863589,
+      "grad_norm": 35.11127353541142,
+      "learning_rate": 2e-06,
+      "loss": 0.4481,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5822990372346596,
+      "grad_norm": 11.140002417533,
+      "learning_rate": 2e-06,
+      "loss": 0.3415,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5825310288829603,
+      "grad_norm": 17.981016984708337,
+      "learning_rate": 2e-06,
+      "loss": 0.3271,
+      "step": 2511
+    },
+    {
+      "epoch": 0.5827630205312608,
+      "grad_norm": 11.72871904063151,
+      "learning_rate": 2e-06,
+      "loss": 0.2438,
+      "step": 2512
+    },
+    {
+      "epoch": 0.5829950121795615,
+      "grad_norm": 5.025027934513667,
+      "learning_rate": 2e-06,
+      "loss": 0.1639,
+      "step": 2513
+    },
+    {
+      "epoch": 0.5832270038278622,
+      "grad_norm": 11.810865289588344,
+      "learning_rate": 2e-06,
+      "loss": 0.2174,
+      "step": 2514
+    },
+    {
+      "epoch": 0.5834589954761629,
+      "grad_norm": 24.26560656264537,
+      "learning_rate": 2e-06,
+      "loss": 0.2419,
+      "step": 2515
+    },
+    {
+      "epoch": 0.5836909871244635,
+      "grad_norm": 18.46991140854295,
+      "learning_rate": 2e-06,
+      "loss": 0.2605,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5839229787727642,
+      "grad_norm": 15.03403044306349,
+      "learning_rate": 2e-06,
+      "loss": 0.3359,
+      "step": 2517
+    },
+    {
+      "epoch": 0.5841549704210648,
+      "grad_norm": 11.645891200490585,
+      "learning_rate": 2e-06,
+      "loss": 0.2115,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5843869620693655,
+      "grad_norm": 14.340377431021409,
+      "learning_rate": 2e-06,
+      "loss": 0.2448,
+      "step": 2519
+    },
+    {
+      "epoch": 0.5846189537176661,
+      "grad_norm": 14.147283918955994,
+      "learning_rate": 2e-06,
+      "loss": 0.2075,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5848509453659668,
+      "grad_norm": 17.865648426932434,
+      "learning_rate": 2e-06,
+      "loss": 0.2673,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5850829370142675,
+      "grad_norm": 14.55836248361842,
+      "learning_rate": 2e-06,
+      "loss": 0.2814,
+      "step": 2522
+    },
+    {
+      "epoch": 0.5853149286625682,
+      "grad_norm": 16.81136985208734,
+      "learning_rate": 2e-06,
+      "loss": 0.3298,
+      "step": 2523
+    },
+    {
+      "epoch": 0.5855469203108689,
+      "grad_norm": 20.565540320996366,
+      "learning_rate": 2e-06,
+      "loss": 0.3327,
+      "step": 2524
+    },
+    {
+      "epoch": 0.5857789119591694,
+      "grad_norm": 24.14059008951668,
+      "learning_rate": 2e-06,
+      "loss": 0.337,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5860109036074701,
+      "grad_norm": 15.991993948265032,
+      "learning_rate": 2e-06,
+      "loss": 0.4677,
+      "step": 2526
+    },
+    {
+      "epoch": 0.5862428952557708,
+      "grad_norm": 12.974831364373912,
+      "learning_rate": 2e-06,
+      "loss": 0.3675,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5864748869040715,
+      "grad_norm": 15.676183243588898,
+      "learning_rate": 2e-06,
+      "loss": 0.3662,
+      "step": 2528
+    },
+    {
+      "epoch": 0.5867068785523721,
+      "grad_norm": 20.18352831596817,
+      "learning_rate": 2e-06,
+      "loss": 0.3298,
+      "step": 2529
+    },
+    {
+      "epoch": 0.5869388702006728,
+      "grad_norm": 11.742845399116613,
+      "learning_rate": 2e-06,
+      "loss": 0.2105,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5871708618489735,
+      "grad_norm": 15.036929135487856,
+      "learning_rate": 2e-06,
+      "loss": 0.3134,
+      "step": 2531
+    },
+    {
+      "epoch": 0.5874028534972741,
+      "grad_norm": 10.653932848203507,
+      "learning_rate": 2e-06,
+      "loss": 0.224,
+      "step": 2532
+    },
+    {
+      "epoch": 0.5876348451455747,
+      "grad_norm": 16.1086614736084,
+      "learning_rate": 2e-06,
+      "loss": 0.3744,
+      "step": 2533
+    },
+    {
+      "epoch": 0.5878668367938754,
+      "grad_norm": 29.101406200313374,
+      "learning_rate": 2e-06,
+      "loss": 0.3329,
+      "step": 2534
+    },
+    {
+      "epoch": 0.5880988284421761,
+      "grad_norm": 38.350231530496075,
+      "learning_rate": 2e-06,
+      "loss": 0.4513,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5883308200904768,
+      "grad_norm": 23.724065742515208,
+      "learning_rate": 2e-06,
+      "loss": 0.3881,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5885628117387774,
+      "grad_norm": 15.37795572922142,
+      "learning_rate": 2e-06,
+      "loss": 0.2551,
+      "step": 2537
+    },
+    {
+      "epoch": 0.588794803387078,
+      "grad_norm": 16.865307773422217,
+      "learning_rate": 2e-06,
+      "loss": 0.315,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5890267950353787,
+      "grad_norm": 17.793328192027396,
+      "learning_rate": 2e-06,
+      "loss": 0.2581,
+      "step": 2539
+    },
+    {
+      "epoch": 0.5892587866836794,
+      "grad_norm": 17.195738504318005,
+      "learning_rate": 2e-06,
+      "loss": 0.2906,
+      "step": 2540
+    },
+    {
+      "epoch": 0.58949077833198,
+      "grad_norm": 10.37882031859556,
+      "learning_rate": 2e-06,
+      "loss": 0.3523,
+      "step": 2541
+    },
+    {
+      "epoch": 0.5897227699802807,
+      "grad_norm": 8.259590272826024,
+      "learning_rate": 2e-06,
+      "loss": 0.2488,
+      "step": 2542
+    },
+    {
+      "epoch": 0.5899547616285814,
+      "grad_norm": 19.193523289503357,
+      "learning_rate": 2e-06,
+      "loss": 0.338,
+      "step": 2543
+    },
+    {
+      "epoch": 0.5901867532768821,
+      "grad_norm": 14.629504299379278,
+      "learning_rate": 2e-06,
+      "loss": 0.254,
+      "step": 2544
+    },
+    {
+      "epoch": 0.5904187449251826,
+      "grad_norm": 8.224400622005064,
+      "learning_rate": 2e-06,
+      "loss": 0.2536,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5906507365734833,
+      "grad_norm": 15.86256323018292,
+      "learning_rate": 2e-06,
+      "loss": 0.2646,
+      "step": 2546
+    },
+    {
+      "epoch": 0.590882728221784,
+      "grad_norm": 18.072618750754682,
+      "learning_rate": 2e-06,
+      "loss": 0.2477,
+      "step": 2547
+    },
+    {
+      "epoch": 0.5911147198700847,
+      "grad_norm": 13.436673286426196,
+      "learning_rate": 2e-06,
+      "loss": 0.299,
+      "step": 2548
+    },
+    {
+      "epoch": 0.5913467115183854,
+      "grad_norm": 6.801983823427113,
+      "learning_rate": 2e-06,
+      "loss": 0.3031,
+      "step": 2549
+    },
+    {
+      "epoch": 0.591578703166686,
+      "grad_norm": 14.892621067621787,
+      "learning_rate": 2e-06,
+      "loss": 0.3393,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5918106948149867,
+      "grad_norm": 15.883135619779361,
+      "learning_rate": 2e-06,
+      "loss": 0.3129,
+      "step": 2551
+    },
+    {
+      "epoch": 0.5920426864632873,
+      "grad_norm": 17.51304692590423,
+      "learning_rate": 2e-06,
+      "loss": 0.3318,
+      "step": 2552
+    },
+    {
+      "epoch": 0.592274678111588,
+      "grad_norm": 13.23396182628087,
+      "learning_rate": 2e-06,
+      "loss": 0.2345,
+      "step": 2553
+    },
+    {
+      "epoch": 0.5925066697598886,
+      "grad_norm": 16.088956436374215,
+      "learning_rate": 2e-06,
+      "loss": 0.3597,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5927386614081893,
+      "grad_norm": 19.618846840750198,
+      "learning_rate": 2e-06,
+      "loss": 0.3431,
+      "step": 2555
+    },
+    {
+      "epoch": 0.59297065305649,
+      "grad_norm": 16.66909534596031,
+      "learning_rate": 2e-06,
+      "loss": 0.2486,
+      "step": 2556
+    },
+    {
+      "epoch": 0.5932026447047907,
+      "grad_norm": 10.560779229199174,
+      "learning_rate": 2e-06,
+      "loss": 0.2887,
+      "step": 2557
+    },
+    {
+      "epoch": 0.5934346363530912,
+      "grad_norm": 9.21356550066053,
+      "learning_rate": 2e-06,
+      "loss": 0.1793,
+      "step": 2558
+    },
+    {
+      "epoch": 0.5936666280013919,
+      "grad_norm": 14.076285578466297,
+      "learning_rate": 2e-06,
+      "loss": 0.3869,
+      "step": 2559
+    },
+    {
+      "epoch": 0.5938986196496926,
+      "grad_norm": 12.847638105175127,
+      "learning_rate": 2e-06,
+      "loss": 0.2315,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5941306112979933,
+      "grad_norm": 14.765732513424698,
+      "learning_rate": 2e-06,
+      "loss": 0.2799,
+      "step": 2561
+    },
+    {
+      "epoch": 0.594362602946294,
+      "grad_norm": 9.942317670085435,
+      "learning_rate": 2e-06,
+      "loss": 0.2714,
+      "step": 2562
+    },
+    {
+      "epoch": 0.5945945945945946,
+      "grad_norm": 6.0178161019588945,
+      "learning_rate": 2e-06,
+      "loss": 0.2055,
+      "step": 2563
+    },
+    {
+      "epoch": 0.5948265862428953,
+      "grad_norm": 16.221366734512934,
+      "learning_rate": 2e-06,
+      "loss": 0.3824,
+      "step": 2564
+    },
+    {
+      "epoch": 0.5950585778911959,
+      "grad_norm": 13.889351920592597,
+      "learning_rate": 2e-06,
+      "loss": 0.303,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5952905695394966,
+      "grad_norm": 23.308845437497098,
+      "learning_rate": 2e-06,
+      "loss": 0.3687,
+      "step": 2566
+    },
+    {
+      "epoch": 0.5955225611877972,
+      "grad_norm": 16.15487734503701,
+      "learning_rate": 2e-06,
+      "loss": 0.2385,
+      "step": 2567
+    },
+    {
+      "epoch": 0.5957545528360979,
+      "grad_norm": 14.492592621249058,
+      "learning_rate": 2e-06,
+      "loss": 0.2203,
+      "step": 2568
+    },
+    {
+      "epoch": 0.5959865444843986,
+      "grad_norm": 14.475746535168714,
+      "learning_rate": 2e-06,
+      "loss": 0.3553,
+      "step": 2569
+    },
+    {
+      "epoch": 0.5962185361326993,
+      "grad_norm": 14.198491742040812,
+      "learning_rate": 2e-06,
+      "loss": 0.2305,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5964505277809999,
+      "grad_norm": 17.549194217988287,
+      "learning_rate": 2e-06,
+      "loss": 0.3198,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5966825194293005,
+      "grad_norm": 13.338985893175025,
+      "learning_rate": 2e-06,
+      "loss": 0.3296,
+      "step": 2572
+    },
+    {
+      "epoch": 0.5969145110776012,
+      "grad_norm": 17.512050505262632,
+      "learning_rate": 2e-06,
+      "loss": 0.2729,
+      "step": 2573
+    },
+    {
+      "epoch": 0.5971465027259019,
+      "grad_norm": 6.628241469541592,
+      "learning_rate": 2e-06,
+      "loss": 0.2512,
+      "step": 2574
+    },
+    {
+      "epoch": 0.5973784943742025,
+      "grad_norm": 21.899371256267624,
+      "learning_rate": 2e-06,
+      "loss": 0.418,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5976104860225032,
+      "grad_norm": 21.141754928031396,
+      "learning_rate": 2e-06,
+      "loss": 0.321,
+      "step": 2576
+    },
+    {
+      "epoch": 0.5978424776708039,
+      "grad_norm": 21.29611023083716,
+      "learning_rate": 2e-06,
+      "loss": 0.3016,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5980744693191045,
+      "grad_norm": 9.439127869148606,
+      "learning_rate": 2e-06,
+      "loss": 0.2466,
+      "step": 2578
+    },
+    {
+      "epoch": 0.5983064609674051,
+      "grad_norm": 12.79725874640436,
+      "learning_rate": 2e-06,
+      "loss": 0.2972,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5985384526157058,
+      "grad_norm": 13.59370150264999,
+      "learning_rate": 2e-06,
+      "loss": 0.2991,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5987704442640065,
+      "grad_norm": 14.290613413702845,
+      "learning_rate": 2e-06,
+      "loss": 0.2741,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5990024359123072,
+      "grad_norm": 14.739111004523165,
+      "learning_rate": 2e-06,
+      "loss": 0.299,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5992344275606079,
+      "grad_norm": 18.43042050151962,
+      "learning_rate": 2e-06,
+      "loss": 0.3581,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5994664192089085,
+      "grad_norm": 16.742572811047022,
+      "learning_rate": 2e-06,
+      "loss": 0.405,
+      "step": 2584
+    },
+    {
+      "epoch": 0.5996984108572091,
+      "grad_norm": 13.803051539755291,
+      "learning_rate": 2e-06,
+      "loss": 0.2549,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5999304025055098,
+      "grad_norm": 18.181730133497524,
+      "learning_rate": 2e-06,
+      "loss": 0.3734,
+      "step": 2586
+    },
+    {
+      "epoch": 0.6001623941538105,
+      "grad_norm": 11.635904384068603,
+      "learning_rate": 2e-06,
+      "loss": 0.3005,
+      "step": 2587
+    },
+    {
+      "epoch": 0.6003943858021111,
+      "grad_norm": 15.347115874466267,
+      "learning_rate": 2e-06,
+      "loss": 0.3835,
+      "step": 2588
+    },
+    {
+      "epoch": 0.6006263774504118,
+      "grad_norm": 8.825379842280917,
+      "learning_rate": 2e-06,
+      "loss": 0.2267,
+      "step": 2589
+    },
+    {
+      "epoch": 0.6008583690987125,
+      "grad_norm": 9.577537283110104,
+      "learning_rate": 2e-06,
+      "loss": 0.3363,
+      "step": 2590
+    },
+    {
+      "epoch": 0.601090360747013,
+      "grad_norm": 27.429490879194134,
+      "learning_rate": 2e-06,
+      "loss": 0.3661,
+      "step": 2591
+    },
+    {
+      "epoch": 0.6013223523953137,
+      "grad_norm": 22.96181063077117,
+      "learning_rate": 2e-06,
+      "loss": 0.3991,
+      "step": 2592
+    },
+    {
+      "epoch": 0.6015543440436144,
+      "grad_norm": 20.831944745474473,
+      "learning_rate": 2e-06,
+      "loss": 0.3794,
+      "step": 2593
+    },
+    {
+      "epoch": 0.6017863356919151,
+      "grad_norm": 24.6377518098682,
+      "learning_rate": 2e-06,
+      "loss": 0.3596,
+      "step": 2594
+    },
+    {
+      "epoch": 0.6020183273402158,
+      "grad_norm": 16.685655535948804,
+      "learning_rate": 2e-06,
+      "loss": 0.3078,
+      "step": 2595
+    },
+    {
+      "epoch": 0.6022503189885164,
+      "grad_norm": 16.949976752704817,
+      "learning_rate": 2e-06,
+      "loss": 0.3985,
+      "step": 2596
+    },
+    {
+      "epoch": 0.6024823106368171,
+      "grad_norm": 10.333889262423241,
+      "learning_rate": 2e-06,
+      "loss": 0.2667,
+      "step": 2597
+    },
+    {
+      "epoch": 0.6027143022851177,
+      "grad_norm": 14.510670341351082,
+      "learning_rate": 2e-06,
+      "loss": 0.3442,
+      "step": 2598
+    },
+    {
+      "epoch": 0.6029462939334184,
+      "grad_norm": 17.69640920886926,
+      "learning_rate": 2e-06,
+      "loss": 0.2401,
+      "step": 2599
+    },
+    {
+      "epoch": 0.603178285581719,
+      "grad_norm": 23.660783062967415,
+      "learning_rate": 2e-06,
+      "loss": 0.3295,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6034102772300197,
+      "grad_norm": 18.057275005324275,
+      "learning_rate": 2e-06,
+      "loss": 0.3407,
+      "step": 2601
+    },
+    {
+      "epoch": 0.6036422688783204,
+      "grad_norm": 24.069099929213795,
+      "learning_rate": 2e-06,
+      "loss": 0.3268,
+      "step": 2602
+    },
+    {
+      "epoch": 0.6038742605266211,
+      "grad_norm": 10.642561428651906,
+      "learning_rate": 2e-06,
+      "loss": 0.3963,
+      "step": 2603
+    },
+    {
+      "epoch": 0.6041062521749218,
+      "grad_norm": 20.174001034192678,
+      "learning_rate": 2e-06,
+      "loss": 0.2681,
+      "step": 2604
+    },
+    {
+      "epoch": 0.6043382438232223,
+      "grad_norm": 9.175946684897225,
+      "learning_rate": 2e-06,
+      "loss": 0.2181,
+      "step": 2605
+    },
+    {
+      "epoch": 0.604570235471523,
+      "grad_norm": 8.166771981839268,
+      "learning_rate": 2e-06,
+      "loss": 0.2716,
+      "step": 2606
+    },
+    {
+      "epoch": 0.6048022271198237,
+      "grad_norm": 16.899229840133128,
+      "learning_rate": 2e-06,
+      "loss": 0.2287,
+      "step": 2607
+    },
+    {
+      "epoch": 0.6050342187681244,
+      "grad_norm": 9.44602089006785,
+      "learning_rate": 2e-06,
+      "loss": 0.2403,
+      "step": 2608
+    },
+    {
+      "epoch": 0.605266210416425,
+      "grad_norm": 13.783306010662567,
+      "learning_rate": 2e-06,
+      "loss": 0.3988,
+      "step": 2609
+    },
+    {
+      "epoch": 0.6054982020647257,
+      "grad_norm": 10.320072678209218,
+      "learning_rate": 2e-06,
+      "loss": 0.2224,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6057301937130263,
+      "grad_norm": 15.997254592009337,
+      "learning_rate": 2e-06,
+      "loss": 0.3589,
+      "step": 2611
+    },
+    {
+      "epoch": 0.605962185361327,
+      "grad_norm": 8.183158067579095,
+      "learning_rate": 2e-06,
+      "loss": 0.168,
+      "step": 2612
+    },
+    {
+      "epoch": 0.6061941770096276,
+      "grad_norm": 10.745161014883454,
+      "learning_rate": 2e-06,
+      "loss": 0.2714,
+      "step": 2613
+    },
+    {
+      "epoch": 0.6064261686579283,
+      "grad_norm": 11.462258671047206,
+      "learning_rate": 2e-06,
+      "loss": 0.2523,
+      "step": 2614
+    },
+    {
+      "epoch": 0.606658160306229,
+      "grad_norm": 12.016709149335806,
+      "learning_rate": 2e-06,
+      "loss": 0.314,
+      "step": 2615
+    },
+    {
+      "epoch": 0.6068901519545297,
+      "grad_norm": 17.155313994573596,
+      "learning_rate": 2e-06,
+      "loss": 0.408,
+      "step": 2616
+    },
+    {
+      "epoch": 0.6071221436028303,
+      "grad_norm": 11.915214369441893,
+      "learning_rate": 2e-06,
+      "loss": 0.3494,
+      "step": 2617
+    },
+    {
+      "epoch": 0.6073541352511309,
+      "grad_norm": 9.603980367813907,
+      "learning_rate": 2e-06,
+      "loss": 0.3015,
+      "step": 2618
+    },
+    {
+      "epoch": 0.6075861268994316,
+      "grad_norm": 14.152671605729422,
+      "learning_rate": 2e-06,
+      "loss": 0.2807,
+      "step": 2619
+    },
+    {
+      "epoch": 0.6078181185477323,
+      "grad_norm": 9.895894107038899,
+      "learning_rate": 2e-06,
+      "loss": 0.2248,
+      "step": 2620
+    },
+    {
+      "epoch": 0.608050110196033,
+      "grad_norm": 21.33095069897468,
+      "learning_rate": 2e-06,
+      "loss": 0.3436,
+      "step": 2621
+    },
+    {
+      "epoch": 0.6082821018443336,
+      "grad_norm": 7.935644587182702,
+      "learning_rate": 2e-06,
+      "loss": 0.3164,
+      "step": 2622
+    },
+    {
+      "epoch": 0.6085140934926343,
+      "grad_norm": 19.81730932480036,
+      "learning_rate": 2e-06,
+      "loss": 0.3868,
+      "step": 2623
+    },
+    {
+      "epoch": 0.608746085140935,
+      "grad_norm": 15.406381840658463,
+      "learning_rate": 2e-06,
+      "loss": 0.2037,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6089780767892355,
+      "grad_norm": 15.622610047217227,
+      "learning_rate": 2e-06,
+      "loss": 0.353,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6092100684375362,
+      "grad_norm": 19.054233656211007,
+      "learning_rate": 2e-06,
+      "loss": 0.3748,
+      "step": 2626
+    },
+    {
+      "epoch": 0.6094420600858369,
+      "grad_norm": 17.01955389711876,
+      "learning_rate": 2e-06,
+      "loss": 0.3669,
+      "step": 2627
+    },
+    {
+      "epoch": 0.6096740517341376,
+      "grad_norm": 13.138403748635863,
+      "learning_rate": 2e-06,
+      "loss": 0.3121,
+      "step": 2628
+    },
+    {
+      "epoch": 0.6099060433824383,
+      "grad_norm": 7.430059477695104,
+      "learning_rate": 2e-06,
+      "loss": 0.2589,
+      "step": 2629
+    },
+    {
+      "epoch": 0.6101380350307389,
+      "grad_norm": 12.017745160492542,
+      "learning_rate": 2e-06,
+      "loss": 0.3337,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6103700266790395,
+      "grad_norm": 21.893420801946824,
+      "learning_rate": 2e-06,
+      "loss": 0.3723,
+      "step": 2631
+    },
+    {
+      "epoch": 0.6106020183273402,
+      "grad_norm": 7.716883732515496,
+      "learning_rate": 2e-06,
+      "loss": 0.246,
+      "step": 2632
+    },
+    {
+      "epoch": 0.6108340099756409,
+      "grad_norm": 12.394263618488978,
+      "learning_rate": 2e-06,
+      "loss": 0.2482,
+      "step": 2633
+    },
+    {
+      "epoch": 0.6110660016239415,
+      "grad_norm": 13.85844630687673,
+      "learning_rate": 2e-06,
+      "loss": 0.3278,
+      "step": 2634
+    },
+    {
+      "epoch": 0.6112979932722422,
+      "grad_norm": 15.23846624738061,
+      "learning_rate": 2e-06,
+      "loss": 0.3521,
+      "step": 2635
+    },
+    {
+      "epoch": 0.6115299849205429,
+      "grad_norm": 17.00492197768118,
+      "learning_rate": 2e-06,
+      "loss": 0.2893,
+      "step": 2636
+    },
+    {
+      "epoch": 0.6117619765688436,
+      "grad_norm": 13.586664525927867,
+      "learning_rate": 2e-06,
+      "loss": 0.2788,
+      "step": 2637
+    },
+    {
+      "epoch": 0.6119939682171441,
+      "grad_norm": 15.014978243660334,
+      "learning_rate": 2e-06,
+      "loss": 0.3178,
+      "step": 2638
+    },
+    {
+      "epoch": 0.6122259598654448,
+      "grad_norm": 13.84349412358929,
+      "learning_rate": 2e-06,
+      "loss": 0.2344,
+      "step": 2639
+    },
+    {
+      "epoch": 0.6124579515137455,
+      "grad_norm": 22.096753673828697,
+      "learning_rate": 2e-06,
+      "loss": 0.3522,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6126899431620462,
+      "grad_norm": 14.059995206076655,
+      "learning_rate": 2e-06,
+      "loss": 0.3093,
+      "step": 2641
+    },
+    {
+      "epoch": 0.6129219348103468,
+      "grad_norm": 8.321542567832232,
+      "learning_rate": 2e-06,
+      "loss": 0.3574,
+      "step": 2642
+    },
+    {
+      "epoch": 0.6131539264586475,
+      "grad_norm": 13.750629896271956,
+      "learning_rate": 2e-06,
+      "loss": 0.1864,
+      "step": 2643
+    },
+    {
+      "epoch": 0.6133859181069482,
+      "grad_norm": 22.7628000844624,
+      "learning_rate": 2e-06,
+      "loss": 0.4686,
+      "step": 2644
+    },
+    {
+      "epoch": 0.6136179097552488,
+      "grad_norm": 9.845866081722727,
+      "learning_rate": 2e-06,
+      "loss": 0.2652,
+      "step": 2645
+    },
+    {
+      "epoch": 0.6138499014035494,
+      "grad_norm": 13.935750749172811,
+      "learning_rate": 2e-06,
+      "loss": 0.3633,
+      "step": 2646
+    },
+    {
+      "epoch": 0.6140818930518501,
+      "grad_norm": 20.494770762962357,
+      "learning_rate": 2e-06,
+      "loss": 0.3229,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6143138847001508,
+      "grad_norm": 16.60093010574271,
+      "learning_rate": 2e-06,
+      "loss": 0.3424,
+      "step": 2648
+    },
+    {
+      "epoch": 0.6145458763484515,
+      "grad_norm": 20.04060203813773,
+      "learning_rate": 2e-06,
+      "loss": 0.4419,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6147778679967522,
+      "grad_norm": 11.737435793683828,
+      "learning_rate": 2e-06,
+      "loss": 0.2781,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6150098596450527,
+      "grad_norm": 13.725621163731567,
+      "learning_rate": 2e-06,
+      "loss": 0.2673,
+      "step": 2651
+    },
+    {
+      "epoch": 0.6152418512933534,
+      "grad_norm": 15.800607780056476,
+      "learning_rate": 2e-06,
+      "loss": 0.2469,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6154738429416541,
+      "grad_norm": 20.152900288357856,
+      "learning_rate": 2e-06,
+      "loss": 0.2142,
+      "step": 2653
+    },
+    {
+      "epoch": 0.6157058345899548,
+      "grad_norm": 16.55339980838737,
+      "learning_rate": 2e-06,
+      "loss": 0.3571,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6159378262382554,
+      "grad_norm": 10.524688730312457,
+      "learning_rate": 2e-06,
+      "loss": 0.3035,
+      "step": 2655
+    },
+    {
+      "epoch": 0.6161698178865561,
+      "grad_norm": 10.31213022805119,
+      "learning_rate": 2e-06,
+      "loss": 0.272,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6164018095348568,
+      "grad_norm": 9.45210475666712,
+      "learning_rate": 2e-06,
+      "loss": 0.2183,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6166338011831574,
+      "grad_norm": 21.582429259346284,
+      "learning_rate": 2e-06,
+      "loss": 0.2551,
+      "step": 2658
+    },
+    {
+      "epoch": 0.616865792831458,
+      "grad_norm": 26.47571112560527,
+      "learning_rate": 2e-06,
+      "loss": 0.3598,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6170977844797587,
+      "grad_norm": 14.710721928558407,
+      "learning_rate": 2e-06,
+      "loss": 0.3941,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6173297761280594,
+      "grad_norm": 10.122986352634644,
+      "learning_rate": 2e-06,
+      "loss": 0.2817,
+      "step": 2661
+    },
+    {
+      "epoch": 0.6175617677763601,
+      "grad_norm": 17.1002915041208,
+      "learning_rate": 2e-06,
+      "loss": 0.2976,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6177937594246607,
+      "grad_norm": 15.228871856863515,
+      "learning_rate": 2e-06,
+      "loss": 0.3329,
+      "step": 2663
+    },
+    {
+      "epoch": 0.6180257510729614,
+      "grad_norm": 10.804501166132074,
+      "learning_rate": 2e-06,
+      "loss": 0.2286,
+      "step": 2664
+    },
+    {
+      "epoch": 0.618257742721262,
+      "grad_norm": 19.23509206707099,
+      "learning_rate": 2e-06,
+      "loss": 0.2952,
+      "step": 2665
+    },
+    {
+      "epoch": 0.6184897343695627,
+      "grad_norm": 19.918949097059865,
+      "learning_rate": 2e-06,
+      "loss": 0.3427,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6187217260178633,
+      "grad_norm": 11.805634939444555,
+      "learning_rate": 2e-06,
+      "loss": 0.3066,
+      "step": 2667
+    },
+    {
+      "epoch": 0.618953717666164,
+      "grad_norm": 15.470640332499393,
+      "learning_rate": 2e-06,
+      "loss": 0.1898,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6191857093144647,
+      "grad_norm": 9.398799386754078,
+      "learning_rate": 2e-06,
+      "loss": 0.2322,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6194177009627654,
+      "grad_norm": 13.784983840873174,
+      "learning_rate": 2e-06,
+      "loss": 0.3268,
+      "step": 2670
+    },
+    {
+      "epoch": 0.619649692611066,
+      "grad_norm": 12.30965234954036,
+      "learning_rate": 2e-06,
+      "loss": 0.2877,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6198816842593666,
+      "grad_norm": 11.794091327712746,
+      "learning_rate": 2e-06,
+      "loss": 0.2644,
+      "step": 2672
+    },
+    {
+      "epoch": 0.6201136759076673,
+      "grad_norm": 21.585226462652304,
+      "learning_rate": 2e-06,
+      "loss": 0.4381,
+      "step": 2673
+    },
+    {
+      "epoch": 0.620345667555968,
+      "grad_norm": 10.362497530620518,
+      "learning_rate": 2e-06,
+      "loss": 0.2232,
+      "step": 2674
+    },
+    {
+      "epoch": 0.6205776592042687,
+      "grad_norm": 12.525941047158943,
+      "learning_rate": 2e-06,
+      "loss": 0.2882,
+      "step": 2675
+    },
+    {
+      "epoch": 0.6208096508525693,
+      "grad_norm": 16.138277764285274,
+      "learning_rate": 2e-06,
+      "loss": 0.3669,
+      "step": 2676
+    },
+    {
+      "epoch": 0.62104164250087,
+      "grad_norm": 15.780768061964427,
+      "learning_rate": 2e-06,
+      "loss": 0.3749,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6212736341491706,
+      "grad_norm": 20.30659677577186,
+      "learning_rate": 2e-06,
+      "loss": 0.3917,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6215056257974713,
+      "grad_norm": 12.25187953805943,
+      "learning_rate": 2e-06,
+      "loss": 0.2449,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6217376174457719,
+      "grad_norm": 12.850671799957505,
+      "learning_rate": 2e-06,
+      "loss": 0.2107,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6219696090940726,
+      "grad_norm": 21.096912590816856,
+      "learning_rate": 2e-06,
+      "loss": 0.2952,
+      "step": 2681
+    },
+    {
+      "epoch": 0.6222016007423733,
+      "grad_norm": 25.320626047184888,
+      "learning_rate": 2e-06,
+      "loss": 0.389,
+      "step": 2682
+    },
+    {
+      "epoch": 0.622433592390674,
+      "grad_norm": 22.577550122053413,
+      "learning_rate": 2e-06,
+      "loss": 0.3165,
+      "step": 2683
+    },
+    {
+      "epoch": 0.6226655840389747,
+      "grad_norm": 12.988787747440956,
+      "learning_rate": 2e-06,
+      "loss": 0.2689,
+      "step": 2684
+    },
+    {
+      "epoch": 0.6228975756872752,
+      "grad_norm": 9.692038170370148,
+      "learning_rate": 2e-06,
+      "loss": 0.2945,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6231295673355759,
+      "grad_norm": 15.284503787530403,
+      "learning_rate": 2e-06,
+      "loss": 0.3389,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6233615589838766,
+      "grad_norm": 23.334498764386655,
+      "learning_rate": 2e-06,
+      "loss": 0.3517,
+      "step": 2687
+    },
+    {
+      "epoch": 0.6235935506321773,
+      "grad_norm": 10.922055258467221,
+      "learning_rate": 2e-06,
+      "loss": 0.2828,
+      "step": 2688
+    },
+    {
+      "epoch": 0.6238255422804779,
+      "grad_norm": 10.065528194947968,
+      "learning_rate": 2e-06,
+      "loss": 0.2002,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6240575339287786,
+      "grad_norm": 18.114588707791558,
+      "learning_rate": 2e-06,
+      "loss": 0.199,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6242895255770792,
+      "grad_norm": 25.19740551477204,
+      "learning_rate": 2e-06,
+      "loss": 0.3786,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6245215172253799,
+      "grad_norm": 12.660265258094135,
+      "learning_rate": 2e-06,
+      "loss": 0.2891,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6247535088736805,
+      "grad_norm": 12.01025971009369,
+      "learning_rate": 2e-06,
+      "loss": 0.218,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6249855005219812,
+      "grad_norm": 19.57290722929422,
+      "learning_rate": 2e-06,
+      "loss": 0.4168,
+      "step": 2694
+    },
+    {
+      "epoch": 0.6252174921702819,
+      "grad_norm": 18.333835413311697,
+      "learning_rate": 2e-06,
+      "loss": 0.3802,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6254494838185826,
+      "grad_norm": 9.452348955554905,
+      "learning_rate": 2e-06,
+      "loss": 0.3192,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6256814754668832,
+      "grad_norm": 19.413813530652142,
+      "learning_rate": 2e-06,
+      "loss": 0.3486,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6259134671151838,
+      "grad_norm": 8.14297216131884,
+      "learning_rate": 2e-06,
+      "loss": 0.1745,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6261454587634845,
+      "grad_norm": 16.460858927880107,
+      "learning_rate": 2e-06,
+      "loss": 0.2898,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6263774504117852,
+      "grad_norm": 23.27528134870079,
+      "learning_rate": 2e-06,
+      "loss": 0.3116,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6266094420600858,
+      "grad_norm": 21.682972803443782,
+      "learning_rate": 2e-06,
+      "loss": 0.3774,
+      "step": 2701
+    },
+    {
+      "epoch": 0.6268414337083865,
+      "grad_norm": 11.274293799473982,
+      "learning_rate": 2e-06,
+      "loss": 0.2252,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6270734253566872,
+      "grad_norm": 14.373437442014854,
+      "learning_rate": 2e-06,
+      "loss": 0.2802,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6273054170049879,
+      "grad_norm": 27.98556682726606,
+      "learning_rate": 2e-06,
+      "loss": 0.3984,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6275374086532884,
+      "grad_norm": 8.125770832169447,
+      "learning_rate": 2e-06,
+      "loss": 0.2483,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6277694003015891,
+      "grad_norm": 14.280645487072892,
+      "learning_rate": 2e-06,
+      "loss": 0.3523,
+      "step": 2706
+    },
+    {
+      "epoch": 0.6280013919498898,
+      "grad_norm": 13.604607657906875,
+      "learning_rate": 2e-06,
+      "loss": 0.2502,
+      "step": 2707
+    },
+    {
+      "epoch": 0.6282333835981905,
+      "grad_norm": 15.215810695283421,
+      "learning_rate": 2e-06,
+      "loss": 0.2886,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6284653752464912,
+      "grad_norm": 10.526388538487529,
+      "learning_rate": 2e-06,
+      "loss": 0.2921,
+      "step": 2709
+    },
+    {
+      "epoch": 0.6286973668947918,
+      "grad_norm": 11.578951665205539,
+      "learning_rate": 2e-06,
+      "loss": 0.2166,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6289293585430924,
+      "grad_norm": 26.76025334672509,
+      "learning_rate": 2e-06,
+      "loss": 0.4258,
+      "step": 2711
+    },
+    {
+      "epoch": 0.6291613501913931,
+      "grad_norm": 5.960315113925062,
+      "learning_rate": 2e-06,
+      "loss": 0.203,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6293933418396938,
+      "grad_norm": 19.098755755319814,
+      "learning_rate": 2e-06,
+      "loss": 0.2265,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6296253334879944,
+      "grad_norm": 11.279047664736554,
+      "learning_rate": 2e-06,
+      "loss": 0.2733,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6298573251362951,
+      "grad_norm": 11.042710355062926,
+      "learning_rate": 2e-06,
+      "loss": 0.2478,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6300893167845958,
+      "grad_norm": 21.186114628597807,
+      "learning_rate": 2e-06,
+      "loss": 0.3966,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6303213084328965,
+      "grad_norm": 12.535985711797327,
+      "learning_rate": 2e-06,
+      "loss": 0.3958,
+      "step": 2717
+    },
+    {
+      "epoch": 0.630553300081197,
+      "grad_norm": 9.076805066853897,
+      "learning_rate": 2e-06,
+      "loss": 0.2573,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6307852917294977,
+      "grad_norm": 19.664137316963803,
+      "learning_rate": 2e-06,
+      "loss": 0.2859,
+      "step": 2719
+    },
+    {
+      "epoch": 0.6310172833777984,
+      "grad_norm": 20.192281362626595,
+      "learning_rate": 2e-06,
+      "loss": 0.3911,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6312492750260991,
+      "grad_norm": 9.34365143008957,
+      "learning_rate": 2e-06,
+      "loss": 0.2553,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6314812666743997,
+      "grad_norm": 24.48111614902375,
+      "learning_rate": 2e-06,
+      "loss": 0.4094,
+      "step": 2722
+    },
+    {
+      "epoch": 0.6317132583227004,
+      "grad_norm": 9.707702530939931,
+      "learning_rate": 2e-06,
+      "loss": 0.2276,
+      "step": 2723
+    },
+    {
+      "epoch": 0.631945249971001,
+      "grad_norm": 10.989348115827445,
+      "learning_rate": 2e-06,
+      "loss": 0.3009,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6321772416193017,
+      "grad_norm": 11.048989765456342,
+      "learning_rate": 2e-06,
+      "loss": 0.2741,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6324092332676023,
+      "grad_norm": 14.44663365989049,
+      "learning_rate": 2e-06,
+      "loss": 0.3041,
+      "step": 2726
+    },
+    {
+      "epoch": 0.632641224915903,
+      "grad_norm": 20.803849251140264,
+      "learning_rate": 2e-06,
+      "loss": 0.335,
+      "step": 2727
+    },
+    {
+      "epoch": 0.6328732165642037,
+      "grad_norm": 37.679247145652994,
+      "learning_rate": 2e-06,
+      "loss": 0.415,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6331052082125044,
+      "grad_norm": 20.184213761179738,
+      "learning_rate": 2e-06,
+      "loss": 0.3163,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6333371998608051,
+      "grad_norm": 11.551312967773589,
+      "learning_rate": 2e-06,
+      "loss": 0.3722,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6335691915091056,
+      "grad_norm": 12.706340951891182,
+      "learning_rate": 2e-06,
+      "loss": 0.262,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6338011831574063,
+      "grad_norm": 7.8792482539519835,
+      "learning_rate": 2e-06,
+      "loss": 0.2088,
+      "step": 2732
+    },
+    {
+      "epoch": 0.634033174805707,
+      "grad_norm": 20.92190277740192,
+      "learning_rate": 2e-06,
+      "loss": 0.4063,
+      "step": 2733
+    },
+    {
+      "epoch": 0.6342651664540077,
+      "grad_norm": 23.1379632657036,
+      "learning_rate": 2e-06,
+      "loss": 0.3856,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6344971581023083,
+      "grad_norm": 10.469166151059296,
+      "learning_rate": 2e-06,
+      "loss": 0.3344,
+      "step": 2735
+    },
+    {
+      "epoch": 0.634729149750609,
+      "grad_norm": 7.87302230405102,
+      "learning_rate": 2e-06,
+      "loss": 0.2252,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6349611413989097,
+      "grad_norm": 10.936525755895664,
+      "learning_rate": 2e-06,
+      "loss": 0.2936,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6351931330472103,
+      "grad_norm": 10.375866159081577,
+      "learning_rate": 2e-06,
+      "loss": 0.2014,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6354251246955109,
+      "grad_norm": 19.10009840905328,
+      "learning_rate": 2e-06,
+      "loss": 0.2933,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6356571163438116,
+      "grad_norm": 10.686041317032329,
+      "learning_rate": 2e-06,
+      "loss": 0.1862,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6358891079921123,
+      "grad_norm": 8.286894335356871,
+      "learning_rate": 2e-06,
+      "loss": 0.2741,
+      "step": 2741
+    },
+    {
+      "epoch": 0.636121099640413,
+      "grad_norm": 14.205426749807465,
+      "learning_rate": 2e-06,
+      "loss": 0.2385,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6363530912887136,
+      "grad_norm": 17.240692772281445,
+      "learning_rate": 2e-06,
+      "loss": 0.2993,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6365850829370142,
+      "grad_norm": 11.042937107655254,
+      "learning_rate": 2e-06,
+      "loss": 0.2914,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6368170745853149,
+      "grad_norm": 11.676379857964525,
+      "learning_rate": 2e-06,
+      "loss": 0.2883,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6370490662336156,
+      "grad_norm": 14.066457991386905,
+      "learning_rate": 2e-06,
+      "loss": 0.3843,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6372810578819162,
+      "grad_norm": 9.493809790836558,
+      "learning_rate": 2e-06,
+      "loss": 0.2383,
+      "step": 2747
+    },
+    {
+      "epoch": 0.6375130495302169,
+      "grad_norm": 18.954488884963048,
+      "learning_rate": 2e-06,
+      "loss": 0.3952,
+      "step": 2748
+    },
+    {
+      "epoch": 0.6377450411785176,
+      "grad_norm": 10.592996727713146,
+      "learning_rate": 2e-06,
+      "loss": 0.2398,
+      "step": 2749
+    },
+    {
+      "epoch": 0.6379770328268183,
+      "grad_norm": 12.26499262139558,
+      "learning_rate": 2e-06,
+      "loss": 0.2301,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6382090244751188,
+      "grad_norm": 16.25474051381953,
+      "learning_rate": 2e-06,
+      "loss": 0.27,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6384410161234195,
+      "grad_norm": 16.255107607009478,
+      "learning_rate": 2e-06,
+      "loss": 0.218,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6386730077717202,
+      "grad_norm": 14.197750750330924,
+      "learning_rate": 2e-06,
+      "loss": 0.3227,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6389049994200209,
+      "grad_norm": 10.311882448604072,
+      "learning_rate": 2e-06,
+      "loss": 0.1804,
+      "step": 2754
+    },
+    {
+      "epoch": 0.6391369910683216,
+      "grad_norm": 19.917667184339937,
+      "learning_rate": 2e-06,
+      "loss": 0.2387,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6393689827166222,
+      "grad_norm": 23.591561126533815,
+      "learning_rate": 2e-06,
+      "loss": 0.2956,
+      "step": 2756
+    },
+    {
+      "epoch": 0.6396009743649229,
+      "grad_norm": 6.848883359950225,
+      "learning_rate": 2e-06,
+      "loss": 0.1989,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6398329660132235,
+      "grad_norm": 30.575828724428405,
+      "learning_rate": 2e-06,
+      "loss": 0.447,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6400649576615242,
+      "grad_norm": 22.80247466822758,
+      "learning_rate": 2e-06,
+      "loss": 0.3561,
+      "step": 2759
+    },
+    {
+      "epoch": 0.6402969493098248,
+      "grad_norm": 16.36344492473047,
+      "learning_rate": 2e-06,
+      "loss": 0.231,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6405289409581255,
+      "grad_norm": 12.587730209883201,
+      "learning_rate": 2e-06,
+      "loss": 0.3164,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6407609326064262,
+      "grad_norm": 10.992985260887487,
+      "learning_rate": 2e-06,
+      "loss": 0.3116,
+      "step": 2762
+    },
+    {
+      "epoch": 0.6409929242547269,
+      "grad_norm": 21.5416745374942,
+      "learning_rate": 2e-06,
+      "loss": 0.363,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6412249159030274,
+      "grad_norm": 10.748047530819289,
+      "learning_rate": 2e-06,
+      "loss": 0.2674,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6414569075513281,
+      "grad_norm": 27.600547677388484,
+      "learning_rate": 2e-06,
+      "loss": 0.4441,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6416888991996288,
+      "grad_norm": 6.163204912631635,
+      "learning_rate": 2e-06,
+      "loss": 0.1463,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6419208908479295,
+      "grad_norm": 21.152287218285384,
+      "learning_rate": 2e-06,
+      "loss": 0.3676,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6421528824962301,
+      "grad_norm": 14.472285140393597,
+      "learning_rate": 2e-06,
+      "loss": 0.2761,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6423848741445308,
+      "grad_norm": 18.049548403993875,
+      "learning_rate": 2e-06,
+      "loss": 0.2679,
+      "step": 2769
+    },
+    {
+      "epoch": 0.6426168657928315,
+      "grad_norm": 18.11985935626689,
+      "learning_rate": 2e-06,
+      "loss": 0.3918,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6428488574411321,
+      "grad_norm": 26.11802174798973,
+      "learning_rate": 2e-06,
+      "loss": 0.3145,
+      "step": 2771
+    },
+    {
+      "epoch": 0.6430808490894327,
+      "grad_norm": 26.13363087877274,
+      "learning_rate": 2e-06,
+      "loss": 0.407,
+      "step": 2772
+    },
+    {
+      "epoch": 0.6433128407377334,
+      "grad_norm": 12.383254109783655,
+      "learning_rate": 2e-06,
+      "loss": 0.2272,
+      "step": 2773
+    },
+    {
+      "epoch": 0.6435448323860341,
+      "grad_norm": 14.884215874916432,
+      "learning_rate": 2e-06,
+      "loss": 0.2583,
+      "step": 2774
+    },
+    {
+      "epoch": 0.6437768240343348,
+      "grad_norm": 18.511544627989096,
+      "learning_rate": 2e-06,
+      "loss": 0.2846,
+      "step": 2775
+    },
+    {
+      "epoch": 0.6440088156826355,
+      "grad_norm": 8.920952494029857,
+      "learning_rate": 2e-06,
+      "loss": 0.2197,
+      "step": 2776
+    },
+    {
+      "epoch": 0.6442408073309361,
+      "grad_norm": 16.965335181403468,
+      "learning_rate": 2e-06,
+      "loss": 0.213,
+      "step": 2777
+    },
+    {
+      "epoch": 0.6444727989792367,
+      "grad_norm": 17.37240735011543,
+      "learning_rate": 2e-06,
+      "loss": 0.2977,
+      "step": 2778
+    },
+    {
+      "epoch": 0.6447047906275374,
+      "grad_norm": 13.824478272806385,
+      "learning_rate": 2e-06,
+      "loss": 0.2514,
+      "step": 2779
+    },
+    {
+      "epoch": 0.6449367822758381,
+      "grad_norm": 12.1812911628817,
+      "learning_rate": 2e-06,
+      "loss": 0.2624,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6451687739241387,
+      "grad_norm": 11.874202270954608,
+      "learning_rate": 2e-06,
+      "loss": 0.3416,
+      "step": 2781
+    },
+    {
+      "epoch": 0.6454007655724394,
+      "grad_norm": 15.154437415027664,
+      "learning_rate": 2e-06,
+      "loss": 0.192,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6456327572207401,
+      "grad_norm": 20.87825469178961,
+      "learning_rate": 2e-06,
+      "loss": 0.3041,
+      "step": 2783
+    },
+    {
+      "epoch": 0.6458647488690407,
+      "grad_norm": 19.52264631822083,
+      "learning_rate": 2e-06,
+      "loss": 0.3867,
+      "step": 2784
+    },
+    {
+      "epoch": 0.6460967405173413,
+      "grad_norm": 9.487212028916371,
+      "learning_rate": 2e-06,
+      "loss": 0.2096,
+      "step": 2785
+    },
+    {
+      "epoch": 0.646328732165642,
+      "grad_norm": 9.733036060848567,
+      "learning_rate": 2e-06,
+      "loss": 0.2419,
+      "step": 2786
+    },
+    {
+      "epoch": 0.6465607238139427,
+      "grad_norm": 14.859036415557348,
+      "learning_rate": 2e-06,
+      "loss": 0.2464,
+      "step": 2787
+    },
+    {
+      "epoch": 0.6467927154622434,
+      "grad_norm": 8.02322631614885,
+      "learning_rate": 2e-06,
+      "loss": 0.1726,
+      "step": 2788
+    },
+    {
+      "epoch": 0.647024707110544,
+      "grad_norm": 16.839870153267437,
+      "learning_rate": 2e-06,
+      "loss": 0.2765,
+      "step": 2789
+    },
+    {
+      "epoch": 0.6472566987588447,
+      "grad_norm": 12.64538932340377,
+      "learning_rate": 2e-06,
+      "loss": 0.233,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6474886904071453,
+      "grad_norm": 8.404400295247754,
+      "learning_rate": 2e-06,
+      "loss": 0.2357,
+      "step": 2791
+    },
+    {
+      "epoch": 0.647720682055446,
+      "grad_norm": 26.048227186813318,
+      "learning_rate": 2e-06,
+      "loss": 0.3968,
+      "step": 2792
+    },
+    {
+      "epoch": 0.6479526737037467,
+      "grad_norm": 21.542478177411446,
+      "learning_rate": 2e-06,
+      "loss": 0.3382,
+      "step": 2793
+    },
+    {
+      "epoch": 0.6481846653520473,
+      "grad_norm": 11.151029842131523,
+      "learning_rate": 2e-06,
+      "loss": 0.3172,
+      "step": 2794
+    },
+    {
+      "epoch": 0.648416657000348,
+      "grad_norm": 11.659646165299003,
+      "learning_rate": 2e-06,
+      "loss": 0.2767,
+      "step": 2795
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 10.997059886039864,
+      "learning_rate": 2e-06,
+      "loss": 0.2589,
+      "step": 2796
+    },
+    {
+      "epoch": 0.6488806402969494,
+      "grad_norm": 18.666221337499135,
+      "learning_rate": 2e-06,
+      "loss": 0.387,
+      "step": 2797
+    },
+    {
+      "epoch": 0.6491126319452499,
+      "grad_norm": 7.395206819823027,
+      "learning_rate": 2e-06,
+      "loss": 0.2303,
+      "step": 2798
+    },
+    {
+      "epoch": 0.6493446235935506,
+      "grad_norm": 11.972980545484056,
+      "learning_rate": 2e-06,
+      "loss": 0.2554,
+      "step": 2799
+    },
+    {
+      "epoch": 0.6495766152418513,
+      "grad_norm": 18.15435582309048,
+      "learning_rate": 2e-06,
+      "loss": 0.2819,
+      "step": 2800
+    },
+    {
+      "epoch": 0.649808606890152,
+      "grad_norm": 14.607467175131468,
+      "learning_rate": 2e-06,
+      "loss": 0.3821,
+      "step": 2801
+    },
+    {
+      "epoch": 0.6500405985384526,
+      "grad_norm": 15.679649143266406,
+      "learning_rate": 2e-06,
+      "loss": 0.2644,
+      "step": 2802
+    },
+    {
+      "epoch": 0.6502725901867533,
+      "grad_norm": 14.006099163322526,
+      "learning_rate": 2e-06,
+      "loss": 0.4045,
+      "step": 2803
+    },
+    {
+      "epoch": 0.6505045818350539,
+      "grad_norm": 18.225064796926716,
+      "learning_rate": 2e-06,
+      "loss": 0.286,
+      "step": 2804
+    },
+    {
+      "epoch": 0.6507365734833546,
+      "grad_norm": 11.06047957203517,
+      "learning_rate": 2e-06,
+      "loss": 0.3596,
+      "step": 2805
+    },
+    {
+      "epoch": 0.6509685651316552,
+      "grad_norm": 16.095596929410195,
+      "learning_rate": 2e-06,
+      "loss": 0.255,
+      "step": 2806
+    },
+    {
+      "epoch": 0.6512005567799559,
+      "grad_norm": 27.95225192803792,
+      "learning_rate": 2e-06,
+      "loss": 0.2987,
+      "step": 2807
+    },
+    {
+      "epoch": 0.6514325484282566,
+      "grad_norm": 19.666700272868894,
+      "learning_rate": 2e-06,
+      "loss": 0.3157,
+      "step": 2808
+    },
+    {
+      "epoch": 0.6516645400765573,
+      "grad_norm": 18.21122294525583,
+      "learning_rate": 2e-06,
+      "loss": 0.2825,
+      "step": 2809
+    },
+    {
+      "epoch": 0.651896531724858,
+      "grad_norm": 8.459447025413976,
+      "learning_rate": 2e-06,
+      "loss": 0.208,
+      "step": 2810
+    },
+    {
+      "epoch": 0.6521285233731585,
+      "grad_norm": 15.6268310899083,
+      "learning_rate": 2e-06,
+      "loss": 0.3198,
+      "step": 2811
+    },
+    {
+      "epoch": 0.6523605150214592,
+      "grad_norm": 28.45664865289904,
+      "learning_rate": 2e-06,
+      "loss": 0.4413,
+      "step": 2812
+    },
+    {
+      "epoch": 0.6525925066697599,
+      "grad_norm": 9.57936608488626,
+      "learning_rate": 2e-06,
+      "loss": 0.1847,
+      "step": 2813
+    },
+    {
+      "epoch": 0.6528244983180606,
+      "grad_norm": 27.51762187086346,
+      "learning_rate": 2e-06,
+      "loss": 0.3609,
+      "step": 2814
+    },
+    {
+      "epoch": 0.6530564899663612,
+      "grad_norm": 10.913821665408513,
+      "learning_rate": 2e-06,
+      "loss": 0.3048,
+      "step": 2815
+    },
+    {
+      "epoch": 0.6532884816146619,
+      "grad_norm": 12.152772036034538,
+      "learning_rate": 2e-06,
+      "loss": 0.2951,
+      "step": 2816
+    },
+    {
+      "epoch": 0.6535204732629626,
+      "grad_norm": 16.723214077755184,
+      "learning_rate": 2e-06,
+      "loss": 0.3622,
+      "step": 2817
+    },
+    {
+      "epoch": 0.6537524649112632,
+      "grad_norm": 16.129821665506423,
+      "learning_rate": 2e-06,
+      "loss": 0.3142,
+      "step": 2818
+    },
+    {
+      "epoch": 0.6539844565595638,
+      "grad_norm": 20.693612400230375,
+      "learning_rate": 2e-06,
+      "loss": 0.3765,
+      "step": 2819
+    },
+    {
+      "epoch": 0.6542164482078645,
+      "grad_norm": 12.490854876318215,
+      "learning_rate": 2e-06,
+      "loss": 0.3044,
+      "step": 2820
+    },
+    {
+      "epoch": 0.6544484398561652,
+      "grad_norm": 14.731393132025028,
+      "learning_rate": 2e-06,
+      "loss": 0.3037,
+      "step": 2821
+    },
+    {
+      "epoch": 0.6546804315044659,
+      "grad_norm": 18.17261097086871,
+      "learning_rate": 2e-06,
+      "loss": 0.3278,
+      "step": 2822
+    },
+    {
+      "epoch": 0.6549124231527665,
+      "grad_norm": 11.349964778383697,
+      "learning_rate": 2e-06,
+      "loss": 0.2254,
+      "step": 2823
+    },
+    {
+      "epoch": 0.6551444148010671,
+      "grad_norm": 11.71263807129237,
+      "learning_rate": 2e-06,
+      "loss": 0.2592,
+      "step": 2824
+    },
+    {
+      "epoch": 0.6553764064493678,
+      "grad_norm": 18.08318611215625,
+      "learning_rate": 2e-06,
+      "loss": 0.2325,
+      "step": 2825
+    },
+    {
+      "epoch": 0.6556083980976685,
+      "grad_norm": 8.138291970407442,
+      "learning_rate": 2e-06,
+      "loss": 0.2515,
+      "step": 2826
+    },
+    {
+      "epoch": 0.6558403897459691,
+      "grad_norm": 19.53165707849859,
+      "learning_rate": 2e-06,
+      "loss": 0.4315,
+      "step": 2827
+    },
+    {
+      "epoch": 0.6560723813942698,
+      "grad_norm": 12.143351381067472,
+      "learning_rate": 2e-06,
+      "loss": 0.2126,
+      "step": 2828
+    },
+    {
+      "epoch": 0.6563043730425705,
+      "grad_norm": 7.720914695591568,
+      "learning_rate": 2e-06,
+      "loss": 0.2538,
+      "step": 2829
+    },
+    {
+      "epoch": 0.6565363646908712,
+      "grad_norm": 21.317140640865798,
+      "learning_rate": 2e-06,
+      "loss": 0.2426,
+      "step": 2830
+    },
+    {
+      "epoch": 0.6567683563391717,
+      "grad_norm": 10.113067962324731,
+      "learning_rate": 2e-06,
+      "loss": 0.2468,
+      "step": 2831
+    },
+    {
+      "epoch": 0.6570003479874724,
+      "grad_norm": 14.102113177194823,
+      "learning_rate": 2e-06,
+      "loss": 0.3143,
+      "step": 2832
+    },
+    {
+      "epoch": 0.6572323396357731,
+      "grad_norm": 13.987038194596758,
+      "learning_rate": 2e-06,
+      "loss": 0.2907,
+      "step": 2833
+    },
+    {
+      "epoch": 0.6574643312840738,
+      "grad_norm": 6.112736996816027,
+      "learning_rate": 2e-06,
+      "loss": 0.2774,
+      "step": 2834
+    },
+    {
+      "epoch": 0.6576963229323745,
+      "grad_norm": 10.449699832658547,
+      "learning_rate": 2e-06,
+      "loss": 0.2477,
+      "step": 2835
+    },
+    {
+      "epoch": 0.6579283145806751,
+      "grad_norm": 11.472365495587425,
+      "learning_rate": 2e-06,
+      "loss": 0.1979,
+      "step": 2836
+    },
+    {
+      "epoch": 0.6581603062289758,
+      "grad_norm": 19.22641768300949,
+      "learning_rate": 2e-06,
+      "loss": 0.4411,
+      "step": 2837
+    },
+    {
+      "epoch": 0.6583922978772764,
+      "grad_norm": 14.055238753422904,
+      "learning_rate": 2e-06,
+      "loss": 0.2558,
+      "step": 2838
+    },
+    {
+      "epoch": 0.6586242895255771,
+      "grad_norm": 10.984608588095922,
+      "learning_rate": 2e-06,
+      "loss": 0.1894,
+      "step": 2839
+    },
+    {
+      "epoch": 0.6588562811738777,
+      "grad_norm": 10.434615531979285,
+      "learning_rate": 2e-06,
+      "loss": 0.3205,
+      "step": 2840
+    },
+    {
+      "epoch": 0.6590882728221784,
+      "grad_norm": 14.644889870932198,
+      "learning_rate": 2e-06,
+      "loss": 0.2957,
+      "step": 2841
+    },
+    {
+      "epoch": 0.6593202644704791,
+      "grad_norm": 14.774480239943642,
+      "learning_rate": 2e-06,
+      "loss": 0.302,
+      "step": 2842
+    },
+    {
+      "epoch": 0.6595522561187798,
+      "grad_norm": 14.71393716092559,
+      "learning_rate": 2e-06,
+      "loss": 0.3118,
+      "step": 2843
+    },
+    {
+      "epoch": 0.6597842477670803,
+      "grad_norm": 11.242429895186634,
+      "learning_rate": 2e-06,
+      "loss": 0.2235,
+      "step": 2844
+    },
+    {
+      "epoch": 0.660016239415381,
+      "grad_norm": 17.995923113107896,
+      "learning_rate": 2e-06,
+      "loss": 0.2863,
+      "step": 2845
+    },
+    {
+      "epoch": 0.6602482310636817,
+      "grad_norm": 17.729162434604135,
+      "learning_rate": 2e-06,
+      "loss": 0.3424,
+      "step": 2846
+    },
+    {
+      "epoch": 0.6604802227119824,
+      "grad_norm": 20.243422288385347,
+      "learning_rate": 2e-06,
+      "loss": 0.3789,
+      "step": 2847
+    },
+    {
+      "epoch": 0.660712214360283,
+      "grad_norm": 18.64732313406746,
+      "learning_rate": 2e-06,
+      "loss": 0.387,
+      "step": 2848
+    },
+    {
+      "epoch": 0.6609442060085837,
+      "grad_norm": 20.458465132834842,
+      "learning_rate": 2e-06,
+      "loss": 0.3071,
+      "step": 2849
+    },
+    {
+      "epoch": 0.6611761976568844,
+      "grad_norm": 11.585734721403155,
+      "learning_rate": 2e-06,
+      "loss": 0.2617,
+      "step": 2850
+    },
+    {
+      "epoch": 0.661408189305185,
+      "grad_norm": 13.461835104531968,
+      "learning_rate": 2e-06,
+      "loss": 0.335,
+      "step": 2851
+    },
+    {
+      "epoch": 0.6616401809534856,
+      "grad_norm": 15.673380657407227,
+      "learning_rate": 2e-06,
+      "loss": 0.2673,
+      "step": 2852
+    },
+    {
+      "epoch": 0.6618721726017863,
+      "grad_norm": 9.921530213879748,
+      "learning_rate": 2e-06,
+      "loss": 0.2912,
+      "step": 2853
+    },
+    {
+      "epoch": 0.662104164250087,
+      "grad_norm": 14.238714967293898,
+      "learning_rate": 2e-06,
+      "loss": 0.3132,
+      "step": 2854
+    },
+    {
+      "epoch": 0.6623361558983877,
+      "grad_norm": 10.974948622348556,
+      "learning_rate": 2e-06,
+      "loss": 0.3454,
+      "step": 2855
+    },
+    {
+      "epoch": 0.6625681475466884,
+      "grad_norm": 14.885775876370817,
+      "learning_rate": 2e-06,
+      "loss": 0.2967,
+      "step": 2856
+    },
+    {
+      "epoch": 0.662800139194989,
+      "grad_norm": 10.72842736132594,
+      "learning_rate": 2e-06,
+      "loss": 0.2827,
+      "step": 2857
+    },
+    {
+      "epoch": 0.6630321308432896,
+      "grad_norm": 14.921435245690917,
+      "learning_rate": 2e-06,
+      "loss": 0.2179,
+      "step": 2858
+    },
+    {
+      "epoch": 0.6632641224915903,
+      "grad_norm": 7.573869874381404,
+      "learning_rate": 2e-06,
+      "loss": 0.2186,
+      "step": 2859
+    },
+    {
+      "epoch": 0.663496114139891,
+      "grad_norm": 20.056549707412877,
+      "learning_rate": 2e-06,
+      "loss": 0.2833,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6637281057881916,
+      "grad_norm": 14.888681037702106,
+      "learning_rate": 2e-06,
+      "loss": 0.2875,
+      "step": 2861
+    },
+    {
+      "epoch": 0.6639600974364923,
+      "grad_norm": 17.421706372364103,
+      "learning_rate": 2e-06,
+      "loss": 0.4869,
+      "step": 2862
+    },
+    {
+      "epoch": 0.664192089084793,
+      "grad_norm": 13.282211399520385,
+      "learning_rate": 2e-06,
+      "loss": 0.3432,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6644240807330936,
+      "grad_norm": 24.679787184942278,
+      "learning_rate": 2e-06,
+      "loss": 0.3197,
+      "step": 2864
+    },
+    {
+      "epoch": 0.6646560723813942,
+      "grad_norm": 19.27889281747182,
+      "learning_rate": 2e-06,
+      "loss": 0.3208,
+      "step": 2865
+    },
+    {
+      "epoch": 0.6648880640296949,
+      "grad_norm": 13.070453246451402,
+      "learning_rate": 2e-06,
+      "loss": 0.2411,
+      "step": 2866
+    },
+    {
+      "epoch": 0.6651200556779956,
+      "grad_norm": 25.468659736645378,
+      "learning_rate": 2e-06,
+      "loss": 0.2217,
+      "step": 2867
+    },
+    {
+      "epoch": 0.6653520473262963,
+      "grad_norm": 13.516564164744764,
+      "learning_rate": 2e-06,
+      "loss": 0.2786,
+      "step": 2868
+    },
+    {
+      "epoch": 0.665584038974597,
+      "grad_norm": 46.16793049596266,
+      "learning_rate": 2e-06,
+      "loss": 0.2839,
+      "step": 2869
+    },
+    {
+      "epoch": 0.6658160306228976,
+      "grad_norm": 15.8472120552767,
+      "learning_rate": 2e-06,
+      "loss": 0.2784,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6660480222711982,
+      "grad_norm": 9.499928057186223,
+      "learning_rate": 2e-06,
+      "loss": 0.2202,
+      "step": 2871
+    },
+    {
+      "epoch": 0.6662800139194989,
+      "grad_norm": 15.961784571679711,
+      "learning_rate": 2e-06,
+      "loss": 0.3505,
+      "step": 2872
+    },
+    {
+      "epoch": 0.6665120055677995,
+      "grad_norm": 14.863888557389888,
+      "learning_rate": 2e-06,
+      "loss": 0.2517,
+      "step": 2873
+    },
+    {
+      "epoch": 0.6667439972161002,
+      "grad_norm": 12.114822736748357,
+      "learning_rate": 2e-06,
+      "loss": 0.266,
+      "step": 2874
+    },
+    {
+      "epoch": 0.6669759888644009,
+      "grad_norm": 18.791475979494606,
+      "learning_rate": 2e-06,
+      "loss": 0.2813,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6672079805127016,
+      "grad_norm": 13.241630128223022,
+      "learning_rate": 2e-06,
+      "loss": 0.1811,
+      "step": 2876
+    },
+    {
+      "epoch": 0.6674399721610021,
+      "grad_norm": 12.684844598060202,
+      "learning_rate": 2e-06,
+      "loss": 0.358,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6676719638093028,
+      "grad_norm": 27.649226549284123,
+      "learning_rate": 2e-06,
+      "loss": 0.3875,
+      "step": 2878
+    },
+    {
+      "epoch": 0.6679039554576035,
+      "grad_norm": 11.450368103403596,
+      "learning_rate": 2e-06,
+      "loss": 0.2175,
+      "step": 2879
+    },
+    {
+      "epoch": 0.6681359471059042,
+      "grad_norm": 15.313099377924349,
+      "learning_rate": 2e-06,
+      "loss": 0.3333,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6683679387542049,
+      "grad_norm": 17.75810877085882,
+      "learning_rate": 2e-06,
+      "loss": 0.2723,
+      "step": 2881
+    },
+    {
+      "epoch": 0.6685999304025055,
+      "grad_norm": 28.161703152994132,
+      "learning_rate": 2e-06,
+      "loss": 0.3317,
+      "step": 2882
+    },
+    {
+      "epoch": 0.6688319220508062,
+      "grad_norm": 9.307605731845863,
+      "learning_rate": 2e-06,
+      "loss": 0.2734,
+      "step": 2883
+    },
+    {
+      "epoch": 0.6690639136991068,
+      "grad_norm": 18.674716530336838,
+      "learning_rate": 2e-06,
+      "loss": 0.3547,
+      "step": 2884
+    },
+    {
+      "epoch": 0.6692959053474075,
+      "grad_norm": 17.731430790847497,
+      "learning_rate": 2e-06,
+      "loss": 0.2739,
+      "step": 2885
+    },
+    {
+      "epoch": 0.6695278969957081,
+      "grad_norm": 15.38332735147208,
+      "learning_rate": 2e-06,
+      "loss": 0.3064,
+      "step": 2886
+    },
+    {
+      "epoch": 0.6697598886440088,
+      "grad_norm": 14.647155898537573,
+      "learning_rate": 2e-06,
+      "loss": 0.1938,
+      "step": 2887
+    },
+    {
+      "epoch": 0.6699918802923095,
+      "grad_norm": 21.958953694103528,
+      "learning_rate": 2e-06,
+      "loss": 0.2576,
+      "step": 2888
+    },
+    {
+      "epoch": 0.6702238719406102,
+      "grad_norm": 19.13698687348818,
+      "learning_rate": 2e-06,
+      "loss": 0.2704,
+      "step": 2889
+    },
+    {
+      "epoch": 0.6704558635889108,
+      "grad_norm": 16.02891183410766,
+      "learning_rate": 2e-06,
+      "loss": 0.3191,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6706878552372114,
+      "grad_norm": 11.89398814623729,
+      "learning_rate": 2e-06,
+      "loss": 0.2183,
+      "step": 2891
+    },
+    {
+      "epoch": 0.6709198468855121,
+      "grad_norm": 27.27437455145165,
+      "learning_rate": 2e-06,
+      "loss": 0.3317,
+      "step": 2892
+    },
+    {
+      "epoch": 0.6711518385338128,
+      "grad_norm": 17.494392318806085,
+      "learning_rate": 2e-06,
+      "loss": 0.2895,
+      "step": 2893
+    },
+    {
+      "epoch": 0.6713838301821135,
+      "grad_norm": 8.532745967737572,
+      "learning_rate": 2e-06,
+      "loss": 0.3024,
+      "step": 2894
+    },
+    {
+      "epoch": 0.6716158218304141,
+      "grad_norm": 15.784287286887945,
+      "learning_rate": 2e-06,
+      "loss": 0.3662,
+      "step": 2895
+    },
+    {
+      "epoch": 0.6718478134787148,
+      "grad_norm": 19.885911995754892,
+      "learning_rate": 2e-06,
+      "loss": 0.2912,
+      "step": 2896
+    },
+    {
+      "epoch": 0.6720798051270154,
+      "grad_norm": 11.502699311807573,
+      "learning_rate": 2e-06,
+      "loss": 0.3606,
+      "step": 2897
+    },
+    {
+      "epoch": 0.672311796775316,
+      "grad_norm": 18.332904766435426,
+      "learning_rate": 2e-06,
+      "loss": 0.2782,
+      "step": 2898
+    },
+    {
+      "epoch": 0.6725437884236167,
+      "grad_norm": 11.192278874423675,
+      "learning_rate": 2e-06,
+      "loss": 0.2895,
+      "step": 2899
+    },
+    {
+      "epoch": 0.6727757800719174,
+      "grad_norm": 10.25836407488176,
+      "learning_rate": 2e-06,
+      "loss": 0.2621,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6730077717202181,
+      "grad_norm": 15.680650605945743,
+      "learning_rate": 2e-06,
+      "loss": 0.3704,
+      "step": 2901
+    },
+    {
+      "epoch": 0.6732397633685188,
+      "grad_norm": 24.10371429779154,
+      "learning_rate": 2e-06,
+      "loss": 0.3392,
+      "step": 2902
+    },
+    {
+      "epoch": 0.6734717550168194,
+      "grad_norm": 8.479237937207667,
+      "learning_rate": 2e-06,
+      "loss": 0.198,
+      "step": 2903
+    },
+    {
+      "epoch": 0.67370374666512,
+      "grad_norm": 14.174717355055936,
+      "learning_rate": 2e-06,
+      "loss": 0.3281,
+      "step": 2904
+    },
+    {
+      "epoch": 0.6739357383134207,
+      "grad_norm": 14.082257018873104,
+      "learning_rate": 2e-06,
+      "loss": 0.2516,
+      "step": 2905
+    },
+    {
+      "epoch": 0.6741677299617214,
+      "grad_norm": 13.382312420174081,
+      "learning_rate": 2e-06,
+      "loss": 0.295,
+      "step": 2906
+    },
+    {
+      "epoch": 0.674399721610022,
+      "grad_norm": 24.6062414245324,
+      "learning_rate": 2e-06,
+      "loss": 0.3857,
+      "step": 2907
+    },
+    {
+      "epoch": 0.6746317132583227,
+      "grad_norm": 27.594212344103287,
+      "learning_rate": 2e-06,
+      "loss": 0.3976,
+      "step": 2908
+    },
+    {
+      "epoch": 0.6748637049066234,
+      "grad_norm": 22.15561527386861,
+      "learning_rate": 2e-06,
+      "loss": 0.4427,
+      "step": 2909
+    },
+    {
+      "epoch": 0.6750956965549241,
+      "grad_norm": 9.280812151914011,
+      "learning_rate": 2e-06,
+      "loss": 0.2409,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6753276882032246,
+      "grad_norm": 14.739297859742182,
+      "learning_rate": 2e-06,
+      "loss": 0.2851,
+      "step": 2911
+    },
+    {
+      "epoch": 0.6755596798515253,
+      "grad_norm": 17.43973773723807,
+      "learning_rate": 2e-06,
+      "loss": 0.2358,
+      "step": 2912
+    },
+    {
+      "epoch": 0.675791671499826,
+      "grad_norm": 21.123065539604433,
+      "learning_rate": 2e-06,
+      "loss": 0.2731,
+      "step": 2913
+    },
+    {
+      "epoch": 0.6760236631481267,
+      "grad_norm": 18.18026849345217,
+      "learning_rate": 2e-06,
+      "loss": 0.2914,
+      "step": 2914
+    },
+    {
+      "epoch": 0.6762556547964274,
+      "grad_norm": 22.75544500204419,
+      "learning_rate": 2e-06,
+      "loss": 0.3933,
+      "step": 2915
+    },
+    {
+      "epoch": 0.676487646444728,
+      "grad_norm": 19.37742185783371,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 2916
+    },
+    {
+      "epoch": 0.6767196380930286,
+      "grad_norm": 16.69819067505101,
+      "learning_rate": 2e-06,
+      "loss": 0.2691,
+      "step": 2917
+    },
+    {
+      "epoch": 0.6769516297413293,
+      "grad_norm": 16.50659193036346,
+      "learning_rate": 2e-06,
+      "loss": 0.2146,
+      "step": 2918
+    },
+    {
+      "epoch": 0.67718362138963,
+      "grad_norm": 7.484957257065774,
+      "learning_rate": 2e-06,
+      "loss": 0.2779,
+      "step": 2919
+    },
+    {
+      "epoch": 0.6774156130379306,
+      "grad_norm": 15.034466055515287,
+      "learning_rate": 2e-06,
+      "loss": 0.2166,
+      "step": 2920
+    },
+    {
+      "epoch": 0.6776476046862313,
+      "grad_norm": 16.417816190692054,
+      "learning_rate": 2e-06,
+      "loss": 0.359,
+      "step": 2921
+    },
+    {
+      "epoch": 0.677879596334532,
+      "grad_norm": 22.66480068571077,
+      "learning_rate": 2e-06,
+      "loss": 0.2557,
+      "step": 2922
+    },
+    {
+      "epoch": 0.6781115879828327,
+      "grad_norm": 9.93097539810531,
+      "learning_rate": 2e-06,
+      "loss": 0.2954,
+      "step": 2923
+    },
+    {
+      "epoch": 0.6783435796311332,
+      "grad_norm": 15.193296735844845,
+      "learning_rate": 2e-06,
+      "loss": 0.2506,
+      "step": 2924
+    },
+    {
+      "epoch": 0.6785755712794339,
+      "grad_norm": 17.52679971654726,
+      "learning_rate": 2e-06,
+      "loss": 0.3571,
+      "step": 2925
+    },
+    {
+      "epoch": 0.6788075629277346,
+      "grad_norm": 7.445586192344902,
+      "learning_rate": 2e-06,
+      "loss": 0.1966,
+      "step": 2926
+    },
+    {
+      "epoch": 0.6790395545760353,
+      "grad_norm": 8.416210239673962,
+      "learning_rate": 2e-06,
+      "loss": 0.2133,
+      "step": 2927
+    },
+    {
+      "epoch": 0.6792715462243359,
+      "grad_norm": 14.313635294079827,
+      "learning_rate": 2e-06,
+      "loss": 0.3515,
+      "step": 2928
+    },
+    {
+      "epoch": 0.6795035378726366,
+      "grad_norm": 24.603948921934705,
+      "learning_rate": 2e-06,
+      "loss": 0.3901,
+      "step": 2929
+    },
+    {
+      "epoch": 0.6797355295209373,
+      "grad_norm": 15.833755249534436,
+      "learning_rate": 2e-06,
+      "loss": 0.3374,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6799675211692379,
+      "grad_norm": 16.04070634218248,
+      "learning_rate": 2e-06,
+      "loss": 0.3736,
+      "step": 2931
+    },
+    {
+      "epoch": 0.6801995128175385,
+      "grad_norm": 20.682720901172534,
+      "learning_rate": 2e-06,
+      "loss": 0.3394,
+      "step": 2932
+    },
+    {
+      "epoch": 0.6804315044658392,
+      "grad_norm": 21.697405182973313,
+      "learning_rate": 2e-06,
+      "loss": 0.3276,
+      "step": 2933
+    },
+    {
+      "epoch": 0.6806634961141399,
+      "grad_norm": 18.865819341538803,
+      "learning_rate": 2e-06,
+      "loss": 0.3264,
+      "step": 2934
+    },
+    {
+      "epoch": 0.6808954877624406,
+      "grad_norm": 10.423027096808157,
+      "learning_rate": 2e-06,
+      "loss": 0.3037,
+      "step": 2935
+    },
+    {
+      "epoch": 0.6811274794107413,
+      "grad_norm": 21.677979146013957,
+      "learning_rate": 2e-06,
+      "loss": 0.367,
+      "step": 2936
+    },
+    {
+      "epoch": 0.6813594710590418,
+      "grad_norm": 26.4302768785951,
+      "learning_rate": 2e-06,
+      "loss": 0.3841,
+      "step": 2937
+    },
+    {
+      "epoch": 0.6815914627073425,
+      "grad_norm": 10.069866494626753,
+      "learning_rate": 2e-06,
+      "loss": 0.2875,
+      "step": 2938
+    },
+    {
+      "epoch": 0.6818234543556432,
+      "grad_norm": 9.805489612875316,
+      "learning_rate": 2e-06,
+      "loss": 0.1803,
+      "step": 2939
+    },
+    {
+      "epoch": 0.6820554460039439,
+      "grad_norm": 13.082756265772101,
+      "learning_rate": 2e-06,
+      "loss": 0.2856,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6822874376522445,
+      "grad_norm": 21.214294775844788,
+      "learning_rate": 2e-06,
+      "loss": 0.3278,
+      "step": 2941
+    },
+    {
+      "epoch": 0.6825194293005452,
+      "grad_norm": 18.98402642369575,
+      "learning_rate": 2e-06,
+      "loss": 0.3717,
+      "step": 2942
+    },
+    {
+      "epoch": 0.6827514209488459,
+      "grad_norm": 18.28594400987843,
+      "learning_rate": 2e-06,
+      "loss": 0.2966,
+      "step": 2943
+    },
+    {
+      "epoch": 0.6829834125971465,
+      "grad_norm": 15.955264733597344,
+      "learning_rate": 2e-06,
+      "loss": 0.2649,
+      "step": 2944
+    },
+    {
+      "epoch": 0.6832154042454471,
+      "grad_norm": 12.143853436389726,
+      "learning_rate": 2e-06,
+      "loss": 0.3043,
+      "step": 2945
+    },
+    {
+      "epoch": 0.6834473958937478,
+      "grad_norm": 33.10313809595552,
+      "learning_rate": 2e-06,
+      "loss": 0.4111,
+      "step": 2946
+    },
+    {
+      "epoch": 0.6836793875420485,
+      "grad_norm": 15.49237897542982,
+      "learning_rate": 2e-06,
+      "loss": 0.3257,
+      "step": 2947
+    },
+    {
+      "epoch": 0.6839113791903492,
+      "grad_norm": 15.234371131142845,
+      "learning_rate": 2e-06,
+      "loss": 0.2847,
+      "step": 2948
+    },
+    {
+      "epoch": 0.6841433708386498,
+      "grad_norm": 14.394745150033899,
+      "learning_rate": 2e-06,
+      "loss": 0.2325,
+      "step": 2949
+    },
+    {
+      "epoch": 0.6843753624869505,
+      "grad_norm": 7.697439348929282,
+      "learning_rate": 2e-06,
+      "loss": 0.2083,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6846073541352511,
+      "grad_norm": 8.347270088398039,
+      "learning_rate": 2e-06,
+      "loss": 0.1849,
+      "step": 2951
+    },
+    {
+      "epoch": 0.6848393457835518,
+      "grad_norm": 8.528135185478353,
+      "learning_rate": 2e-06,
+      "loss": 0.2895,
+      "step": 2952
+    },
+    {
+      "epoch": 0.6850713374318524,
+      "grad_norm": 11.440085523850653,
+      "learning_rate": 2e-06,
+      "loss": 0.3013,
+      "step": 2953
+    },
+    {
+      "epoch": 0.6853033290801531,
+      "grad_norm": 14.240300923564252,
+      "learning_rate": 2e-06,
+      "loss": 0.2868,
+      "step": 2954
+    },
+    {
+      "epoch": 0.6855353207284538,
+      "grad_norm": 15.298243190826206,
+      "learning_rate": 2e-06,
+      "loss": 0.2526,
+      "step": 2955
+    },
+    {
+      "epoch": 0.6857673123767545,
+      "grad_norm": 16.287029088480868,
+      "learning_rate": 2e-06,
+      "loss": 0.2477,
+      "step": 2956
+    },
+    {
+      "epoch": 0.685999304025055,
+      "grad_norm": 16.330072651646685,
+      "learning_rate": 2e-06,
+      "loss": 0.2338,
+      "step": 2957
+    },
+    {
+      "epoch": 0.6862312956733557,
+      "grad_norm": 10.144289299404209,
+      "learning_rate": 2e-06,
+      "loss": 0.2366,
+      "step": 2958
+    },
+    {
+      "epoch": 0.6864632873216564,
+      "grad_norm": 15.970485036072468,
+      "learning_rate": 2e-06,
+      "loss": 0.2597,
+      "step": 2959
+    },
+    {
+      "epoch": 0.6866952789699571,
+      "grad_norm": 20.059300974805478,
+      "learning_rate": 2e-06,
+      "loss": 0.2755,
+      "step": 2960
+    },
+    {
+      "epoch": 0.6869272706182578,
+      "grad_norm": 13.123594011226873,
+      "learning_rate": 2e-06,
+      "loss": 0.2808,
+      "step": 2961
+    },
+    {
+      "epoch": 0.6871592622665584,
+      "grad_norm": 11.967794532068869,
+      "learning_rate": 2e-06,
+      "loss": 0.2214,
+      "step": 2962
+    },
+    {
+      "epoch": 0.6873912539148591,
+      "grad_norm": 13.279811333305938,
+      "learning_rate": 2e-06,
+      "loss": 0.2894,
+      "step": 2963
+    },
+    {
+      "epoch": 0.6876232455631597,
+      "grad_norm": 15.978601053972245,
+      "learning_rate": 2e-06,
+      "loss": 0.3233,
+      "step": 2964
+    },
+    {
+      "epoch": 0.6878552372114604,
+      "grad_norm": 19.482813281061436,
+      "learning_rate": 2e-06,
+      "loss": 0.3722,
+      "step": 2965
+    },
+    {
+      "epoch": 0.688087228859761,
+      "grad_norm": 16.022926928632646,
+      "learning_rate": 2e-06,
+      "loss": 0.4637,
+      "step": 2966
+    },
+    {
+      "epoch": 0.6883192205080617,
+      "grad_norm": 25.90689576178664,
+      "learning_rate": 2e-06,
+      "loss": 0.3036,
+      "step": 2967
+    },
+    {
+      "epoch": 0.6885512121563624,
+      "grad_norm": 15.81243335582368,
+      "learning_rate": 2e-06,
+      "loss": 0.257,
+      "step": 2968
+    },
+    {
+      "epoch": 0.6887832038046631,
+      "grad_norm": 18.661608358707515,
+      "learning_rate": 2e-06,
+      "loss": 0.3841,
+      "step": 2969
+    },
+    {
+      "epoch": 0.6890151954529637,
+      "grad_norm": 20.789753229496526,
+      "learning_rate": 2e-06,
+      "loss": 0.3484,
+      "step": 2970
+    },
+    {
+      "epoch": 0.6892471871012643,
+      "grad_norm": 11.855384758128736,
+      "learning_rate": 2e-06,
+      "loss": 0.2881,
+      "step": 2971
+    },
+    {
+      "epoch": 0.689479178749565,
+      "grad_norm": 17.159193927611167,
+      "learning_rate": 2e-06,
+      "loss": 0.2717,
+      "step": 2972
+    },
+    {
+      "epoch": 0.6897111703978657,
+      "grad_norm": 13.992237879814896,
+      "learning_rate": 2e-06,
+      "loss": 0.4286,
+      "step": 2973
+    },
+    {
+      "epoch": 0.6899431620461663,
+      "grad_norm": 24.35252878929027,
+      "learning_rate": 2e-06,
+      "loss": 0.3378,
+      "step": 2974
+    },
+    {
+      "epoch": 0.690175153694467,
+      "grad_norm": 16.71016085700087,
+      "learning_rate": 2e-06,
+      "loss": 0.2689,
+      "step": 2975
+    },
+    {
+      "epoch": 0.6904071453427677,
+      "grad_norm": 9.784394283410494,
+      "learning_rate": 2e-06,
+      "loss": 0.2627,
+      "step": 2976
+    },
+    {
+      "epoch": 0.6906391369910683,
+      "grad_norm": 8.624407375319688,
+      "learning_rate": 2e-06,
+      "loss": 0.2334,
+      "step": 2977
+    },
+    {
+      "epoch": 0.690871128639369,
+      "grad_norm": 6.348981967459156,
+      "learning_rate": 2e-06,
+      "loss": 0.2202,
+      "step": 2978
+    },
+    {
+      "epoch": 0.6911031202876696,
+      "grad_norm": 19.035352735373152,
+      "learning_rate": 2e-06,
+      "loss": 0.3631,
+      "step": 2979
+    },
+    {
+      "epoch": 0.6913351119359703,
+      "grad_norm": 7.198738788917573,
+      "learning_rate": 2e-06,
+      "loss": 0.2179,
+      "step": 2980
+    },
+    {
+      "epoch": 0.691567103584271,
+      "grad_norm": 18.931518041058506,
+      "learning_rate": 2e-06,
+      "loss": 0.3189,
+      "step": 2981
+    },
+    {
+      "epoch": 0.6917990952325717,
+      "grad_norm": 23.18489676907909,
+      "learning_rate": 2e-06,
+      "loss": 0.2713,
+      "step": 2982
+    },
+    {
+      "epoch": 0.6920310868808723,
+      "grad_norm": 15.316381457085425,
+      "learning_rate": 2e-06,
+      "loss": 0.3127,
+      "step": 2983
+    },
+    {
+      "epoch": 0.6922630785291729,
+      "grad_norm": 11.503904532399071,
+      "learning_rate": 2e-06,
+      "loss": 0.2112,
+      "step": 2984
+    },
+    {
+      "epoch": 0.6924950701774736,
+      "grad_norm": 23.105103132210573,
+      "learning_rate": 2e-06,
+      "loss": 0.35,
+      "step": 2985
+    },
+    {
+      "epoch": 0.6927270618257743,
+      "grad_norm": 8.934461565490222,
+      "learning_rate": 2e-06,
+      "loss": 0.1694,
+      "step": 2986
+    },
+    {
+      "epoch": 0.6929590534740749,
+      "grad_norm": 15.460634718345247,
+      "learning_rate": 2e-06,
+      "loss": 0.327,
+      "step": 2987
+    },
+    {
+      "epoch": 0.6931910451223756,
+      "grad_norm": 10.533426557156625,
+      "learning_rate": 2e-06,
+      "loss": 0.2937,
+      "step": 2988
+    },
+    {
+      "epoch": 0.6934230367706763,
+      "grad_norm": 12.54384265691961,
+      "learning_rate": 2e-06,
+      "loss": 0.2716,
+      "step": 2989
+    },
+    {
+      "epoch": 0.693655028418977,
+      "grad_norm": 17.7973153093463,
+      "learning_rate": 2e-06,
+      "loss": 0.2359,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6938870200672775,
+      "grad_norm": 7.855860871585307,
+      "learning_rate": 2e-06,
+      "loss": 0.214,
+      "step": 2991
+    },
+    {
+      "epoch": 0.6941190117155782,
+      "grad_norm": 23.44117120120443,
+      "learning_rate": 2e-06,
+      "loss": 0.3826,
+      "step": 2992
+    },
+    {
+      "epoch": 0.6943510033638789,
+      "grad_norm": 21.587237505595454,
+      "learning_rate": 2e-06,
+      "loss": 0.3726,
+      "step": 2993
+    },
+    {
+      "epoch": 0.6945829950121796,
+      "grad_norm": 8.966649034172018,
+      "learning_rate": 2e-06,
+      "loss": 0.3061,
+      "step": 2994
+    },
+    {
+      "epoch": 0.6948149866604802,
+      "grad_norm": 10.394999748177568,
+      "learning_rate": 2e-06,
+      "loss": 0.2836,
+      "step": 2995
+    },
+    {
+      "epoch": 0.6950469783087809,
+      "grad_norm": 16.36252544489369,
+      "learning_rate": 2e-06,
+      "loss": 0.2719,
+      "step": 2996
+    },
+    {
+      "epoch": 0.6952789699570815,
+      "grad_norm": 15.73753046229693,
+      "learning_rate": 2e-06,
+      "loss": 0.2106,
+      "step": 2997
+    },
+    {
+      "epoch": 0.6955109616053822,
+      "grad_norm": 10.482261401826955,
+      "learning_rate": 2e-06,
+      "loss": 0.2225,
+      "step": 2998
+    },
+    {
+      "epoch": 0.6957429532536828,
+      "grad_norm": 16.162687708016666,
+      "learning_rate": 2e-06,
+      "loss": 0.3063,
+      "step": 2999
+    },
+    {
+      "epoch": 0.6959749449019835,
+      "grad_norm": 10.163603053519628,
+      "learning_rate": 2e-06,
+      "loss": 0.2532,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6962069365502842,
+      "grad_norm": 13.518454150870886,
+      "learning_rate": 2e-06,
+      "loss": 0.3016,
+      "step": 3001
+    },
+    {
+      "epoch": 0.6964389281985849,
+      "grad_norm": 18.165892745018137,
+      "learning_rate": 2e-06,
+      "loss": 0.3698,
+      "step": 3002
+    },
+    {
+      "epoch": 0.6966709198468856,
+      "grad_norm": 17.90105884966254,
+      "learning_rate": 2e-06,
+      "loss": 0.3063,
+      "step": 3003
+    },
+    {
+      "epoch": 0.6969029114951861,
+      "grad_norm": 12.788468740614176,
+      "learning_rate": 2e-06,
+      "loss": 0.3079,
+      "step": 3004
+    },
+    {
+      "epoch": 0.6971349031434868,
+      "grad_norm": 15.957661767789833,
+      "learning_rate": 2e-06,
+      "loss": 0.2346,
+      "step": 3005
+    },
+    {
+      "epoch": 0.6973668947917875,
+      "grad_norm": 18.210358640908733,
+      "learning_rate": 2e-06,
+      "loss": 0.2909,
+      "step": 3006
+    },
+    {
+      "epoch": 0.6975988864400882,
+      "grad_norm": 13.050628004493788,
+      "learning_rate": 2e-06,
+      "loss": 0.3312,
+      "step": 3007
+    },
+    {
+      "epoch": 0.6978308780883888,
+      "grad_norm": 13.177363294059413,
+      "learning_rate": 2e-06,
+      "loss": 0.2819,
+      "step": 3008
+    },
+    {
+      "epoch": 0.6980628697366895,
+      "grad_norm": 8.645162144450342,
+      "learning_rate": 2e-06,
+      "loss": 0.2241,
+      "step": 3009
+    },
+    {
+      "epoch": 0.6982948613849902,
+      "grad_norm": 12.456599894548935,
+      "learning_rate": 2e-06,
+      "loss": 0.2404,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6985268530332908,
+      "grad_norm": 16.152022904207286,
+      "learning_rate": 2e-06,
+      "loss": 0.2267,
+      "step": 3011
+    },
+    {
+      "epoch": 0.6987588446815914,
+      "grad_norm": 15.475853041190083,
+      "learning_rate": 2e-06,
+      "loss": 0.2555,
+      "step": 3012
+    },
+    {
+      "epoch": 0.6989908363298921,
+      "grad_norm": 17.594340371945172,
+      "learning_rate": 2e-06,
+      "loss": 0.3743,
+      "step": 3013
+    },
+    {
+      "epoch": 0.6992228279781928,
+      "grad_norm": 13.891850067454452,
+      "learning_rate": 2e-06,
+      "loss": 0.333,
+      "step": 3014
+    },
+    {
+      "epoch": 0.6994548196264935,
+      "grad_norm": 9.746005875818053,
+      "learning_rate": 2e-06,
+      "loss": 0.2134,
+      "step": 3015
+    },
+    {
+      "epoch": 0.6996868112747942,
+      "grad_norm": 13.990318327075137,
+      "learning_rate": 2e-06,
+      "loss": 0.3343,
+      "step": 3016
+    },
+    {
+      "epoch": 0.6999188029230947,
+      "grad_norm": 23.025914502524685,
+      "learning_rate": 2e-06,
+      "loss": 0.4144,
+      "step": 3017
+    },
+    {
+      "epoch": 0.7001507945713954,
+      "grad_norm": 22.778332248585592,
+      "learning_rate": 2e-06,
+      "loss": 0.1918,
+      "step": 3018
+    },
+    {
+      "epoch": 0.7003827862196961,
+      "grad_norm": 15.430801616526118,
+      "learning_rate": 2e-06,
+      "loss": 0.3679,
+      "step": 3019
+    },
+    {
+      "epoch": 0.7006147778679968,
+      "grad_norm": 16.207361770057013,
+      "learning_rate": 2e-06,
+      "loss": 0.2704,
+      "step": 3020
+    },
+    {
+      "epoch": 0.7008467695162974,
+      "grad_norm": 16.574744397317918,
+      "learning_rate": 2e-06,
+      "loss": 0.3118,
+      "step": 3021
+    },
+    {
+      "epoch": 0.7010787611645981,
+      "grad_norm": 9.771455601212924,
+      "learning_rate": 2e-06,
+      "loss": 0.2042,
+      "step": 3022
+    },
+    {
+      "epoch": 0.7013107528128988,
+      "grad_norm": 16.352935531849965,
+      "learning_rate": 2e-06,
+      "loss": 0.3103,
+      "step": 3023
+    },
+    {
+      "epoch": 0.7015427444611994,
+      "grad_norm": 16.012828911522448,
+      "learning_rate": 2e-06,
+      "loss": 0.3218,
+      "step": 3024
+    },
+    {
+      "epoch": 0.7017747361095,
+      "grad_norm": 31.336384254208152,
+      "learning_rate": 2e-06,
+      "loss": 0.3786,
+      "step": 3025
+    },
+    {
+      "epoch": 0.7020067277578007,
+      "grad_norm": 11.076684492126956,
+      "learning_rate": 2e-06,
+      "loss": 0.2038,
+      "step": 3026
+    },
+    {
+      "epoch": 0.7022387194061014,
+      "grad_norm": 19.254274788008026,
+      "learning_rate": 2e-06,
+      "loss": 0.3597,
+      "step": 3027
+    },
+    {
+      "epoch": 0.7024707110544021,
+      "grad_norm": 22.845025573712128,
+      "learning_rate": 2e-06,
+      "loss": 0.3286,
+      "step": 3028
+    },
+    {
+      "epoch": 0.7027027027027027,
+      "grad_norm": 10.265590830213393,
+      "learning_rate": 2e-06,
+      "loss": 0.2768,
+      "step": 3029
+    },
+    {
+      "epoch": 0.7029346943510033,
+      "grad_norm": 13.635127225963343,
+      "learning_rate": 2e-06,
+      "loss": 0.2435,
+      "step": 3030
+    },
+    {
+      "epoch": 0.703166685999304,
+      "grad_norm": 21.43039852701912,
+      "learning_rate": 2e-06,
+      "loss": 0.3317,
+      "step": 3031
+    },
+    {
+      "epoch": 0.7033986776476047,
+      "grad_norm": 20.249225557418523,
+      "learning_rate": 2e-06,
+      "loss": 0.3132,
+      "step": 3032
+    },
+    {
+      "epoch": 0.7036306692959053,
+      "grad_norm": 10.437849106321048,
+      "learning_rate": 2e-06,
+      "loss": 0.2831,
+      "step": 3033
+    },
+    {
+      "epoch": 0.703862660944206,
+      "grad_norm": 14.921942445808748,
+      "learning_rate": 2e-06,
+      "loss": 0.3055,
+      "step": 3034
+    },
+    {
+      "epoch": 0.7040946525925067,
+      "grad_norm": 13.427974898300857,
+      "learning_rate": 2e-06,
+      "loss": 0.2818,
+      "step": 3035
+    },
+    {
+      "epoch": 0.7043266442408074,
+      "grad_norm": 18.31794009982437,
+      "learning_rate": 2e-06,
+      "loss": 0.3405,
+      "step": 3036
+    },
+    {
+      "epoch": 0.7045586358891079,
+      "grad_norm": 31.05526528803719,
+      "learning_rate": 2e-06,
+      "loss": 0.5122,
+      "step": 3037
+    },
+    {
+      "epoch": 0.7047906275374086,
+      "grad_norm": 12.254876345659827,
+      "learning_rate": 2e-06,
+      "loss": 0.3137,
+      "step": 3038
+    },
+    {
+      "epoch": 0.7050226191857093,
+      "grad_norm": 15.370854458054378,
+      "learning_rate": 2e-06,
+      "loss": 0.3237,
+      "step": 3039
+    },
+    {
+      "epoch": 0.70525461083401,
+      "grad_norm": 18.637823728977246,
+      "learning_rate": 2e-06,
+      "loss": 0.2775,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7054866024823107,
+      "grad_norm": 12.873597948913517,
+      "learning_rate": 2e-06,
+      "loss": 0.3299,
+      "step": 3041
+    },
+    {
+      "epoch": 0.7057185941306113,
+      "grad_norm": 17.18921170254753,
+      "learning_rate": 2e-06,
+      "loss": 0.3152,
+      "step": 3042
+    },
+    {
+      "epoch": 0.705950585778912,
+      "grad_norm": 15.409492736211421,
+      "learning_rate": 2e-06,
+      "loss": 0.2607,
+      "step": 3043
+    },
+    {
+      "epoch": 0.7061825774272126,
+      "grad_norm": 9.975829678915247,
+      "learning_rate": 2e-06,
+      "loss": 0.3616,
+      "step": 3044
+    },
+    {
+      "epoch": 0.7064145690755133,
+      "grad_norm": 15.207969039612548,
+      "learning_rate": 2e-06,
+      "loss": 0.3872,
+      "step": 3045
+    },
+    {
+      "epoch": 0.7066465607238139,
+      "grad_norm": 12.70874179865131,
+      "learning_rate": 2e-06,
+      "loss": 0.2803,
+      "step": 3046
+    },
+    {
+      "epoch": 0.7068785523721146,
+      "grad_norm": 10.777788004616392,
+      "learning_rate": 2e-06,
+      "loss": 0.3013,
+      "step": 3047
+    },
+    {
+      "epoch": 0.7071105440204153,
+      "grad_norm": 9.71067971397703,
+      "learning_rate": 2e-06,
+      "loss": 0.231,
+      "step": 3048
+    },
+    {
+      "epoch": 0.707342535668716,
+      "grad_norm": 16.182320295977373,
+      "learning_rate": 2e-06,
+      "loss": 0.2541,
+      "step": 3049
+    },
+    {
+      "epoch": 0.7075745273170165,
+      "grad_norm": 14.24194784140107,
+      "learning_rate": 2e-06,
+      "loss": 0.3133,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7078065189653172,
+      "grad_norm": 8.116297025441758,
+      "learning_rate": 2e-06,
+      "loss": 0.2765,
+      "step": 3051
+    },
+    {
+      "epoch": 0.7080385106136179,
+      "grad_norm": 15.408449590028416,
+      "learning_rate": 2e-06,
+      "loss": 0.3464,
+      "step": 3052
+    },
+    {
+      "epoch": 0.7082705022619186,
+      "grad_norm": 10.126091809211045,
+      "learning_rate": 2e-06,
+      "loss": 0.2563,
+      "step": 3053
+    },
+    {
+      "epoch": 0.7085024939102192,
+      "grad_norm": 15.76931845699468,
+      "learning_rate": 2e-06,
+      "loss": 0.3641,
+      "step": 3054
+    },
+    {
+      "epoch": 0.7087344855585199,
+      "grad_norm": 14.3181566502542,
+      "learning_rate": 2e-06,
+      "loss": 0.2404,
+      "step": 3055
+    },
+    {
+      "epoch": 0.7089664772068206,
+      "grad_norm": 11.518059207296266,
+      "learning_rate": 2e-06,
+      "loss": 0.3092,
+      "step": 3056
+    },
+    {
+      "epoch": 0.7091984688551212,
+      "grad_norm": 11.075548644625707,
+      "learning_rate": 2e-06,
+      "loss": 0.2368,
+      "step": 3057
+    },
+    {
+      "epoch": 0.7094304605034218,
+      "grad_norm": 8.931630979263996,
+      "learning_rate": 2e-06,
+      "loss": 0.2094,
+      "step": 3058
+    },
+    {
+      "epoch": 0.7096624521517225,
+      "grad_norm": 10.015903169009626,
+      "learning_rate": 2e-06,
+      "loss": 0.3026,
+      "step": 3059
+    },
+    {
+      "epoch": 0.7098944438000232,
+      "grad_norm": 9.001640545817724,
+      "learning_rate": 2e-06,
+      "loss": 0.2358,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7101264354483239,
+      "grad_norm": 9.884731171167504,
+      "learning_rate": 2e-06,
+      "loss": 0.2066,
+      "step": 3061
+    },
+    {
+      "epoch": 0.7103584270966246,
+      "grad_norm": 8.79796649865471,
+      "learning_rate": 2e-06,
+      "loss": 0.303,
+      "step": 3062
+    },
+    {
+      "epoch": 0.7105904187449252,
+      "grad_norm": 12.035283919663504,
+      "learning_rate": 2e-06,
+      "loss": 0.3186,
+      "step": 3063
+    },
+    {
+      "epoch": 0.7108224103932258,
+      "grad_norm": 5.610850717951971,
+      "learning_rate": 2e-06,
+      "loss": 0.1643,
+      "step": 3064
+    },
+    {
+      "epoch": 0.7110544020415265,
+      "grad_norm": 20.315813498268504,
+      "learning_rate": 2e-06,
+      "loss": 0.2786,
+      "step": 3065
+    },
+    {
+      "epoch": 0.7112863936898272,
+      "grad_norm": 15.024208154539762,
+      "learning_rate": 2e-06,
+      "loss": 0.358,
+      "step": 3066
+    },
+    {
+      "epoch": 0.7115183853381278,
+      "grad_norm": 18.763448531490877,
+      "learning_rate": 2e-06,
+      "loss": 0.3619,
+      "step": 3067
+    },
+    {
+      "epoch": 0.7117503769864285,
+      "grad_norm": 8.850772694364398,
+      "learning_rate": 2e-06,
+      "loss": 0.3079,
+      "step": 3068
+    },
+    {
+      "epoch": 0.7119823686347292,
+      "grad_norm": 12.558505567041202,
+      "learning_rate": 2e-06,
+      "loss": 0.2539,
+      "step": 3069
+    },
+    {
+      "epoch": 0.7122143602830298,
+      "grad_norm": 9.413887213263452,
+      "learning_rate": 2e-06,
+      "loss": 0.2578,
+      "step": 3070
+    },
+    {
+      "epoch": 0.7124463519313304,
+      "grad_norm": 8.268247015403908,
+      "learning_rate": 2e-06,
+      "loss": 0.3173,
+      "step": 3071
+    },
+    {
+      "epoch": 0.7126783435796311,
+      "grad_norm": 8.897679021367162,
+      "learning_rate": 2e-06,
+      "loss": 0.2035,
+      "step": 3072
+    },
+    {
+      "epoch": 0.7129103352279318,
+      "grad_norm": 6.774172208125166,
+      "learning_rate": 2e-06,
+      "loss": 0.2567,
+      "step": 3073
+    },
+    {
+      "epoch": 0.7131423268762325,
+      "grad_norm": 15.784809665803815,
+      "learning_rate": 2e-06,
+      "loss": 0.3427,
+      "step": 3074
+    },
+    {
+      "epoch": 0.7133743185245331,
+      "grad_norm": 9.490956131277786,
+      "learning_rate": 2e-06,
+      "loss": 0.305,
+      "step": 3075
+    },
+    {
+      "epoch": 0.7136063101728338,
+      "grad_norm": 12.768478698546604,
+      "learning_rate": 2e-06,
+      "loss": 0.2892,
+      "step": 3076
+    },
+    {
+      "epoch": 0.7138383018211344,
+      "grad_norm": 14.733261060077966,
+      "learning_rate": 2e-06,
+      "loss": 0.2683,
+      "step": 3077
+    },
+    {
+      "epoch": 0.7140702934694351,
+      "grad_norm": 16.361146756468674,
+      "learning_rate": 2e-06,
+      "loss": 0.2763,
+      "step": 3078
+    },
+    {
+      "epoch": 0.7143022851177357,
+      "grad_norm": 20.56834735004541,
+      "learning_rate": 2e-06,
+      "loss": 0.3263,
+      "step": 3079
+    },
+    {
+      "epoch": 0.7145342767660364,
+      "grad_norm": 19.139723749178586,
+      "learning_rate": 2e-06,
+      "loss": 0.337,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7147662684143371,
+      "grad_norm": 6.061627605458326,
+      "learning_rate": 2e-06,
+      "loss": 0.1807,
+      "step": 3081
+    },
+    {
+      "epoch": 0.7149982600626378,
+      "grad_norm": 17.17408580688916,
+      "learning_rate": 2e-06,
+      "loss": 0.2639,
+      "step": 3082
+    },
+    {
+      "epoch": 0.7152302517109385,
+      "grad_norm": 21.46351281558989,
+      "learning_rate": 2e-06,
+      "loss": 0.3246,
+      "step": 3083
+    },
+    {
+      "epoch": 0.715462243359239,
+      "grad_norm": 15.413561020002579,
+      "learning_rate": 2e-06,
+      "loss": 0.2781,
+      "step": 3084
+    },
+    {
+      "epoch": 0.7156942350075397,
+      "grad_norm": 17.441447535031255,
+      "learning_rate": 2e-06,
+      "loss": 0.345,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7159262266558404,
+      "grad_norm": 14.122717322632662,
+      "learning_rate": 2e-06,
+      "loss": 0.2305,
+      "step": 3086
+    },
+    {
+      "epoch": 0.7161582183041411,
+      "grad_norm": 11.784198991709928,
+      "learning_rate": 2e-06,
+      "loss": 0.31,
+      "step": 3087
+    },
+    {
+      "epoch": 0.7163902099524417,
+      "grad_norm": 16.435399537046283,
+      "learning_rate": 2e-06,
+      "loss": 0.2253,
+      "step": 3088
+    },
+    {
+      "epoch": 0.7166222016007424,
+      "grad_norm": 20.087543297417806,
+      "learning_rate": 2e-06,
+      "loss": 0.3557,
+      "step": 3089
+    },
+    {
+      "epoch": 0.716854193249043,
+      "grad_norm": 9.136332789225563,
+      "learning_rate": 2e-06,
+      "loss": 0.2596,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7170861848973437,
+      "grad_norm": 17.3524058908081,
+      "learning_rate": 2e-06,
+      "loss": 0.2313,
+      "step": 3091
+    },
+    {
+      "epoch": 0.7173181765456443,
+      "grad_norm": 17.29438646410513,
+      "learning_rate": 2e-06,
+      "loss": 0.2101,
+      "step": 3092
+    },
+    {
+      "epoch": 0.717550168193945,
+      "grad_norm": 12.382294379053754,
+      "learning_rate": 2e-06,
+      "loss": 0.2556,
+      "step": 3093
+    },
+    {
+      "epoch": 0.7177821598422457,
+      "grad_norm": 13.213939787408052,
+      "learning_rate": 2e-06,
+      "loss": 0.2625,
+      "step": 3094
+    },
+    {
+      "epoch": 0.7180141514905464,
+      "grad_norm": 8.476230812288442,
+      "learning_rate": 2e-06,
+      "loss": 0.2527,
+      "step": 3095
+    },
+    {
+      "epoch": 0.718246143138847,
+      "grad_norm": 19.212254605560727,
+      "learning_rate": 2e-06,
+      "loss": 0.2904,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7184781347871476,
+      "grad_norm": 10.343472922864036,
+      "learning_rate": 2e-06,
+      "loss": 0.3471,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7187101264354483,
+      "grad_norm": 13.725070104455183,
+      "learning_rate": 2e-06,
+      "loss": 0.361,
+      "step": 3098
+    },
+    {
+      "epoch": 0.718942118083749,
+      "grad_norm": 16.651969325920845,
+      "learning_rate": 2e-06,
+      "loss": 0.3317,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7191741097320496,
+      "grad_norm": 9.55939511575531,
+      "learning_rate": 2e-06,
+      "loss": 0.3012,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7194061013803503,
+      "grad_norm": 15.3184086976544,
+      "learning_rate": 2e-06,
+      "loss": 0.2286,
+      "step": 3101
+    },
+    {
+      "epoch": 0.719638093028651,
+      "grad_norm": 12.566094977158508,
+      "learning_rate": 2e-06,
+      "loss": 0.2944,
+      "step": 3102
+    },
+    {
+      "epoch": 0.7198700846769517,
+      "grad_norm": 6.97831787840303,
+      "learning_rate": 2e-06,
+      "loss": 0.1917,
+      "step": 3103
+    },
+    {
+      "epoch": 0.7201020763252522,
+      "grad_norm": 16.899575716149464,
+      "learning_rate": 2e-06,
+      "loss": 0.2448,
+      "step": 3104
+    },
+    {
+      "epoch": 0.7203340679735529,
+      "grad_norm": 17.976110532970768,
+      "learning_rate": 2e-06,
+      "loss": 0.3163,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7205660596218536,
+      "grad_norm": 18.232864963298365,
+      "learning_rate": 2e-06,
+      "loss": 0.4656,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7207980512701543,
+      "grad_norm": 9.588755633507239,
+      "learning_rate": 2e-06,
+      "loss": 0.321,
+      "step": 3107
+    },
+    {
+      "epoch": 0.721030042918455,
+      "grad_norm": 9.184369414779091,
+      "learning_rate": 2e-06,
+      "loss": 0.185,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7212620345667556,
+      "grad_norm": 8.18229606602893,
+      "learning_rate": 2e-06,
+      "loss": 0.2636,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7214940262150562,
+      "grad_norm": 9.565619460509582,
+      "learning_rate": 2e-06,
+      "loss": 0.3021,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7217260178633569,
+      "grad_norm": 8.042176091017213,
+      "learning_rate": 2e-06,
+      "loss": 0.1867,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7219580095116576,
+      "grad_norm": 13.345881386328118,
+      "learning_rate": 2e-06,
+      "loss": 0.2084,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7221900011599582,
+      "grad_norm": 15.26049198288833,
+      "learning_rate": 2e-06,
+      "loss": 0.3182,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7224219928082589,
+      "grad_norm": 12.405214439820826,
+      "learning_rate": 2e-06,
+      "loss": 0.3215,
+      "step": 3114
+    },
+    {
+      "epoch": 0.7226539844565596,
+      "grad_norm": 8.279027687701165,
+      "learning_rate": 2e-06,
+      "loss": 0.2029,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7228859761048603,
+      "grad_norm": 9.23346549546795,
+      "learning_rate": 2e-06,
+      "loss": 0.2842,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7231179677531608,
+      "grad_norm": 13.90968760840073,
+      "learning_rate": 2e-06,
+      "loss": 0.2085,
+      "step": 3117
+    },
+    {
+      "epoch": 0.7233499594014615,
+      "grad_norm": 13.869798608630939,
+      "learning_rate": 2e-06,
+      "loss": 0.2016,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7235819510497622,
+      "grad_norm": 7.716027129845056,
+      "learning_rate": 2e-06,
+      "loss": 0.2188,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7238139426980629,
+      "grad_norm": 10.440026382109156,
+      "learning_rate": 2e-06,
+      "loss": 0.2281,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7240459343463636,
+      "grad_norm": 17.4379176105839,
+      "learning_rate": 2e-06,
+      "loss": 0.3783,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7242779259946642,
+      "grad_norm": 17.53811782802477,
+      "learning_rate": 2e-06,
+      "loss": 0.2852,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7245099176429649,
+      "grad_norm": 12.403149884465117,
+      "learning_rate": 2e-06,
+      "loss": 0.2562,
+      "step": 3123
+    },
+    {
+      "epoch": 0.7247419092912655,
+      "grad_norm": 7.189055185689566,
+      "learning_rate": 2e-06,
+      "loss": 0.2395,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7249739009395662,
+      "grad_norm": 13.366602934084645,
+      "learning_rate": 2e-06,
+      "loss": 0.3757,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7252058925878668,
+      "grad_norm": 20.723002926799648,
+      "learning_rate": 2e-06,
+      "loss": 0.3032,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7254378842361675,
+      "grad_norm": 14.345531287053802,
+      "learning_rate": 2e-06,
+      "loss": 0.2848,
+      "step": 3127
+    },
+    {
+      "epoch": 0.7256698758844682,
+      "grad_norm": 15.277830898750224,
+      "learning_rate": 2e-06,
+      "loss": 0.2698,
+      "step": 3128
+    },
+    {
+      "epoch": 0.7259018675327689,
+      "grad_norm": 14.343649187551526,
+      "learning_rate": 2e-06,
+      "loss": 0.2137,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7261338591810694,
+      "grad_norm": 11.927520687935251,
+      "learning_rate": 2e-06,
+      "loss": 0.2584,
+      "step": 3130
+    },
+    {
+      "epoch": 0.7263658508293701,
+      "grad_norm": 7.569851105062665,
+      "learning_rate": 2e-06,
+      "loss": 0.2569,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7265978424776708,
+      "grad_norm": 15.411236027620829,
+      "learning_rate": 2e-06,
+      "loss": 0.2988,
+      "step": 3132
+    },
+    {
+      "epoch": 0.7268298341259715,
+      "grad_norm": 17.315685638615545,
+      "learning_rate": 2e-06,
+      "loss": 0.3518,
+      "step": 3133
+    },
+    {
+      "epoch": 0.7270618257742721,
+      "grad_norm": 12.187758686052286,
+      "learning_rate": 2e-06,
+      "loss": 0.2202,
+      "step": 3134
+    },
+    {
+      "epoch": 0.7272938174225728,
+      "grad_norm": 9.516509125547941,
+      "learning_rate": 2e-06,
+      "loss": 0.2373,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7275258090708735,
+      "grad_norm": 13.633348693158915,
+      "learning_rate": 2e-06,
+      "loss": 0.244,
+      "step": 3136
+    },
+    {
+      "epoch": 0.7277578007191741,
+      "grad_norm": 18.169940194464743,
+      "learning_rate": 2e-06,
+      "loss": 0.295,
+      "step": 3137
+    },
+    {
+      "epoch": 0.7279897923674747,
+      "grad_norm": 16.600460469573456,
+      "learning_rate": 2e-06,
+      "loss": 0.236,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7282217840157754,
+      "grad_norm": 13.243093105767455,
+      "learning_rate": 2e-06,
+      "loss": 0.2635,
+      "step": 3139
+    },
+    {
+      "epoch": 0.7284537756640761,
+      "grad_norm": 18.433591061729665,
+      "learning_rate": 2e-06,
+      "loss": 0.3301,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7286857673123768,
+      "grad_norm": 9.66065462790196,
+      "learning_rate": 2e-06,
+      "loss": 0.2138,
+      "step": 3141
+    },
+    {
+      "epoch": 0.7289177589606775,
+      "grad_norm": 13.209312050697344,
+      "learning_rate": 2e-06,
+      "loss": 0.2658,
+      "step": 3142
+    },
+    {
+      "epoch": 0.7291497506089781,
+      "grad_norm": 19.082935008032695,
+      "learning_rate": 2e-06,
+      "loss": 0.291,
+      "step": 3143
+    },
+    {
+      "epoch": 0.7293817422572787,
+      "grad_norm": 7.89564746805161,
+      "learning_rate": 2e-06,
+      "loss": 0.223,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7296137339055794,
+      "grad_norm": 16.304437940242224,
+      "learning_rate": 2e-06,
+      "loss": 0.2943,
+      "step": 3145
+    },
+    {
+      "epoch": 0.72984572555388,
+      "grad_norm": 22.612646825330938,
+      "learning_rate": 2e-06,
+      "loss": 0.421,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7300777172021807,
+      "grad_norm": 10.784007437530533,
+      "learning_rate": 2e-06,
+      "loss": 0.2252,
+      "step": 3147
+    },
+    {
+      "epoch": 0.7303097088504814,
+      "grad_norm": 22.407086484125312,
+      "learning_rate": 2e-06,
+      "loss": 0.3782,
+      "step": 3148
+    },
+    {
+      "epoch": 0.7305417004987821,
+      "grad_norm": 22.865257080954436,
+      "learning_rate": 2e-06,
+      "loss": 0.3711,
+      "step": 3149
+    },
+    {
+      "epoch": 0.7307736921470827,
+      "grad_norm": 16.632205181733372,
+      "learning_rate": 2e-06,
+      "loss": 0.2656,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7310056837953833,
+      "grad_norm": 12.773540189237295,
+      "learning_rate": 2e-06,
+      "loss": 0.2561,
+      "step": 3151
+    },
+    {
+      "epoch": 0.731237675443684,
+      "grad_norm": 19.910063057592886,
+      "learning_rate": 2e-06,
+      "loss": 0.3787,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7314696670919847,
+      "grad_norm": 11.60355812504079,
+      "learning_rate": 2e-06,
+      "loss": 0.248,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7317016587402854,
+      "grad_norm": 11.24794631426084,
+      "learning_rate": 2e-06,
+      "loss": 0.2551,
+      "step": 3154
+    },
+    {
+      "epoch": 0.731933650388586,
+      "grad_norm": 17.127338422807483,
+      "learning_rate": 2e-06,
+      "loss": 0.3598,
+      "step": 3155
+    },
+    {
+      "epoch": 0.7321656420368867,
+      "grad_norm": 21.97871303967765,
+      "learning_rate": 2e-06,
+      "loss": 0.4009,
+      "step": 3156
+    },
+    {
+      "epoch": 0.7323976336851873,
+      "grad_norm": 15.615835043718105,
+      "learning_rate": 2e-06,
+      "loss": 0.3075,
+      "step": 3157
+    },
+    {
+      "epoch": 0.732629625333488,
+      "grad_norm": 17.522596850256367,
+      "learning_rate": 2e-06,
+      "loss": 0.2784,
+      "step": 3158
+    },
+    {
+      "epoch": 0.7328616169817886,
+      "grad_norm": 18.739187983179765,
+      "learning_rate": 2e-06,
+      "loss": 0.3305,
+      "step": 3159
+    },
+    {
+      "epoch": 0.7330936086300893,
+      "grad_norm": 6.660399942455831,
+      "learning_rate": 2e-06,
+      "loss": 0.2451,
+      "step": 3160
+    },
+    {
+      "epoch": 0.73332560027839,
+      "grad_norm": 15.690373574466072,
+      "learning_rate": 2e-06,
+      "loss": 0.2543,
+      "step": 3161
+    },
+    {
+      "epoch": 0.7335575919266907,
+      "grad_norm": 10.188035897846122,
+      "learning_rate": 2e-06,
+      "loss": 0.2615,
+      "step": 3162
+    },
+    {
+      "epoch": 0.7337895835749914,
+      "grad_norm": 23.18149272189231,
+      "learning_rate": 2e-06,
+      "loss": 0.2897,
+      "step": 3163
+    },
+    {
+      "epoch": 0.7340215752232919,
+      "grad_norm": 14.122180502228332,
+      "learning_rate": 2e-06,
+      "loss": 0.2234,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7342535668715926,
+      "grad_norm": 7.541936206004345,
+      "learning_rate": 2e-06,
+      "loss": 0.2265,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7344855585198933,
+      "grad_norm": 12.774313254828565,
+      "learning_rate": 2e-06,
+      "loss": 0.2341,
+      "step": 3166
+    },
+    {
+      "epoch": 0.734717550168194,
+      "grad_norm": 17.641684894723078,
+      "learning_rate": 2e-06,
+      "loss": 0.2906,
+      "step": 3167
+    },
+    {
+      "epoch": 0.7349495418164946,
+      "grad_norm": 19.12104028237049,
+      "learning_rate": 2e-06,
+      "loss": 0.3829,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7351815334647953,
+      "grad_norm": 22.529089396316817,
+      "learning_rate": 2e-06,
+      "loss": 0.2992,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7354135251130959,
+      "grad_norm": 19.771390540790247,
+      "learning_rate": 2e-06,
+      "loss": 0.276,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7356455167613966,
+      "grad_norm": 16.61134813442355,
+      "learning_rate": 2e-06,
+      "loss": 0.2635,
+      "step": 3171
+    },
+    {
+      "epoch": 0.7358775084096972,
+      "grad_norm": 18.759166483704387,
+      "learning_rate": 2e-06,
+      "loss": 0.3144,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7361095000579979,
+      "grad_norm": 10.461120170280791,
+      "learning_rate": 2e-06,
+      "loss": 0.326,
+      "step": 3173
+    },
+    {
+      "epoch": 0.7363414917062986,
+      "grad_norm": 11.874710306707069,
+      "learning_rate": 2e-06,
+      "loss": 0.3141,
+      "step": 3174
+    },
+    {
+      "epoch": 0.7365734833545993,
+      "grad_norm": 24.942525916485767,
+      "learning_rate": 2e-06,
+      "loss": 0.2926,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7368054750029,
+      "grad_norm": 15.934880410013454,
+      "learning_rate": 2e-06,
+      "loss": 0.3051,
+      "step": 3176
+    },
+    {
+      "epoch": 0.7370374666512005,
+      "grad_norm": 9.61878494372916,
+      "learning_rate": 2e-06,
+      "loss": 0.2273,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7372694582995012,
+      "grad_norm": 17.32211608727347,
+      "learning_rate": 2e-06,
+      "loss": 0.2992,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7375014499478019,
+      "grad_norm": 19.998321939059572,
+      "learning_rate": 2e-06,
+      "loss": 0.1923,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7377334415961025,
+      "grad_norm": 10.703880883805544,
+      "learning_rate": 2e-06,
+      "loss": 0.2353,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7379654332444032,
+      "grad_norm": 19.34032775586904,
+      "learning_rate": 2e-06,
+      "loss": 0.3466,
+      "step": 3181
+    },
+    {
+      "epoch": 0.7381974248927039,
+      "grad_norm": 9.371863932024695,
+      "learning_rate": 2e-06,
+      "loss": 0.2022,
+      "step": 3182
+    },
+    {
+      "epoch": 0.7384294165410045,
+      "grad_norm": 12.088369944261673,
+      "learning_rate": 2e-06,
+      "loss": 0.3356,
+      "step": 3183
+    },
+    {
+      "epoch": 0.7386614081893051,
+      "grad_norm": 13.208820592992916,
+      "learning_rate": 2e-06,
+      "loss": 0.3584,
+      "step": 3184
+    },
+    {
+      "epoch": 0.7388933998376058,
+      "grad_norm": 14.48019733506012,
+      "learning_rate": 2e-06,
+      "loss": 0.2866,
+      "step": 3185
+    },
+    {
+      "epoch": 0.7391253914859065,
+      "grad_norm": 9.297669852631856,
+      "learning_rate": 2e-06,
+      "loss": 0.285,
+      "step": 3186
+    },
+    {
+      "epoch": 0.7393573831342072,
+      "grad_norm": 13.59255319093113,
+      "learning_rate": 2e-06,
+      "loss": 0.2217,
+      "step": 3187
+    },
+    {
+      "epoch": 0.7395893747825079,
+      "grad_norm": 9.63516912508189,
+      "learning_rate": 2e-06,
+      "loss": 0.2913,
+      "step": 3188
+    },
+    {
+      "epoch": 0.7398213664308085,
+      "grad_norm": 12.207878815290089,
+      "learning_rate": 2e-06,
+      "loss": 0.1737,
+      "step": 3189
+    },
+    {
+      "epoch": 0.7400533580791091,
+      "grad_norm": 7.723048500107733,
+      "learning_rate": 2e-06,
+      "loss": 0.1719,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7402853497274098,
+      "grad_norm": 15.372306937561344,
+      "learning_rate": 2e-06,
+      "loss": 0.3438,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7405173413757105,
+      "grad_norm": 10.693950673258227,
+      "learning_rate": 2e-06,
+      "loss": 0.2514,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7407493330240111,
+      "grad_norm": 14.196738573156889,
+      "learning_rate": 2e-06,
+      "loss": 0.3165,
+      "step": 3193
+    },
+    {
+      "epoch": 0.7409813246723118,
+      "grad_norm": 266.23074189192454,
+      "learning_rate": 2e-06,
+      "loss": 0.3074,
+      "step": 3194
+    },
+    {
+      "epoch": 0.7412133163206125,
+      "grad_norm": 11.658793115573301,
+      "learning_rate": 2e-06,
+      "loss": 0.2975,
+      "step": 3195
+    },
+    {
+      "epoch": 0.7414453079689132,
+      "grad_norm": 18.86350667547308,
+      "learning_rate": 2e-06,
+      "loss": 0.4451,
+      "step": 3196
+    },
+    {
+      "epoch": 0.7416772996172137,
+      "grad_norm": 10.675248689307919,
+      "learning_rate": 2e-06,
+      "loss": 0.2638,
+      "step": 3197
+    },
+    {
+      "epoch": 0.7419092912655144,
+      "grad_norm": 10.46535396830821,
+      "learning_rate": 2e-06,
+      "loss": 0.1983,
+      "step": 3198
+    },
+    {
+      "epoch": 0.7421412829138151,
+      "grad_norm": 14.436103478643547,
+      "learning_rate": 2e-06,
+      "loss": 0.2665,
+      "step": 3199
+    },
+    {
+      "epoch": 0.7423732745621158,
+      "grad_norm": 18.35508453929755,
+      "learning_rate": 2e-06,
+      "loss": 0.3326,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7426052662104164,
+      "grad_norm": 13.92213201139075,
+      "learning_rate": 2e-06,
+      "loss": 0.3001,
+      "step": 3201
+    },
+    {
+      "epoch": 0.7428372578587171,
+      "grad_norm": 24.057651507458285,
+      "learning_rate": 2e-06,
+      "loss": 0.4314,
+      "step": 3202
+    },
+    {
+      "epoch": 0.7430692495070177,
+      "grad_norm": 13.864093476181436,
+      "learning_rate": 2e-06,
+      "loss": 0.2738,
+      "step": 3203
+    },
+    {
+      "epoch": 0.7433012411553184,
+      "grad_norm": 12.054702304883937,
+      "learning_rate": 2e-06,
+      "loss": 0.3038,
+      "step": 3204
+    },
+    {
+      "epoch": 0.743533232803619,
+      "grad_norm": 12.60991002298068,
+      "learning_rate": 2e-06,
+      "loss": 0.196,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7437652244519197,
+      "grad_norm": 12.127707235247618,
+      "learning_rate": 2e-06,
+      "loss": 0.1946,
+      "step": 3206
+    },
+    {
+      "epoch": 0.7439972161002204,
+      "grad_norm": 23.30100209188892,
+      "learning_rate": 2e-06,
+      "loss": 0.273,
+      "step": 3207
+    },
+    {
+      "epoch": 0.7442292077485211,
+      "grad_norm": 13.99161322820209,
+      "learning_rate": 2e-06,
+      "loss": 0.2144,
+      "step": 3208
+    },
+    {
+      "epoch": 0.7444611993968218,
+      "grad_norm": 13.887700930199868,
+      "learning_rate": 2e-06,
+      "loss": 0.2683,
+      "step": 3209
+    },
+    {
+      "epoch": 0.7446931910451223,
+      "grad_norm": 26.819487750849106,
+      "learning_rate": 2e-06,
+      "loss": 0.4258,
+      "step": 3210
+    },
+    {
+      "epoch": 0.744925182693423,
+      "grad_norm": 9.733351998525075,
+      "learning_rate": 2e-06,
+      "loss": 0.2155,
+      "step": 3211
+    },
+    {
+      "epoch": 0.7451571743417237,
+      "grad_norm": 14.340214523192815,
+      "learning_rate": 2e-06,
+      "loss": 0.2755,
+      "step": 3212
+    },
+    {
+      "epoch": 0.7453891659900244,
+      "grad_norm": 11.54245757950112,
+      "learning_rate": 2e-06,
+      "loss": 0.3347,
+      "step": 3213
+    },
+    {
+      "epoch": 0.745621157638325,
+      "grad_norm": 23.06041982558424,
+      "learning_rate": 2e-06,
+      "loss": 0.3852,
+      "step": 3214
+    },
+    {
+      "epoch": 0.7458531492866257,
+      "grad_norm": 15.579761828443129,
+      "learning_rate": 2e-06,
+      "loss": 0.1818,
+      "step": 3215
+    },
+    {
+      "epoch": 0.7460851409349264,
+      "grad_norm": 27.921307898526475,
+      "learning_rate": 2e-06,
+      "loss": 0.3919,
+      "step": 3216
+    },
+    {
+      "epoch": 0.746317132583227,
+      "grad_norm": 10.451501045988298,
+      "learning_rate": 2e-06,
+      "loss": 0.2935,
+      "step": 3217
+    },
+    {
+      "epoch": 0.7465491242315276,
+      "grad_norm": 18.411670307096102,
+      "learning_rate": 2e-06,
+      "loss": 0.3829,
+      "step": 3218
+    },
+    {
+      "epoch": 0.7467811158798283,
+      "grad_norm": 17.096497754766254,
+      "learning_rate": 2e-06,
+      "loss": 0.2315,
+      "step": 3219
+    },
+    {
+      "epoch": 0.747013107528129,
+      "grad_norm": 19.05129410304447,
+      "learning_rate": 2e-06,
+      "loss": 0.3125,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7472450991764297,
+      "grad_norm": 7.424645567567219,
+      "learning_rate": 2e-06,
+      "loss": 0.2718,
+      "step": 3221
+    },
+    {
+      "epoch": 0.7474770908247304,
+      "grad_norm": 11.603587562641335,
+      "learning_rate": 2e-06,
+      "loss": 0.2976,
+      "step": 3222
+    },
+    {
+      "epoch": 0.7477090824730309,
+      "grad_norm": 9.346124673462597,
+      "learning_rate": 2e-06,
+      "loss": 0.2669,
+      "step": 3223
+    },
+    {
+      "epoch": 0.7479410741213316,
+      "grad_norm": 13.97160979948989,
+      "learning_rate": 2e-06,
+      "loss": 0.2398,
+      "step": 3224
+    },
+    {
+      "epoch": 0.7481730657696323,
+      "grad_norm": 25.922000700573506,
+      "learning_rate": 2e-06,
+      "loss": 0.3631,
+      "step": 3225
+    },
+    {
+      "epoch": 0.748405057417933,
+      "grad_norm": 10.308629661553015,
+      "learning_rate": 2e-06,
+      "loss": 0.2849,
+      "step": 3226
+    },
+    {
+      "epoch": 0.7486370490662336,
+      "grad_norm": 16.028103541976417,
+      "learning_rate": 2e-06,
+      "loss": 0.2732,
+      "step": 3227
+    },
+    {
+      "epoch": 0.7488690407145343,
+      "grad_norm": 11.867317699477844,
+      "learning_rate": 2e-06,
+      "loss": 0.3506,
+      "step": 3228
+    },
+    {
+      "epoch": 0.749101032362835,
+      "grad_norm": 27.591524237265336,
+      "learning_rate": 2e-06,
+      "loss": 0.4672,
+      "step": 3229
+    },
+    {
+      "epoch": 0.7493330240111356,
+      "grad_norm": 16.75662670018917,
+      "learning_rate": 2e-06,
+      "loss": 0.3969,
+      "step": 3230
+    },
+    {
+      "epoch": 0.7495650156594362,
+      "grad_norm": 21.443112852719207,
+      "learning_rate": 2e-06,
+      "loss": 0.3879,
+      "step": 3231
+    },
+    {
+      "epoch": 0.7497970073077369,
+      "grad_norm": 18.57519074076872,
+      "learning_rate": 2e-06,
+      "loss": 0.3253,
+      "step": 3232
+    },
+    {
+      "epoch": 0.7500289989560376,
+      "grad_norm": 9.228484999772808,
+      "learning_rate": 2e-06,
+      "loss": 0.2678,
+      "step": 3233
+    },
+    {
+      "epoch": 0.7502609906043383,
+      "grad_norm": 20.858727268557686,
+      "learning_rate": 2e-06,
+      "loss": 0.2909,
+      "step": 3234
+    },
+    {
+      "epoch": 0.7504929822526389,
+      "grad_norm": 26.914560888617398,
+      "learning_rate": 2e-06,
+      "loss": 0.3719,
+      "step": 3235
+    },
+    {
+      "epoch": 0.7507249739009396,
+      "grad_norm": 15.709843888605175,
+      "learning_rate": 2e-06,
+      "loss": 0.2176,
+      "step": 3236
+    },
+    {
+      "epoch": 0.7509569655492402,
+      "grad_norm": 19.16131357159073,
+      "learning_rate": 2e-06,
+      "loss": 0.2419,
+      "step": 3237
+    },
+    {
+      "epoch": 0.7511889571975409,
+      "grad_norm": 9.892834755241338,
+      "learning_rate": 2e-06,
+      "loss": 0.2641,
+      "step": 3238
+    },
+    {
+      "epoch": 0.7514209488458415,
+      "grad_norm": 12.710909928586199,
+      "learning_rate": 2e-06,
+      "loss": 0.3617,
+      "step": 3239
+    },
+    {
+      "epoch": 0.7516529404941422,
+      "grad_norm": 14.861685648929491,
+      "learning_rate": 2e-06,
+      "loss": 0.2903,
+      "step": 3240
+    },
+    {
+      "epoch": 0.7518849321424429,
+      "grad_norm": 10.636791599844129,
+      "learning_rate": 2e-06,
+      "loss": 0.2154,
+      "step": 3241
+    },
+    {
+      "epoch": 0.7521169237907436,
+      "grad_norm": 10.824860251051467,
+      "learning_rate": 2e-06,
+      "loss": 0.3095,
+      "step": 3242
+    },
+    {
+      "epoch": 0.7523489154390441,
+      "grad_norm": 11.258032137291424,
+      "learning_rate": 2e-06,
+      "loss": 0.2784,
+      "step": 3243
+    },
+    {
+      "epoch": 0.7525809070873448,
+      "grad_norm": 12.216580409438913,
+      "learning_rate": 2e-06,
+      "loss": 0.3381,
+      "step": 3244
+    },
+    {
+      "epoch": 0.7528128987356455,
+      "grad_norm": 16.46023841396873,
+      "learning_rate": 2e-06,
+      "loss": 0.2966,
+      "step": 3245
+    },
+    {
+      "epoch": 0.7530448903839462,
+      "grad_norm": 10.773669247612482,
+      "learning_rate": 2e-06,
+      "loss": 0.2822,
+      "step": 3246
+    },
+    {
+      "epoch": 0.7532768820322469,
+      "grad_norm": 15.47427621377556,
+      "learning_rate": 2e-06,
+      "loss": 0.3528,
+      "step": 3247
+    },
+    {
+      "epoch": 0.7535088736805475,
+      "grad_norm": 9.437936233561318,
+      "learning_rate": 2e-06,
+      "loss": 0.3515,
+      "step": 3248
+    },
+    {
+      "epoch": 0.7537408653288482,
+      "grad_norm": 15.060815916367174,
+      "learning_rate": 2e-06,
+      "loss": 0.369,
+      "step": 3249
+    },
+    {
+      "epoch": 0.7539728569771488,
+      "grad_norm": 10.051667383454044,
+      "learning_rate": 2e-06,
+      "loss": 0.2607,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7542048486254495,
+      "grad_norm": 15.210897124764472,
+      "learning_rate": 2e-06,
+      "loss": 0.3493,
+      "step": 3251
+    },
+    {
+      "epoch": 0.7544368402737501,
+      "grad_norm": 16.097661441306023,
+      "learning_rate": 2e-06,
+      "loss": 0.2453,
+      "step": 3252
+    },
+    {
+      "epoch": 0.7546688319220508,
+      "grad_norm": 9.117636788720601,
+      "learning_rate": 2e-06,
+      "loss": 0.2432,
+      "step": 3253
+    },
+    {
+      "epoch": 0.7549008235703515,
+      "grad_norm": 9.98191377832381,
+      "learning_rate": 2e-06,
+      "loss": 0.2017,
+      "step": 3254
+    },
+    {
+      "epoch": 0.7551328152186522,
+      "grad_norm": 7.908774046591722,
+      "learning_rate": 2e-06,
+      "loss": 0.2378,
+      "step": 3255
+    },
+    {
+      "epoch": 0.7553648068669528,
+      "grad_norm": 23.628388777108437,
+      "learning_rate": 2e-06,
+      "loss": 0.3264,
+      "step": 3256
+    },
+    {
+      "epoch": 0.7555967985152534,
+      "grad_norm": 13.1515153464434,
+      "learning_rate": 2e-06,
+      "loss": 0.28,
+      "step": 3257
+    },
+    {
+      "epoch": 0.7558287901635541,
+      "grad_norm": 16.609710434365738,
+      "learning_rate": 2e-06,
+      "loss": 0.2632,
+      "step": 3258
+    },
+    {
+      "epoch": 0.7560607818118548,
+      "grad_norm": 19.25268556355,
+      "learning_rate": 2e-06,
+      "loss": 0.2485,
+      "step": 3259
+    },
+    {
+      "epoch": 0.7562927734601554,
+      "grad_norm": 8.798902479176105,
+      "learning_rate": 2e-06,
+      "loss": 0.2291,
+      "step": 3260
+    },
+    {
+      "epoch": 0.7565247651084561,
+      "grad_norm": 15.102364203693176,
+      "learning_rate": 2e-06,
+      "loss": 0.3814,
+      "step": 3261
+    },
+    {
+      "epoch": 0.7567567567567568,
+      "grad_norm": 12.251104054670868,
+      "learning_rate": 2e-06,
+      "loss": 0.205,
+      "step": 3262
+    },
+    {
+      "epoch": 0.7569887484050574,
+      "grad_norm": 17.830770306606684,
+      "learning_rate": 2e-06,
+      "loss": 0.3803,
+      "step": 3263
+    },
+    {
+      "epoch": 0.757220740053358,
+      "grad_norm": 9.898043465896796,
+      "learning_rate": 2e-06,
+      "loss": 0.2717,
+      "step": 3264
+    },
+    {
+      "epoch": 0.7574527317016587,
+      "grad_norm": 10.869198655689024,
+      "learning_rate": 2e-06,
+      "loss": 0.2438,
+      "step": 3265
+    },
+    {
+      "epoch": 0.7576847233499594,
+      "grad_norm": 9.569985886548965,
+      "learning_rate": 2e-06,
+      "loss": 0.2298,
+      "step": 3266
+    },
+    {
+      "epoch": 0.7579167149982601,
+      "grad_norm": 8.895433458214619,
+      "learning_rate": 2e-06,
+      "loss": 0.2976,
+      "step": 3267
+    },
+    {
+      "epoch": 0.7581487066465608,
+      "grad_norm": 10.838387016596009,
+      "learning_rate": 2e-06,
+      "loss": 0.1739,
+      "step": 3268
+    },
+    {
+      "epoch": 0.7583806982948614,
+      "grad_norm": 15.143001781351865,
+      "learning_rate": 2e-06,
+      "loss": 0.2297,
+      "step": 3269
+    },
+    {
+      "epoch": 0.758612689943162,
+      "grad_norm": 15.826963259618264,
+      "learning_rate": 2e-06,
+      "loss": 0.2124,
+      "step": 3270
+    },
+    {
+      "epoch": 0.7588446815914627,
+      "grad_norm": 18.73429903645318,
+      "learning_rate": 2e-06,
+      "loss": 0.3329,
+      "step": 3271
+    },
+    {
+      "epoch": 0.7590766732397634,
+      "grad_norm": 8.054226271696935,
+      "learning_rate": 2e-06,
+      "loss": 0.2594,
+      "step": 3272
+    },
+    {
+      "epoch": 0.759308664888064,
+      "grad_norm": 18.537441641456375,
+      "learning_rate": 2e-06,
+      "loss": 0.4262,
+      "step": 3273
+    },
+    {
+      "epoch": 0.7595406565363647,
+      "grad_norm": 18.060962169878092,
+      "learning_rate": 2e-06,
+      "loss": 0.329,
+      "step": 3274
+    },
+    {
+      "epoch": 0.7597726481846654,
+      "grad_norm": 19.176654897756123,
+      "learning_rate": 2e-06,
+      "loss": 0.2127,
+      "step": 3275
+    },
+    {
+      "epoch": 0.7600046398329661,
+      "grad_norm": 40.19397033267575,
+      "learning_rate": 2e-06,
+      "loss": 0.4408,
+      "step": 3276
+    },
+    {
+      "epoch": 0.7602366314812666,
+      "grad_norm": 14.183460089274515,
+      "learning_rate": 2e-06,
+      "loss": 0.2787,
+      "step": 3277
+    },
+    {
+      "epoch": 0.7604686231295673,
+      "grad_norm": 12.614565362170005,
+      "learning_rate": 2e-06,
+      "loss": 0.2865,
+      "step": 3278
+    },
+    {
+      "epoch": 0.760700614777868,
+      "grad_norm": 15.294760785557783,
+      "learning_rate": 2e-06,
+      "loss": 0.3313,
+      "step": 3279
+    },
+    {
+      "epoch": 0.7609326064261687,
+      "grad_norm": 14.678094198469996,
+      "learning_rate": 2e-06,
+      "loss": 0.2752,
+      "step": 3280
+    },
+    {
+      "epoch": 0.7611645980744693,
+      "grad_norm": 16.512603314388123,
+      "learning_rate": 2e-06,
+      "loss": 0.2736,
+      "step": 3281
+    },
+    {
+      "epoch": 0.76139658972277,
+      "grad_norm": 21.798372206513797,
+      "learning_rate": 2e-06,
+      "loss": 0.3324,
+      "step": 3282
+    },
+    {
+      "epoch": 0.7616285813710706,
+      "grad_norm": 13.556144644099309,
+      "learning_rate": 2e-06,
+      "loss": 0.313,
+      "step": 3283
+    },
+    {
+      "epoch": 0.7618605730193713,
+      "grad_norm": 12.0913929886167,
+      "learning_rate": 2e-06,
+      "loss": 0.2542,
+      "step": 3284
+    },
+    {
+      "epoch": 0.762092564667672,
+      "grad_norm": 8.532666772364651,
+      "learning_rate": 2e-06,
+      "loss": 0.2037,
+      "step": 3285
+    },
+    {
+      "epoch": 0.7623245563159726,
+      "grad_norm": 15.608666752534837,
+      "learning_rate": 2e-06,
+      "loss": 0.277,
+      "step": 3286
+    },
+    {
+      "epoch": 0.7625565479642733,
+      "grad_norm": 12.648901474399997,
+      "learning_rate": 2e-06,
+      "loss": 0.403,
+      "step": 3287
+    },
+    {
+      "epoch": 0.762788539612574,
+      "grad_norm": 15.6219832778702,
+      "learning_rate": 2e-06,
+      "loss": 0.3407,
+      "step": 3288
+    },
+    {
+      "epoch": 0.7630205312608747,
+      "grad_norm": 25.88285292749361,
+      "learning_rate": 2e-06,
+      "loss": 0.4021,
+      "step": 3289
+    },
+    {
+      "epoch": 0.7632525229091752,
+      "grad_norm": 14.51290497078572,
+      "learning_rate": 2e-06,
+      "loss": 0.3171,
+      "step": 3290
+    },
+    {
+      "epoch": 0.7634845145574759,
+      "grad_norm": 15.373419614063504,
+      "learning_rate": 2e-06,
+      "loss": 0.3018,
+      "step": 3291
+    },
+    {
+      "epoch": 0.7637165062057766,
+      "grad_norm": 7.465941023334532,
+      "learning_rate": 2e-06,
+      "loss": 0.1827,
+      "step": 3292
+    },
+    {
+      "epoch": 0.7639484978540773,
+      "grad_norm": 22.68478128905687,
+      "learning_rate": 2e-06,
+      "loss": 0.3353,
+      "step": 3293
+    },
+    {
+      "epoch": 0.7641804895023779,
+      "grad_norm": 12.719013375955988,
+      "learning_rate": 2e-06,
+      "loss": 0.2751,
+      "step": 3294
+    },
+    {
+      "epoch": 0.7644124811506786,
+      "grad_norm": 12.201266141777696,
+      "learning_rate": 2e-06,
+      "loss": 0.2502,
+      "step": 3295
+    },
+    {
+      "epoch": 0.7646444727989793,
+      "grad_norm": 12.401332055108211,
+      "learning_rate": 2e-06,
+      "loss": 0.3048,
+      "step": 3296
+    },
+    {
+      "epoch": 0.7648764644472799,
+      "grad_norm": 9.101908144365112,
+      "learning_rate": 2e-06,
+      "loss": 0.2507,
+      "step": 3297
+    },
+    {
+      "epoch": 0.7651084560955805,
+      "grad_norm": 15.446972836287449,
+      "learning_rate": 2e-06,
+      "loss": 0.3506,
+      "step": 3298
+    },
+    {
+      "epoch": 0.7653404477438812,
+      "grad_norm": 10.270501438242315,
+      "learning_rate": 2e-06,
+      "loss": 0.2518,
+      "step": 3299
+    },
+    {
+      "epoch": 0.7655724393921819,
+      "grad_norm": 10.880237503429663,
+      "learning_rate": 2e-06,
+      "loss": 0.2826,
+      "step": 3300
+    },
+    {
+      "epoch": 0.7658044310404826,
+      "grad_norm": 10.411867222827498,
+      "learning_rate": 2e-06,
+      "loss": 0.2021,
+      "step": 3301
+    },
+    {
+      "epoch": 0.7660364226887832,
+      "grad_norm": 6.641121711480695,
+      "learning_rate": 2e-06,
+      "loss": 0.1922,
+      "step": 3302
+    },
+    {
+      "epoch": 0.7662684143370838,
+      "grad_norm": 16.986816691328375,
+      "learning_rate": 2e-06,
+      "loss": 0.3354,
+      "step": 3303
+    },
+    {
+      "epoch": 0.7665004059853845,
+      "grad_norm": 7.048691651075567,
+      "learning_rate": 2e-06,
+      "loss": 0.227,
+      "step": 3304
+    },
+    {
+      "epoch": 0.7667323976336852,
+      "grad_norm": 10.118835944738485,
+      "learning_rate": 2e-06,
+      "loss": 0.1769,
+      "step": 3305
+    },
+    {
+      "epoch": 0.7669643892819858,
+      "grad_norm": 20.612598747313648,
+      "learning_rate": 2e-06,
+      "loss": 0.3865,
+      "step": 3306
+    },
+    {
+      "epoch": 0.7671963809302865,
+      "grad_norm": 11.739597678261555,
+      "learning_rate": 2e-06,
+      "loss": 0.1937,
+      "step": 3307
+    },
+    {
+      "epoch": 0.7674283725785872,
+      "grad_norm": 13.091498954890323,
+      "learning_rate": 2e-06,
+      "loss": 0.3293,
+      "step": 3308
+    },
+    {
+      "epoch": 0.7676603642268879,
+      "grad_norm": 12.450404437779614,
+      "learning_rate": 2e-06,
+      "loss": 0.2431,
+      "step": 3309
+    },
+    {
+      "epoch": 0.7678923558751884,
+      "grad_norm": 12.381029099949977,
+      "learning_rate": 2e-06,
+      "loss": 0.3628,
+      "step": 3310
+    },
+    {
+      "epoch": 0.7681243475234891,
+      "grad_norm": 12.71570898337696,
+      "learning_rate": 2e-06,
+      "loss": 0.1727,
+      "step": 3311
+    },
+    {
+      "epoch": 0.7683563391717898,
+      "grad_norm": 19.630121385148918,
+      "learning_rate": 2e-06,
+      "loss": 0.3437,
+      "step": 3312
+    },
+    {
+      "epoch": 0.7685883308200905,
+      "grad_norm": 15.42711658110915,
+      "learning_rate": 2e-06,
+      "loss": 0.2281,
+      "step": 3313
+    },
+    {
+      "epoch": 0.7688203224683912,
+      "grad_norm": 6.885955436112426,
+      "learning_rate": 2e-06,
+      "loss": 0.1984,
+      "step": 3314
+    },
+    {
+      "epoch": 0.7690523141166918,
+      "grad_norm": 20.38462034150911,
+      "learning_rate": 2e-06,
+      "loss": 0.3107,
+      "step": 3315
+    },
+    {
+      "epoch": 0.7692843057649925,
+      "grad_norm": 14.158397888232825,
+      "learning_rate": 2e-06,
+      "loss": 0.2736,
+      "step": 3316
+    },
+    {
+      "epoch": 0.7695162974132931,
+      "grad_norm": 18.52903730581829,
+      "learning_rate": 2e-06,
+      "loss": 0.2503,
+      "step": 3317
+    },
+    {
+      "epoch": 0.7697482890615938,
+      "grad_norm": 24.06934932448499,
+      "learning_rate": 2e-06,
+      "loss": 0.3461,
+      "step": 3318
+    },
+    {
+      "epoch": 0.7699802807098944,
+      "grad_norm": 11.039495942718224,
+      "learning_rate": 2e-06,
+      "loss": 0.3537,
+      "step": 3319
+    },
+    {
+      "epoch": 0.7702122723581951,
+      "grad_norm": 17.35805561822558,
+      "learning_rate": 2e-06,
+      "loss": 0.3575,
+      "step": 3320
+    },
+    {
+      "epoch": 0.7704442640064958,
+      "grad_norm": 14.312621737355462,
+      "learning_rate": 2e-06,
+      "loss": 0.3117,
+      "step": 3321
+    },
+    {
+      "epoch": 0.7706762556547965,
+      "grad_norm": 14.30444248042403,
+      "learning_rate": 2e-06,
+      "loss": 0.3536,
+      "step": 3322
+    },
+    {
+      "epoch": 0.770908247303097,
+      "grad_norm": 21.15912940860675,
+      "learning_rate": 2e-06,
+      "loss": 0.433,
+      "step": 3323
+    },
+    {
+      "epoch": 0.7711402389513977,
+      "grad_norm": 12.043006519589039,
+      "learning_rate": 2e-06,
+      "loss": 0.2539,
+      "step": 3324
+    },
+    {
+      "epoch": 0.7713722305996984,
+      "grad_norm": 8.225323237276257,
+      "learning_rate": 2e-06,
+      "loss": 0.1483,
+      "step": 3325
+    },
+    {
+      "epoch": 0.7716042222479991,
+      "grad_norm": 12.721679821297943,
+      "learning_rate": 2e-06,
+      "loss": 0.266,
+      "step": 3326
+    },
+    {
+      "epoch": 0.7718362138962997,
+      "grad_norm": 18.161270535697973,
+      "learning_rate": 2e-06,
+      "loss": 0.2539,
+      "step": 3327
+    },
+    {
+      "epoch": 0.7720682055446004,
+      "grad_norm": 16.406188851303185,
+      "learning_rate": 2e-06,
+      "loss": 0.275,
+      "step": 3328
+    },
+    {
+      "epoch": 0.7723001971929011,
+      "grad_norm": 12.093543887689135,
+      "learning_rate": 2e-06,
+      "loss": 0.2418,
+      "step": 3329
+    },
+    {
+      "epoch": 0.7725321888412017,
+      "grad_norm": 17.715551132036744,
+      "learning_rate": 2e-06,
+      "loss": 0.3897,
+      "step": 3330
+    },
+    {
+      "epoch": 0.7727641804895024,
+      "grad_norm": 12.83841974804794,
+      "learning_rate": 2e-06,
+      "loss": 0.3194,
+      "step": 3331
+    },
+    {
+      "epoch": 0.772996172137803,
+      "grad_norm": 10.728654064148865,
+      "learning_rate": 2e-06,
+      "loss": 0.2545,
+      "step": 3332
+    },
+    {
+      "epoch": 0.7732281637861037,
+      "grad_norm": 10.098533183865884,
+      "learning_rate": 2e-06,
+      "loss": 0.354,
+      "step": 3333
+    },
+    {
+      "epoch": 0.7734601554344044,
+      "grad_norm": 15.94271340760764,
+      "learning_rate": 2e-06,
+      "loss": 0.2849,
+      "step": 3334
+    },
+    {
+      "epoch": 0.7736921470827051,
+      "grad_norm": 15.753903381136727,
+      "learning_rate": 2e-06,
+      "loss": 0.3928,
+      "step": 3335
+    },
+    {
+      "epoch": 0.7739241387310056,
+      "grad_norm": 13.636682401677106,
+      "learning_rate": 2e-06,
+      "loss": 0.2834,
+      "step": 3336
+    },
+    {
+      "epoch": 0.7741561303793063,
+      "grad_norm": 21.367904185410392,
+      "learning_rate": 2e-06,
+      "loss": 0.4015,
+      "step": 3337
+    },
+    {
+      "epoch": 0.774388122027607,
+      "grad_norm": 15.817210585659566,
+      "learning_rate": 2e-06,
+      "loss": 0.3041,
+      "step": 3338
+    },
+    {
+      "epoch": 0.7746201136759077,
+      "grad_norm": 10.298483366955015,
+      "learning_rate": 2e-06,
+      "loss": 0.2287,
+      "step": 3339
+    },
+    {
+      "epoch": 0.7748521053242083,
+      "grad_norm": 19.063689770985306,
+      "learning_rate": 2e-06,
+      "loss": 0.4006,
+      "step": 3340
+    },
+    {
+      "epoch": 0.775084096972509,
+      "grad_norm": 15.485253568225815,
+      "learning_rate": 2e-06,
+      "loss": 0.3251,
+      "step": 3341
+    },
+    {
+      "epoch": 0.7753160886208097,
+      "grad_norm": 8.78843640841125,
+      "learning_rate": 2e-06,
+      "loss": 0.4089,
+      "step": 3342
+    },
+    {
+      "epoch": 0.7755480802691103,
+      "grad_norm": 10.77846730692596,
+      "learning_rate": 2e-06,
+      "loss": 0.2857,
+      "step": 3343
+    },
+    {
+      "epoch": 0.7757800719174109,
+      "grad_norm": 13.253812636990132,
+      "learning_rate": 2e-06,
+      "loss": 0.1985,
+      "step": 3344
+    },
+    {
+      "epoch": 0.7760120635657116,
+      "grad_norm": 10.455512361721096,
+      "learning_rate": 2e-06,
+      "loss": 0.3522,
+      "step": 3345
+    },
+    {
+      "epoch": 0.7762440552140123,
+      "grad_norm": 20.698183046351787,
+      "learning_rate": 2e-06,
+      "loss": 0.3496,
+      "step": 3346
+    },
+    {
+      "epoch": 0.776476046862313,
+      "grad_norm": 7.3385363802913535,
+      "learning_rate": 2e-06,
+      "loss": 0.2692,
+      "step": 3347
+    },
+    {
+      "epoch": 0.7767080385106137,
+      "grad_norm": 14.080623288768594,
+      "learning_rate": 2e-06,
+      "loss": 0.2952,
+      "step": 3348
+    },
+    {
+      "epoch": 0.7769400301589143,
+      "grad_norm": 7.9727325221498155,
+      "learning_rate": 2e-06,
+      "loss": 0.2831,
+      "step": 3349
+    },
+    {
+      "epoch": 0.7771720218072149,
+      "grad_norm": 19.27180239056616,
+      "learning_rate": 2e-06,
+      "loss": 0.2993,
+      "step": 3350
+    },
+    {
+      "epoch": 0.7774040134555156,
+      "grad_norm": 17.39930094442428,
+      "learning_rate": 2e-06,
+      "loss": 0.4456,
+      "step": 3351
+    },
+    {
+      "epoch": 0.7776360051038163,
+      "grad_norm": 8.830695174435625,
+      "learning_rate": 2e-06,
+      "loss": 0.2363,
+      "step": 3352
+    },
+    {
+      "epoch": 0.7778679967521169,
+      "grad_norm": 12.427463571267143,
+      "learning_rate": 2e-06,
+      "loss": 0.2699,
+      "step": 3353
+    },
+    {
+      "epoch": 0.7780999884004176,
+      "grad_norm": 14.715522465947913,
+      "learning_rate": 2e-06,
+      "loss": 0.2928,
+      "step": 3354
+    },
+    {
+      "epoch": 0.7783319800487183,
+      "grad_norm": 9.18348706874619,
+      "learning_rate": 2e-06,
+      "loss": 0.333,
+      "step": 3355
+    },
+    {
+      "epoch": 0.7785639716970189,
+      "grad_norm": 11.394303641825907,
+      "learning_rate": 2e-06,
+      "loss": 0.3114,
+      "step": 3356
+    },
+    {
+      "epoch": 0.7787959633453195,
+      "grad_norm": 8.11159833896116,
+      "learning_rate": 2e-06,
+      "loss": 0.2845,
+      "step": 3357
+    },
+    {
+      "epoch": 0.7790279549936202,
+      "grad_norm": 21.21505875288533,
+      "learning_rate": 2e-06,
+      "loss": 0.2917,
+      "step": 3358
+    },
+    {
+      "epoch": 0.7792599466419209,
+      "grad_norm": 14.613621762182552,
+      "learning_rate": 2e-06,
+      "loss": 0.2917,
+      "step": 3359
+    },
+    {
+      "epoch": 0.7794919382902216,
+      "grad_norm": 8.027194232523648,
+      "learning_rate": 2e-06,
+      "loss": 0.1708,
+      "step": 3360
+    },
+    {
+      "epoch": 0.7797239299385222,
+      "grad_norm": 15.20719004188822,
+      "learning_rate": 2e-06,
+      "loss": 0.3193,
+      "step": 3361
+    },
+    {
+      "epoch": 0.7799559215868229,
+      "grad_norm": 17.25766357716426,
+      "learning_rate": 2e-06,
+      "loss": 0.2882,
+      "step": 3362
+    },
+    {
+      "epoch": 0.7801879132351235,
+      "grad_norm": 11.739053354244797,
+      "learning_rate": 2e-06,
+      "loss": 0.1788,
+      "step": 3363
+    },
+    {
+      "epoch": 0.7804199048834242,
+      "grad_norm": 12.986409445697113,
+      "learning_rate": 2e-06,
+      "loss": 0.2961,
+      "step": 3364
+    },
+    {
+      "epoch": 0.7806518965317248,
+      "grad_norm": 10.988482359288097,
+      "learning_rate": 2e-06,
+      "loss": 0.1862,
+      "step": 3365
+    },
+    {
+      "epoch": 0.7808838881800255,
+      "grad_norm": 15.64859142014504,
+      "learning_rate": 2e-06,
+      "loss": 0.2246,
+      "step": 3366
+    },
+    {
+      "epoch": 0.7811158798283262,
+      "grad_norm": 16.32751338857044,
+      "learning_rate": 2e-06,
+      "loss": 0.2849,
+      "step": 3367
+    },
+    {
+      "epoch": 0.7813478714766269,
+      "grad_norm": 13.76485487728567,
+      "learning_rate": 2e-06,
+      "loss": 0.2635,
+      "step": 3368
+    },
+    {
+      "epoch": 0.7815798631249276,
+      "grad_norm": 10.305451819845539,
+      "learning_rate": 2e-06,
+      "loss": 0.2086,
+      "step": 3369
+    },
+    {
+      "epoch": 0.7818118547732281,
+      "grad_norm": 8.387552087751217,
+      "learning_rate": 2e-06,
+      "loss": 0.2816,
+      "step": 3370
+    },
+    {
+      "epoch": 0.7820438464215288,
+      "grad_norm": 17.442401998483444,
+      "learning_rate": 2e-06,
+      "loss": 0.387,
+      "step": 3371
+    },
+    {
+      "epoch": 0.7822758380698295,
+      "grad_norm": 16.706704163997404,
+      "learning_rate": 2e-06,
+      "loss": 0.3759,
+      "step": 3372
+    },
+    {
+      "epoch": 0.7825078297181302,
+      "grad_norm": 13.969522490290837,
+      "learning_rate": 2e-06,
+      "loss": 0.2574,
+      "step": 3373
+    },
+    {
+      "epoch": 0.7827398213664308,
+      "grad_norm": 7.801153663110549,
+      "learning_rate": 2e-06,
+      "loss": 0.2363,
+      "step": 3374
+    },
+    {
+      "epoch": 0.7829718130147315,
+      "grad_norm": 16.480451323794128,
+      "learning_rate": 2e-06,
+      "loss": 0.2805,
+      "step": 3375
+    },
+    {
+      "epoch": 0.7832038046630321,
+      "grad_norm": 14.186735457312171,
+      "learning_rate": 2e-06,
+      "loss": 0.2595,
+      "step": 3376
+    },
+    {
+      "epoch": 0.7834357963113328,
+      "grad_norm": 11.646776179140373,
+      "learning_rate": 2e-06,
+      "loss": 0.2839,
+      "step": 3377
+    },
+    {
+      "epoch": 0.7836677879596334,
+      "grad_norm": 11.631828438973786,
+      "learning_rate": 2e-06,
+      "loss": 0.2361,
+      "step": 3378
+    },
+    {
+      "epoch": 0.7838997796079341,
+      "grad_norm": 14.183205529064336,
+      "learning_rate": 2e-06,
+      "loss": 0.3969,
+      "step": 3379
+    },
+    {
+      "epoch": 0.7841317712562348,
+      "grad_norm": 10.892634668274297,
+      "learning_rate": 2e-06,
+      "loss": 0.2694,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7843637629045355,
+      "grad_norm": 11.98108815181279,
+      "learning_rate": 2e-06,
+      "loss": 0.273,
+      "step": 3381
+    },
+    {
+      "epoch": 0.7845957545528361,
+      "grad_norm": 8.18435288456861,
+      "learning_rate": 2e-06,
+      "loss": 0.2099,
+      "step": 3382
+    },
+    {
+      "epoch": 0.7848277462011367,
+      "grad_norm": 12.308328485458324,
+      "learning_rate": 2e-06,
+      "loss": 0.2194,
+      "step": 3383
+    },
+    {
+      "epoch": 0.7850597378494374,
+      "grad_norm": 14.88690096912976,
+      "learning_rate": 2e-06,
+      "loss": 0.3388,
+      "step": 3384
+    },
+    {
+      "epoch": 0.7852917294977381,
+      "grad_norm": 23.606043344795978,
+      "learning_rate": 2e-06,
+      "loss": 0.3365,
+      "step": 3385
+    },
+    {
+      "epoch": 0.7855237211460387,
+      "grad_norm": 17.447552640446613,
+      "learning_rate": 2e-06,
+      "loss": 0.3042,
+      "step": 3386
+    },
+    {
+      "epoch": 0.7857557127943394,
+      "grad_norm": 15.865962850844356,
+      "learning_rate": 2e-06,
+      "loss": 0.3659,
+      "step": 3387
+    },
+    {
+      "epoch": 0.7859877044426401,
+      "grad_norm": 11.920441370957631,
+      "learning_rate": 2e-06,
+      "loss": 0.2215,
+      "step": 3388
+    },
+    {
+      "epoch": 0.7862196960909408,
+      "grad_norm": 16.597232700936654,
+      "learning_rate": 2e-06,
+      "loss": 0.3375,
+      "step": 3389
+    },
+    {
+      "epoch": 0.7864516877392413,
+      "grad_norm": 18.77556256589631,
+      "learning_rate": 2e-06,
+      "loss": 0.2961,
+      "step": 3390
+    },
+    {
+      "epoch": 0.786683679387542,
+      "grad_norm": 12.901281129823335,
+      "learning_rate": 2e-06,
+      "loss": 0.2619,
+      "step": 3391
+    },
+    {
+      "epoch": 0.7869156710358427,
+      "grad_norm": 12.44922517671116,
+      "learning_rate": 2e-06,
+      "loss": 0.1882,
+      "step": 3392
+    },
+    {
+      "epoch": 0.7871476626841434,
+      "grad_norm": 14.76842717352895,
+      "learning_rate": 2e-06,
+      "loss": 0.2978,
+      "step": 3393
+    },
+    {
+      "epoch": 0.7873796543324441,
+      "grad_norm": 14.152124730026602,
+      "learning_rate": 2e-06,
+      "loss": 0.2882,
+      "step": 3394
+    },
+    {
+      "epoch": 0.7876116459807447,
+      "grad_norm": 18.40012721570145,
+      "learning_rate": 2e-06,
+      "loss": 0.3215,
+      "step": 3395
+    },
+    {
+      "epoch": 0.7878436376290453,
+      "grad_norm": 11.181603808392303,
+      "learning_rate": 2e-06,
+      "loss": 0.2649,
+      "step": 3396
+    },
+    {
+      "epoch": 0.788075629277346,
+      "grad_norm": 18.509419869703052,
+      "learning_rate": 2e-06,
+      "loss": 0.3095,
+      "step": 3397
+    },
+    {
+      "epoch": 0.7883076209256467,
+      "grad_norm": 14.184333048504568,
+      "learning_rate": 2e-06,
+      "loss": 0.2602,
+      "step": 3398
+    },
+    {
+      "epoch": 0.7885396125739473,
+      "grad_norm": 12.587887359899812,
+      "learning_rate": 2e-06,
+      "loss": 0.2427,
+      "step": 3399
+    },
+    {
+      "epoch": 0.788771604222248,
+      "grad_norm": 24.507560493039886,
+      "learning_rate": 2e-06,
+      "loss": 0.335,
+      "step": 3400
+    },
+    {
+      "epoch": 0.7890035958705487,
+      "grad_norm": 4.819863587109384,
+      "learning_rate": 2e-06,
+      "loss": 0.1711,
+      "step": 3401
+    },
+    {
+      "epoch": 0.7892355875188494,
+      "grad_norm": 22.111328616473372,
+      "learning_rate": 2e-06,
+      "loss": 0.3374,
+      "step": 3402
+    },
+    {
+      "epoch": 0.7894675791671499,
+      "grad_norm": 11.906781767275788,
+      "learning_rate": 2e-06,
+      "loss": 0.207,
+      "step": 3403
+    },
+    {
+      "epoch": 0.7896995708154506,
+      "grad_norm": 18.47361798133603,
+      "learning_rate": 2e-06,
+      "loss": 0.3547,
+      "step": 3404
+    },
+    {
+      "epoch": 0.7899315624637513,
+      "grad_norm": 15.688065948218464,
+      "learning_rate": 2e-06,
+      "loss": 0.2639,
+      "step": 3405
+    },
+    {
+      "epoch": 0.790163554112052,
+      "grad_norm": 26.116619607587307,
+      "learning_rate": 2e-06,
+      "loss": 0.3287,
+      "step": 3406
+    },
+    {
+      "epoch": 0.7903955457603526,
+      "grad_norm": 8.882349560205764,
+      "learning_rate": 2e-06,
+      "loss": 0.2585,
+      "step": 3407
+    },
+    {
+      "epoch": 0.7906275374086533,
+      "grad_norm": 9.446665087382272,
+      "learning_rate": 2e-06,
+      "loss": 0.2615,
+      "step": 3408
+    },
+    {
+      "epoch": 0.790859529056954,
+      "grad_norm": 18.048231253616922,
+      "learning_rate": 2e-06,
+      "loss": 0.3177,
+      "step": 3409
+    },
+    {
+      "epoch": 0.7910915207052546,
+      "grad_norm": 14.29683313119664,
+      "learning_rate": 2e-06,
+      "loss": 0.2627,
+      "step": 3410
+    },
+    {
+      "epoch": 0.7913235123535552,
+      "grad_norm": 10.825488918220556,
+      "learning_rate": 2e-06,
+      "loss": 0.2825,
+      "step": 3411
+    },
+    {
+      "epoch": 0.7915555040018559,
+      "grad_norm": 8.743391123103203,
+      "learning_rate": 2e-06,
+      "loss": 0.3564,
+      "step": 3412
+    },
+    {
+      "epoch": 0.7917874956501566,
+      "grad_norm": 8.390341529464372,
+      "learning_rate": 2e-06,
+      "loss": 0.2535,
+      "step": 3413
+    },
+    {
+      "epoch": 0.7920194872984573,
+      "grad_norm": 11.615387497887594,
+      "learning_rate": 2e-06,
+      "loss": 0.2495,
+      "step": 3414
+    },
+    {
+      "epoch": 0.792251478946758,
+      "grad_norm": 10.339780478199742,
+      "learning_rate": 2e-06,
+      "loss": 0.2124,
+      "step": 3415
+    },
+    {
+      "epoch": 0.7924834705950585,
+      "grad_norm": 19.814645086976675,
+      "learning_rate": 2e-06,
+      "loss": 0.3676,
+      "step": 3416
+    },
+    {
+      "epoch": 0.7927154622433592,
+      "grad_norm": 11.85489858812581,
+      "learning_rate": 2e-06,
+      "loss": 0.2406,
+      "step": 3417
+    },
+    {
+      "epoch": 0.7929474538916599,
+      "grad_norm": 16.299759592564307,
+      "learning_rate": 2e-06,
+      "loss": 0.2692,
+      "step": 3418
+    },
+    {
+      "epoch": 0.7931794455399606,
+      "grad_norm": 13.872791910011749,
+      "learning_rate": 2e-06,
+      "loss": 0.305,
+      "step": 3419
+    },
+    {
+      "epoch": 0.7934114371882612,
+      "grad_norm": 19.750414611149914,
+      "learning_rate": 2e-06,
+      "loss": 0.2335,
+      "step": 3420
+    },
+    {
+      "epoch": 0.7936434288365619,
+      "grad_norm": 19.956545241264873,
+      "learning_rate": 2e-06,
+      "loss": 0.3085,
+      "step": 3421
+    },
+    {
+      "epoch": 0.7938754204848626,
+      "grad_norm": 8.149993746924585,
+      "learning_rate": 2e-06,
+      "loss": 0.2059,
+      "step": 3422
+    },
+    {
+      "epoch": 0.7941074121331632,
+      "grad_norm": 10.028973054761785,
+      "learning_rate": 2e-06,
+      "loss": 0.2432,
+      "step": 3423
+    },
+    {
+      "epoch": 0.7943394037814638,
+      "grad_norm": 18.778972827943914,
+      "learning_rate": 2e-06,
+      "loss": 0.3628,
+      "step": 3424
+    },
+    {
+      "epoch": 0.7945713954297645,
+      "grad_norm": 13.930116536144318,
+      "learning_rate": 2e-06,
+      "loss": 0.2445,
+      "step": 3425
+    },
+    {
+      "epoch": 0.7948033870780652,
+      "grad_norm": 14.63089032989506,
+      "learning_rate": 2e-06,
+      "loss": 0.2343,
+      "step": 3426
+    },
+    {
+      "epoch": 0.7950353787263659,
+      "grad_norm": 15.72743340478428,
+      "learning_rate": 2e-06,
+      "loss": 0.3289,
+      "step": 3427
+    },
+    {
+      "epoch": 0.7952673703746665,
+      "grad_norm": 12.471700775785143,
+      "learning_rate": 2e-06,
+      "loss": 0.2463,
+      "step": 3428
+    },
+    {
+      "epoch": 0.7954993620229672,
+      "grad_norm": 15.784691045568465,
+      "learning_rate": 2e-06,
+      "loss": 0.3693,
+      "step": 3429
+    },
+    {
+      "epoch": 0.7957313536712678,
+      "grad_norm": 11.487372968049451,
+      "learning_rate": 2e-06,
+      "loss": 0.2886,
+      "step": 3430
+    },
+    {
+      "epoch": 0.7959633453195685,
+      "grad_norm": 23.592938069770618,
+      "learning_rate": 2e-06,
+      "loss": 0.413,
+      "step": 3431
+    },
+    {
+      "epoch": 0.7961953369678691,
+      "grad_norm": 13.804563113914236,
+      "learning_rate": 2e-06,
+      "loss": 0.2107,
+      "step": 3432
+    },
+    {
+      "epoch": 0.7964273286161698,
+      "grad_norm": 11.229386196040094,
+      "learning_rate": 2e-06,
+      "loss": 0.3498,
+      "step": 3433
+    },
+    {
+      "epoch": 0.7966593202644705,
+      "grad_norm": 11.510512004956807,
+      "learning_rate": 2e-06,
+      "loss": 0.2175,
+      "step": 3434
+    },
+    {
+      "epoch": 0.7968913119127712,
+      "grad_norm": 16.424219271937943,
+      "learning_rate": 2e-06,
+      "loss": 0.2835,
+      "step": 3435
+    },
+    {
+      "epoch": 0.7971233035610717,
+      "grad_norm": 11.513067182158752,
+      "learning_rate": 2e-06,
+      "loss": 0.2623,
+      "step": 3436
+    },
+    {
+      "epoch": 0.7973552952093724,
+      "grad_norm": 9.572539910406347,
+      "learning_rate": 2e-06,
+      "loss": 0.1982,
+      "step": 3437
+    },
+    {
+      "epoch": 0.7975872868576731,
+      "grad_norm": 17.580798746144982,
+      "learning_rate": 2e-06,
+      "loss": 0.2625,
+      "step": 3438
+    },
+    {
+      "epoch": 0.7978192785059738,
+      "grad_norm": 12.457497591236477,
+      "learning_rate": 2e-06,
+      "loss": 0.2578,
+      "step": 3439
+    },
+    {
+      "epoch": 0.7980512701542745,
+      "grad_norm": 14.722234583713858,
+      "learning_rate": 2e-06,
+      "loss": 0.3125,
+      "step": 3440
+    },
+    {
+      "epoch": 0.7982832618025751,
+      "grad_norm": 8.417471859774857,
+      "learning_rate": 2e-06,
+      "loss": 0.279,
+      "step": 3441
+    },
+    {
+      "epoch": 0.7985152534508758,
+      "grad_norm": 13.0848253183779,
+      "learning_rate": 2e-06,
+      "loss": 0.3323,
+      "step": 3442
+    },
+    {
+      "epoch": 0.7987472450991764,
+      "grad_norm": 13.163029759028204,
+      "learning_rate": 2e-06,
+      "loss": 0.2724,
+      "step": 3443
+    },
+    {
+      "epoch": 0.7989792367474771,
+      "grad_norm": 7.058717519417359,
+      "learning_rate": 2e-06,
+      "loss": 0.157,
+      "step": 3444
+    },
+    {
+      "epoch": 0.7992112283957777,
+      "grad_norm": 15.644201443704182,
+      "learning_rate": 2e-06,
+      "loss": 0.3231,
+      "step": 3445
+    },
+    {
+      "epoch": 0.7994432200440784,
+      "grad_norm": 16.464898783661045,
+      "learning_rate": 2e-06,
+      "loss": 0.3016,
+      "step": 3446
+    },
+    {
+      "epoch": 0.7996752116923791,
+      "grad_norm": 13.152454899489452,
+      "learning_rate": 2e-06,
+      "loss": 0.2524,
+      "step": 3447
+    },
+    {
+      "epoch": 0.7999072033406798,
+      "grad_norm": 16.487316256533344,
+      "learning_rate": 2e-06,
+      "loss": 0.3709,
+      "step": 3448
+    },
+    {
+      "epoch": 0.8001391949889805,
+      "grad_norm": 9.283139468973133,
+      "learning_rate": 2e-06,
+      "loss": 0.2256,
+      "step": 3449
+    },
+    {
+      "epoch": 0.800371186637281,
+      "grad_norm": 18.27553759166784,
+      "learning_rate": 2e-06,
+      "loss": 0.309,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8006031782855817,
+      "grad_norm": 17.61707193996193,
+      "learning_rate": 2e-06,
+      "loss": 0.2662,
+      "step": 3451
+    },
+    {
+      "epoch": 0.8008351699338824,
+      "grad_norm": 13.457161435100732,
+      "learning_rate": 2e-06,
+      "loss": 0.2381,
+      "step": 3452
+    },
+    {
+      "epoch": 0.801067161582183,
+      "grad_norm": 13.991405109514043,
+      "learning_rate": 2e-06,
+      "loss": 0.2798,
+      "step": 3453
+    },
+    {
+      "epoch": 0.8012991532304837,
+      "grad_norm": 18.77836921343787,
+      "learning_rate": 2e-06,
+      "loss": 0.3454,
+      "step": 3454
+    },
+    {
+      "epoch": 0.8015311448787844,
+      "grad_norm": 24.73440836115384,
+      "learning_rate": 2e-06,
+      "loss": 0.3345,
+      "step": 3455
+    },
+    {
+      "epoch": 0.801763136527085,
+      "grad_norm": 19.19064961801089,
+      "learning_rate": 2e-06,
+      "loss": 0.2585,
+      "step": 3456
+    },
+    {
+      "epoch": 0.8019951281753857,
+      "grad_norm": 10.14125178150145,
+      "learning_rate": 2e-06,
+      "loss": 0.2333,
+      "step": 3457
+    },
+    {
+      "epoch": 0.8022271198236863,
+      "grad_norm": 7.075252080603709,
+      "learning_rate": 2e-06,
+      "loss": 0.1803,
+      "step": 3458
+    },
+    {
+      "epoch": 0.802459111471987,
+      "grad_norm": 15.228216202553677,
+      "learning_rate": 2e-06,
+      "loss": 0.2749,
+      "step": 3459
+    },
+    {
+      "epoch": 0.8026911031202877,
+      "grad_norm": 15.90276123435102,
+      "learning_rate": 2e-06,
+      "loss": 0.3938,
+      "step": 3460
+    },
+    {
+      "epoch": 0.8029230947685884,
+      "grad_norm": 12.67302214729661,
+      "learning_rate": 2e-06,
+      "loss": 0.2562,
+      "step": 3461
+    },
+    {
+      "epoch": 0.803155086416889,
+      "grad_norm": 14.535054771013746,
+      "learning_rate": 2e-06,
+      "loss": 0.3209,
+      "step": 3462
+    },
+    {
+      "epoch": 0.8033870780651896,
+      "grad_norm": 10.919756951774275,
+      "learning_rate": 2e-06,
+      "loss": 0.2377,
+      "step": 3463
+    },
+    {
+      "epoch": 0.8036190697134903,
+      "grad_norm": 14.211186237722966,
+      "learning_rate": 2e-06,
+      "loss": 0.303,
+      "step": 3464
+    },
+    {
+      "epoch": 0.803851061361791,
+      "grad_norm": 10.324439146983549,
+      "learning_rate": 2e-06,
+      "loss": 0.2724,
+      "step": 3465
+    },
+    {
+      "epoch": 0.8040830530100916,
+      "grad_norm": 18.212276946809332,
+      "learning_rate": 2e-06,
+      "loss": 0.2857,
+      "step": 3466
+    },
+    {
+      "epoch": 0.8043150446583923,
+      "grad_norm": 9.08491619843411,
+      "learning_rate": 2e-06,
+      "loss": 0.3044,
+      "step": 3467
+    },
+    {
+      "epoch": 0.804547036306693,
+      "grad_norm": 13.555167646674509,
+      "learning_rate": 2e-06,
+      "loss": 0.3019,
+      "step": 3468
+    },
+    {
+      "epoch": 0.8047790279549937,
+      "grad_norm": 14.173546709054419,
+      "learning_rate": 2e-06,
+      "loss": 0.2405,
+      "step": 3469
+    },
+    {
+      "epoch": 0.8050110196032942,
+      "grad_norm": 13.906666800937307,
+      "learning_rate": 2e-06,
+      "loss": 0.293,
+      "step": 3470
+    },
+    {
+      "epoch": 0.8052430112515949,
+      "grad_norm": 20.61968237180272,
+      "learning_rate": 2e-06,
+      "loss": 0.3068,
+      "step": 3471
+    },
+    {
+      "epoch": 0.8054750028998956,
+      "grad_norm": 7.016349103407426,
+      "learning_rate": 2e-06,
+      "loss": 0.2895,
+      "step": 3472
+    },
+    {
+      "epoch": 0.8057069945481963,
+      "grad_norm": 9.27839396191916,
+      "learning_rate": 2e-06,
+      "loss": 0.259,
+      "step": 3473
+    },
+    {
+      "epoch": 0.805938986196497,
+      "grad_norm": 9.333666595921253,
+      "learning_rate": 2e-06,
+      "loss": 0.2843,
+      "step": 3474
+    },
+    {
+      "epoch": 0.8061709778447976,
+      "grad_norm": 12.925200076949773,
+      "learning_rate": 2e-06,
+      "loss": 0.2524,
+      "step": 3475
+    },
+    {
+      "epoch": 0.8064029694930982,
+      "grad_norm": 23.82557788438642,
+      "learning_rate": 2e-06,
+      "loss": 0.3742,
+      "step": 3476
+    },
+    {
+      "epoch": 0.8066349611413989,
+      "grad_norm": 12.30874586445278,
+      "learning_rate": 2e-06,
+      "loss": 0.2475,
+      "step": 3477
+    },
+    {
+      "epoch": 0.8068669527896996,
+      "grad_norm": 17.45344333498659,
+      "learning_rate": 2e-06,
+      "loss": 0.2469,
+      "step": 3478
+    },
+    {
+      "epoch": 0.8070989444380002,
+      "grad_norm": 20.44619451931843,
+      "learning_rate": 2e-06,
+      "loss": 0.4079,
+      "step": 3479
+    },
+    {
+      "epoch": 0.8073309360863009,
+      "grad_norm": 16.5515046014371,
+      "learning_rate": 2e-06,
+      "loss": 0.3359,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8075629277346016,
+      "grad_norm": 24.519338165710717,
+      "learning_rate": 2e-06,
+      "loss": 0.3352,
+      "step": 3481
+    },
+    {
+      "epoch": 0.8077949193829023,
+      "grad_norm": 14.495647774820247,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 3482
+    },
+    {
+      "epoch": 0.8080269110312028,
+      "grad_norm": 35.31943301373369,
+      "learning_rate": 2e-06,
+      "loss": 0.5115,
+      "step": 3483
+    },
+    {
+      "epoch": 0.8082589026795035,
+      "grad_norm": 13.811477132735103,
+      "learning_rate": 2e-06,
+      "loss": 0.3339,
+      "step": 3484
+    },
+    {
+      "epoch": 0.8084908943278042,
+      "grad_norm": 11.395040916060081,
+      "learning_rate": 2e-06,
+      "loss": 0.3566,
+      "step": 3485
+    },
+    {
+      "epoch": 0.8087228859761049,
+      "grad_norm": 6.689994663714095,
+      "learning_rate": 2e-06,
+      "loss": 0.2268,
+      "step": 3486
+    },
+    {
+      "epoch": 0.8089548776244055,
+      "grad_norm": 7.131946165209261,
+      "learning_rate": 2e-06,
+      "loss": 0.2644,
+      "step": 3487
+    },
+    {
+      "epoch": 0.8091868692727062,
+      "grad_norm": 13.33439706490419,
+      "learning_rate": 2e-06,
+      "loss": 0.3359,
+      "step": 3488
+    },
+    {
+      "epoch": 0.8094188609210068,
+      "grad_norm": 7.313970577613089,
+      "learning_rate": 2e-06,
+      "loss": 0.2157,
+      "step": 3489
+    },
+    {
+      "epoch": 0.8096508525693075,
+      "grad_norm": 8.979295973783886,
+      "learning_rate": 2e-06,
+      "loss": 0.2594,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8098828442176081,
+      "grad_norm": 7.810909631076036,
+      "learning_rate": 2e-06,
+      "loss": 0.2591,
+      "step": 3491
+    },
+    {
+      "epoch": 0.8101148358659088,
+      "grad_norm": 12.246774385443148,
+      "learning_rate": 2e-06,
+      "loss": 0.3403,
+      "step": 3492
+    },
+    {
+      "epoch": 0.8103468275142095,
+      "grad_norm": 9.947901848029089,
+      "learning_rate": 2e-06,
+      "loss": 0.2852,
+      "step": 3493
+    },
+    {
+      "epoch": 0.8105788191625102,
+      "grad_norm": 11.094416640136869,
+      "learning_rate": 2e-06,
+      "loss": 0.2473,
+      "step": 3494
+    },
+    {
+      "epoch": 0.8108108108108109,
+      "grad_norm": 11.774081114699563,
+      "learning_rate": 2e-06,
+      "loss": 0.3085,
+      "step": 3495
+    },
+    {
+      "epoch": 0.8110428024591114,
+      "grad_norm": 10.698679440962588,
+      "learning_rate": 2e-06,
+      "loss": 0.2418,
+      "step": 3496
+    },
+    {
+      "epoch": 0.8112747941074121,
+      "grad_norm": 15.764409636004656,
+      "learning_rate": 2e-06,
+      "loss": 0.2487,
+      "step": 3497
+    },
+    {
+      "epoch": 0.8115067857557128,
+      "grad_norm": 15.765164065297412,
+      "learning_rate": 2e-06,
+      "loss": 0.3738,
+      "step": 3498
+    },
+    {
+      "epoch": 0.8117387774040135,
+      "grad_norm": 13.187289856086876,
+      "learning_rate": 2e-06,
+      "loss": 0.3151,
+      "step": 3499
+    },
+    {
+      "epoch": 0.8119707690523141,
+      "grad_norm": 13.801830481066148,
+      "learning_rate": 2e-06,
+      "loss": 0.2247,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8122027607006148,
+      "grad_norm": 11.524118770924476,
+      "learning_rate": 2e-06,
+      "loss": 0.2742,
+      "step": 3501
+    },
+    {
+      "epoch": 0.8124347523489155,
+      "grad_norm": 20.522591212945343,
+      "learning_rate": 2e-06,
+      "loss": 0.3449,
+      "step": 3502
+    },
+    {
+      "epoch": 0.8126667439972161,
+      "grad_norm": 11.509315641920494,
+      "learning_rate": 2e-06,
+      "loss": 0.2796,
+      "step": 3503
+    },
+    {
+      "epoch": 0.8128987356455167,
+      "grad_norm": 5.488506283368135,
+      "learning_rate": 2e-06,
+      "loss": 0.1703,
+      "step": 3504
+    },
+    {
+      "epoch": 0.8131307272938174,
+      "grad_norm": 11.409597576924604,
+      "learning_rate": 2e-06,
+      "loss": 0.2738,
+      "step": 3505
+    },
+    {
+      "epoch": 0.8133627189421181,
+      "grad_norm": 14.455097672576871,
+      "learning_rate": 2e-06,
+      "loss": 0.3249,
+      "step": 3506
+    },
+    {
+      "epoch": 0.8135947105904188,
+      "grad_norm": 15.071206169953461,
+      "learning_rate": 2e-06,
+      "loss": 0.3254,
+      "step": 3507
+    },
+    {
+      "epoch": 0.8138267022387194,
+      "grad_norm": 10.482927998182806,
+      "learning_rate": 2e-06,
+      "loss": 0.2376,
+      "step": 3508
+    },
+    {
+      "epoch": 0.81405869388702,
+      "grad_norm": 25.693406081649588,
+      "learning_rate": 2e-06,
+      "loss": 0.3781,
+      "step": 3509
+    },
+    {
+      "epoch": 0.8142906855353207,
+      "grad_norm": 13.608869586175004,
+      "learning_rate": 2e-06,
+      "loss": 0.2643,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8145226771836214,
+      "grad_norm": 12.004444029483697,
+      "learning_rate": 2e-06,
+      "loss": 0.2325,
+      "step": 3511
+    },
+    {
+      "epoch": 0.814754668831922,
+      "grad_norm": 12.940201273993516,
+      "learning_rate": 2e-06,
+      "loss": 0.2498,
+      "step": 3512
+    },
+    {
+      "epoch": 0.8149866604802227,
+      "grad_norm": 18.105886390268846,
+      "learning_rate": 2e-06,
+      "loss": 0.2966,
+      "step": 3513
+    },
+    {
+      "epoch": 0.8152186521285234,
+      "grad_norm": 9.755817402936005,
+      "learning_rate": 2e-06,
+      "loss": 0.2896,
+      "step": 3514
+    },
+    {
+      "epoch": 0.8154506437768241,
+      "grad_norm": 20.685876534208933,
+      "learning_rate": 2e-06,
+      "loss": 0.2989,
+      "step": 3515
+    },
+    {
+      "epoch": 0.8156826354251246,
+      "grad_norm": 10.772642056094304,
+      "learning_rate": 2e-06,
+      "loss": 0.2951,
+      "step": 3516
+    },
+    {
+      "epoch": 0.8159146270734253,
+      "grad_norm": 22.421591893301528,
+      "learning_rate": 2e-06,
+      "loss": 0.3378,
+      "step": 3517
+    },
+    {
+      "epoch": 0.816146618721726,
+      "grad_norm": 15.838181455870886,
+      "learning_rate": 2e-06,
+      "loss": 0.3841,
+      "step": 3518
+    },
+    {
+      "epoch": 0.8163786103700267,
+      "grad_norm": 10.795058033101041,
+      "learning_rate": 2e-06,
+      "loss": 0.2946,
+      "step": 3519
+    },
+    {
+      "epoch": 0.8166106020183274,
+      "grad_norm": 10.703899141780543,
+      "learning_rate": 2e-06,
+      "loss": 0.3325,
+      "step": 3520
+    },
+    {
+      "epoch": 0.816842593666628,
+      "grad_norm": 8.945582349508102,
+      "learning_rate": 2e-06,
+      "loss": 0.2288,
+      "step": 3521
+    },
+    {
+      "epoch": 0.8170745853149287,
+      "grad_norm": 16.457137415765683,
+      "learning_rate": 2e-06,
+      "loss": 0.3724,
+      "step": 3522
+    },
+    {
+      "epoch": 0.8173065769632293,
+      "grad_norm": 14.915024776601184,
+      "learning_rate": 2e-06,
+      "loss": 0.2437,
+      "step": 3523
+    },
+    {
+      "epoch": 0.81753856861153,
+      "grad_norm": 19.130665774689742,
+      "learning_rate": 2e-06,
+      "loss": 0.42,
+      "step": 3524
+    },
+    {
+      "epoch": 0.8177705602598306,
+      "grad_norm": 11.793326568443652,
+      "learning_rate": 2e-06,
+      "loss": 0.2295,
+      "step": 3525
+    },
+    {
+      "epoch": 0.8180025519081313,
+      "grad_norm": 6.473724516177291,
+      "learning_rate": 2e-06,
+      "loss": 0.1777,
+      "step": 3526
+    },
+    {
+      "epoch": 0.818234543556432,
+      "grad_norm": 13.25425715941778,
+      "learning_rate": 2e-06,
+      "loss": 0.26,
+      "step": 3527
+    },
+    {
+      "epoch": 0.8184665352047327,
+      "grad_norm": 10.245529294799168,
+      "learning_rate": 2e-06,
+      "loss": 0.3425,
+      "step": 3528
+    },
+    {
+      "epoch": 0.8186985268530332,
+      "grad_norm": 6.669639477632707,
+      "learning_rate": 2e-06,
+      "loss": 0.2318,
+      "step": 3529
+    },
+    {
+      "epoch": 0.8189305185013339,
+      "grad_norm": 18.227102749489706,
+      "learning_rate": 2e-06,
+      "loss": 0.2112,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8191625101496346,
+      "grad_norm": 9.03996463876371,
+      "learning_rate": 2e-06,
+      "loss": 0.296,
+      "step": 3531
+    },
+    {
+      "epoch": 0.8193945017979353,
+      "grad_norm": 14.128603749034852,
+      "learning_rate": 2e-06,
+      "loss": 0.228,
+      "step": 3532
+    },
+    {
+      "epoch": 0.819626493446236,
+      "grad_norm": 19.581050800333305,
+      "learning_rate": 2e-06,
+      "loss": 0.2736,
+      "step": 3533
+    },
+    {
+      "epoch": 0.8198584850945366,
+      "grad_norm": 10.830024818169084,
+      "learning_rate": 2e-06,
+      "loss": 0.4177,
+      "step": 3534
+    },
+    {
+      "epoch": 0.8200904767428373,
+      "grad_norm": 10.547522500539598,
+      "learning_rate": 2e-06,
+      "loss": 0.2547,
+      "step": 3535
+    },
+    {
+      "epoch": 0.8203224683911379,
+      "grad_norm": 32.3173599240258,
+      "learning_rate": 2e-06,
+      "loss": 0.3983,
+      "step": 3536
+    },
+    {
+      "epoch": 0.8205544600394385,
+      "grad_norm": 13.226067374404431,
+      "learning_rate": 2e-06,
+      "loss": 0.2506,
+      "step": 3537
+    },
+    {
+      "epoch": 0.8207864516877392,
+      "grad_norm": 10.462376223619785,
+      "learning_rate": 2e-06,
+      "loss": 0.2232,
+      "step": 3538
+    },
+    {
+      "epoch": 0.8210184433360399,
+      "grad_norm": 32.82054766950882,
+      "learning_rate": 2e-06,
+      "loss": 0.3654,
+      "step": 3539
+    },
+    {
+      "epoch": 0.8212504349843406,
+      "grad_norm": 10.860846241032947,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 3540
+    },
+    {
+      "epoch": 0.8214824266326413,
+      "grad_norm": 8.995251473292827,
+      "learning_rate": 2e-06,
+      "loss": 0.3209,
+      "step": 3541
+    },
+    {
+      "epoch": 0.8217144182809419,
+      "grad_norm": 15.77422954469678,
+      "learning_rate": 2e-06,
+      "loss": 0.3464,
+      "step": 3542
+    },
+    {
+      "epoch": 0.8219464099292425,
+      "grad_norm": 10.039797919380941,
+      "learning_rate": 2e-06,
+      "loss": 0.3015,
+      "step": 3543
+    },
+    {
+      "epoch": 0.8221784015775432,
+      "grad_norm": 9.594735943914506,
+      "learning_rate": 2e-06,
+      "loss": 0.2629,
+      "step": 3544
+    },
+    {
+      "epoch": 0.8224103932258439,
+      "grad_norm": 13.127890377106954,
+      "learning_rate": 2e-06,
+      "loss": 0.2683,
+      "step": 3545
+    },
+    {
+      "epoch": 0.8226423848741445,
+      "grad_norm": 6.018401443740743,
+      "learning_rate": 2e-06,
+      "loss": 0.1822,
+      "step": 3546
+    },
+    {
+      "epoch": 0.8228743765224452,
+      "grad_norm": 10.239555624879175,
+      "learning_rate": 2e-06,
+      "loss": 0.2981,
+      "step": 3547
+    },
+    {
+      "epoch": 0.8231063681707459,
+      "grad_norm": 8.44416471941474,
+      "learning_rate": 2e-06,
+      "loss": 0.1853,
+      "step": 3548
+    },
+    {
+      "epoch": 0.8233383598190465,
+      "grad_norm": 11.946867911578675,
+      "learning_rate": 2e-06,
+      "loss": 0.2552,
+      "step": 3549
+    },
+    {
+      "epoch": 0.8235703514673471,
+      "grad_norm": 17.157645834778513,
+      "learning_rate": 2e-06,
+      "loss": 0.3434,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8238023431156478,
+      "grad_norm": 15.446833564674439,
+      "learning_rate": 2e-06,
+      "loss": 0.2505,
+      "step": 3551
+    },
+    {
+      "epoch": 0.8240343347639485,
+      "grad_norm": 16.49241567046502,
+      "learning_rate": 2e-06,
+      "loss": 0.2732,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8242663264122492,
+      "grad_norm": 15.670109551955258,
+      "learning_rate": 2e-06,
+      "loss": 0.3211,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8244983180605499,
+      "grad_norm": 16.140530370283813,
+      "learning_rate": 2e-06,
+      "loss": 0.2206,
+      "step": 3554
+    },
+    {
+      "epoch": 0.8247303097088505,
+      "grad_norm": 10.511695486802429,
+      "learning_rate": 2e-06,
+      "loss": 0.2428,
+      "step": 3555
+    },
+    {
+      "epoch": 0.8249623013571511,
+      "grad_norm": 10.306828334456238,
+      "learning_rate": 2e-06,
+      "loss": 0.2865,
+      "step": 3556
+    },
+    {
+      "epoch": 0.8251942930054518,
+      "grad_norm": 15.412563379487153,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 3557
+    },
+    {
+      "epoch": 0.8254262846537525,
+      "grad_norm": 14.469061271601475,
+      "learning_rate": 2e-06,
+      "loss": 0.3926,
+      "step": 3558
+    },
+    {
+      "epoch": 0.8256582763020531,
+      "grad_norm": 12.314557820247007,
+      "learning_rate": 2e-06,
+      "loss": 0.285,
+      "step": 3559
+    },
+    {
+      "epoch": 0.8258902679503538,
+      "grad_norm": 20.344101930227662,
+      "learning_rate": 2e-06,
+      "loss": 0.2161,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8261222595986545,
+      "grad_norm": 14.321188193323712,
+      "learning_rate": 2e-06,
+      "loss": 0.282,
+      "step": 3561
+    },
+    {
+      "epoch": 0.8263542512469552,
+      "grad_norm": 10.369535133540206,
+      "learning_rate": 2e-06,
+      "loss": 0.2756,
+      "step": 3562
+    },
+    {
+      "epoch": 0.8265862428952557,
+      "grad_norm": 14.405162531020206,
+      "learning_rate": 2e-06,
+      "loss": 0.2895,
+      "step": 3563
+    },
+    {
+      "epoch": 0.8268182345435564,
+      "grad_norm": 27.51068067463199,
+      "learning_rate": 2e-06,
+      "loss": 0.3162,
+      "step": 3564
+    },
+    {
+      "epoch": 0.8270502261918571,
+      "grad_norm": 9.47044995543265,
+      "learning_rate": 2e-06,
+      "loss": 0.1824,
+      "step": 3565
+    },
+    {
+      "epoch": 0.8272822178401578,
+      "grad_norm": 13.374539645458363,
+      "learning_rate": 2e-06,
+      "loss": 0.2139,
+      "step": 3566
+    },
+    {
+      "epoch": 0.8275142094884584,
+      "grad_norm": 17.735043818984096,
+      "learning_rate": 2e-06,
+      "loss": 0.2542,
+      "step": 3567
+    },
+    {
+      "epoch": 0.8277462011367591,
+      "grad_norm": 9.242686802461979,
+      "learning_rate": 2e-06,
+      "loss": 0.1826,
+      "step": 3568
+    },
+    {
+      "epoch": 0.8279781927850597,
+      "grad_norm": 5.677185399446114,
+      "learning_rate": 2e-06,
+      "loss": 0.1462,
+      "step": 3569
+    },
+    {
+      "epoch": 0.8282101844333604,
+      "grad_norm": 12.844155490278478,
+      "learning_rate": 2e-06,
+      "loss": 0.4044,
+      "step": 3570
+    },
+    {
+      "epoch": 0.828442176081661,
+      "grad_norm": 15.018609163790327,
+      "learning_rate": 2e-06,
+      "loss": 0.4075,
+      "step": 3571
+    },
+    {
+      "epoch": 0.8286741677299617,
+      "grad_norm": 28.60480917901526,
+      "learning_rate": 2e-06,
+      "loss": 0.3256,
+      "step": 3572
+    },
+    {
+      "epoch": 0.8289061593782624,
+      "grad_norm": 11.797828274897043,
+      "learning_rate": 2e-06,
+      "loss": 0.2946,
+      "step": 3573
+    },
+    {
+      "epoch": 0.8291381510265631,
+      "grad_norm": 12.151158127436506,
+      "learning_rate": 2e-06,
+      "loss": 0.3089,
+      "step": 3574
+    },
+    {
+      "epoch": 0.8293701426748638,
+      "grad_norm": 10.69731287572852,
+      "learning_rate": 2e-06,
+      "loss": 0.1956,
+      "step": 3575
+    },
+    {
+      "epoch": 0.8296021343231643,
+      "grad_norm": 15.209091661657826,
+      "learning_rate": 2e-06,
+      "loss": 0.3467,
+      "step": 3576
+    },
+    {
+      "epoch": 0.829834125971465,
+      "grad_norm": 9.48025969073124,
+      "learning_rate": 2e-06,
+      "loss": 0.3034,
+      "step": 3577
+    },
+    {
+      "epoch": 0.8300661176197657,
+      "grad_norm": 22.156690040220127,
+      "learning_rate": 2e-06,
+      "loss": 0.2798,
+      "step": 3578
+    },
+    {
+      "epoch": 0.8302981092680664,
+      "grad_norm": 17.120131547477015,
+      "learning_rate": 2e-06,
+      "loss": 0.4078,
+      "step": 3579
+    },
+    {
+      "epoch": 0.830530100916367,
+      "grad_norm": 13.727506476081903,
+      "learning_rate": 2e-06,
+      "loss": 0.2804,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8307620925646677,
+      "grad_norm": 9.66740188676226,
+      "learning_rate": 2e-06,
+      "loss": 0.2391,
+      "step": 3581
+    },
+    {
+      "epoch": 0.8309940842129684,
+      "grad_norm": 17.59160175368584,
+      "learning_rate": 2e-06,
+      "loss": 0.3679,
+      "step": 3582
+    },
+    {
+      "epoch": 0.831226075861269,
+      "grad_norm": 17.56972129711204,
+      "learning_rate": 2e-06,
+      "loss": 0.3698,
+      "step": 3583
+    },
+    {
+      "epoch": 0.8314580675095696,
+      "grad_norm": 13.197156971473062,
+      "learning_rate": 2e-06,
+      "loss": 0.2521,
+      "step": 3584
+    },
+    {
+      "epoch": 0.8316900591578703,
+      "grad_norm": 14.553013737894982,
+      "learning_rate": 2e-06,
+      "loss": 0.2363,
+      "step": 3585
+    },
+    {
+      "epoch": 0.831922050806171,
+      "grad_norm": 8.26424690417724,
+      "learning_rate": 2e-06,
+      "loss": 0.1732,
+      "step": 3586
+    },
+    {
+      "epoch": 0.8321540424544717,
+      "grad_norm": 16.971467765915932,
+      "learning_rate": 2e-06,
+      "loss": 0.2755,
+      "step": 3587
+    },
+    {
+      "epoch": 0.8323860341027723,
+      "grad_norm": 9.547199083280193,
+      "learning_rate": 2e-06,
+      "loss": 0.2821,
+      "step": 3588
+    },
+    {
+      "epoch": 0.8326180257510729,
+      "grad_norm": 8.19466444893716,
+      "learning_rate": 2e-06,
+      "loss": 0.1939,
+      "step": 3589
+    },
+    {
+      "epoch": 0.8328500173993736,
+      "grad_norm": 13.773792857129889,
+      "learning_rate": 2e-06,
+      "loss": 0.2627,
+      "step": 3590
+    },
+    {
+      "epoch": 0.8330820090476743,
+      "grad_norm": 25.164178842438982,
+      "learning_rate": 2e-06,
+      "loss": 0.1979,
+      "step": 3591
+    },
+    {
+      "epoch": 0.8333140006959749,
+      "grad_norm": 19.67131566725128,
+      "learning_rate": 2e-06,
+      "loss": 0.4116,
+      "step": 3592
+    },
+    {
+      "epoch": 0.8335459923442756,
+      "grad_norm": 11.797629439099808,
+      "learning_rate": 2e-06,
+      "loss": 0.2482,
+      "step": 3593
+    },
+    {
+      "epoch": 0.8337779839925763,
+      "grad_norm": 13.672341117810163,
+      "learning_rate": 2e-06,
+      "loss": 0.26,
+      "step": 3594
+    },
+    {
+      "epoch": 0.834009975640877,
+      "grad_norm": 13.89021157924789,
+      "learning_rate": 2e-06,
+      "loss": 0.3536,
+      "step": 3595
+    },
+    {
+      "epoch": 0.8342419672891775,
+      "grad_norm": 12.734349467657891,
+      "learning_rate": 2e-06,
+      "loss": 0.3681,
+      "step": 3596
+    },
+    {
+      "epoch": 0.8344739589374782,
+      "grad_norm": 18.167801879618423,
+      "learning_rate": 2e-06,
+      "loss": 0.2395,
+      "step": 3597
+    },
+    {
+      "epoch": 0.8347059505857789,
+      "grad_norm": 15.522534985390315,
+      "learning_rate": 2e-06,
+      "loss": 0.3314,
+      "step": 3598
+    },
+    {
+      "epoch": 0.8349379422340796,
+      "grad_norm": 14.490683555228062,
+      "learning_rate": 2e-06,
+      "loss": 0.2978,
+      "step": 3599
+    },
+    {
+      "epoch": 0.8351699338823803,
+      "grad_norm": 29.838603940900047,
+      "learning_rate": 2e-06,
+      "loss": 0.3134,
+      "step": 3600
+    },
+    {
+      "epoch": 0.8354019255306809,
+      "grad_norm": 13.116967226198804,
+      "learning_rate": 2e-06,
+      "loss": 0.2744,
+      "step": 3601
+    },
+    {
+      "epoch": 0.8356339171789816,
+      "grad_norm": 25.33370211615091,
+      "learning_rate": 2e-06,
+      "loss": 0.4356,
+      "step": 3602
+    },
+    {
+      "epoch": 0.8358659088272822,
+      "grad_norm": 16.302885312128048,
+      "learning_rate": 2e-06,
+      "loss": 0.3364,
+      "step": 3603
+    },
+    {
+      "epoch": 0.8360979004755829,
+      "grad_norm": 12.653664618715865,
+      "learning_rate": 2e-06,
+      "loss": 0.2252,
+      "step": 3604
+    },
+    {
+      "epoch": 0.8363298921238835,
+      "grad_norm": 11.957415500851111,
+      "learning_rate": 2e-06,
+      "loss": 0.2408,
+      "step": 3605
+    },
+    {
+      "epoch": 0.8365618837721842,
+      "grad_norm": 26.471764142537726,
+      "learning_rate": 2e-06,
+      "loss": 0.3768,
+      "step": 3606
+    },
+    {
+      "epoch": 0.8367938754204849,
+      "grad_norm": 17.280954916011552,
+      "learning_rate": 2e-06,
+      "loss": 0.3322,
+      "step": 3607
+    },
+    {
+      "epoch": 0.8370258670687856,
+      "grad_norm": 12.912623624553863,
+      "learning_rate": 2e-06,
+      "loss": 0.2971,
+      "step": 3608
+    },
+    {
+      "epoch": 0.8372578587170861,
+      "grad_norm": 33.409709385748855,
+      "learning_rate": 2e-06,
+      "loss": 0.3724,
+      "step": 3609
+    },
+    {
+      "epoch": 0.8374898503653868,
+      "grad_norm": 13.12016105283151,
+      "learning_rate": 2e-06,
+      "loss": 0.3083,
+      "step": 3610
+    },
+    {
+      "epoch": 0.8377218420136875,
+      "grad_norm": 21.16503577999807,
+      "learning_rate": 2e-06,
+      "loss": 0.4303,
+      "step": 3611
+    },
+    {
+      "epoch": 0.8379538336619882,
+      "grad_norm": 10.40878601387677,
+      "learning_rate": 2e-06,
+      "loss": 0.2138,
+      "step": 3612
+    },
+    {
+      "epoch": 0.8381858253102888,
+      "grad_norm": 19.2833065138601,
+      "learning_rate": 2e-06,
+      "loss": 0.269,
+      "step": 3613
+    },
+    {
+      "epoch": 0.8384178169585895,
+      "grad_norm": 17.850717804608752,
+      "learning_rate": 2e-06,
+      "loss": 0.3181,
+      "step": 3614
+    },
+    {
+      "epoch": 0.8386498086068902,
+      "grad_norm": 13.76177943878717,
+      "learning_rate": 2e-06,
+      "loss": 0.2724,
+      "step": 3615
+    },
+    {
+      "epoch": 0.8388818002551908,
+      "grad_norm": 13.036284766357339,
+      "learning_rate": 2e-06,
+      "loss": 0.2476,
+      "step": 3616
+    },
+    {
+      "epoch": 0.8391137919034914,
+      "grad_norm": 17.38148826319962,
+      "learning_rate": 2e-06,
+      "loss": 0.2757,
+      "step": 3617
+    },
+    {
+      "epoch": 0.8393457835517921,
+      "grad_norm": 10.903695405275533,
+      "learning_rate": 2e-06,
+      "loss": 0.2743,
+      "step": 3618
+    },
+    {
+      "epoch": 0.8395777752000928,
+      "grad_norm": 11.87283401552707,
+      "learning_rate": 2e-06,
+      "loss": 0.292,
+      "step": 3619
+    },
+    {
+      "epoch": 0.8398097668483935,
+      "grad_norm": 16.771758918256218,
+      "learning_rate": 2e-06,
+      "loss": 0.3851,
+      "step": 3620
+    },
+    {
+      "epoch": 0.8400417584966942,
+      "grad_norm": 13.719924046998493,
+      "learning_rate": 2e-06,
+      "loss": 0.2639,
+      "step": 3621
+    },
+    {
+      "epoch": 0.8402737501449948,
+      "grad_norm": 12.03787222608202,
+      "learning_rate": 2e-06,
+      "loss": 0.2875,
+      "step": 3622
+    },
+    {
+      "epoch": 0.8405057417932954,
+      "grad_norm": 12.36045663543364,
+      "learning_rate": 2e-06,
+      "loss": 0.199,
+      "step": 3623
+    },
+    {
+      "epoch": 0.8407377334415961,
+      "grad_norm": 11.3673385044902,
+      "learning_rate": 2e-06,
+      "loss": 0.2191,
+      "step": 3624
+    },
+    {
+      "epoch": 0.8409697250898968,
+      "grad_norm": 12.709416450778644,
+      "learning_rate": 2e-06,
+      "loss": 0.1901,
+      "step": 3625
+    },
+    {
+      "epoch": 0.8412017167381974,
+      "grad_norm": 9.564064452472941,
+      "learning_rate": 2e-06,
+      "loss": 0.2829,
+      "step": 3626
+    },
+    {
+      "epoch": 0.8414337083864981,
+      "grad_norm": 11.667053315828442,
+      "learning_rate": 2e-06,
+      "loss": 0.3535,
+      "step": 3627
+    },
+    {
+      "epoch": 0.8416657000347988,
+      "grad_norm": 18.12528031498164,
+      "learning_rate": 2e-06,
+      "loss": 0.2946,
+      "step": 3628
+    },
+    {
+      "epoch": 0.8418976916830994,
+      "grad_norm": 23.08275506339047,
+      "learning_rate": 2e-06,
+      "loss": 0.4012,
+      "step": 3629
+    },
+    {
+      "epoch": 0.8421296833314,
+      "grad_norm": 9.646394367795933,
+      "learning_rate": 2e-06,
+      "loss": 0.2934,
+      "step": 3630
+    },
+    {
+      "epoch": 0.8423616749797007,
+      "grad_norm": 20.793361796856317,
+      "learning_rate": 2e-06,
+      "loss": 0.3444,
+      "step": 3631
+    },
+    {
+      "epoch": 0.8425936666280014,
+      "grad_norm": 14.113988911611878,
+      "learning_rate": 2e-06,
+      "loss": 0.2877,
+      "step": 3632
+    },
+    {
+      "epoch": 0.8428256582763021,
+      "grad_norm": 8.786341654974642,
+      "learning_rate": 2e-06,
+      "loss": 0.2987,
+      "step": 3633
+    },
+    {
+      "epoch": 0.8430576499246027,
+      "grad_norm": 18.030693899144588,
+      "learning_rate": 2e-06,
+      "loss": 0.3672,
+      "step": 3634
+    },
+    {
+      "epoch": 0.8432896415729034,
+      "grad_norm": 13.25992214773615,
+      "learning_rate": 2e-06,
+      "loss": 0.3816,
+      "step": 3635
+    },
+    {
+      "epoch": 0.843521633221204,
+      "grad_norm": 12.693332638873756,
+      "learning_rate": 2e-06,
+      "loss": 0.2074,
+      "step": 3636
+    },
+    {
+      "epoch": 0.8437536248695047,
+      "grad_norm": 17.1773251818126,
+      "learning_rate": 2e-06,
+      "loss": 0.3135,
+      "step": 3637
+    },
+    {
+      "epoch": 0.8439856165178053,
+      "grad_norm": 16.218314383885915,
+      "learning_rate": 2e-06,
+      "loss": 0.288,
+      "step": 3638
+    },
+    {
+      "epoch": 0.844217608166106,
+      "grad_norm": 16.041706947386142,
+      "learning_rate": 2e-06,
+      "loss": 0.2734,
+      "step": 3639
+    },
+    {
+      "epoch": 0.8444495998144067,
+      "grad_norm": 9.535551188493375,
+      "learning_rate": 2e-06,
+      "loss": 0.248,
+      "step": 3640
+    },
+    {
+      "epoch": 0.8446815914627074,
+      "grad_norm": 18.716928330818007,
+      "learning_rate": 2e-06,
+      "loss": 0.3895,
+      "step": 3641
+    },
+    {
+      "epoch": 0.844913583111008,
+      "grad_norm": 11.504353512182753,
+      "learning_rate": 2e-06,
+      "loss": 0.2513,
+      "step": 3642
+    },
+    {
+      "epoch": 0.8451455747593086,
+      "grad_norm": 18.269519731683634,
+      "learning_rate": 2e-06,
+      "loss": 0.303,
+      "step": 3643
+    },
+    {
+      "epoch": 0.8453775664076093,
+      "grad_norm": 15.2935008617942,
+      "learning_rate": 2e-06,
+      "loss": 0.349,
+      "step": 3644
+    },
+    {
+      "epoch": 0.84560955805591,
+      "grad_norm": 11.958395285026274,
+      "learning_rate": 2e-06,
+      "loss": 0.2466,
+      "step": 3645
+    },
+    {
+      "epoch": 0.8458415497042107,
+      "grad_norm": 11.5462305491056,
+      "learning_rate": 2e-06,
+      "loss": 0.2421,
+      "step": 3646
+    },
+    {
+      "epoch": 0.8460735413525113,
+      "grad_norm": 9.021273192696786,
+      "learning_rate": 2e-06,
+      "loss": 0.2478,
+      "step": 3647
+    },
+    {
+      "epoch": 0.846305533000812,
+      "grad_norm": 18.029921292402488,
+      "learning_rate": 2e-06,
+      "loss": 0.3454,
+      "step": 3648
+    },
+    {
+      "epoch": 0.8465375246491126,
+      "grad_norm": 9.117078760672142,
+      "learning_rate": 2e-06,
+      "loss": 0.256,
+      "step": 3649
+    },
+    {
+      "epoch": 0.8467695162974133,
+      "grad_norm": 8.027104481169884,
+      "learning_rate": 2e-06,
+      "loss": 0.2244,
+      "step": 3650
+    },
+    {
+      "epoch": 0.8470015079457139,
+      "grad_norm": 18.36707474383036,
+      "learning_rate": 2e-06,
+      "loss": 0.3189,
+      "step": 3651
+    },
+    {
+      "epoch": 0.8472334995940146,
+      "grad_norm": 11.373520346001401,
+      "learning_rate": 2e-06,
+      "loss": 0.2491,
+      "step": 3652
+    },
+    {
+      "epoch": 0.8474654912423153,
+      "grad_norm": 9.023667037281498,
+      "learning_rate": 2e-06,
+      "loss": 0.236,
+      "step": 3653
+    },
+    {
+      "epoch": 0.847697482890616,
+      "grad_norm": 12.47557691705421,
+      "learning_rate": 2e-06,
+      "loss": 0.3199,
+      "step": 3654
+    },
+    {
+      "epoch": 0.8479294745389166,
+      "grad_norm": 11.849709139019977,
+      "learning_rate": 2e-06,
+      "loss": 0.312,
+      "step": 3655
+    },
+    {
+      "epoch": 0.8481614661872172,
+      "grad_norm": 16.08275240994812,
+      "learning_rate": 2e-06,
+      "loss": 0.2989,
+      "step": 3656
+    },
+    {
+      "epoch": 0.8483934578355179,
+      "grad_norm": 15.974269786013867,
+      "learning_rate": 2e-06,
+      "loss": 0.3295,
+      "step": 3657
+    },
+    {
+      "epoch": 0.8486254494838186,
+      "grad_norm": 12.325930745153997,
+      "learning_rate": 2e-06,
+      "loss": 0.2483,
+      "step": 3658
+    },
+    {
+      "epoch": 0.8488574411321193,
+      "grad_norm": 15.839887071663428,
+      "learning_rate": 2e-06,
+      "loss": 0.31,
+      "step": 3659
+    },
+    {
+      "epoch": 0.8490894327804199,
+      "grad_norm": 15.718046583011358,
+      "learning_rate": 2e-06,
+      "loss": 0.3434,
+      "step": 3660
+    },
+    {
+      "epoch": 0.8493214244287206,
+      "grad_norm": 21.360025144175662,
+      "learning_rate": 2e-06,
+      "loss": 0.3237,
+      "step": 3661
+    },
+    {
+      "epoch": 0.8495534160770212,
+      "grad_norm": 21.407456834013093,
+      "learning_rate": 2e-06,
+      "loss": 0.3511,
+      "step": 3662
+    },
+    {
+      "epoch": 0.8497854077253219,
+      "grad_norm": 17.516957487751252,
+      "learning_rate": 2e-06,
+      "loss": 0.2002,
+      "step": 3663
+    },
+    {
+      "epoch": 0.8500173993736225,
+      "grad_norm": 10.00063509585583,
+      "learning_rate": 2e-06,
+      "loss": 0.3101,
+      "step": 3664
+    },
+    {
+      "epoch": 0.8502493910219232,
+      "grad_norm": 16.61146615224927,
+      "learning_rate": 2e-06,
+      "loss": 0.2991,
+      "step": 3665
+    },
+    {
+      "epoch": 0.8504813826702239,
+      "grad_norm": 14.226566986205597,
+      "learning_rate": 2e-06,
+      "loss": 0.217,
+      "step": 3666
+    },
+    {
+      "epoch": 0.8507133743185246,
+      "grad_norm": 13.941877038153388,
+      "learning_rate": 2e-06,
+      "loss": 0.4054,
+      "step": 3667
+    },
+    {
+      "epoch": 0.8509453659668252,
+      "grad_norm": 20.923056394448633,
+      "learning_rate": 2e-06,
+      "loss": 0.2737,
+      "step": 3668
+    },
+    {
+      "epoch": 0.8511773576151258,
+      "grad_norm": 11.327595266683401,
+      "learning_rate": 2e-06,
+      "loss": 0.2025,
+      "step": 3669
+    },
+    {
+      "epoch": 0.8514093492634265,
+      "grad_norm": 13.082863132805368,
+      "learning_rate": 2e-06,
+      "loss": 0.2532,
+      "step": 3670
+    },
+    {
+      "epoch": 0.8516413409117272,
+      "grad_norm": 11.390712446630442,
+      "learning_rate": 2e-06,
+      "loss": 0.254,
+      "step": 3671
+    },
+    {
+      "epoch": 0.8518733325600278,
+      "grad_norm": 21.27549243664026,
+      "learning_rate": 2e-06,
+      "loss": 0.3924,
+      "step": 3672
+    },
+    {
+      "epoch": 0.8521053242083285,
+      "grad_norm": 21.3979323864136,
+      "learning_rate": 2e-06,
+      "loss": 0.2268,
+      "step": 3673
+    },
+    {
+      "epoch": 0.8523373158566292,
+      "grad_norm": 8.612003270380857,
+      "learning_rate": 2e-06,
+      "loss": 0.1924,
+      "step": 3674
+    },
+    {
+      "epoch": 0.8525693075049299,
+      "grad_norm": 19.820763966955873,
+      "learning_rate": 2e-06,
+      "loss": 0.4264,
+      "step": 3675
+    },
+    {
+      "epoch": 0.8528012991532304,
+      "grad_norm": 20.045730155740298,
+      "learning_rate": 2e-06,
+      "loss": 0.3346,
+      "step": 3676
+    },
+    {
+      "epoch": 0.8530332908015311,
+      "grad_norm": 13.650897189920228,
+      "learning_rate": 2e-06,
+      "loss": 0.3416,
+      "step": 3677
+    },
+    {
+      "epoch": 0.8532652824498318,
+      "grad_norm": 10.14256766403005,
+      "learning_rate": 2e-06,
+      "loss": 0.2986,
+      "step": 3678
+    },
+    {
+      "epoch": 0.8534972740981325,
+      "grad_norm": 10.037657539343313,
+      "learning_rate": 2e-06,
+      "loss": 0.2429,
+      "step": 3679
+    },
+    {
+      "epoch": 0.8537292657464332,
+      "grad_norm": 12.66315273931505,
+      "learning_rate": 2e-06,
+      "loss": 0.2382,
+      "step": 3680
+    },
+    {
+      "epoch": 0.8539612573947338,
+      "grad_norm": 9.312231073579792,
+      "learning_rate": 2e-06,
+      "loss": 0.2034,
+      "step": 3681
+    },
+    {
+      "epoch": 0.8541932490430344,
+      "grad_norm": 8.761478461055901,
+      "learning_rate": 2e-06,
+      "loss": 0.216,
+      "step": 3682
+    },
+    {
+      "epoch": 0.8544252406913351,
+      "grad_norm": 13.743741778322043,
+      "learning_rate": 2e-06,
+      "loss": 0.2603,
+      "step": 3683
+    },
+    {
+      "epoch": 0.8546572323396358,
+      "grad_norm": 8.56315291962295,
+      "learning_rate": 2e-06,
+      "loss": 0.1987,
+      "step": 3684
+    },
+    {
+      "epoch": 0.8548892239879364,
+      "grad_norm": 9.797116980529367,
+      "learning_rate": 2e-06,
+      "loss": 0.3647,
+      "step": 3685
+    },
+    {
+      "epoch": 0.8551212156362371,
+      "grad_norm": 13.622610342375989,
+      "learning_rate": 2e-06,
+      "loss": 0.2278,
+      "step": 3686
+    },
+    {
+      "epoch": 0.8553532072845378,
+      "grad_norm": 9.724405842358887,
+      "learning_rate": 2e-06,
+      "loss": 0.2339,
+      "step": 3687
+    },
+    {
+      "epoch": 0.8555851989328385,
+      "grad_norm": 9.427475360484458,
+      "learning_rate": 2e-06,
+      "loss": 0.1833,
+      "step": 3688
+    },
+    {
+      "epoch": 0.855817190581139,
+      "grad_norm": 23.056746456756965,
+      "learning_rate": 2e-06,
+      "loss": 0.3746,
+      "step": 3689
+    },
+    {
+      "epoch": 0.8560491822294397,
+      "grad_norm": 8.420394491891004,
+      "learning_rate": 2e-06,
+      "loss": 0.2763,
+      "step": 3690
+    },
+    {
+      "epoch": 0.8562811738777404,
+      "grad_norm": 17.380813276857875,
+      "learning_rate": 2e-06,
+      "loss": 0.2846,
+      "step": 3691
+    },
+    {
+      "epoch": 0.8565131655260411,
+      "grad_norm": 13.749892547660071,
+      "learning_rate": 2e-06,
+      "loss": 0.2235,
+      "step": 3692
+    },
+    {
+      "epoch": 0.8567451571743417,
+      "grad_norm": 13.50948624508527,
+      "learning_rate": 2e-06,
+      "loss": 0.3027,
+      "step": 3693
+    },
+    {
+      "epoch": 0.8569771488226424,
+      "grad_norm": 16.78402828019823,
+      "learning_rate": 2e-06,
+      "loss": 0.361,
+      "step": 3694
+    },
+    {
+      "epoch": 0.8572091404709431,
+      "grad_norm": 10.7775988342208,
+      "learning_rate": 2e-06,
+      "loss": 0.2344,
+      "step": 3695
+    },
+    {
+      "epoch": 0.8574411321192437,
+      "grad_norm": 17.91160614351676,
+      "learning_rate": 2e-06,
+      "loss": 0.4287,
+      "step": 3696
+    },
+    {
+      "epoch": 0.8576731237675443,
+      "grad_norm": 10.208952391466227,
+      "learning_rate": 2e-06,
+      "loss": 0.1963,
+      "step": 3697
+    },
+    {
+      "epoch": 0.857905115415845,
+      "grad_norm": 12.492443606261487,
+      "learning_rate": 2e-06,
+      "loss": 0.1823,
+      "step": 3698
+    },
+    {
+      "epoch": 0.8581371070641457,
+      "grad_norm": 7.964969511028492,
+      "learning_rate": 2e-06,
+      "loss": 0.1687,
+      "step": 3699
+    },
+    {
+      "epoch": 0.8583690987124464,
+      "grad_norm": 7.738033256851267,
+      "learning_rate": 2e-06,
+      "loss": 0.1877,
+      "step": 3700
+    },
+    {
+      "epoch": 0.858601090360747,
+      "grad_norm": 13.255320400154975,
+      "learning_rate": 2e-06,
+      "loss": 0.3013,
+      "step": 3701
+    },
+    {
+      "epoch": 0.8588330820090476,
+      "grad_norm": 12.671403201642306,
+      "learning_rate": 2e-06,
+      "loss": 0.2109,
+      "step": 3702
+    },
+    {
+      "epoch": 0.8590650736573483,
+      "grad_norm": 11.814382103135062,
+      "learning_rate": 2e-06,
+      "loss": 0.1372,
+      "step": 3703
+    },
+    {
+      "epoch": 0.859297065305649,
+      "grad_norm": 13.982761604963033,
+      "learning_rate": 2e-06,
+      "loss": 0.32,
+      "step": 3704
+    },
+    {
+      "epoch": 0.8595290569539497,
+      "grad_norm": 14.133755063373435,
+      "learning_rate": 2e-06,
+      "loss": 0.2347,
+      "step": 3705
+    },
+    {
+      "epoch": 0.8597610486022503,
+      "grad_norm": 8.462385181045686,
+      "learning_rate": 2e-06,
+      "loss": 0.2493,
+      "step": 3706
+    },
+    {
+      "epoch": 0.859993040250551,
+      "grad_norm": 14.085965020035298,
+      "learning_rate": 2e-06,
+      "loss": 0.3511,
+      "step": 3707
+    },
+    {
+      "epoch": 0.8602250318988517,
+      "grad_norm": 58.65048537341041,
+      "learning_rate": 2e-06,
+      "loss": 0.3468,
+      "step": 3708
+    },
+    {
+      "epoch": 0.8604570235471523,
+      "grad_norm": 13.616616688476421,
+      "learning_rate": 2e-06,
+      "loss": 0.2393,
+      "step": 3709
+    },
+    {
+      "epoch": 0.8606890151954529,
+      "grad_norm": 12.441324389887878,
+      "learning_rate": 2e-06,
+      "loss": 0.2537,
+      "step": 3710
+    },
+    {
+      "epoch": 0.8609210068437536,
+      "grad_norm": 14.305380497619064,
+      "learning_rate": 2e-06,
+      "loss": 0.2861,
+      "step": 3711
+    },
+    {
+      "epoch": 0.8611529984920543,
+      "grad_norm": 16.613861586038418,
+      "learning_rate": 2e-06,
+      "loss": 0.2142,
+      "step": 3712
+    },
+    {
+      "epoch": 0.861384990140355,
+      "grad_norm": 10.199542141395073,
+      "learning_rate": 2e-06,
+      "loss": 0.31,
+      "step": 3713
+    },
+    {
+      "epoch": 0.8616169817886556,
+      "grad_norm": 12.55464480155109,
+      "learning_rate": 2e-06,
+      "loss": 0.1862,
+      "step": 3714
+    },
+    {
+      "epoch": 0.8618489734369563,
+      "grad_norm": 16.10525020587473,
+      "learning_rate": 2e-06,
+      "loss": 0.3392,
+      "step": 3715
+    },
+    {
+      "epoch": 0.8620809650852569,
+      "grad_norm": 13.438956564440527,
+      "learning_rate": 2e-06,
+      "loss": 0.2419,
+      "step": 3716
+    },
+    {
+      "epoch": 0.8623129567335576,
+      "grad_norm": 9.808457425512904,
+      "learning_rate": 2e-06,
+      "loss": 0.2457,
+      "step": 3717
+    },
+    {
+      "epoch": 0.8625449483818582,
+      "grad_norm": 14.98954005495591,
+      "learning_rate": 2e-06,
+      "loss": 0.2695,
+      "step": 3718
+    },
+    {
+      "epoch": 0.8627769400301589,
+      "grad_norm": 14.064247216948907,
+      "learning_rate": 2e-06,
+      "loss": 0.2397,
+      "step": 3719
+    },
+    {
+      "epoch": 0.8630089316784596,
+      "grad_norm": 31.850410637664915,
+      "learning_rate": 2e-06,
+      "loss": 0.5098,
+      "step": 3720
+    },
+    {
+      "epoch": 0.8632409233267603,
+      "grad_norm": 27.13004682039094,
+      "learning_rate": 2e-06,
+      "loss": 0.2238,
+      "step": 3721
+    },
+    {
+      "epoch": 0.8634729149750608,
+      "grad_norm": 12.81414870338563,
+      "learning_rate": 2e-06,
+      "loss": 0.28,
+      "step": 3722
+    },
+    {
+      "epoch": 0.8637049066233615,
+      "grad_norm": 14.876898239075308,
+      "learning_rate": 2e-06,
+      "loss": 0.3276,
+      "step": 3723
+    },
+    {
+      "epoch": 0.8639368982716622,
+      "grad_norm": 13.039596946514822,
+      "learning_rate": 2e-06,
+      "loss": 0.3539,
+      "step": 3724
+    },
+    {
+      "epoch": 0.8641688899199629,
+      "grad_norm": 11.520663900550497,
+      "learning_rate": 2e-06,
+      "loss": 0.2403,
+      "step": 3725
+    },
+    {
+      "epoch": 0.8644008815682636,
+      "grad_norm": 20.113972396436466,
+      "learning_rate": 2e-06,
+      "loss": 0.2682,
+      "step": 3726
+    },
+    {
+      "epoch": 0.8646328732165642,
+      "grad_norm": 9.768363407171256,
+      "learning_rate": 2e-06,
+      "loss": 0.3367,
+      "step": 3727
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 9.754462011322826,
+      "learning_rate": 2e-06,
+      "loss": 0.1976,
+      "step": 3728
+    },
+    {
+      "epoch": 0.8650968565131655,
+      "grad_norm": 15.897468817722482,
+      "learning_rate": 2e-06,
+      "loss": 0.3516,
+      "step": 3729
+    },
+    {
+      "epoch": 0.8653288481614662,
+      "grad_norm": 12.539672362894047,
+      "learning_rate": 2e-06,
+      "loss": 0.2137,
+      "step": 3730
+    },
+    {
+      "epoch": 0.8655608398097668,
+      "grad_norm": 22.675675973677986,
+      "learning_rate": 2e-06,
+      "loss": 0.3773,
+      "step": 3731
+    },
+    {
+      "epoch": 0.8657928314580675,
+      "grad_norm": 21.084484439088442,
+      "learning_rate": 2e-06,
+      "loss": 0.2725,
+      "step": 3732
+    },
+    {
+      "epoch": 0.8660248231063682,
+      "grad_norm": 12.945612595748198,
+      "learning_rate": 2e-06,
+      "loss": 0.2171,
+      "step": 3733
+    },
+    {
+      "epoch": 0.8662568147546689,
+      "grad_norm": 10.168197656317433,
+      "learning_rate": 2e-06,
+      "loss": 0.2676,
+      "step": 3734
+    },
+    {
+      "epoch": 0.8664888064029695,
+      "grad_norm": 12.095214911594292,
+      "learning_rate": 2e-06,
+      "loss": 0.2778,
+      "step": 3735
+    },
+    {
+      "epoch": 0.8667207980512701,
+      "grad_norm": 17.38157046620352,
+      "learning_rate": 2e-06,
+      "loss": 0.3717,
+      "step": 3736
+    },
+    {
+      "epoch": 0.8669527896995708,
+      "grad_norm": 21.62919673063253,
+      "learning_rate": 2e-06,
+      "loss": 0.2156,
+      "step": 3737
+    },
+    {
+      "epoch": 0.8671847813478715,
+      "grad_norm": 11.91250814736637,
+      "learning_rate": 2e-06,
+      "loss": 0.2699,
+      "step": 3738
+    },
+    {
+      "epoch": 0.8674167729961721,
+      "grad_norm": 14.675029900401139,
+      "learning_rate": 2e-06,
+      "loss": 0.1935,
+      "step": 3739
+    },
+    {
+      "epoch": 0.8676487646444728,
+      "grad_norm": 15.028454890183673,
+      "learning_rate": 2e-06,
+      "loss": 0.2753,
+      "step": 3740
+    },
+    {
+      "epoch": 0.8678807562927735,
+      "grad_norm": 12.685941107430299,
+      "learning_rate": 2e-06,
+      "loss": 0.3368,
+      "step": 3741
+    },
+    {
+      "epoch": 0.8681127479410741,
+      "grad_norm": 14.086959911298255,
+      "learning_rate": 2e-06,
+      "loss": 0.3123,
+      "step": 3742
+    },
+    {
+      "epoch": 0.8683447395893747,
+      "grad_norm": 29.59069253850969,
+      "learning_rate": 2e-06,
+      "loss": 0.3462,
+      "step": 3743
+    },
+    {
+      "epoch": 0.8685767312376754,
+      "grad_norm": 15.70500076917416,
+      "learning_rate": 2e-06,
+      "loss": 0.2672,
+      "step": 3744
+    },
+    {
+      "epoch": 0.8688087228859761,
+      "grad_norm": 10.885835318229043,
+      "learning_rate": 2e-06,
+      "loss": 0.2552,
+      "step": 3745
+    },
+    {
+      "epoch": 0.8690407145342768,
+      "grad_norm": 36.85881948138301,
+      "learning_rate": 2e-06,
+      "loss": 0.3901,
+      "step": 3746
+    },
+    {
+      "epoch": 0.8692727061825775,
+      "grad_norm": 16.53745340862496,
+      "learning_rate": 2e-06,
+      "loss": 0.2888,
+      "step": 3747
+    },
+    {
+      "epoch": 0.8695046978308781,
+      "grad_norm": 16.261524994880595,
+      "learning_rate": 2e-06,
+      "loss": 0.4434,
+      "step": 3748
+    },
+    {
+      "epoch": 0.8697366894791787,
+      "grad_norm": 15.966881686175501,
+      "learning_rate": 2e-06,
+      "loss": 0.2976,
+      "step": 3749
+    },
+    {
+      "epoch": 0.8699686811274794,
+      "grad_norm": 24.370815496080688,
+      "learning_rate": 2e-06,
+      "loss": 0.3686,
+      "step": 3750
+    },
+    {
+      "epoch": 0.8702006727757801,
+      "grad_norm": 14.66934237706017,
+      "learning_rate": 2e-06,
+      "loss": 0.2677,
+      "step": 3751
+    },
+    {
+      "epoch": 0.8704326644240807,
+      "grad_norm": 13.86784165636144,
+      "learning_rate": 2e-06,
+      "loss": 0.2662,
+      "step": 3752
+    },
+    {
+      "epoch": 0.8706646560723814,
+      "grad_norm": 22.526612910044197,
+      "learning_rate": 2e-06,
+      "loss": 0.3681,
+      "step": 3753
+    },
+    {
+      "epoch": 0.8708966477206821,
+      "grad_norm": 10.304405082274162,
+      "learning_rate": 2e-06,
+      "loss": 0.3004,
+      "step": 3754
+    },
+    {
+      "epoch": 0.8711286393689828,
+      "grad_norm": 19.197447189792538,
+      "learning_rate": 2e-06,
+      "loss": 0.3221,
+      "step": 3755
+    },
+    {
+      "epoch": 0.8713606310172833,
+      "grad_norm": 12.457718761366353,
+      "learning_rate": 2e-06,
+      "loss": 0.2205,
+      "step": 3756
+    },
+    {
+      "epoch": 0.871592622665584,
+      "grad_norm": 12.306067228478893,
+      "learning_rate": 2e-06,
+      "loss": 0.3971,
+      "step": 3757
+    },
+    {
+      "epoch": 0.8718246143138847,
+      "grad_norm": 18.56590997063515,
+      "learning_rate": 2e-06,
+      "loss": 0.2905,
+      "step": 3758
+    },
+    {
+      "epoch": 0.8720566059621854,
+      "grad_norm": 18.02352978933951,
+      "learning_rate": 2e-06,
+      "loss": 0.3457,
+      "step": 3759
+    },
+    {
+      "epoch": 0.872288597610486,
+      "grad_norm": 14.420142710853485,
+      "learning_rate": 2e-06,
+      "loss": 0.2826,
+      "step": 3760
+    },
+    {
+      "epoch": 0.8725205892587867,
+      "grad_norm": 6.153364462060262,
+      "learning_rate": 2e-06,
+      "loss": 0.2808,
+      "step": 3761
+    },
+    {
+      "epoch": 0.8727525809070873,
+      "grad_norm": 7.722426387420128,
+      "learning_rate": 2e-06,
+      "loss": 0.2435,
+      "step": 3762
+    },
+    {
+      "epoch": 0.872984572555388,
+      "grad_norm": 12.084046865494983,
+      "learning_rate": 2e-06,
+      "loss": 0.297,
+      "step": 3763
+    },
+    {
+      "epoch": 0.8732165642036886,
+      "grad_norm": 14.772204290356598,
+      "learning_rate": 2e-06,
+      "loss": 0.2052,
+      "step": 3764
+    },
+    {
+      "epoch": 0.8734485558519893,
+      "grad_norm": 18.298133333640326,
+      "learning_rate": 2e-06,
+      "loss": 0.3385,
+      "step": 3765
+    },
+    {
+      "epoch": 0.87368054750029,
+      "grad_norm": 8.033540187695454,
+      "learning_rate": 2e-06,
+      "loss": 0.2035,
+      "step": 3766
+    },
+    {
+      "epoch": 0.8739125391485907,
+      "grad_norm": 14.501022464289226,
+      "learning_rate": 2e-06,
+      "loss": 0.3331,
+      "step": 3767
+    },
+    {
+      "epoch": 0.8741445307968914,
+      "grad_norm": 8.580235421829364,
+      "learning_rate": 2e-06,
+      "loss": 0.2644,
+      "step": 3768
+    },
+    {
+      "epoch": 0.8743765224451919,
+      "grad_norm": 8.855536509852117,
+      "learning_rate": 2e-06,
+      "loss": 0.2153,
+      "step": 3769
+    },
+    {
+      "epoch": 0.8746085140934926,
+      "grad_norm": 11.36016770749889,
+      "learning_rate": 2e-06,
+      "loss": 0.2991,
+      "step": 3770
+    },
+    {
+      "epoch": 0.8748405057417933,
+      "grad_norm": 13.635951353515276,
+      "learning_rate": 2e-06,
+      "loss": 0.3342,
+      "step": 3771
+    },
+    {
+      "epoch": 0.875072497390094,
+      "grad_norm": 7.944159507447317,
+      "learning_rate": 2e-06,
+      "loss": 0.253,
+      "step": 3772
+    },
+    {
+      "epoch": 0.8753044890383946,
+      "grad_norm": 11.172510325456205,
+      "learning_rate": 2e-06,
+      "loss": 0.2876,
+      "step": 3773
+    },
+    {
+      "epoch": 0.8755364806866953,
+      "grad_norm": 18.73017899852579,
+      "learning_rate": 2e-06,
+      "loss": 0.3334,
+      "step": 3774
+    },
+    {
+      "epoch": 0.875768472334996,
+      "grad_norm": 10.47454924116766,
+      "learning_rate": 2e-06,
+      "loss": 0.2581,
+      "step": 3775
+    },
+    {
+      "epoch": 0.8760004639832966,
+      "grad_norm": 7.591698553448666,
+      "learning_rate": 2e-06,
+      "loss": 0.2232,
+      "step": 3776
+    },
+    {
+      "epoch": 0.8762324556315972,
+      "grad_norm": 16.266482505616455,
+      "learning_rate": 2e-06,
+      "loss": 0.3319,
+      "step": 3777
+    },
+    {
+      "epoch": 0.8764644472798979,
+      "grad_norm": 7.2554270371657354,
+      "learning_rate": 2e-06,
+      "loss": 0.1495,
+      "step": 3778
+    },
+    {
+      "epoch": 0.8766964389281986,
+      "grad_norm": 19.825695220731603,
+      "learning_rate": 2e-06,
+      "loss": 0.3339,
+      "step": 3779
+    },
+    {
+      "epoch": 0.8769284305764993,
+      "grad_norm": 13.50315788016503,
+      "learning_rate": 2e-06,
+      "loss": 0.2265,
+      "step": 3780
+    },
+    {
+      "epoch": 0.8771604222248,
+      "grad_norm": 9.889392871091157,
+      "learning_rate": 2e-06,
+      "loss": 0.3026,
+      "step": 3781
+    },
+    {
+      "epoch": 0.8773924138731005,
+      "grad_norm": 15.48773175623725,
+      "learning_rate": 2e-06,
+      "loss": 0.2651,
+      "step": 3782
+    },
+    {
+      "epoch": 0.8776244055214012,
+      "grad_norm": 14.822191847325213,
+      "learning_rate": 2e-06,
+      "loss": 0.2641,
+      "step": 3783
+    },
+    {
+      "epoch": 0.8778563971697019,
+      "grad_norm": 13.547907543836468,
+      "learning_rate": 2e-06,
+      "loss": 0.3234,
+      "step": 3784
+    },
+    {
+      "epoch": 0.8780883888180026,
+      "grad_norm": 14.875134126943225,
+      "learning_rate": 2e-06,
+      "loss": 0.2694,
+      "step": 3785
+    },
+    {
+      "epoch": 0.8783203804663032,
+      "grad_norm": 20.39451897252101,
+      "learning_rate": 2e-06,
+      "loss": 0.2972,
+      "step": 3786
+    },
+    {
+      "epoch": 0.8785523721146039,
+      "grad_norm": 17.90206206999193,
+      "learning_rate": 2e-06,
+      "loss": 0.3046,
+      "step": 3787
+    },
+    {
+      "epoch": 0.8787843637629046,
+      "grad_norm": 12.566470812398759,
+      "learning_rate": 2e-06,
+      "loss": 0.2508,
+      "step": 3788
+    },
+    {
+      "epoch": 0.8790163554112052,
+      "grad_norm": 8.893414440354205,
+      "learning_rate": 2e-06,
+      "loss": 0.294,
+      "step": 3789
+    },
+    {
+      "epoch": 0.8792483470595058,
+      "grad_norm": 19.14245934946466,
+      "learning_rate": 2e-06,
+      "loss": 0.3494,
+      "step": 3790
+    },
+    {
+      "epoch": 0.8794803387078065,
+      "grad_norm": 15.10092811690474,
+      "learning_rate": 2e-06,
+      "loss": 0.3559,
+      "step": 3791
+    },
+    {
+      "epoch": 0.8797123303561072,
+      "grad_norm": 10.508681437132152,
+      "learning_rate": 2e-06,
+      "loss": 0.2268,
+      "step": 3792
+    },
+    {
+      "epoch": 0.8799443220044079,
+      "grad_norm": 21.07249450103903,
+      "learning_rate": 2e-06,
+      "loss": 0.245,
+      "step": 3793
+    },
+    {
+      "epoch": 0.8801763136527085,
+      "grad_norm": 17.65752075276022,
+      "learning_rate": 2e-06,
+      "loss": 0.269,
+      "step": 3794
+    },
+    {
+      "epoch": 0.8804083053010091,
+      "grad_norm": 17.466559161419354,
+      "learning_rate": 2e-06,
+      "loss": 0.2853,
+      "step": 3795
+    },
+    {
+      "epoch": 0.8806402969493098,
+      "grad_norm": 17.14933191795175,
+      "learning_rate": 2e-06,
+      "loss": 0.2767,
+      "step": 3796
+    },
+    {
+      "epoch": 0.8808722885976105,
+      "grad_norm": 14.78273941840289,
+      "learning_rate": 2e-06,
+      "loss": 0.3052,
+      "step": 3797
+    },
+    {
+      "epoch": 0.8811042802459111,
+      "grad_norm": 12.53725963656686,
+      "learning_rate": 2e-06,
+      "loss": 0.2303,
+      "step": 3798
+    },
+    {
+      "epoch": 0.8813362718942118,
+      "grad_norm": 11.924772982478647,
+      "learning_rate": 2e-06,
+      "loss": 0.3213,
+      "step": 3799
+    },
+    {
+      "epoch": 0.8815682635425125,
+      "grad_norm": 11.76335039639468,
+      "learning_rate": 2e-06,
+      "loss": 0.2632,
+      "step": 3800
+    },
+    {
+      "epoch": 0.8818002551908132,
+      "grad_norm": 19.85232422623424,
+      "learning_rate": 2e-06,
+      "loss": 0.3021,
+      "step": 3801
+    },
+    {
+      "epoch": 0.8820322468391137,
+      "grad_norm": 23.65456243506742,
+      "learning_rate": 2e-06,
+      "loss": 0.3246,
+      "step": 3802
+    },
+    {
+      "epoch": 0.8822642384874144,
+      "grad_norm": 5.716607208165604,
+      "learning_rate": 2e-06,
+      "loss": 0.2176,
+      "step": 3803
+    },
+    {
+      "epoch": 0.8824962301357151,
+      "grad_norm": 6.022036608263883,
+      "learning_rate": 2e-06,
+      "loss": 0.1139,
+      "step": 3804
+    },
+    {
+      "epoch": 0.8827282217840158,
+      "grad_norm": 16.350486842879306,
+      "learning_rate": 2e-06,
+      "loss": 0.3059,
+      "step": 3805
+    },
+    {
+      "epoch": 0.8829602134323165,
+      "grad_norm": 15.069408373253298,
+      "learning_rate": 2e-06,
+      "loss": 0.2275,
+      "step": 3806
+    },
+    {
+      "epoch": 0.8831922050806171,
+      "grad_norm": 10.384160567512737,
+      "learning_rate": 2e-06,
+      "loss": 0.2738,
+      "step": 3807
+    },
+    {
+      "epoch": 0.8834241967289178,
+      "grad_norm": 16.660880271426983,
+      "learning_rate": 2e-06,
+      "loss": 0.2554,
+      "step": 3808
+    },
+    {
+      "epoch": 0.8836561883772184,
+      "grad_norm": 18.374192009215328,
+      "learning_rate": 2e-06,
+      "loss": 0.3591,
+      "step": 3809
+    },
+    {
+      "epoch": 0.883888180025519,
+      "grad_norm": 12.872310982621023,
+      "learning_rate": 2e-06,
+      "loss": 0.2772,
+      "step": 3810
+    },
+    {
+      "epoch": 0.8841201716738197,
+      "grad_norm": 16.019644899210274,
+      "learning_rate": 2e-06,
+      "loss": 0.5156,
+      "step": 3811
+    },
+    {
+      "epoch": 0.8843521633221204,
+      "grad_norm": 12.442141234475963,
+      "learning_rate": 2e-06,
+      "loss": 0.2239,
+      "step": 3812
+    },
+    {
+      "epoch": 0.8845841549704211,
+      "grad_norm": 22.525745572070075,
+      "learning_rate": 2e-06,
+      "loss": 0.3765,
+      "step": 3813
+    },
+    {
+      "epoch": 0.8848161466187218,
+      "grad_norm": 16.304838685551868,
+      "learning_rate": 2e-06,
+      "loss": 0.4056,
+      "step": 3814
+    },
+    {
+      "epoch": 0.8850481382670223,
+      "grad_norm": 7.3709776680881705,
+      "learning_rate": 2e-06,
+      "loss": 0.1974,
+      "step": 3815
+    },
+    {
+      "epoch": 0.885280129915323,
+      "grad_norm": 13.210776449554263,
+      "learning_rate": 2e-06,
+      "loss": 0.225,
+      "step": 3816
+    },
+    {
+      "epoch": 0.8855121215636237,
+      "grad_norm": 16.99200282899661,
+      "learning_rate": 2e-06,
+      "loss": 0.2187,
+      "step": 3817
+    },
+    {
+      "epoch": 0.8857441132119244,
+      "grad_norm": 14.484274148768987,
+      "learning_rate": 2e-06,
+      "loss": 0.2841,
+      "step": 3818
+    },
+    {
+      "epoch": 0.885976104860225,
+      "grad_norm": 13.690872050877267,
+      "learning_rate": 2e-06,
+      "loss": 0.3322,
+      "step": 3819
+    },
+    {
+      "epoch": 0.8862080965085257,
+      "grad_norm": 15.21118667258607,
+      "learning_rate": 2e-06,
+      "loss": 0.3085,
+      "step": 3820
+    },
+    {
+      "epoch": 0.8864400881568264,
+      "grad_norm": 11.234375561403914,
+      "learning_rate": 2e-06,
+      "loss": 0.221,
+      "step": 3821
+    },
+    {
+      "epoch": 0.886672079805127,
+      "grad_norm": 11.309069358024804,
+      "learning_rate": 2e-06,
+      "loss": 0.2239,
+      "step": 3822
+    },
+    {
+      "epoch": 0.8869040714534276,
+      "grad_norm": 9.480559997389527,
+      "learning_rate": 2e-06,
+      "loss": 0.4035,
+      "step": 3823
+    },
+    {
+      "epoch": 0.8871360631017283,
+      "grad_norm": 12.818376026225353,
+      "learning_rate": 2e-06,
+      "loss": 0.2866,
+      "step": 3824
+    },
+    {
+      "epoch": 0.887368054750029,
+      "grad_norm": 13.600698803419629,
+      "learning_rate": 2e-06,
+      "loss": 0.2284,
+      "step": 3825
+    },
+    {
+      "epoch": 0.8876000463983297,
+      "grad_norm": 12.323299185200682,
+      "learning_rate": 2e-06,
+      "loss": 0.2643,
+      "step": 3826
+    },
+    {
+      "epoch": 0.8878320380466304,
+      "grad_norm": 16.262470843816768,
+      "learning_rate": 2e-06,
+      "loss": 0.3144,
+      "step": 3827
+    },
+    {
+      "epoch": 0.888064029694931,
+      "grad_norm": 15.8268984032703,
+      "learning_rate": 2e-06,
+      "loss": 0.2297,
+      "step": 3828
+    },
+    {
+      "epoch": 0.8882960213432316,
+      "grad_norm": 15.451969149107263,
+      "learning_rate": 2e-06,
+      "loss": 0.1857,
+      "step": 3829
+    },
+    {
+      "epoch": 0.8885280129915323,
+      "grad_norm": 13.444047438579203,
+      "learning_rate": 2e-06,
+      "loss": 0.2971,
+      "step": 3830
+    },
+    {
+      "epoch": 0.888760004639833,
+      "grad_norm": 11.468701419479576,
+      "learning_rate": 2e-06,
+      "loss": 0.2492,
+      "step": 3831
+    },
+    {
+      "epoch": 0.8889919962881336,
+      "grad_norm": 21.788790413025257,
+      "learning_rate": 2e-06,
+      "loss": 0.2497,
+      "step": 3832
+    },
+    {
+      "epoch": 0.8892239879364343,
+      "grad_norm": 8.534053601552444,
+      "learning_rate": 2e-06,
+      "loss": 0.2748,
+      "step": 3833
+    },
+    {
+      "epoch": 0.889455979584735,
+      "grad_norm": 17.14812910368447,
+      "learning_rate": 2e-06,
+      "loss": 0.3216,
+      "step": 3834
+    },
+    {
+      "epoch": 0.8896879712330356,
+      "grad_norm": 10.518702926059186,
+      "learning_rate": 2e-06,
+      "loss": 0.1831,
+      "step": 3835
+    },
+    {
+      "epoch": 0.8899199628813362,
+      "grad_norm": 15.381986422849188,
+      "learning_rate": 2e-06,
+      "loss": 0.2893,
+      "step": 3836
+    },
+    {
+      "epoch": 0.8901519545296369,
+      "grad_norm": 21.466599789047734,
+      "learning_rate": 2e-06,
+      "loss": 0.4283,
+      "step": 3837
+    },
+    {
+      "epoch": 0.8903839461779376,
+      "grad_norm": 18.883265852737733,
+      "learning_rate": 2e-06,
+      "loss": 0.3159,
+      "step": 3838
+    },
+    {
+      "epoch": 0.8906159378262383,
+      "grad_norm": 9.604683531122912,
+      "learning_rate": 2e-06,
+      "loss": 0.293,
+      "step": 3839
+    },
+    {
+      "epoch": 0.890847929474539,
+      "grad_norm": 10.466214318640201,
+      "learning_rate": 2e-06,
+      "loss": 0.2093,
+      "step": 3840
+    },
+    {
+      "epoch": 0.8910799211228396,
+      "grad_norm": 30.045014770740092,
+      "learning_rate": 2e-06,
+      "loss": 0.3721,
+      "step": 3841
+    },
+    {
+      "epoch": 0.8913119127711402,
+      "grad_norm": 17.520621297975964,
+      "learning_rate": 2e-06,
+      "loss": 0.3136,
+      "step": 3842
+    },
+    {
+      "epoch": 0.8915439044194409,
+      "grad_norm": 10.430427280788878,
+      "learning_rate": 2e-06,
+      "loss": 0.3048,
+      "step": 3843
+    },
+    {
+      "epoch": 0.8917758960677415,
+      "grad_norm": 16.52127291757037,
+      "learning_rate": 2e-06,
+      "loss": 0.2781,
+      "step": 3844
+    },
+    {
+      "epoch": 0.8920078877160422,
+      "grad_norm": 14.03793238580409,
+      "learning_rate": 2e-06,
+      "loss": 0.3499,
+      "step": 3845
+    },
+    {
+      "epoch": 0.8922398793643429,
+      "grad_norm": 13.465497188671403,
+      "learning_rate": 2e-06,
+      "loss": 0.2369,
+      "step": 3846
+    },
+    {
+      "epoch": 0.8924718710126436,
+      "grad_norm": 16.203513528553657,
+      "learning_rate": 2e-06,
+      "loss": 0.3037,
+      "step": 3847
+    },
+    {
+      "epoch": 0.8927038626609443,
+      "grad_norm": 16.91252227614217,
+      "learning_rate": 2e-06,
+      "loss": 0.3807,
+      "step": 3848
+    },
+    {
+      "epoch": 0.8929358543092448,
+      "grad_norm": 15.13921064541899,
+      "learning_rate": 2e-06,
+      "loss": 0.2367,
+      "step": 3849
+    },
+    {
+      "epoch": 0.8931678459575455,
+      "grad_norm": 17.470809701462834,
+      "learning_rate": 2e-06,
+      "loss": 0.4561,
+      "step": 3850
+    },
+    {
+      "epoch": 0.8933998376058462,
+      "grad_norm": 17.55558022042597,
+      "learning_rate": 2e-06,
+      "loss": 0.2008,
+      "step": 3851
+    },
+    {
+      "epoch": 0.8936318292541469,
+      "grad_norm": 15.181061039277026,
+      "learning_rate": 2e-06,
+      "loss": 0.312,
+      "step": 3852
+    },
+    {
+      "epoch": 0.8938638209024475,
+      "grad_norm": 16.306178080300985,
+      "learning_rate": 2e-06,
+      "loss": 0.3254,
+      "step": 3853
+    },
+    {
+      "epoch": 0.8940958125507482,
+      "grad_norm": 13.315390126795844,
+      "learning_rate": 2e-06,
+      "loss": 0.1952,
+      "step": 3854
+    },
+    {
+      "epoch": 0.8943278041990488,
+      "grad_norm": 14.600753372967226,
+      "learning_rate": 2e-06,
+      "loss": 0.2792,
+      "step": 3855
+    },
+    {
+      "epoch": 0.8945597958473495,
+      "grad_norm": 25.411581518678336,
+      "learning_rate": 2e-06,
+      "loss": 0.3914,
+      "step": 3856
+    },
+    {
+      "epoch": 0.8947917874956501,
+      "grad_norm": 14.153992756140097,
+      "learning_rate": 2e-06,
+      "loss": 0.3199,
+      "step": 3857
+    },
+    {
+      "epoch": 0.8950237791439508,
+      "grad_norm": 30.60125428323226,
+      "learning_rate": 2e-06,
+      "loss": 0.3962,
+      "step": 3858
+    },
+    {
+      "epoch": 0.8952557707922515,
+      "grad_norm": 8.730276553996347,
+      "learning_rate": 2e-06,
+      "loss": 0.2931,
+      "step": 3859
+    },
+    {
+      "epoch": 0.8954877624405522,
+      "grad_norm": 23.977248564892673,
+      "learning_rate": 2e-06,
+      "loss": 0.3746,
+      "step": 3860
+    },
+    {
+      "epoch": 0.8957197540888528,
+      "grad_norm": 12.272894635641682,
+      "learning_rate": 2e-06,
+      "loss": 0.1781,
+      "step": 3861
+    },
+    {
+      "epoch": 0.8959517457371534,
+      "grad_norm": 10.266127690588375,
+      "learning_rate": 2e-06,
+      "loss": 0.198,
+      "step": 3862
+    },
+    {
+      "epoch": 0.8961837373854541,
+      "grad_norm": 28.403463540976567,
+      "learning_rate": 2e-06,
+      "loss": 0.3536,
+      "step": 3863
+    },
+    {
+      "epoch": 0.8964157290337548,
+      "grad_norm": 11.053863738445942,
+      "learning_rate": 2e-06,
+      "loss": 0.2466,
+      "step": 3864
+    },
+    {
+      "epoch": 0.8966477206820554,
+      "grad_norm": 9.177659242675198,
+      "learning_rate": 2e-06,
+      "loss": 0.2325,
+      "step": 3865
+    },
+    {
+      "epoch": 0.8968797123303561,
+      "grad_norm": 9.30944371626104,
+      "learning_rate": 2e-06,
+      "loss": 0.2986,
+      "step": 3866
+    },
+    {
+      "epoch": 0.8971117039786568,
+      "grad_norm": 12.882120895306459,
+      "learning_rate": 2e-06,
+      "loss": 0.3005,
+      "step": 3867
+    },
+    {
+      "epoch": 0.8973436956269575,
+      "grad_norm": 8.934266590413841,
+      "learning_rate": 2e-06,
+      "loss": 0.1734,
+      "step": 3868
+    },
+    {
+      "epoch": 0.897575687275258,
+      "grad_norm": 5.311613108640557,
+      "learning_rate": 2e-06,
+      "loss": 0.2151,
+      "step": 3869
+    },
+    {
+      "epoch": 0.8978076789235587,
+      "grad_norm": 24.19912099032846,
+      "learning_rate": 2e-06,
+      "loss": 0.3605,
+      "step": 3870
+    },
+    {
+      "epoch": 0.8980396705718594,
+      "grad_norm": 17.4060760039238,
+      "learning_rate": 2e-06,
+      "loss": 0.36,
+      "step": 3871
+    },
+    {
+      "epoch": 0.8982716622201601,
+      "grad_norm": 9.24006750676526,
+      "learning_rate": 2e-06,
+      "loss": 0.2765,
+      "step": 3872
+    },
+    {
+      "epoch": 0.8985036538684608,
+      "grad_norm": 12.557047617052072,
+      "learning_rate": 2e-06,
+      "loss": 0.3256,
+      "step": 3873
+    },
+    {
+      "epoch": 0.8987356455167614,
+      "grad_norm": 16.55716231109906,
+      "learning_rate": 2e-06,
+      "loss": 0.3307,
+      "step": 3874
+    },
+    {
+      "epoch": 0.898967637165062,
+      "grad_norm": 14.257984695745476,
+      "learning_rate": 2e-06,
+      "loss": 0.2903,
+      "step": 3875
+    },
+    {
+      "epoch": 0.8991996288133627,
+      "grad_norm": 6.765606224454405,
+      "learning_rate": 2e-06,
+      "loss": 0.1147,
+      "step": 3876
+    },
+    {
+      "epoch": 0.8994316204616634,
+      "grad_norm": 19.732475611334035,
+      "learning_rate": 2e-06,
+      "loss": 0.3762,
+      "step": 3877
+    },
+    {
+      "epoch": 0.899663612109964,
+      "grad_norm": 10.076354836054715,
+      "learning_rate": 2e-06,
+      "loss": 0.2457,
+      "step": 3878
+    },
+    {
+      "epoch": 0.8998956037582647,
+      "grad_norm": 22.692936353918086,
+      "learning_rate": 2e-06,
+      "loss": 0.3286,
+      "step": 3879
+    },
+    {
+      "epoch": 0.9001275954065654,
+      "grad_norm": 18.87531924416284,
+      "learning_rate": 2e-06,
+      "loss": 0.3505,
+      "step": 3880
+    },
+    {
+      "epoch": 0.9003595870548661,
+      "grad_norm": 21.461426922667258,
+      "learning_rate": 2e-06,
+      "loss": 0.4371,
+      "step": 3881
+    },
+    {
+      "epoch": 0.9005915787031666,
+      "grad_norm": 12.672233116120237,
+      "learning_rate": 2e-06,
+      "loss": 0.2552,
+      "step": 3882
+    },
+    {
+      "epoch": 0.9008235703514673,
+      "grad_norm": 14.92757839909907,
+      "learning_rate": 2e-06,
+      "loss": 0.2595,
+      "step": 3883
+    },
+    {
+      "epoch": 0.901055561999768,
+      "grad_norm": 9.864217379715534,
+      "learning_rate": 2e-06,
+      "loss": 0.2253,
+      "step": 3884
+    },
+    {
+      "epoch": 0.9012875536480687,
+      "grad_norm": 22.137721500164403,
+      "learning_rate": 2e-06,
+      "loss": 0.316,
+      "step": 3885
+    },
+    {
+      "epoch": 0.9015195452963694,
+      "grad_norm": 8.633971401134554,
+      "learning_rate": 2e-06,
+      "loss": 0.2195,
+      "step": 3886
+    },
+    {
+      "epoch": 0.90175153694467,
+      "grad_norm": 11.99123929014511,
+      "learning_rate": 2e-06,
+      "loss": 0.2543,
+      "step": 3887
+    },
+    {
+      "epoch": 0.9019835285929707,
+      "grad_norm": 23.05153162087007,
+      "learning_rate": 2e-06,
+      "loss": 0.2595,
+      "step": 3888
+    },
+    {
+      "epoch": 0.9022155202412713,
+      "grad_norm": 15.378579184780568,
+      "learning_rate": 2e-06,
+      "loss": 0.2613,
+      "step": 3889
+    },
+    {
+      "epoch": 0.902447511889572,
+      "grad_norm": 9.358759181548711,
+      "learning_rate": 2e-06,
+      "loss": 0.1977,
+      "step": 3890
+    },
+    {
+      "epoch": 0.9026795035378726,
+      "grad_norm": 14.072218217487547,
+      "learning_rate": 2e-06,
+      "loss": 0.314,
+      "step": 3891
+    },
+    {
+      "epoch": 0.9029114951861733,
+      "grad_norm": 8.650515094354374,
+      "learning_rate": 2e-06,
+      "loss": 0.2392,
+      "step": 3892
+    },
+    {
+      "epoch": 0.903143486834474,
+      "grad_norm": 13.973869182199863,
+      "learning_rate": 2e-06,
+      "loss": 0.2385,
+      "step": 3893
+    },
+    {
+      "epoch": 0.9033754784827747,
+      "grad_norm": 28.02108106068224,
+      "learning_rate": 2e-06,
+      "loss": 0.4868,
+      "step": 3894
+    },
+    {
+      "epoch": 0.9036074701310752,
+      "grad_norm": 10.126726222892133,
+      "learning_rate": 2e-06,
+      "loss": 0.2712,
+      "step": 3895
+    },
+    {
+      "epoch": 0.9038394617793759,
+      "grad_norm": 14.446733796997036,
+      "learning_rate": 2e-06,
+      "loss": 0.2759,
+      "step": 3896
+    },
+    {
+      "epoch": 0.9040714534276766,
+      "grad_norm": 9.142437995803487,
+      "learning_rate": 2e-06,
+      "loss": 0.2935,
+      "step": 3897
+    },
+    {
+      "epoch": 0.9043034450759773,
+      "grad_norm": 16.347152291076053,
+      "learning_rate": 2e-06,
+      "loss": 0.3192,
+      "step": 3898
+    },
+    {
+      "epoch": 0.9045354367242779,
+      "grad_norm": 10.663833545657894,
+      "learning_rate": 2e-06,
+      "loss": 0.2734,
+      "step": 3899
+    },
+    {
+      "epoch": 0.9047674283725786,
+      "grad_norm": 18.334678168407894,
+      "learning_rate": 2e-06,
+      "loss": 0.2272,
+      "step": 3900
+    },
+    {
+      "epoch": 0.9049994200208793,
+      "grad_norm": 9.98864088999583,
+      "learning_rate": 2e-06,
+      "loss": 0.2761,
+      "step": 3901
+    },
+    {
+      "epoch": 0.9052314116691799,
+      "grad_norm": 16.88195086485774,
+      "learning_rate": 2e-06,
+      "loss": 0.2828,
+      "step": 3902
+    },
+    {
+      "epoch": 0.9054634033174805,
+      "grad_norm": 14.193321843242153,
+      "learning_rate": 2e-06,
+      "loss": 0.3519,
+      "step": 3903
+    },
+    {
+      "epoch": 0.9056953949657812,
+      "grad_norm": 24.72235772205411,
+      "learning_rate": 2e-06,
+      "loss": 0.3367,
+      "step": 3904
+    },
+    {
+      "epoch": 0.9059273866140819,
+      "grad_norm": 20.661583880476023,
+      "learning_rate": 2e-06,
+      "loss": 0.3861,
+      "step": 3905
+    },
+    {
+      "epoch": 0.9061593782623826,
+      "grad_norm": 10.309625510352685,
+      "learning_rate": 2e-06,
+      "loss": 0.2169,
+      "step": 3906
+    },
+    {
+      "epoch": 0.9063913699106833,
+      "grad_norm": 15.189679441464554,
+      "learning_rate": 2e-06,
+      "loss": 0.1656,
+      "step": 3907
+    },
+    {
+      "epoch": 0.9066233615589839,
+      "grad_norm": 7.763584174957461,
+      "learning_rate": 2e-06,
+      "loss": 0.2384,
+      "step": 3908
+    },
+    {
+      "epoch": 0.9068553532072845,
+      "grad_norm": 32.802394565514405,
+      "learning_rate": 2e-06,
+      "loss": 0.418,
+      "step": 3909
+    },
+    {
+      "epoch": 0.9070873448555852,
+      "grad_norm": 19.839585714773417,
+      "learning_rate": 2e-06,
+      "loss": 0.2461,
+      "step": 3910
+    },
+    {
+      "epoch": 0.9073193365038859,
+      "grad_norm": 13.767551143679269,
+      "learning_rate": 2e-06,
+      "loss": 0.2116,
+      "step": 3911
+    },
+    {
+      "epoch": 0.9075513281521865,
+      "grad_norm": 20.153545824967566,
+      "learning_rate": 2e-06,
+      "loss": 0.3376,
+      "step": 3912
+    },
+    {
+      "epoch": 0.9077833198004872,
+      "grad_norm": 8.939747154687687,
+      "learning_rate": 2e-06,
+      "loss": 0.2546,
+      "step": 3913
+    },
+    {
+      "epoch": 0.9080153114487879,
+      "grad_norm": 32.43756975476725,
+      "learning_rate": 2e-06,
+      "loss": 0.3785,
+      "step": 3914
+    },
+    {
+      "epoch": 0.9082473030970885,
+      "grad_norm": 11.848078031401903,
+      "learning_rate": 2e-06,
+      "loss": 0.2395,
+      "step": 3915
+    },
+    {
+      "epoch": 0.9084792947453891,
+      "grad_norm": 11.22352972732074,
+      "learning_rate": 2e-06,
+      "loss": 0.215,
+      "step": 3916
+    },
+    {
+      "epoch": 0.9087112863936898,
+      "grad_norm": 17.184953012684105,
+      "learning_rate": 2e-06,
+      "loss": 0.3557,
+      "step": 3917
+    },
+    {
+      "epoch": 0.9089432780419905,
+      "grad_norm": 8.829476518236506,
+      "learning_rate": 2e-06,
+      "loss": 0.2368,
+      "step": 3918
+    },
+    {
+      "epoch": 0.9091752696902912,
+      "grad_norm": 25.986430156365554,
+      "learning_rate": 2e-06,
+      "loss": 0.2737,
+      "step": 3919
+    },
+    {
+      "epoch": 0.9094072613385918,
+      "grad_norm": 8.631894893578531,
+      "learning_rate": 2e-06,
+      "loss": 0.2415,
+      "step": 3920
+    },
+    {
+      "epoch": 0.9096392529868925,
+      "grad_norm": 14.692330088194996,
+      "learning_rate": 2e-06,
+      "loss": 0.2917,
+      "step": 3921
+    },
+    {
+      "epoch": 0.9098712446351931,
+      "grad_norm": 10.057081616113766,
+      "learning_rate": 2e-06,
+      "loss": 0.3515,
+      "step": 3922
+    },
+    {
+      "epoch": 0.9101032362834938,
+      "grad_norm": 18.242861421349204,
+      "learning_rate": 2e-06,
+      "loss": 0.3514,
+      "step": 3923
+    },
+    {
+      "epoch": 0.9103352279317944,
+      "grad_norm": 12.00313556221313,
+      "learning_rate": 2e-06,
+      "loss": 0.3548,
+      "step": 3924
+    },
+    {
+      "epoch": 0.9105672195800951,
+      "grad_norm": 8.759182341190806,
+      "learning_rate": 2e-06,
+      "loss": 0.2854,
+      "step": 3925
+    },
+    {
+      "epoch": 0.9107992112283958,
+      "grad_norm": 21.635245268730696,
+      "learning_rate": 2e-06,
+      "loss": 0.261,
+      "step": 3926
+    },
+    {
+      "epoch": 0.9110312028766965,
+      "grad_norm": 14.39027864566995,
+      "learning_rate": 2e-06,
+      "loss": 0.2331,
+      "step": 3927
+    },
+    {
+      "epoch": 0.9112631945249972,
+      "grad_norm": 17.651616795395558,
+      "learning_rate": 2e-06,
+      "loss": 0.3345,
+      "step": 3928
+    },
+    {
+      "epoch": 0.9114951861732977,
+      "grad_norm": 17.026843525740297,
+      "learning_rate": 2e-06,
+      "loss": 0.2664,
+      "step": 3929
+    },
+    {
+      "epoch": 0.9117271778215984,
+      "grad_norm": 15.703856377194578,
+      "learning_rate": 2e-06,
+      "loss": 0.3038,
+      "step": 3930
+    },
+    {
+      "epoch": 0.9119591694698991,
+      "grad_norm": 19.510509020521962,
+      "learning_rate": 2e-06,
+      "loss": 0.3946,
+      "step": 3931
+    },
+    {
+      "epoch": 0.9121911611181998,
+      "grad_norm": 21.184958616085613,
+      "learning_rate": 2e-06,
+      "loss": 0.2226,
+      "step": 3932
+    },
+    {
+      "epoch": 0.9124231527665004,
+      "grad_norm": 22.910390703702443,
+      "learning_rate": 2e-06,
+      "loss": 0.3747,
+      "step": 3933
+    },
+    {
+      "epoch": 0.9126551444148011,
+      "grad_norm": 21.79356223730189,
+      "learning_rate": 2e-06,
+      "loss": 0.2648,
+      "step": 3934
+    },
+    {
+      "epoch": 0.9128871360631017,
+      "grad_norm": 16.719709200320167,
+      "learning_rate": 2e-06,
+      "loss": 0.3129,
+      "step": 3935
+    },
+    {
+      "epoch": 0.9131191277114024,
+      "grad_norm": 15.75587390438053,
+      "learning_rate": 2e-06,
+      "loss": 0.2771,
+      "step": 3936
+    },
+    {
+      "epoch": 0.913351119359703,
+      "grad_norm": 20.445984418989443,
+      "learning_rate": 2e-06,
+      "loss": 0.3287,
+      "step": 3937
+    },
+    {
+      "epoch": 0.9135831110080037,
+      "grad_norm": 9.295243164429655,
+      "learning_rate": 2e-06,
+      "loss": 0.2542,
+      "step": 3938
+    },
+    {
+      "epoch": 0.9138151026563044,
+      "grad_norm": 16.194199891364974,
+      "learning_rate": 2e-06,
+      "loss": 0.2801,
+      "step": 3939
+    },
+    {
+      "epoch": 0.9140470943046051,
+      "grad_norm": 20.491671606951503,
+      "learning_rate": 2e-06,
+      "loss": 0.2353,
+      "step": 3940
+    },
+    {
+      "epoch": 0.9142790859529057,
+      "grad_norm": 10.452287640229645,
+      "learning_rate": 2e-06,
+      "loss": 0.2174,
+      "step": 3941
+    },
+    {
+      "epoch": 0.9145110776012063,
+      "grad_norm": 15.182715386927697,
+      "learning_rate": 2e-06,
+      "loss": 0.2636,
+      "step": 3942
+    },
+    {
+      "epoch": 0.914743069249507,
+      "grad_norm": 14.727316457709057,
+      "learning_rate": 2e-06,
+      "loss": 0.2375,
+      "step": 3943
+    },
+    {
+      "epoch": 0.9149750608978077,
+      "grad_norm": 14.991906354041783,
+      "learning_rate": 2e-06,
+      "loss": 0.2735,
+      "step": 3944
+    },
+    {
+      "epoch": 0.9152070525461083,
+      "grad_norm": 20.877355785804873,
+      "learning_rate": 2e-06,
+      "loss": 0.3779,
+      "step": 3945
+    },
+    {
+      "epoch": 0.915439044194409,
+      "grad_norm": 15.027199298330427,
+      "learning_rate": 2e-06,
+      "loss": 0.2824,
+      "step": 3946
+    },
+    {
+      "epoch": 0.9156710358427097,
+      "grad_norm": 8.425894451610421,
+      "learning_rate": 2e-06,
+      "loss": 0.1924,
+      "step": 3947
+    },
+    {
+      "epoch": 0.9159030274910103,
+      "grad_norm": 25.349503788018012,
+      "learning_rate": 2e-06,
+      "loss": 0.3346,
+      "step": 3948
+    },
+    {
+      "epoch": 0.916135019139311,
+      "grad_norm": 10.193314799347108,
+      "learning_rate": 2e-06,
+      "loss": 0.2686,
+      "step": 3949
+    },
+    {
+      "epoch": 0.9163670107876116,
+      "grad_norm": 18.574048790419745,
+      "learning_rate": 2e-06,
+      "loss": 0.4334,
+      "step": 3950
+    },
+    {
+      "epoch": 0.9165990024359123,
+      "grad_norm": 6.773484630130486,
+      "learning_rate": 2e-06,
+      "loss": 0.1877,
+      "step": 3951
+    },
+    {
+      "epoch": 0.916830994084213,
+      "grad_norm": 13.643189811756903,
+      "learning_rate": 2e-06,
+      "loss": 0.3131,
+      "step": 3952
+    },
+    {
+      "epoch": 0.9170629857325137,
+      "grad_norm": 19.924213913482326,
+      "learning_rate": 2e-06,
+      "loss": 0.3825,
+      "step": 3953
+    },
+    {
+      "epoch": 0.9172949773808143,
+      "grad_norm": 12.958783220911148,
+      "learning_rate": 2e-06,
+      "loss": 0.3044,
+      "step": 3954
+    },
+    {
+      "epoch": 0.9175269690291149,
+      "grad_norm": 12.0595184813058,
+      "learning_rate": 2e-06,
+      "loss": 0.2276,
+      "step": 3955
+    },
+    {
+      "epoch": 0.9177589606774156,
+      "grad_norm": 19.21193616308311,
+      "learning_rate": 2e-06,
+      "loss": 0.2676,
+      "step": 3956
+    },
+    {
+      "epoch": 0.9179909523257163,
+      "grad_norm": 11.476061295294725,
+      "learning_rate": 2e-06,
+      "loss": 0.2979,
+      "step": 3957
+    },
+    {
+      "epoch": 0.9182229439740169,
+      "grad_norm": 19.28669197764272,
+      "learning_rate": 2e-06,
+      "loss": 0.4399,
+      "step": 3958
+    },
+    {
+      "epoch": 0.9184549356223176,
+      "grad_norm": 7.265364825229757,
+      "learning_rate": 2e-06,
+      "loss": 0.2276,
+      "step": 3959
+    },
+    {
+      "epoch": 0.9186869272706183,
+      "grad_norm": 15.89433824195223,
+      "learning_rate": 2e-06,
+      "loss": 0.2828,
+      "step": 3960
+    },
+    {
+      "epoch": 0.918918918918919,
+      "grad_norm": 9.074347573516029,
+      "learning_rate": 2e-06,
+      "loss": 0.2016,
+      "step": 3961
+    },
+    {
+      "epoch": 0.9191509105672195,
+      "grad_norm": 16.415884335639657,
+      "learning_rate": 2e-06,
+      "loss": 0.2457,
+      "step": 3962
+    },
+    {
+      "epoch": 0.9193829022155202,
+      "grad_norm": 26.616894388131367,
+      "learning_rate": 2e-06,
+      "loss": 0.2135,
+      "step": 3963
+    },
+    {
+      "epoch": 0.9196148938638209,
+      "grad_norm": 6.792574370355875,
+      "learning_rate": 2e-06,
+      "loss": 0.215,
+      "step": 3964
+    },
+    {
+      "epoch": 0.9198468855121216,
+      "grad_norm": 14.404411259491736,
+      "learning_rate": 2e-06,
+      "loss": 0.3734,
+      "step": 3965
+    },
+    {
+      "epoch": 0.9200788771604222,
+      "grad_norm": 28.16018432451622,
+      "learning_rate": 2e-06,
+      "loss": 0.4642,
+      "step": 3966
+    },
+    {
+      "epoch": 0.9203108688087229,
+      "grad_norm": 19.233118360271835,
+      "learning_rate": 2e-06,
+      "loss": 0.3211,
+      "step": 3967
+    },
+    {
+      "epoch": 0.9205428604570235,
+      "grad_norm": 13.944355764448762,
+      "learning_rate": 2e-06,
+      "loss": 0.2451,
+      "step": 3968
+    },
+    {
+      "epoch": 0.9207748521053242,
+      "grad_norm": 12.641243085346089,
+      "learning_rate": 2e-06,
+      "loss": 0.2858,
+      "step": 3969
+    },
+    {
+      "epoch": 0.9210068437536248,
+      "grad_norm": 18.809098112566154,
+      "learning_rate": 2e-06,
+      "loss": 0.3246,
+      "step": 3970
+    },
+    {
+      "epoch": 0.9212388354019255,
+      "grad_norm": 21.782444054054192,
+      "learning_rate": 2e-06,
+      "loss": 0.3959,
+      "step": 3971
+    },
+    {
+      "epoch": 0.9214708270502262,
+      "grad_norm": 13.814108123062622,
+      "learning_rate": 2e-06,
+      "loss": 0.4264,
+      "step": 3972
+    },
+    {
+      "epoch": 0.9217028186985269,
+      "grad_norm": 21.81885611322097,
+      "learning_rate": 2e-06,
+      "loss": 0.3552,
+      "step": 3973
+    },
+    {
+      "epoch": 0.9219348103468276,
+      "grad_norm": 14.101735582385354,
+      "learning_rate": 2e-06,
+      "loss": 0.3584,
+      "step": 3974
+    },
+    {
+      "epoch": 0.9221668019951281,
+      "grad_norm": 11.043363753344781,
+      "learning_rate": 2e-06,
+      "loss": 0.2681,
+      "step": 3975
+    },
+    {
+      "epoch": 0.9223987936434288,
+      "grad_norm": 13.593291215016803,
+      "learning_rate": 2e-06,
+      "loss": 0.2515,
+      "step": 3976
+    },
+    {
+      "epoch": 0.9226307852917295,
+      "grad_norm": 12.362049676839284,
+      "learning_rate": 2e-06,
+      "loss": 0.257,
+      "step": 3977
+    },
+    {
+      "epoch": 0.9228627769400302,
+      "grad_norm": 13.270066216692763,
+      "learning_rate": 2e-06,
+      "loss": 0.3117,
+      "step": 3978
+    },
+    {
+      "epoch": 0.9230947685883308,
+      "grad_norm": 14.579060042594897,
+      "learning_rate": 2e-06,
+      "loss": 0.3214,
+      "step": 3979
+    },
+    {
+      "epoch": 0.9233267602366315,
+      "grad_norm": 14.720997537346065,
+      "learning_rate": 2e-06,
+      "loss": 0.2896,
+      "step": 3980
+    },
+    {
+      "epoch": 0.9235587518849322,
+      "grad_norm": 27.34929272148167,
+      "learning_rate": 2e-06,
+      "loss": 0.3681,
+      "step": 3981
+    },
+    {
+      "epoch": 0.9237907435332328,
+      "grad_norm": 6.8723669925715605,
+      "learning_rate": 2e-06,
+      "loss": 0.2386,
+      "step": 3982
+    },
+    {
+      "epoch": 0.9240227351815334,
+      "grad_norm": 15.46405296753594,
+      "learning_rate": 2e-06,
+      "loss": 0.2891,
+      "step": 3983
+    },
+    {
+      "epoch": 0.9242547268298341,
+      "grad_norm": 21.133738829314755,
+      "learning_rate": 2e-06,
+      "loss": 0.4005,
+      "step": 3984
+    },
+    {
+      "epoch": 0.9244867184781348,
+      "grad_norm": 23.69066601838177,
+      "learning_rate": 2e-06,
+      "loss": 0.2624,
+      "step": 3985
+    },
+    {
+      "epoch": 0.9247187101264355,
+      "grad_norm": 11.928606304200803,
+      "learning_rate": 2e-06,
+      "loss": 0.2461,
+      "step": 3986
+    },
+    {
+      "epoch": 0.9249507017747362,
+      "grad_norm": 7.763964912210666,
+      "learning_rate": 2e-06,
+      "loss": 0.299,
+      "step": 3987
+    },
+    {
+      "epoch": 0.9251826934230367,
+      "grad_norm": 14.448936450722345,
+      "learning_rate": 2e-06,
+      "loss": 0.2497,
+      "step": 3988
+    },
+    {
+      "epoch": 0.9254146850713374,
+      "grad_norm": 6.527725086557093,
+      "learning_rate": 2e-06,
+      "loss": 0.231,
+      "step": 3989
+    },
+    {
+      "epoch": 0.9256466767196381,
+      "grad_norm": 22.37946721451997,
+      "learning_rate": 2e-06,
+      "loss": 0.3802,
+      "step": 3990
+    },
+    {
+      "epoch": 0.9258786683679388,
+      "grad_norm": 12.629177421202952,
+      "learning_rate": 2e-06,
+      "loss": 0.3255,
+      "step": 3991
+    },
+    {
+      "epoch": 0.9261106600162394,
+      "grad_norm": 15.587491757728799,
+      "learning_rate": 2e-06,
+      "loss": 0.2531,
+      "step": 3992
+    },
+    {
+      "epoch": 0.9263426516645401,
+      "grad_norm": 11.847664403920094,
+      "learning_rate": 2e-06,
+      "loss": 0.269,
+      "step": 3993
+    },
+    {
+      "epoch": 0.9265746433128408,
+      "grad_norm": 13.322943271190882,
+      "learning_rate": 2e-06,
+      "loss": 0.278,
+      "step": 3994
+    },
+    {
+      "epoch": 0.9268066349611414,
+      "grad_norm": 14.495165691483393,
+      "learning_rate": 2e-06,
+      "loss": 0.3801,
+      "step": 3995
+    },
+    {
+      "epoch": 0.927038626609442,
+      "grad_norm": 14.3677196682021,
+      "learning_rate": 2e-06,
+      "loss": 0.2334,
+      "step": 3996
+    },
+    {
+      "epoch": 0.9272706182577427,
+      "grad_norm": 10.218329276484413,
+      "learning_rate": 2e-06,
+      "loss": 0.2553,
+      "step": 3997
+    },
+    {
+      "epoch": 0.9275026099060434,
+      "grad_norm": 34.637262702418646,
+      "learning_rate": 2e-06,
+      "loss": 0.4718,
+      "step": 3998
+    },
+    {
+      "epoch": 0.9277346015543441,
+      "grad_norm": 9.237413082217557,
+      "learning_rate": 2e-06,
+      "loss": 0.2473,
+      "step": 3999
+    },
+    {
+      "epoch": 0.9279665932026447,
+      "grad_norm": 16.090490051135355,
+      "learning_rate": 2e-06,
+      "loss": 0.2933,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9281985848509454,
+      "grad_norm": 18.131571394205444,
+      "learning_rate": 2e-06,
+      "loss": 0.3476,
+      "step": 4001
+    },
+    {
+      "epoch": 0.928430576499246,
+      "grad_norm": 13.298876519414382,
+      "learning_rate": 2e-06,
+      "loss": 0.2335,
+      "step": 4002
+    },
+    {
+      "epoch": 0.9286625681475467,
+      "grad_norm": 13.195268495451675,
+      "learning_rate": 2e-06,
+      "loss": 0.2911,
+      "step": 4003
+    },
+    {
+      "epoch": 0.9288945597958473,
+      "grad_norm": 8.619850603564261,
+      "learning_rate": 2e-06,
+      "loss": 0.1825,
+      "step": 4004
+    },
+    {
+      "epoch": 0.929126551444148,
+      "grad_norm": 9.168772392177162,
+      "learning_rate": 2e-06,
+      "loss": 0.2824,
+      "step": 4005
+    },
+    {
+      "epoch": 0.9293585430924487,
+      "grad_norm": 18.2102243935687,
+      "learning_rate": 2e-06,
+      "loss": 0.2815,
+      "step": 4006
+    },
+    {
+      "epoch": 0.9295905347407494,
+      "grad_norm": 10.502444860681994,
+      "learning_rate": 2e-06,
+      "loss": 0.3784,
+      "step": 4007
+    },
+    {
+      "epoch": 0.9298225263890499,
+      "grad_norm": 19.634937250194543,
+      "learning_rate": 2e-06,
+      "loss": 0.3284,
+      "step": 4008
+    },
+    {
+      "epoch": 0.9300545180373506,
+      "grad_norm": 17.777781446560336,
+      "learning_rate": 2e-06,
+      "loss": 0.3177,
+      "step": 4009
+    },
+    {
+      "epoch": 0.9302865096856513,
+      "grad_norm": 15.66459952879304,
+      "learning_rate": 2e-06,
+      "loss": 0.3439,
+      "step": 4010
+    },
+    {
+      "epoch": 0.930518501333952,
+      "grad_norm": 12.296224082963853,
+      "learning_rate": 2e-06,
+      "loss": 0.2705,
+      "step": 4011
+    },
+    {
+      "epoch": 0.9307504929822527,
+      "grad_norm": 13.958049366132693,
+      "learning_rate": 2e-06,
+      "loss": 0.2072,
+      "step": 4012
+    },
+    {
+      "epoch": 0.9309824846305533,
+      "grad_norm": 11.377406989138178,
+      "learning_rate": 2e-06,
+      "loss": 0.2086,
+      "step": 4013
+    },
+    {
+      "epoch": 0.931214476278854,
+      "grad_norm": 18.73659133228136,
+      "learning_rate": 2e-06,
+      "loss": 0.3641,
+      "step": 4014
+    },
+    {
+      "epoch": 0.9314464679271546,
+      "grad_norm": 9.579161039043823,
+      "learning_rate": 2e-06,
+      "loss": 0.1496,
+      "step": 4015
+    },
+    {
+      "epoch": 0.9316784595754553,
+      "grad_norm": 20.95385075541004,
+      "learning_rate": 2e-06,
+      "loss": 0.4974,
+      "step": 4016
+    },
+    {
+      "epoch": 0.9319104512237559,
+      "grad_norm": 15.665155596272864,
+      "learning_rate": 2e-06,
+      "loss": 0.3812,
+      "step": 4017
+    },
+    {
+      "epoch": 0.9321424428720566,
+      "grad_norm": 10.504501557374716,
+      "learning_rate": 2e-06,
+      "loss": 0.1418,
+      "step": 4018
+    },
+    {
+      "epoch": 0.9323744345203573,
+      "grad_norm": 10.85342891308318,
+      "learning_rate": 2e-06,
+      "loss": 0.319,
+      "step": 4019
+    },
+    {
+      "epoch": 0.932606426168658,
+      "grad_norm": 8.10601244212646,
+      "learning_rate": 2e-06,
+      "loss": 0.2395,
+      "step": 4020
+    },
+    {
+      "epoch": 0.9328384178169586,
+      "grad_norm": 13.022680006869349,
+      "learning_rate": 2e-06,
+      "loss": 0.2419,
+      "step": 4021
+    },
+    {
+      "epoch": 0.9330704094652592,
+      "grad_norm": 12.13270254820567,
+      "learning_rate": 2e-06,
+      "loss": 0.2383,
+      "step": 4022
+    },
+    {
+      "epoch": 0.9333024011135599,
+      "grad_norm": 16.32468558517697,
+      "learning_rate": 2e-06,
+      "loss": 0.2664,
+      "step": 4023
+    },
+    {
+      "epoch": 0.9335343927618606,
+      "grad_norm": 12.246989823743979,
+      "learning_rate": 2e-06,
+      "loss": 0.3144,
+      "step": 4024
+    },
+    {
+      "epoch": 0.9337663844101612,
+      "grad_norm": 15.224846029260464,
+      "learning_rate": 2e-06,
+      "loss": 0.2838,
+      "step": 4025
+    },
+    {
+      "epoch": 0.9339983760584619,
+      "grad_norm": 15.018031125518513,
+      "learning_rate": 2e-06,
+      "loss": 0.2506,
+      "step": 4026
+    },
+    {
+      "epoch": 0.9342303677067626,
+      "grad_norm": 7.434044658596272,
+      "learning_rate": 2e-06,
+      "loss": 0.2986,
+      "step": 4027
+    },
+    {
+      "epoch": 0.9344623593550632,
+      "grad_norm": 20.494236364736892,
+      "learning_rate": 2e-06,
+      "loss": 0.2571,
+      "step": 4028
+    },
+    {
+      "epoch": 0.9346943510033638,
+      "grad_norm": 6.418952262961799,
+      "learning_rate": 2e-06,
+      "loss": 0.2598,
+      "step": 4029
+    },
+    {
+      "epoch": 0.9349263426516645,
+      "grad_norm": 21.616245646302445,
+      "learning_rate": 2e-06,
+      "loss": 0.2834,
+      "step": 4030
+    },
+    {
+      "epoch": 0.9351583342999652,
+      "grad_norm": 18.3306232616933,
+      "learning_rate": 2e-06,
+      "loss": 0.3145,
+      "step": 4031
+    },
+    {
+      "epoch": 0.9353903259482659,
+      "grad_norm": 10.352480111360936,
+      "learning_rate": 2e-06,
+      "loss": 0.2459,
+      "step": 4032
+    },
+    {
+      "epoch": 0.9356223175965666,
+      "grad_norm": 11.204629468614156,
+      "learning_rate": 2e-06,
+      "loss": 0.2544,
+      "step": 4033
+    },
+    {
+      "epoch": 0.9358543092448672,
+      "grad_norm": 16.492577310753635,
+      "learning_rate": 2e-06,
+      "loss": 0.2445,
+      "step": 4034
+    },
+    {
+      "epoch": 0.9360863008931678,
+      "grad_norm": 22.371945371319974,
+      "learning_rate": 2e-06,
+      "loss": 0.2821,
+      "step": 4035
+    },
+    {
+      "epoch": 0.9363182925414685,
+      "grad_norm": 8.859132258593608,
+      "learning_rate": 2e-06,
+      "loss": 0.3099,
+      "step": 4036
+    },
+    {
+      "epoch": 0.9365502841897692,
+      "grad_norm": 17.014776481789077,
+      "learning_rate": 2e-06,
+      "loss": 0.3262,
+      "step": 4037
+    },
+    {
+      "epoch": 0.9367822758380698,
+      "grad_norm": 10.625597532800239,
+      "learning_rate": 2e-06,
+      "loss": 0.2873,
+      "step": 4038
+    },
+    {
+      "epoch": 0.9370142674863705,
+      "grad_norm": 14.122822330175744,
+      "learning_rate": 2e-06,
+      "loss": 0.38,
+      "step": 4039
+    },
+    {
+      "epoch": 0.9372462591346712,
+      "grad_norm": 10.328734624379049,
+      "learning_rate": 2e-06,
+      "loss": 0.2192,
+      "step": 4040
+    },
+    {
+      "epoch": 0.9374782507829719,
+      "grad_norm": 9.09386762470707,
+      "learning_rate": 2e-06,
+      "loss": 0.2266,
+      "step": 4041
+    },
+    {
+      "epoch": 0.9377102424312724,
+      "grad_norm": 18.684735772439982,
+      "learning_rate": 2e-06,
+      "loss": 0.2231,
+      "step": 4042
+    },
+    {
+      "epoch": 0.9379422340795731,
+      "grad_norm": 9.957849633530989,
+      "learning_rate": 2e-06,
+      "loss": 0.2933,
+      "step": 4043
+    },
+    {
+      "epoch": 0.9381742257278738,
+      "grad_norm": 13.49504899297702,
+      "learning_rate": 2e-06,
+      "loss": 0.1928,
+      "step": 4044
+    },
+    {
+      "epoch": 0.9384062173761745,
+      "grad_norm": 11.66253981574796,
+      "learning_rate": 2e-06,
+      "loss": 0.2601,
+      "step": 4045
+    },
+    {
+      "epoch": 0.9386382090244751,
+      "grad_norm": 10.106589206845713,
+      "learning_rate": 2e-06,
+      "loss": 0.2958,
+      "step": 4046
+    },
+    {
+      "epoch": 0.9388702006727758,
+      "grad_norm": 12.192365381188443,
+      "learning_rate": 2e-06,
+      "loss": 0.2948,
+      "step": 4047
+    },
+    {
+      "epoch": 0.9391021923210764,
+      "grad_norm": 11.08645100908848,
+      "learning_rate": 2e-06,
+      "loss": 0.2181,
+      "step": 4048
+    },
+    {
+      "epoch": 0.9393341839693771,
+      "grad_norm": 10.461564499933814,
+      "learning_rate": 2e-06,
+      "loss": 0.2554,
+      "step": 4049
+    },
+    {
+      "epoch": 0.9395661756176777,
+      "grad_norm": 12.516633499592912,
+      "learning_rate": 2e-06,
+      "loss": 0.3163,
+      "step": 4050
+    },
+    {
+      "epoch": 0.9397981672659784,
+      "grad_norm": 17.879976322757418,
+      "learning_rate": 2e-06,
+      "loss": 0.3279,
+      "step": 4051
+    },
+    {
+      "epoch": 0.9400301589142791,
+      "grad_norm": 10.951916309028428,
+      "learning_rate": 2e-06,
+      "loss": 0.2341,
+      "step": 4052
+    },
+    {
+      "epoch": 0.9402621505625798,
+      "grad_norm": 8.976366702731776,
+      "learning_rate": 2e-06,
+      "loss": 0.1948,
+      "step": 4053
+    },
+    {
+      "epoch": 0.9404941422108805,
+      "grad_norm": 9.076599604883869,
+      "learning_rate": 2e-06,
+      "loss": 0.3378,
+      "step": 4054
+    },
+    {
+      "epoch": 0.940726133859181,
+      "grad_norm": 15.382474393870892,
+      "learning_rate": 2e-06,
+      "loss": 0.2488,
+      "step": 4055
+    },
+    {
+      "epoch": 0.9409581255074817,
+      "grad_norm": 14.329098715179613,
+      "learning_rate": 2e-06,
+      "loss": 0.3046,
+      "step": 4056
+    },
+    {
+      "epoch": 0.9411901171557824,
+      "grad_norm": 9.195493737277706,
+      "learning_rate": 2e-06,
+      "loss": 0.2264,
+      "step": 4057
+    },
+    {
+      "epoch": 0.9414221088040831,
+      "grad_norm": 10.882441981983053,
+      "learning_rate": 2e-06,
+      "loss": 0.2679,
+      "step": 4058
+    },
+    {
+      "epoch": 0.9416541004523837,
+      "grad_norm": 7.640288467837494,
+      "learning_rate": 2e-06,
+      "loss": 0.1782,
+      "step": 4059
+    },
+    {
+      "epoch": 0.9418860921006844,
+      "grad_norm": 7.9402748004771295,
+      "learning_rate": 2e-06,
+      "loss": 0.2199,
+      "step": 4060
+    },
+    {
+      "epoch": 0.9421180837489851,
+      "grad_norm": 14.183099572940574,
+      "learning_rate": 2e-06,
+      "loss": 0.197,
+      "step": 4061
+    },
+    {
+      "epoch": 0.9423500753972857,
+      "grad_norm": 14.253871053319683,
+      "learning_rate": 2e-06,
+      "loss": 0.2358,
+      "step": 4062
+    },
+    {
+      "epoch": 0.9425820670455863,
+      "grad_norm": 14.87910110838926,
+      "learning_rate": 2e-06,
+      "loss": 0.249,
+      "step": 4063
+    },
+    {
+      "epoch": 0.942814058693887,
+      "grad_norm": 18.414766910385318,
+      "learning_rate": 2e-06,
+      "loss": 0.2946,
+      "step": 4064
+    },
+    {
+      "epoch": 0.9430460503421877,
+      "grad_norm": 19.484331898866966,
+      "learning_rate": 2e-06,
+      "loss": 0.3497,
+      "step": 4065
+    },
+    {
+      "epoch": 0.9432780419904884,
+      "grad_norm": 7.831919172941622,
+      "learning_rate": 2e-06,
+      "loss": 0.2108,
+      "step": 4066
+    },
+    {
+      "epoch": 0.943510033638789,
+      "grad_norm": 13.970992782476834,
+      "learning_rate": 2e-06,
+      "loss": 0.2523,
+      "step": 4067
+    },
+    {
+      "epoch": 0.9437420252870896,
+      "grad_norm": 15.572274581276728,
+      "learning_rate": 2e-06,
+      "loss": 0.3366,
+      "step": 4068
+    },
+    {
+      "epoch": 0.9439740169353903,
+      "grad_norm": 18.73910628418021,
+      "learning_rate": 2e-06,
+      "loss": 0.3037,
+      "step": 4069
+    },
+    {
+      "epoch": 0.944206008583691,
+      "grad_norm": 14.560548536813151,
+      "learning_rate": 2e-06,
+      "loss": 0.2343,
+      "step": 4070
+    },
+    {
+      "epoch": 0.9444380002319916,
+      "grad_norm": 13.014495432933003,
+      "learning_rate": 2e-06,
+      "loss": 0.2704,
+      "step": 4071
+    },
+    {
+      "epoch": 0.9446699918802923,
+      "grad_norm": 16.70216711953269,
+      "learning_rate": 2e-06,
+      "loss": 0.2593,
+      "step": 4072
+    },
+    {
+      "epoch": 0.944901983528593,
+      "grad_norm": 12.274314755806715,
+      "learning_rate": 2e-06,
+      "loss": 0.334,
+      "step": 4073
+    },
+    {
+      "epoch": 0.9451339751768937,
+      "grad_norm": 12.610739079859846,
+      "learning_rate": 2e-06,
+      "loss": 0.261,
+      "step": 4074
+    },
+    {
+      "epoch": 0.9453659668251942,
+      "grad_norm": 7.652569686562883,
+      "learning_rate": 2e-06,
+      "loss": 0.253,
+      "step": 4075
+    },
+    {
+      "epoch": 0.9455979584734949,
+      "grad_norm": 9.82109278599335,
+      "learning_rate": 2e-06,
+      "loss": 0.3167,
+      "step": 4076
+    },
+    {
+      "epoch": 0.9458299501217956,
+      "grad_norm": 15.045805903494685,
+      "learning_rate": 2e-06,
+      "loss": 0.391,
+      "step": 4077
+    },
+    {
+      "epoch": 0.9460619417700963,
+      "grad_norm": 20.756814333126798,
+      "learning_rate": 2e-06,
+      "loss": 0.3532,
+      "step": 4078
+    },
+    {
+      "epoch": 0.946293933418397,
+      "grad_norm": 11.520740221476641,
+      "learning_rate": 2e-06,
+      "loss": 0.2224,
+      "step": 4079
+    },
+    {
+      "epoch": 0.9465259250666976,
+      "grad_norm": 20.332249191152876,
+      "learning_rate": 2e-06,
+      "loss": 0.3074,
+      "step": 4080
+    },
+    {
+      "epoch": 0.9467579167149983,
+      "grad_norm": 9.704537942099549,
+      "learning_rate": 2e-06,
+      "loss": 0.2567,
+      "step": 4081
+    },
+    {
+      "epoch": 0.9469899083632989,
+      "grad_norm": 12.313089663784114,
+      "learning_rate": 2e-06,
+      "loss": 0.3028,
+      "step": 4082
+    },
+    {
+      "epoch": 0.9472219000115996,
+      "grad_norm": 21.2792596586522,
+      "learning_rate": 2e-06,
+      "loss": 0.2978,
+      "step": 4083
+    },
+    {
+      "epoch": 0.9474538916599002,
+      "grad_norm": 10.101328052308142,
+      "learning_rate": 2e-06,
+      "loss": 0.2412,
+      "step": 4084
+    },
+    {
+      "epoch": 0.9476858833082009,
+      "grad_norm": 17.19211419820723,
+      "learning_rate": 2e-06,
+      "loss": 0.3151,
+      "step": 4085
+    },
+    {
+      "epoch": 0.9479178749565016,
+      "grad_norm": 16.302835135561683,
+      "learning_rate": 2e-06,
+      "loss": 0.2125,
+      "step": 4086
+    },
+    {
+      "epoch": 0.9481498666048023,
+      "grad_norm": 11.946190807136137,
+      "learning_rate": 2e-06,
+      "loss": 0.3241,
+      "step": 4087
+    },
+    {
+      "epoch": 0.9483818582531028,
+      "grad_norm": 13.151658952123395,
+      "learning_rate": 2e-06,
+      "loss": 0.2848,
+      "step": 4088
+    },
+    {
+      "epoch": 0.9486138499014035,
+      "grad_norm": 23.02715333790368,
+      "learning_rate": 2e-06,
+      "loss": 0.2935,
+      "step": 4089
+    },
+    {
+      "epoch": 0.9488458415497042,
+      "grad_norm": 8.503461364810859,
+      "learning_rate": 2e-06,
+      "loss": 0.1989,
+      "step": 4090
+    },
+    {
+      "epoch": 0.9490778331980049,
+      "grad_norm": 18.427826665139552,
+      "learning_rate": 2e-06,
+      "loss": 0.2229,
+      "step": 4091
+    },
+    {
+      "epoch": 0.9493098248463055,
+      "grad_norm": 14.303047249938743,
+      "learning_rate": 2e-06,
+      "loss": 0.2439,
+      "step": 4092
+    },
+    {
+      "epoch": 0.9495418164946062,
+      "grad_norm": 19.143796564075636,
+      "learning_rate": 2e-06,
+      "loss": 0.2807,
+      "step": 4093
+    },
+    {
+      "epoch": 0.9497738081429069,
+      "grad_norm": 13.488732980205729,
+      "learning_rate": 2e-06,
+      "loss": 0.257,
+      "step": 4094
+    },
+    {
+      "epoch": 0.9500057997912075,
+      "grad_norm": 8.526817701556933,
+      "learning_rate": 2e-06,
+      "loss": 0.193,
+      "step": 4095
+    },
+    {
+      "epoch": 0.9502377914395082,
+      "grad_norm": 15.559353236992328,
+      "learning_rate": 2e-06,
+      "loss": 0.3272,
+      "step": 4096
+    },
+    {
+      "epoch": 0.9504697830878088,
+      "grad_norm": 24.587515021857655,
+      "learning_rate": 2e-06,
+      "loss": 0.3034,
+      "step": 4097
+    },
+    {
+      "epoch": 0.9507017747361095,
+      "grad_norm": 13.200634807361052,
+      "learning_rate": 2e-06,
+      "loss": 0.2902,
+      "step": 4098
+    },
+    {
+      "epoch": 0.9509337663844102,
+      "grad_norm": 20.434562905867345,
+      "learning_rate": 2e-06,
+      "loss": 0.3055,
+      "step": 4099
+    },
+    {
+      "epoch": 0.9511657580327109,
+      "grad_norm": 18.417630148667733,
+      "learning_rate": 2e-06,
+      "loss": 0.3549,
+      "step": 4100
+    },
+    {
+      "epoch": 0.9513977496810114,
+      "grad_norm": 13.31683770770985,
+      "learning_rate": 2e-06,
+      "loss": 0.2553,
+      "step": 4101
+    },
+    {
+      "epoch": 0.9516297413293121,
+      "grad_norm": 11.843045183770629,
+      "learning_rate": 2e-06,
+      "loss": 0.252,
+      "step": 4102
+    },
+    {
+      "epoch": 0.9518617329776128,
+      "grad_norm": 11.666980315192783,
+      "learning_rate": 2e-06,
+      "loss": 0.2539,
+      "step": 4103
+    },
+    {
+      "epoch": 0.9520937246259135,
+      "grad_norm": 19.913932697088207,
+      "learning_rate": 2e-06,
+      "loss": 0.3337,
+      "step": 4104
+    },
+    {
+      "epoch": 0.9523257162742141,
+      "grad_norm": 13.36391426480302,
+      "learning_rate": 2e-06,
+      "loss": 0.2737,
+      "step": 4105
+    },
+    {
+      "epoch": 0.9525577079225148,
+      "grad_norm": 20.073464429920577,
+      "learning_rate": 2e-06,
+      "loss": 0.4001,
+      "step": 4106
+    },
+    {
+      "epoch": 0.9527896995708155,
+      "grad_norm": 15.891488742906676,
+      "learning_rate": 2e-06,
+      "loss": 0.3502,
+      "step": 4107
+    },
+    {
+      "epoch": 0.9530216912191161,
+      "grad_norm": 21.231285427185863,
+      "learning_rate": 2e-06,
+      "loss": 0.3465,
+      "step": 4108
+    },
+    {
+      "epoch": 0.9532536828674167,
+      "grad_norm": 13.94812781849407,
+      "learning_rate": 2e-06,
+      "loss": 0.2747,
+      "step": 4109
+    },
+    {
+      "epoch": 0.9534856745157174,
+      "grad_norm": 7.340569324720874,
+      "learning_rate": 2e-06,
+      "loss": 0.2401,
+      "step": 4110
+    },
+    {
+      "epoch": 0.9537176661640181,
+      "grad_norm": 15.352599805848575,
+      "learning_rate": 2e-06,
+      "loss": 0.338,
+      "step": 4111
+    },
+    {
+      "epoch": 0.9539496578123188,
+      "grad_norm": 12.305729093870848,
+      "learning_rate": 2e-06,
+      "loss": 0.2499,
+      "step": 4112
+    },
+    {
+      "epoch": 0.9541816494606195,
+      "grad_norm": 9.690251495539805,
+      "learning_rate": 2e-06,
+      "loss": 0.2528,
+      "step": 4113
+    },
+    {
+      "epoch": 0.9544136411089201,
+      "grad_norm": 15.161578406160716,
+      "learning_rate": 2e-06,
+      "loss": 0.3333,
+      "step": 4114
+    },
+    {
+      "epoch": 0.9546456327572207,
+      "grad_norm": 28.285554380523788,
+      "learning_rate": 2e-06,
+      "loss": 0.2778,
+      "step": 4115
+    },
+    {
+      "epoch": 0.9548776244055214,
+      "grad_norm": 16.19964193709801,
+      "learning_rate": 2e-06,
+      "loss": 0.2748,
+      "step": 4116
+    },
+    {
+      "epoch": 0.955109616053822,
+      "grad_norm": 12.83960278879911,
+      "learning_rate": 2e-06,
+      "loss": 0.2952,
+      "step": 4117
+    },
+    {
+      "epoch": 0.9553416077021227,
+      "grad_norm": 8.782977231717306,
+      "learning_rate": 2e-06,
+      "loss": 0.2084,
+      "step": 4118
+    },
+    {
+      "epoch": 0.9555735993504234,
+      "grad_norm": 15.30528731331944,
+      "learning_rate": 2e-06,
+      "loss": 0.326,
+      "step": 4119
+    },
+    {
+      "epoch": 0.9558055909987241,
+      "grad_norm": 13.290100729515132,
+      "learning_rate": 2e-06,
+      "loss": 0.272,
+      "step": 4120
+    },
+    {
+      "epoch": 0.9560375826470247,
+      "grad_norm": 10.406690132651857,
+      "learning_rate": 2e-06,
+      "loss": 0.2514,
+      "step": 4121
+    },
+    {
+      "epoch": 0.9562695742953253,
+      "grad_norm": 26.04862314136393,
+      "learning_rate": 2e-06,
+      "loss": 0.3575,
+      "step": 4122
+    },
+    {
+      "epoch": 0.956501565943626,
+      "grad_norm": 7.815336659949993,
+      "learning_rate": 2e-06,
+      "loss": 0.1821,
+      "step": 4123
+    },
+    {
+      "epoch": 0.9567335575919267,
+      "grad_norm": 9.946114361385957,
+      "learning_rate": 2e-06,
+      "loss": 0.2606,
+      "step": 4124
+    },
+    {
+      "epoch": 0.9569655492402274,
+      "grad_norm": 8.008143895713363,
+      "learning_rate": 2e-06,
+      "loss": 0.258,
+      "step": 4125
+    },
+    {
+      "epoch": 0.957197540888528,
+      "grad_norm": 12.47252896155183,
+      "learning_rate": 2e-06,
+      "loss": 0.2618,
+      "step": 4126
+    },
+    {
+      "epoch": 0.9574295325368287,
+      "grad_norm": 8.102956281530695,
+      "learning_rate": 2e-06,
+      "loss": 0.3792,
+      "step": 4127
+    },
+    {
+      "epoch": 0.9576615241851293,
+      "grad_norm": 8.2663318549362,
+      "learning_rate": 2e-06,
+      "loss": 0.3004,
+      "step": 4128
+    },
+    {
+      "epoch": 0.95789351583343,
+      "grad_norm": 12.246000066279436,
+      "learning_rate": 2e-06,
+      "loss": 0.2644,
+      "step": 4129
+    },
+    {
+      "epoch": 0.9581255074817306,
+      "grad_norm": 11.872467722008956,
+      "learning_rate": 2e-06,
+      "loss": 0.2984,
+      "step": 4130
+    },
+    {
+      "epoch": 0.9583574991300313,
+      "grad_norm": 21.614820482071732,
+      "learning_rate": 2e-06,
+      "loss": 0.2752,
+      "step": 4131
+    },
+    {
+      "epoch": 0.958589490778332,
+      "grad_norm": 14.068048318367484,
+      "learning_rate": 2e-06,
+      "loss": 0.3651,
+      "step": 4132
+    },
+    {
+      "epoch": 0.9588214824266327,
+      "grad_norm": 15.711678066427929,
+      "learning_rate": 2e-06,
+      "loss": 0.2752,
+      "step": 4133
+    },
+    {
+      "epoch": 0.9590534740749334,
+      "grad_norm": 5.923996766497805,
+      "learning_rate": 2e-06,
+      "loss": 0.2064,
+      "step": 4134
+    },
+    {
+      "epoch": 0.9592854657232339,
+      "grad_norm": 9.744245529377244,
+      "learning_rate": 2e-06,
+      "loss": 0.3068,
+      "step": 4135
+    },
+    {
+      "epoch": 0.9595174573715346,
+      "grad_norm": 8.475805972629109,
+      "learning_rate": 2e-06,
+      "loss": 0.2627,
+      "step": 4136
+    },
+    {
+      "epoch": 0.9597494490198353,
+      "grad_norm": 17.447438317238976,
+      "learning_rate": 2e-06,
+      "loss": 0.3531,
+      "step": 4137
+    },
+    {
+      "epoch": 0.959981440668136,
+      "grad_norm": 10.248204605683624,
+      "learning_rate": 2e-06,
+      "loss": 0.3174,
+      "step": 4138
+    },
+    {
+      "epoch": 0.9602134323164366,
+      "grad_norm": 9.349030328082655,
+      "learning_rate": 2e-06,
+      "loss": 0.2251,
+      "step": 4139
+    },
+    {
+      "epoch": 0.9604454239647373,
+      "grad_norm": 10.040785549242996,
+      "learning_rate": 2e-06,
+      "loss": 0.3646,
+      "step": 4140
+    },
+    {
+      "epoch": 0.9606774156130379,
+      "grad_norm": 17.99881429560681,
+      "learning_rate": 2e-06,
+      "loss": 0.3179,
+      "step": 4141
+    },
+    {
+      "epoch": 0.9609094072613386,
+      "grad_norm": 23.15821161884971,
+      "learning_rate": 2e-06,
+      "loss": 0.3449,
+      "step": 4142
+    },
+    {
+      "epoch": 0.9611413989096392,
+      "grad_norm": 11.559981381311555,
+      "learning_rate": 2e-06,
+      "loss": 0.2579,
+      "step": 4143
+    },
+    {
+      "epoch": 0.9613733905579399,
+      "grad_norm": 16.29055174473012,
+      "learning_rate": 2e-06,
+      "loss": 0.2503,
+      "step": 4144
+    },
+    {
+      "epoch": 0.9616053822062406,
+      "grad_norm": 11.7947484697283,
+      "learning_rate": 2e-06,
+      "loss": 0.2859,
+      "step": 4145
+    },
+    {
+      "epoch": 0.9618373738545413,
+      "grad_norm": 19.286142596191475,
+      "learning_rate": 2e-06,
+      "loss": 0.3261,
+      "step": 4146
+    },
+    {
+      "epoch": 0.9620693655028419,
+      "grad_norm": 9.755826226950822,
+      "learning_rate": 2e-06,
+      "loss": 0.2265,
+      "step": 4147
+    },
+    {
+      "epoch": 0.9623013571511425,
+      "grad_norm": 16.999414195147992,
+      "learning_rate": 2e-06,
+      "loss": 0.2519,
+      "step": 4148
+    },
+    {
+      "epoch": 0.9625333487994432,
+      "grad_norm": 9.47299950537098,
+      "learning_rate": 2e-06,
+      "loss": 0.1947,
+      "step": 4149
+    },
+    {
+      "epoch": 0.9627653404477439,
+      "grad_norm": 7.826711210558996,
+      "learning_rate": 2e-06,
+      "loss": 0.2812,
+      "step": 4150
+    },
+    {
+      "epoch": 0.9629973320960445,
+      "grad_norm": 10.426702042786996,
+      "learning_rate": 2e-06,
+      "loss": 0.2236,
+      "step": 4151
+    },
+    {
+      "epoch": 0.9632293237443452,
+      "grad_norm": 15.108961521422483,
+      "learning_rate": 2e-06,
+      "loss": 0.3395,
+      "step": 4152
+    },
+    {
+      "epoch": 0.9634613153926459,
+      "grad_norm": 9.802328672252278,
+      "learning_rate": 2e-06,
+      "loss": 0.2706,
+      "step": 4153
+    },
+    {
+      "epoch": 0.9636933070409466,
+      "grad_norm": 18.037924430831286,
+      "learning_rate": 2e-06,
+      "loss": 0.2694,
+      "step": 4154
+    },
+    {
+      "epoch": 0.9639252986892471,
+      "grad_norm": 8.017284974062889,
+      "learning_rate": 2e-06,
+      "loss": 0.2999,
+      "step": 4155
+    },
+    {
+      "epoch": 0.9641572903375478,
+      "grad_norm": 12.34805634741516,
+      "learning_rate": 2e-06,
+      "loss": 0.255,
+      "step": 4156
+    },
+    {
+      "epoch": 0.9643892819858485,
+      "grad_norm": 16.752460181250132,
+      "learning_rate": 2e-06,
+      "loss": 0.2908,
+      "step": 4157
+    },
+    {
+      "epoch": 0.9646212736341492,
+      "grad_norm": 16.975203492078528,
+      "learning_rate": 2e-06,
+      "loss": 0.3314,
+      "step": 4158
+    },
+    {
+      "epoch": 0.9648532652824499,
+      "grad_norm": 18.677633298719076,
+      "learning_rate": 2e-06,
+      "loss": 0.2746,
+      "step": 4159
+    },
+    {
+      "epoch": 0.9650852569307505,
+      "grad_norm": 16.169374674291078,
+      "learning_rate": 2e-06,
+      "loss": 0.3081,
+      "step": 4160
+    },
+    {
+      "epoch": 0.9653172485790511,
+      "grad_norm": 8.690820570400597,
+      "learning_rate": 2e-06,
+      "loss": 0.2812,
+      "step": 4161
+    },
+    {
+      "epoch": 0.9655492402273518,
+      "grad_norm": 18.24990491642374,
+      "learning_rate": 2e-06,
+      "loss": 0.2228,
+      "step": 4162
+    },
+    {
+      "epoch": 0.9657812318756525,
+      "grad_norm": 7.2511972849280015,
+      "learning_rate": 2e-06,
+      "loss": 0.294,
+      "step": 4163
+    },
+    {
+      "epoch": 0.9660132235239531,
+      "grad_norm": 14.82011859195889,
+      "learning_rate": 2e-06,
+      "loss": 0.2779,
+      "step": 4164
+    },
+    {
+      "epoch": 0.9662452151722538,
+      "grad_norm": 8.815089700662222,
+      "learning_rate": 2e-06,
+      "loss": 0.1979,
+      "step": 4165
+    },
+    {
+      "epoch": 0.9664772068205545,
+      "grad_norm": 11.440596644880184,
+      "learning_rate": 2e-06,
+      "loss": 0.2995,
+      "step": 4166
+    },
+    {
+      "epoch": 0.9667091984688552,
+      "grad_norm": 21.152456518163454,
+      "learning_rate": 2e-06,
+      "loss": 0.3035,
+      "step": 4167
+    },
+    {
+      "epoch": 0.9669411901171557,
+      "grad_norm": 22.36876306470267,
+      "learning_rate": 2e-06,
+      "loss": 0.3341,
+      "step": 4168
+    },
+    {
+      "epoch": 0.9671731817654564,
+      "grad_norm": 9.240470420717966,
+      "learning_rate": 2e-06,
+      "loss": 0.2977,
+      "step": 4169
+    },
+    {
+      "epoch": 0.9674051734137571,
+      "grad_norm": 12.167277516648443,
+      "learning_rate": 2e-06,
+      "loss": 0.2749,
+      "step": 4170
+    },
+    {
+      "epoch": 0.9676371650620578,
+      "grad_norm": 5.054791610066648,
+      "learning_rate": 2e-06,
+      "loss": 0.2408,
+      "step": 4171
+    },
+    {
+      "epoch": 0.9678691567103584,
+      "grad_norm": 6.7646904021890455,
+      "learning_rate": 2e-06,
+      "loss": 0.2495,
+      "step": 4172
+    },
+    {
+      "epoch": 0.9681011483586591,
+      "grad_norm": 16.408927035270644,
+      "learning_rate": 2e-06,
+      "loss": 0.2613,
+      "step": 4173
+    },
+    {
+      "epoch": 0.9683331400069598,
+      "grad_norm": 14.037063486759456,
+      "learning_rate": 2e-06,
+      "loss": 0.3029,
+      "step": 4174
+    },
+    {
+      "epoch": 0.9685651316552604,
+      "grad_norm": 13.446123483781111,
+      "learning_rate": 2e-06,
+      "loss": 0.2195,
+      "step": 4175
+    },
+    {
+      "epoch": 0.968797123303561,
+      "grad_norm": 12.642700280341838,
+      "learning_rate": 2e-06,
+      "loss": 0.2894,
+      "step": 4176
+    },
+    {
+      "epoch": 0.9690291149518617,
+      "grad_norm": 10.132367629156734,
+      "learning_rate": 2e-06,
+      "loss": 0.253,
+      "step": 4177
+    },
+    {
+      "epoch": 0.9692611066001624,
+      "grad_norm": 8.498917805369992,
+      "learning_rate": 2e-06,
+      "loss": 0.3635,
+      "step": 4178
+    },
+    {
+      "epoch": 0.9694930982484631,
+      "grad_norm": 13.209622449831079,
+      "learning_rate": 2e-06,
+      "loss": 0.2891,
+      "step": 4179
+    },
+    {
+      "epoch": 0.9697250898967638,
+      "grad_norm": 12.29051439378758,
+      "learning_rate": 2e-06,
+      "loss": 0.3119,
+      "step": 4180
+    },
+    {
+      "epoch": 0.9699570815450643,
+      "grad_norm": 21.12653821260995,
+      "learning_rate": 2e-06,
+      "loss": 0.3552,
+      "step": 4181
+    },
+    {
+      "epoch": 0.970189073193365,
+      "grad_norm": 5.3595942600030115,
+      "learning_rate": 2e-06,
+      "loss": 0.1905,
+      "step": 4182
+    },
+    {
+      "epoch": 0.9704210648416657,
+      "grad_norm": 10.134794135814035,
+      "learning_rate": 2e-06,
+      "loss": 0.2624,
+      "step": 4183
+    },
+    {
+      "epoch": 0.9706530564899664,
+      "grad_norm": 21.330682910796718,
+      "learning_rate": 2e-06,
+      "loss": 0.311,
+      "step": 4184
+    },
+    {
+      "epoch": 0.970885048138267,
+      "grad_norm": 14.964958740540329,
+      "learning_rate": 2e-06,
+      "loss": 0.2662,
+      "step": 4185
+    },
+    {
+      "epoch": 0.9711170397865677,
+      "grad_norm": 11.493977780236726,
+      "learning_rate": 2e-06,
+      "loss": 0.2423,
+      "step": 4186
+    },
+    {
+      "epoch": 0.9713490314348684,
+      "grad_norm": 16.30578846347033,
+      "learning_rate": 2e-06,
+      "loss": 0.2295,
+      "step": 4187
+    },
+    {
+      "epoch": 0.971581023083169,
+      "grad_norm": 21.62575639230587,
+      "learning_rate": 2e-06,
+      "loss": 0.2544,
+      "step": 4188
+    },
+    {
+      "epoch": 0.9718130147314696,
+      "grad_norm": 14.537087718010312,
+      "learning_rate": 2e-06,
+      "loss": 0.2882,
+      "step": 4189
+    },
+    {
+      "epoch": 0.9720450063797703,
+      "grad_norm": 13.378301810685581,
+      "learning_rate": 2e-06,
+      "loss": 0.326,
+      "step": 4190
+    },
+    {
+      "epoch": 0.972276998028071,
+      "grad_norm": 14.564670167831924,
+      "learning_rate": 2e-06,
+      "loss": 0.2357,
+      "step": 4191
+    },
+    {
+      "epoch": 0.9725089896763717,
+      "grad_norm": 11.895665121896728,
+      "learning_rate": 2e-06,
+      "loss": 0.2513,
+      "step": 4192
+    },
+    {
+      "epoch": 0.9727409813246723,
+      "grad_norm": 29.173504877419095,
+      "learning_rate": 2e-06,
+      "loss": 0.2729,
+      "step": 4193
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 18.29224069723669,
+      "learning_rate": 2e-06,
+      "loss": 0.2807,
+      "step": 4194
+    },
+    {
+      "epoch": 0.9732049646212736,
+      "grad_norm": 10.68192300084348,
+      "learning_rate": 2e-06,
+      "loss": 0.2298,
+      "step": 4195
+    },
+    {
+      "epoch": 0.9734369562695743,
+      "grad_norm": 21.485573797924964,
+      "learning_rate": 2e-06,
+      "loss": 0.3444,
+      "step": 4196
+    },
+    {
+      "epoch": 0.973668947917875,
+      "grad_norm": 7.255065307915454,
+      "learning_rate": 2e-06,
+      "loss": 0.2094,
+      "step": 4197
+    },
+    {
+      "epoch": 0.9739009395661756,
+      "grad_norm": 11.98053928489005,
+      "learning_rate": 2e-06,
+      "loss": 0.2398,
+      "step": 4198
+    },
+    {
+      "epoch": 0.9741329312144763,
+      "grad_norm": 16.306959457810798,
+      "learning_rate": 2e-06,
+      "loss": 0.2924,
+      "step": 4199
+    },
+    {
+      "epoch": 0.974364922862777,
+      "grad_norm": 15.066723005130221,
+      "learning_rate": 2e-06,
+      "loss": 0.2981,
+      "step": 4200
+    },
+    {
+      "epoch": 0.9745969145110776,
+      "grad_norm": 17.8311141003004,
+      "learning_rate": 2e-06,
+      "loss": 0.2809,
+      "step": 4201
+    },
+    {
+      "epoch": 0.9748289061593782,
+      "grad_norm": 17.330070730465206,
+      "learning_rate": 2e-06,
+      "loss": 0.3275,
+      "step": 4202
+    },
+    {
+      "epoch": 0.9750608978076789,
+      "grad_norm": 13.899796439940532,
+      "learning_rate": 2e-06,
+      "loss": 0.2049,
+      "step": 4203
+    },
+    {
+      "epoch": 0.9752928894559796,
+      "grad_norm": 12.515907162025195,
+      "learning_rate": 2e-06,
+      "loss": 0.2258,
+      "step": 4204
+    },
+    {
+      "epoch": 0.9755248811042803,
+      "grad_norm": 14.070388956916347,
+      "learning_rate": 2e-06,
+      "loss": 0.3706,
+      "step": 4205
+    },
+    {
+      "epoch": 0.9757568727525809,
+      "grad_norm": 9.998273054313804,
+      "learning_rate": 2e-06,
+      "loss": 0.2585,
+      "step": 4206
+    },
+    {
+      "epoch": 0.9759888644008816,
+      "grad_norm": 17.59697288259026,
+      "learning_rate": 2e-06,
+      "loss": 0.2893,
+      "step": 4207
+    },
+    {
+      "epoch": 0.9762208560491822,
+      "grad_norm": 10.29582133296254,
+      "learning_rate": 2e-06,
+      "loss": 0.1864,
+      "step": 4208
+    },
+    {
+      "epoch": 0.9764528476974829,
+      "grad_norm": 15.152312481568451,
+      "learning_rate": 2e-06,
+      "loss": 0.2662,
+      "step": 4209
+    },
+    {
+      "epoch": 0.9766848393457835,
+      "grad_norm": 13.491237354588291,
+      "learning_rate": 2e-06,
+      "loss": 0.2564,
+      "step": 4210
+    },
+    {
+      "epoch": 0.9769168309940842,
+      "grad_norm": 22.77241362575545,
+      "learning_rate": 2e-06,
+      "loss": 0.3766,
+      "step": 4211
+    },
+    {
+      "epoch": 0.9771488226423849,
+      "grad_norm": 24.502660636251704,
+      "learning_rate": 2e-06,
+      "loss": 0.3272,
+      "step": 4212
+    },
+    {
+      "epoch": 0.9773808142906856,
+      "grad_norm": 24.914401016942282,
+      "learning_rate": 2e-06,
+      "loss": 0.288,
+      "step": 4213
+    },
+    {
+      "epoch": 0.9776128059389863,
+      "grad_norm": 11.111214277761023,
+      "learning_rate": 2e-06,
+      "loss": 0.2549,
+      "step": 4214
+    },
+    {
+      "epoch": 0.9778447975872868,
+      "grad_norm": 11.49523120194123,
+      "learning_rate": 2e-06,
+      "loss": 0.317,
+      "step": 4215
+    },
+    {
+      "epoch": 0.9780767892355875,
+      "grad_norm": 9.055681111603517,
+      "learning_rate": 2e-06,
+      "loss": 0.2275,
+      "step": 4216
+    },
+    {
+      "epoch": 0.9783087808838882,
+      "grad_norm": 15.174180810545446,
+      "learning_rate": 2e-06,
+      "loss": 0.2826,
+      "step": 4217
+    },
+    {
+      "epoch": 0.9785407725321889,
+      "grad_norm": 13.401080527534248,
+      "learning_rate": 2e-06,
+      "loss": 0.2857,
+      "step": 4218
+    },
+    {
+      "epoch": 0.9787727641804895,
+      "grad_norm": 6.164648447458631,
+      "learning_rate": 2e-06,
+      "loss": 0.1799,
+      "step": 4219
+    },
+    {
+      "epoch": 0.9790047558287902,
+      "grad_norm": 27.775235280358498,
+      "learning_rate": 2e-06,
+      "loss": 0.3474,
+      "step": 4220
+    },
+    {
+      "epoch": 0.9792367474770908,
+      "grad_norm": 18.37992954725875,
+      "learning_rate": 2e-06,
+      "loss": 0.229,
+      "step": 4221
+    },
+    {
+      "epoch": 0.9794687391253915,
+      "grad_norm": 13.19515503304382,
+      "learning_rate": 2e-06,
+      "loss": 0.3051,
+      "step": 4222
+    },
+    {
+      "epoch": 0.9797007307736921,
+      "grad_norm": 11.132635370674684,
+      "learning_rate": 2e-06,
+      "loss": 0.2385,
+      "step": 4223
+    },
+    {
+      "epoch": 0.9799327224219928,
+      "grad_norm": 8.405337802905594,
+      "learning_rate": 2e-06,
+      "loss": 0.1608,
+      "step": 4224
+    },
+    {
+      "epoch": 0.9801647140702935,
+      "grad_norm": 13.917574725560614,
+      "learning_rate": 2e-06,
+      "loss": 0.3205,
+      "step": 4225
+    },
+    {
+      "epoch": 0.9803967057185942,
+      "grad_norm": 6.015763156205712,
+      "learning_rate": 2e-06,
+      "loss": 0.1753,
+      "step": 4226
+    },
+    {
+      "epoch": 0.9806286973668948,
+      "grad_norm": 15.86064516096754,
+      "learning_rate": 2e-06,
+      "loss": 0.3311,
+      "step": 4227
+    },
+    {
+      "epoch": 0.9808606890151954,
+      "grad_norm": 13.07809649383298,
+      "learning_rate": 2e-06,
+      "loss": 0.3762,
+      "step": 4228
+    },
+    {
+      "epoch": 0.9810926806634961,
+      "grad_norm": 22.12385581572482,
+      "learning_rate": 2e-06,
+      "loss": 0.3983,
+      "step": 4229
+    },
+    {
+      "epoch": 0.9813246723117968,
+      "grad_norm": 21.61215045671309,
+      "learning_rate": 2e-06,
+      "loss": 0.2543,
+      "step": 4230
+    },
+    {
+      "epoch": 0.9815566639600974,
+      "grad_norm": 11.469648361559175,
+      "learning_rate": 2e-06,
+      "loss": 0.2123,
+      "step": 4231
+    },
+    {
+      "epoch": 0.9817886556083981,
+      "grad_norm": 22.827779221927173,
+      "learning_rate": 2e-06,
+      "loss": 0.4152,
+      "step": 4232
+    },
+    {
+      "epoch": 0.9820206472566988,
+      "grad_norm": 16.671441093942356,
+      "learning_rate": 2e-06,
+      "loss": 0.2653,
+      "step": 4233
+    },
+    {
+      "epoch": 0.9822526389049995,
+      "grad_norm": 15.742088488552337,
+      "learning_rate": 2e-06,
+      "loss": 0.1953,
+      "step": 4234
+    },
+    {
+      "epoch": 0.9824846305533,
+      "grad_norm": 6.266727832106582,
+      "learning_rate": 2e-06,
+      "loss": 0.1748,
+      "step": 4235
+    },
+    {
+      "epoch": 0.9827166222016007,
+      "grad_norm": 15.052647218778917,
+      "learning_rate": 2e-06,
+      "loss": 0.2619,
+      "step": 4236
+    },
+    {
+      "epoch": 0.9829486138499014,
+      "grad_norm": 11.529712668140277,
+      "learning_rate": 2e-06,
+      "loss": 0.2717,
+      "step": 4237
+    },
+    {
+      "epoch": 0.9831806054982021,
+      "grad_norm": 11.626580721336351,
+      "learning_rate": 2e-06,
+      "loss": 0.3329,
+      "step": 4238
+    },
+    {
+      "epoch": 0.9834125971465028,
+      "grad_norm": 17.82590025916798,
+      "learning_rate": 2e-06,
+      "loss": 0.3335,
+      "step": 4239
+    },
+    {
+      "epoch": 0.9836445887948034,
+      "grad_norm": 13.926504229230394,
+      "learning_rate": 2e-06,
+      "loss": 0.3854,
+      "step": 4240
+    },
+    {
+      "epoch": 0.983876580443104,
+      "grad_norm": 7.309050762390744,
+      "learning_rate": 2e-06,
+      "loss": 0.2475,
+      "step": 4241
+    },
+    {
+      "epoch": 0.9841085720914047,
+      "grad_norm": 10.44914747577233,
+      "learning_rate": 2e-06,
+      "loss": 0.1732,
+      "step": 4242
+    },
+    {
+      "epoch": 0.9843405637397054,
+      "grad_norm": 17.92649994514198,
+      "learning_rate": 2e-06,
+      "loss": 0.2721,
+      "step": 4243
+    },
+    {
+      "epoch": 0.984572555388006,
+      "grad_norm": 12.515934818615193,
+      "learning_rate": 2e-06,
+      "loss": 0.2515,
+      "step": 4244
+    },
+    {
+      "epoch": 0.9848045470363067,
+      "grad_norm": 11.349447658048096,
+      "learning_rate": 2e-06,
+      "loss": 0.2519,
+      "step": 4245
+    },
+    {
+      "epoch": 0.9850365386846074,
+      "grad_norm": 17.278232190175704,
+      "learning_rate": 2e-06,
+      "loss": 0.2798,
+      "step": 4246
+    },
+    {
+      "epoch": 0.9852685303329081,
+      "grad_norm": 19.564700818775094,
+      "learning_rate": 2e-06,
+      "loss": 0.3114,
+      "step": 4247
+    },
+    {
+      "epoch": 0.9855005219812086,
+      "grad_norm": 11.075353127881275,
+      "learning_rate": 2e-06,
+      "loss": 0.2909,
+      "step": 4248
+    },
+    {
+      "epoch": 0.9857325136295093,
+      "grad_norm": 12.89124213988945,
+      "learning_rate": 2e-06,
+      "loss": 0.1592,
+      "step": 4249
+    },
+    {
+      "epoch": 0.98596450527781,
+      "grad_norm": 16.98800196949275,
+      "learning_rate": 2e-06,
+      "loss": 0.3969,
+      "step": 4250
+    },
+    {
+      "epoch": 0.9861964969261107,
+      "grad_norm": 13.578942902506654,
+      "learning_rate": 2e-06,
+      "loss": 0.2777,
+      "step": 4251
+    },
+    {
+      "epoch": 0.9864284885744113,
+      "grad_norm": 8.1871697074597,
+      "learning_rate": 2e-06,
+      "loss": 0.203,
+      "step": 4252
+    },
+    {
+      "epoch": 0.986660480222712,
+      "grad_norm": 7.040253276385697,
+      "learning_rate": 2e-06,
+      "loss": 0.224,
+      "step": 4253
+    },
+    {
+      "epoch": 0.9868924718710126,
+      "grad_norm": 14.405757347393758,
+      "learning_rate": 2e-06,
+      "loss": 0.2295,
+      "step": 4254
+    },
+    {
+      "epoch": 0.9871244635193133,
+      "grad_norm": 9.033447010628546,
+      "learning_rate": 2e-06,
+      "loss": 0.1497,
+      "step": 4255
+    },
+    {
+      "epoch": 0.9873564551676139,
+      "grad_norm": 20.97446599435015,
+      "learning_rate": 2e-06,
+      "loss": 0.1794,
+      "step": 4256
+    },
+    {
+      "epoch": 0.9875884468159146,
+      "grad_norm": 6.904763491483926,
+      "learning_rate": 2e-06,
+      "loss": 0.1891,
+      "step": 4257
+    },
+    {
+      "epoch": 0.9878204384642153,
+      "grad_norm": 18.688009180195195,
+      "learning_rate": 2e-06,
+      "loss": 0.3001,
+      "step": 4258
+    },
+    {
+      "epoch": 0.988052430112516,
+      "grad_norm": 13.364486799681748,
+      "learning_rate": 2e-06,
+      "loss": 0.2617,
+      "step": 4259
+    },
+    {
+      "epoch": 0.9882844217608167,
+      "grad_norm": 15.973372169182674,
+      "learning_rate": 2e-06,
+      "loss": 0.2303,
+      "step": 4260
+    },
+    {
+      "epoch": 0.9885164134091172,
+      "grad_norm": 10.556847505773627,
+      "learning_rate": 2e-06,
+      "loss": 0.3017,
+      "step": 4261
+    },
+    {
+      "epoch": 0.9887484050574179,
+      "grad_norm": 23.634222130250084,
+      "learning_rate": 2e-06,
+      "loss": 0.4199,
+      "step": 4262
+    },
+    {
+      "epoch": 0.9889803967057186,
+      "grad_norm": 16.572984667619835,
+      "learning_rate": 2e-06,
+      "loss": 0.3701,
+      "step": 4263
+    },
+    {
+      "epoch": 0.9892123883540193,
+      "grad_norm": 14.404180757195148,
+      "learning_rate": 2e-06,
+      "loss": 0.332,
+      "step": 4264
+    },
+    {
+      "epoch": 0.9894443800023199,
+      "grad_norm": 13.811304035079404,
+      "learning_rate": 2e-06,
+      "loss": 0.2003,
+      "step": 4265
+    },
+    {
+      "epoch": 0.9896763716506206,
+      "grad_norm": 11.8826651967571,
+      "learning_rate": 2e-06,
+      "loss": 0.333,
+      "step": 4266
+    },
+    {
+      "epoch": 0.9899083632989213,
+      "grad_norm": 14.923848968766313,
+      "learning_rate": 2e-06,
+      "loss": 0.4666,
+      "step": 4267
+    },
+    {
+      "epoch": 0.9901403549472219,
+      "grad_norm": 19.934166664100598,
+      "learning_rate": 2e-06,
+      "loss": 0.3621,
+      "step": 4268
+    },
+    {
+      "epoch": 0.9903723465955225,
+      "grad_norm": 10.548852262130172,
+      "learning_rate": 2e-06,
+      "loss": 0.2605,
+      "step": 4269
+    },
+    {
+      "epoch": 0.9906043382438232,
+      "grad_norm": 17.044978616656284,
+      "learning_rate": 2e-06,
+      "loss": 0.2738,
+      "step": 4270
+    },
+    {
+      "epoch": 0.9908363298921239,
+      "grad_norm": 7.3678955030967,
+      "learning_rate": 2e-06,
+      "loss": 0.1474,
+      "step": 4271
+    },
+    {
+      "epoch": 0.9910683215404246,
+      "grad_norm": 18.214716715347542,
+      "learning_rate": 2e-06,
+      "loss": 0.4379,
+      "step": 4272
+    },
+    {
+      "epoch": 0.9913003131887252,
+      "grad_norm": 12.945968711053286,
+      "learning_rate": 2e-06,
+      "loss": 0.2377,
+      "step": 4273
+    },
+    {
+      "epoch": 0.9915323048370258,
+      "grad_norm": 17.728922143117316,
+      "learning_rate": 2e-06,
+      "loss": 0.3382,
+      "step": 4274
+    },
+    {
+      "epoch": 0.9917642964853265,
+      "grad_norm": 18.646348539918346,
+      "learning_rate": 2e-06,
+      "loss": 0.3737,
+      "step": 4275
+    },
+    {
+      "epoch": 0.9919962881336272,
+      "grad_norm": 17.22778162656779,
+      "learning_rate": 2e-06,
+      "loss": 0.281,
+      "step": 4276
+    },
+    {
+      "epoch": 0.9922282797819278,
+      "grad_norm": 12.540238388456121,
+      "learning_rate": 2e-06,
+      "loss": 0.3496,
+      "step": 4277
+    },
+    {
+      "epoch": 0.9924602714302285,
+      "grad_norm": 9.732713018493623,
+      "learning_rate": 2e-06,
+      "loss": 0.2186,
+      "step": 4278
+    },
+    {
+      "epoch": 0.9926922630785292,
+      "grad_norm": 7.953560031755493,
+      "learning_rate": 2e-06,
+      "loss": 0.2391,
+      "step": 4279
+    },
+    {
+      "epoch": 0.9929242547268299,
+      "grad_norm": 11.722711090296285,
+      "learning_rate": 2e-06,
+      "loss": 0.2226,
+      "step": 4280
+    },
+    {
+      "epoch": 0.9931562463751304,
+      "grad_norm": 18.63265751918661,
+      "learning_rate": 2e-06,
+      "loss": 0.2999,
+      "step": 4281
+    },
+    {
+      "epoch": 0.9933882380234311,
+      "grad_norm": 10.19449522121469,
+      "learning_rate": 2e-06,
+      "loss": 0.2203,
+      "step": 4282
+    },
+    {
+      "epoch": 0.9936202296717318,
+      "grad_norm": 18.11672937369853,
+      "learning_rate": 2e-06,
+      "loss": 0.3204,
+      "step": 4283
+    },
+    {
+      "epoch": 0.9938522213200325,
+      "grad_norm": 16.01830303614881,
+      "learning_rate": 2e-06,
+      "loss": 0.3197,
+      "step": 4284
+    },
+    {
+      "epoch": 0.9940842129683332,
+      "grad_norm": 17.3963087347634,
+      "learning_rate": 2e-06,
+      "loss": 0.3376,
+      "step": 4285
+    },
+    {
+      "epoch": 0.9943162046166338,
+      "grad_norm": 17.667883565032273,
+      "learning_rate": 2e-06,
+      "loss": 0.3051,
+      "step": 4286
+    },
+    {
+      "epoch": 0.9945481962649345,
+      "grad_norm": 8.313509916883348,
+      "learning_rate": 2e-06,
+      "loss": 0.2424,
+      "step": 4287
+    },
+    {
+      "epoch": 0.9947801879132351,
+      "grad_norm": 21.853369792188385,
+      "learning_rate": 2e-06,
+      "loss": 0.3898,
+      "step": 4288
+    },
+    {
+      "epoch": 0.9950121795615358,
+      "grad_norm": 12.909500013417489,
+      "learning_rate": 2e-06,
+      "loss": 0.2194,
+      "step": 4289
+    },
+    {
+      "epoch": 0.9952441712098364,
+      "grad_norm": 15.270128959373558,
+      "learning_rate": 2e-06,
+      "loss": 0.2795,
+      "step": 4290
+    },
+    {
+      "epoch": 0.9954761628581371,
+      "grad_norm": 21.79462981369358,
+      "learning_rate": 2e-06,
+      "loss": 0.339,
+      "step": 4291
+    },
+    {
+      "epoch": 0.9957081545064378,
+      "grad_norm": 19.88564636383995,
+      "learning_rate": 2e-06,
+      "loss": 0.3024,
+      "step": 4292
+    },
+    {
+      "epoch": 0.9959401461547385,
+      "grad_norm": 14.399655142947225,
+      "learning_rate": 2e-06,
+      "loss": 0.2895,
+      "step": 4293
+    },
+    {
+      "epoch": 0.996172137803039,
+      "grad_norm": 19.85949576833494,
+      "learning_rate": 2e-06,
+      "loss": 0.3431,
+      "step": 4294
+    },
+    {
+      "epoch": 0.9964041294513397,
+      "grad_norm": 11.683854205528311,
+      "learning_rate": 2e-06,
+      "loss": 0.2807,
+      "step": 4295
+    },
+    {
+      "epoch": 0.9966361210996404,
+      "grad_norm": 17.951758210842055,
+      "learning_rate": 2e-06,
+      "loss": 0.3739,
+      "step": 4296
+    },
+    {
+      "epoch": 0.9968681127479411,
+      "grad_norm": 21.711123070741547,
+      "learning_rate": 2e-06,
+      "loss": 0.2775,
+      "step": 4297
+    },
+    {
+      "epoch": 0.9971001043962417,
+      "grad_norm": 12.616016808487865,
+      "learning_rate": 2e-06,
+      "loss": 0.2993,
+      "step": 4298
+    },
+    {
+      "epoch": 0.9973320960445424,
+      "grad_norm": 20.280139049312588,
+      "learning_rate": 2e-06,
+      "loss": 0.2819,
+      "step": 4299
+    },
+    {
+      "epoch": 0.9975640876928431,
+      "grad_norm": 12.354640230223774,
+      "learning_rate": 2e-06,
+      "loss": 0.264,
+      "step": 4300
+    },
+    {
+      "epoch": 0.9977960793411437,
+      "grad_norm": 8.04778189340334,
+      "learning_rate": 2e-06,
+      "loss": 0.3385,
+      "step": 4301
+    },
+    {
+      "epoch": 0.9980280709894443,
+      "grad_norm": 13.08075846529473,
+      "learning_rate": 2e-06,
+      "loss": 0.1973,
+      "step": 4302
+    },
+    {
+      "epoch": 0.998260062637745,
+      "grad_norm": 12.520305141646622,
+      "learning_rate": 2e-06,
+      "loss": 0.3255,
+      "step": 4303
+    },
+    {
+      "epoch": 0.9984920542860457,
+      "grad_norm": 11.281683034711,
+      "learning_rate": 2e-06,
+      "loss": 0.2641,
+      "step": 4304
+    },
+    {
+      "epoch": 0.9987240459343464,
+      "grad_norm": 9.467801053894949,
+      "learning_rate": 2e-06,
+      "loss": 0.2862,
+      "step": 4305
+    },
+    {
+      "epoch": 0.9989560375826471,
+      "grad_norm": 10.738481898576978,
+      "learning_rate": 2e-06,
+      "loss": 0.2807,
+      "step": 4306
+    },
+    {
+      "epoch": 0.9991880292309477,
+      "grad_norm": 12.40324963497283,
+      "learning_rate": 2e-06,
+      "loss": 0.2784,
+      "step": 4307
+    },
+    {
+      "epoch": 0.9994200208792483,
+      "grad_norm": 17.279743062019485,
+      "learning_rate": 2e-06,
+      "loss": 0.3601,
+      "step": 4308
+    },
+    {
+      "epoch": 0.999652012527549,
+      "grad_norm": 12.254255842119658,
+      "learning_rate": 2e-06,
+      "loss": 0.2333,
+      "step": 4309
+    },
+    {
+      "epoch": 0.9998840041758497,
+      "grad_norm": 15.644995614708913,
+      "learning_rate": 2e-06,
+      "loss": 0.301,
+      "step": 4310
+    },
+    {
+      "epoch": 1.0001159958241503,
+      "grad_norm": 16.718302350093527,
+      "learning_rate": 2e-06,
+      "loss": 0.3431,
+      "step": 4311
+    },
+    {
+      "epoch": 1.000347987472451,
+      "grad_norm": 5.981456267975129,
+      "learning_rate": 2e-06,
+      "loss": 0.278,
+      "step": 4312
+    },
+    {
+      "epoch": 1.0005799791207517,
+      "grad_norm": 11.01508633788959,
+      "learning_rate": 2e-06,
+      "loss": 0.2755,
+      "step": 4313
+    },
+    {
+      "epoch": 1.0008119707690524,
+      "grad_norm": 10.28879898371332,
+      "learning_rate": 2e-06,
+      "loss": 0.2578,
+      "step": 4314
+    },
+    {
+      "epoch": 1.001043962417353,
+      "grad_norm": 19.773665483140853,
+      "learning_rate": 2e-06,
+      "loss": 0.427,
+      "step": 4315
+    },
+    {
+      "epoch": 1.0012759540656537,
+      "grad_norm": 14.035676093485598,
+      "learning_rate": 2e-06,
+      "loss": 0.3006,
+      "step": 4316
+    },
+    {
+      "epoch": 1.0015079457139544,
+      "grad_norm": 13.1091865357948,
+      "learning_rate": 2e-06,
+      "loss": 0.2829,
+      "step": 4317
+    },
+    {
+      "epoch": 1.0017399373622549,
+      "grad_norm": 19.786693874665463,
+      "learning_rate": 2e-06,
+      "loss": 0.319,
+      "step": 4318
+    },
+    {
+      "epoch": 1.0019719290105555,
+      "grad_norm": 7.270669487731423,
+      "learning_rate": 2e-06,
+      "loss": 0.2108,
+      "step": 4319
+    },
+    {
+      "epoch": 1.0022039206588562,
+      "grad_norm": 20.17286026961784,
+      "learning_rate": 2e-06,
+      "loss": 0.3011,
+      "step": 4320
+    },
+    {
+      "epoch": 1.002435912307157,
+      "grad_norm": 12.587046558200004,
+      "learning_rate": 2e-06,
+      "loss": 0.2785,
+      "step": 4321
+    },
+    {
+      "epoch": 1.0026679039554576,
+      "grad_norm": 8.869916800554623,
+      "learning_rate": 2e-06,
+      "loss": 0.2481,
+      "step": 4322
+    },
+    {
+      "epoch": 1.0028998956037583,
+      "grad_norm": 12.136751295143798,
+      "learning_rate": 2e-06,
+      "loss": 0.264,
+      "step": 4323
+    },
+    {
+      "epoch": 1.003131887252059,
+      "grad_norm": 8.667913267118324,
+      "learning_rate": 2e-06,
+      "loss": 0.1941,
+      "step": 4324
+    },
+    {
+      "epoch": 1.0033638789003596,
+      "grad_norm": 14.403161655669786,
+      "learning_rate": 2e-06,
+      "loss": 0.2535,
+      "step": 4325
+    },
+    {
+      "epoch": 1.0035958705486603,
+      "grad_norm": 18.612318169636975,
+      "learning_rate": 2e-06,
+      "loss": 0.2766,
+      "step": 4326
+    },
+    {
+      "epoch": 1.003827862196961,
+      "grad_norm": 16.99645141292192,
+      "learning_rate": 2e-06,
+      "loss": 0.363,
+      "step": 4327
+    },
+    {
+      "epoch": 1.0040598538452616,
+      "grad_norm": 13.751830433854773,
+      "learning_rate": 2e-06,
+      "loss": 0.155,
+      "step": 4328
+    },
+    {
+      "epoch": 1.0042918454935623,
+      "grad_norm": 13.671771418605278,
+      "learning_rate": 2e-06,
+      "loss": 0.2716,
+      "step": 4329
+    },
+    {
+      "epoch": 1.004523837141863,
+      "grad_norm": 18.752642833061604,
+      "learning_rate": 2e-06,
+      "loss": 0.3661,
+      "step": 4330
+    },
+    {
+      "epoch": 1.0047558287901635,
+      "grad_norm": 14.847783747948847,
+      "learning_rate": 2e-06,
+      "loss": 0.2702,
+      "step": 4331
+    },
+    {
+      "epoch": 1.0049878204384641,
+      "grad_norm": 17.885233787267723,
+      "learning_rate": 2e-06,
+      "loss": 0.3698,
+      "step": 4332
+    },
+    {
+      "epoch": 1.0052198120867648,
+      "grad_norm": 27.13789320974805,
+      "learning_rate": 2e-06,
+      "loss": 0.3145,
+      "step": 4333
+    },
+    {
+      "epoch": 1.0054518037350655,
+      "grad_norm": 13.304508645892248,
+      "learning_rate": 2e-06,
+      "loss": 0.2326,
+      "step": 4334
+    },
+    {
+      "epoch": 1.0056837953833662,
+      "grad_norm": 9.862556153578181,
+      "learning_rate": 2e-06,
+      "loss": 0.1547,
+      "step": 4335
+    },
+    {
+      "epoch": 1.0059157870316668,
+      "grad_norm": 10.683912066124044,
+      "learning_rate": 2e-06,
+      "loss": 0.3168,
+      "step": 4336
+    },
+    {
+      "epoch": 1.0061477786799675,
+      "grad_norm": 12.134907278341815,
+      "learning_rate": 2e-06,
+      "loss": 0.2847,
+      "step": 4337
+    },
+    {
+      "epoch": 1.0063797703282682,
+      "grad_norm": 16.990495829795368,
+      "learning_rate": 2e-06,
+      "loss": 0.1884,
+      "step": 4338
+    },
+    {
+      "epoch": 1.0066117619765689,
+      "grad_norm": 11.49411437654733,
+      "learning_rate": 2e-06,
+      "loss": 0.2447,
+      "step": 4339
+    },
+    {
+      "epoch": 1.0068437536248696,
+      "grad_norm": 14.995630381719709,
+      "learning_rate": 2e-06,
+      "loss": 0.2222,
+      "step": 4340
+    },
+    {
+      "epoch": 1.0070757452731702,
+      "grad_norm": 13.971606657098535,
+      "learning_rate": 2e-06,
+      "loss": 0.2872,
+      "step": 4341
+    },
+    {
+      "epoch": 1.007307736921471,
+      "grad_norm": 20.96371526573696,
+      "learning_rate": 2e-06,
+      "loss": 0.2905,
+      "step": 4342
+    },
+    {
+      "epoch": 1.0075397285697716,
+      "grad_norm": 9.266262013323754,
+      "learning_rate": 2e-06,
+      "loss": 0.1939,
+      "step": 4343
+    },
+    {
+      "epoch": 1.007771720218072,
+      "grad_norm": 19.77717418896224,
+      "learning_rate": 2e-06,
+      "loss": 0.3568,
+      "step": 4344
+    },
+    {
+      "epoch": 1.0080037118663727,
+      "grad_norm": 12.26230275832593,
+      "learning_rate": 2e-06,
+      "loss": 0.3573,
+      "step": 4345
+    },
+    {
+      "epoch": 1.0082357035146734,
+      "grad_norm": 14.19253024029564,
+      "learning_rate": 2e-06,
+      "loss": 0.3354,
+      "step": 4346
+    },
+    {
+      "epoch": 1.008467695162974,
+      "grad_norm": 10.803220157659359,
+      "learning_rate": 2e-06,
+      "loss": 0.27,
+      "step": 4347
+    },
+    {
+      "epoch": 1.0086996868112748,
+      "grad_norm": 23.124982254178747,
+      "learning_rate": 2e-06,
+      "loss": 0.3364,
+      "step": 4348
+    },
+    {
+      "epoch": 1.0089316784595754,
+      "grad_norm": 9.731324037618686,
+      "learning_rate": 2e-06,
+      "loss": 0.1988,
+      "step": 4349
+    },
+    {
+      "epoch": 1.009163670107876,
+      "grad_norm": 7.93294592056329,
+      "learning_rate": 2e-06,
+      "loss": 0.2416,
+      "step": 4350
+    },
+    {
+      "epoch": 1.0093956617561768,
+      "grad_norm": 14.057267162792947,
+      "learning_rate": 2e-06,
+      "loss": 0.32,
+      "step": 4351
+    },
+    {
+      "epoch": 1.0096276534044775,
+      "grad_norm": 13.900748666564924,
+      "learning_rate": 2e-06,
+      "loss": 0.2171,
+      "step": 4352
+    },
+    {
+      "epoch": 1.0098596450527781,
+      "grad_norm": 8.245172013413677,
+      "learning_rate": 2e-06,
+      "loss": 0.2147,
+      "step": 4353
+    },
+    {
+      "epoch": 1.0100916367010788,
+      "grad_norm": 10.271433355787496,
+      "learning_rate": 2e-06,
+      "loss": 0.1919,
+      "step": 4354
+    },
+    {
+      "epoch": 1.0103236283493795,
+      "grad_norm": 8.194394625100914,
+      "learning_rate": 2e-06,
+      "loss": 0.1986,
+      "step": 4355
+    },
+    {
+      "epoch": 1.0105556199976802,
+      "grad_norm": 15.079051609451382,
+      "learning_rate": 2e-06,
+      "loss": 0.2572,
+      "step": 4356
+    },
+    {
+      "epoch": 1.0107876116459809,
+      "grad_norm": 15.291543050990853,
+      "learning_rate": 2e-06,
+      "loss": 0.4196,
+      "step": 4357
+    },
+    {
+      "epoch": 1.0110196032942813,
+      "grad_norm": 17.82830434225328,
+      "learning_rate": 2e-06,
+      "loss": 0.2908,
+      "step": 4358
+    },
+    {
+      "epoch": 1.011251594942582,
+      "grad_norm": 12.837214688662794,
+      "learning_rate": 2e-06,
+      "loss": 0.3404,
+      "step": 4359
+    },
+    {
+      "epoch": 1.0114835865908827,
+      "grad_norm": 11.901047019390361,
+      "learning_rate": 2e-06,
+      "loss": 0.2712,
+      "step": 4360
+    },
+    {
+      "epoch": 1.0117155782391833,
+      "grad_norm": 23.758163883570887,
+      "learning_rate": 2e-06,
+      "loss": 0.3621,
+      "step": 4361
+    },
+    {
+      "epoch": 1.011947569887484,
+      "grad_norm": 11.482949711438708,
+      "learning_rate": 2e-06,
+      "loss": 0.147,
+      "step": 4362
+    },
+    {
+      "epoch": 1.0121795615357847,
+      "grad_norm": 10.940010722924,
+      "learning_rate": 2e-06,
+      "loss": 0.317,
+      "step": 4363
+    },
+    {
+      "epoch": 1.0124115531840854,
+      "grad_norm": 9.369929809574218,
+      "learning_rate": 2e-06,
+      "loss": 0.1951,
+      "step": 4364
+    },
+    {
+      "epoch": 1.012643544832386,
+      "grad_norm": 9.403910759231197,
+      "learning_rate": 2e-06,
+      "loss": 0.2646,
+      "step": 4365
+    },
+    {
+      "epoch": 1.0128755364806867,
+      "grad_norm": 11.168686934520977,
+      "learning_rate": 2e-06,
+      "loss": 0.2605,
+      "step": 4366
+    },
+    {
+      "epoch": 1.0131075281289874,
+      "grad_norm": 11.041500841883662,
+      "learning_rate": 2e-06,
+      "loss": 0.3602,
+      "step": 4367
+    },
+    {
+      "epoch": 1.013339519777288,
+      "grad_norm": 9.369178993030669,
+      "learning_rate": 2e-06,
+      "loss": 0.2316,
+      "step": 4368
+    },
+    {
+      "epoch": 1.0135715114255888,
+      "grad_norm": 12.33230547811149,
+      "learning_rate": 2e-06,
+      "loss": 0.1931,
+      "step": 4369
+    },
+    {
+      "epoch": 1.0138035030738894,
+      "grad_norm": 13.487586445637827,
+      "learning_rate": 2e-06,
+      "loss": 0.3596,
+      "step": 4370
+    },
+    {
+      "epoch": 1.01403549472219,
+      "grad_norm": 15.733338825367728,
+      "learning_rate": 2e-06,
+      "loss": 0.214,
+      "step": 4371
+    },
+    {
+      "epoch": 1.0142674863704906,
+      "grad_norm": 14.042948766828657,
+      "learning_rate": 2e-06,
+      "loss": 0.2801,
+      "step": 4372
+    },
+    {
+      "epoch": 1.0144994780187913,
+      "grad_norm": 10.908660927650285,
+      "learning_rate": 2e-06,
+      "loss": 0.2215,
+      "step": 4373
+    },
+    {
+      "epoch": 1.014731469667092,
+      "grad_norm": 12.053128384428641,
+      "learning_rate": 2e-06,
+      "loss": 0.2202,
+      "step": 4374
+    },
+    {
+      "epoch": 1.0149634613153926,
+      "grad_norm": 15.94150176034634,
+      "learning_rate": 2e-06,
+      "loss": 0.34,
+      "step": 4375
+    },
+    {
+      "epoch": 1.0151954529636933,
+      "grad_norm": 30.488268098804266,
+      "learning_rate": 2e-06,
+      "loss": 0.3275,
+      "step": 4376
+    },
+    {
+      "epoch": 1.015427444611994,
+      "grad_norm": 16.418756536831573,
+      "learning_rate": 2e-06,
+      "loss": 0.2529,
+      "step": 4377
+    },
+    {
+      "epoch": 1.0156594362602946,
+      "grad_norm": 26.929167976037675,
+      "learning_rate": 2e-06,
+      "loss": 0.3632,
+      "step": 4378
+    },
+    {
+      "epoch": 1.0158914279085953,
+      "grad_norm": 13.390221016403952,
+      "learning_rate": 2e-06,
+      "loss": 0.1912,
+      "step": 4379
+    },
+    {
+      "epoch": 1.016123419556896,
+      "grad_norm": 11.176161477331638,
+      "learning_rate": 2e-06,
+      "loss": 0.2135,
+      "step": 4380
+    },
+    {
+      "epoch": 1.0163554112051967,
+      "grad_norm": 19.321719328090147,
+      "learning_rate": 2e-06,
+      "loss": 0.38,
+      "step": 4381
+    },
+    {
+      "epoch": 1.0165874028534974,
+      "grad_norm": 8.73358529268335,
+      "learning_rate": 2e-06,
+      "loss": 0.2182,
+      "step": 4382
+    },
+    {
+      "epoch": 1.016819394501798,
+      "grad_norm": 11.346499255352734,
+      "learning_rate": 2e-06,
+      "loss": 0.2417,
+      "step": 4383
+    },
+    {
+      "epoch": 1.0170513861500985,
+      "grad_norm": 9.501895122254266,
+      "learning_rate": 2e-06,
+      "loss": 0.2326,
+      "step": 4384
+    },
+    {
+      "epoch": 1.0172833777983992,
+      "grad_norm": 12.229807304921541,
+      "learning_rate": 2e-06,
+      "loss": 0.2692,
+      "step": 4385
+    },
+    {
+      "epoch": 1.0175153694466998,
+      "grad_norm": 11.847609278185644,
+      "learning_rate": 2e-06,
+      "loss": 0.2968,
+      "step": 4386
+    },
+    {
+      "epoch": 1.0177473610950005,
+      "grad_norm": 9.549411930106775,
+      "learning_rate": 2e-06,
+      "loss": 0.1805,
+      "step": 4387
+    },
+    {
+      "epoch": 1.0179793527433012,
+      "grad_norm": 9.267107386625788,
+      "learning_rate": 2e-06,
+      "loss": 0.2132,
+      "step": 4388
+    },
+    {
+      "epoch": 1.0182113443916019,
+      "grad_norm": 8.690027982071577,
+      "learning_rate": 2e-06,
+      "loss": 0.2782,
+      "step": 4389
+    },
+    {
+      "epoch": 1.0184433360399026,
+      "grad_norm": 12.795072421966484,
+      "learning_rate": 2e-06,
+      "loss": 0.274,
+      "step": 4390
+    },
+    {
+      "epoch": 1.0186753276882032,
+      "grad_norm": 5.842035700725516,
+      "learning_rate": 2e-06,
+      "loss": 0.1685,
+      "step": 4391
+    },
+    {
+      "epoch": 1.018907319336504,
+      "grad_norm": 14.051283930385816,
+      "learning_rate": 2e-06,
+      "loss": 0.254,
+      "step": 4392
+    },
+    {
+      "epoch": 1.0191393109848046,
+      "grad_norm": 13.348897262207288,
+      "learning_rate": 2e-06,
+      "loss": 0.2763,
+      "step": 4393
+    },
+    {
+      "epoch": 1.0193713026331053,
+      "grad_norm": 22.540889105334582,
+      "learning_rate": 2e-06,
+      "loss": 0.3685,
+      "step": 4394
+    },
+    {
+      "epoch": 1.019603294281406,
+      "grad_norm": 11.525100788667038,
+      "learning_rate": 2e-06,
+      "loss": 0.2691,
+      "step": 4395
+    },
+    {
+      "epoch": 1.0198352859297066,
+      "grad_norm": 18.453868704349077,
+      "learning_rate": 2e-06,
+      "loss": 0.3714,
+      "step": 4396
+    },
+    {
+      "epoch": 1.0200672775780073,
+      "grad_norm": 14.36172001747481,
+      "learning_rate": 2e-06,
+      "loss": 0.3594,
+      "step": 4397
+    },
+    {
+      "epoch": 1.0202992692263078,
+      "grad_norm": 11.360356324324796,
+      "learning_rate": 2e-06,
+      "loss": 0.2799,
+      "step": 4398
+    },
+    {
+      "epoch": 1.0205312608746084,
+      "grad_norm": 15.46130705982928,
+      "learning_rate": 2e-06,
+      "loss": 0.2868,
+      "step": 4399
+    },
+    {
+      "epoch": 1.0207632525229091,
+      "grad_norm": 18.75112773586128,
+      "learning_rate": 2e-06,
+      "loss": 0.2558,
+      "step": 4400
+    },
+    {
+      "epoch": 1.0209952441712098,
+      "grad_norm": 14.047846349720592,
+      "learning_rate": 2e-06,
+      "loss": 0.2228,
+      "step": 4401
+    },
+    {
+      "epoch": 1.0212272358195105,
+      "grad_norm": 8.51175892802764,
+      "learning_rate": 2e-06,
+      "loss": 0.2604,
+      "step": 4402
+    },
+    {
+      "epoch": 1.0214592274678111,
+      "grad_norm": 10.553828364794468,
+      "learning_rate": 2e-06,
+      "loss": 0.1464,
+      "step": 4403
+    },
+    {
+      "epoch": 1.0216912191161118,
+      "grad_norm": 16.277207470334066,
+      "learning_rate": 2e-06,
+      "loss": 0.2669,
+      "step": 4404
+    },
+    {
+      "epoch": 1.0219232107644125,
+      "grad_norm": 14.650722863677906,
+      "learning_rate": 2e-06,
+      "loss": 0.4093,
+      "step": 4405
+    },
+    {
+      "epoch": 1.0221552024127132,
+      "grad_norm": 20.686933586265983,
+      "learning_rate": 2e-06,
+      "loss": 0.3219,
+      "step": 4406
+    },
+    {
+      "epoch": 1.0223871940610139,
+      "grad_norm": 18.149022562966458,
+      "learning_rate": 2e-06,
+      "loss": 0.2369,
+      "step": 4407
+    },
+    {
+      "epoch": 1.0226191857093145,
+      "grad_norm": 13.441727627483365,
+      "learning_rate": 2e-06,
+      "loss": 0.2794,
+      "step": 4408
+    },
+    {
+      "epoch": 1.0228511773576152,
+      "grad_norm": 9.002417763935158,
+      "learning_rate": 2e-06,
+      "loss": 0.1919,
+      "step": 4409
+    },
+    {
+      "epoch": 1.023083169005916,
+      "grad_norm": 9.604553834199644,
+      "learning_rate": 2e-06,
+      "loss": 0.1538,
+      "step": 4410
+    },
+    {
+      "epoch": 1.0233151606542163,
+      "grad_norm": 7.693559744151872,
+      "learning_rate": 2e-06,
+      "loss": 0.1894,
+      "step": 4411
+    },
+    {
+      "epoch": 1.023547152302517,
+      "grad_norm": 13.679783666333623,
+      "learning_rate": 2e-06,
+      "loss": 0.2397,
+      "step": 4412
+    },
+    {
+      "epoch": 1.0237791439508177,
+      "grad_norm": 16.107777090519455,
+      "learning_rate": 2e-06,
+      "loss": 0.2707,
+      "step": 4413
+    },
+    {
+      "epoch": 1.0240111355991184,
+      "grad_norm": 22.498111337105726,
+      "learning_rate": 2e-06,
+      "loss": 0.235,
+      "step": 4414
+    },
+    {
+      "epoch": 1.024243127247419,
+      "grad_norm": 11.17453992197758,
+      "learning_rate": 2e-06,
+      "loss": 0.3602,
+      "step": 4415
+    },
+    {
+      "epoch": 1.0244751188957197,
+      "grad_norm": 13.490648517081917,
+      "learning_rate": 2e-06,
+      "loss": 0.2619,
+      "step": 4416
+    },
+    {
+      "epoch": 1.0247071105440204,
+      "grad_norm": 14.101245578112804,
+      "learning_rate": 2e-06,
+      "loss": 0.3315,
+      "step": 4417
+    },
+    {
+      "epoch": 1.024939102192321,
+      "grad_norm": 14.941819487142078,
+      "learning_rate": 2e-06,
+      "loss": 0.2541,
+      "step": 4418
+    },
+    {
+      "epoch": 1.0251710938406218,
+      "grad_norm": 15.85648435924746,
+      "learning_rate": 2e-06,
+      "loss": 0.3405,
+      "step": 4419
+    },
+    {
+      "epoch": 1.0254030854889224,
+      "grad_norm": 17.142409104258302,
+      "learning_rate": 2e-06,
+      "loss": 0.2756,
+      "step": 4420
+    },
+    {
+      "epoch": 1.0256350771372231,
+      "grad_norm": 14.136208692909497,
+      "learning_rate": 2e-06,
+      "loss": 0.3226,
+      "step": 4421
+    },
+    {
+      "epoch": 1.0258670687855238,
+      "grad_norm": 11.199660197653973,
+      "learning_rate": 2e-06,
+      "loss": 0.1979,
+      "step": 4422
+    },
+    {
+      "epoch": 1.0260990604338245,
+      "grad_norm": 27.70601825277928,
+      "learning_rate": 2e-06,
+      "loss": 0.374,
+      "step": 4423
+    },
+    {
+      "epoch": 1.026331052082125,
+      "grad_norm": 20.308926895318287,
+      "learning_rate": 2e-06,
+      "loss": 0.2835,
+      "step": 4424
+    },
+    {
+      "epoch": 1.0265630437304256,
+      "grad_norm": 11.668644087020201,
+      "learning_rate": 2e-06,
+      "loss": 0.159,
+      "step": 4425
+    },
+    {
+      "epoch": 1.0267950353787263,
+      "grad_norm": 14.05495343536885,
+      "learning_rate": 2e-06,
+      "loss": 0.2125,
+      "step": 4426
+    },
+    {
+      "epoch": 1.027027027027027,
+      "grad_norm": 9.009071699795452,
+      "learning_rate": 2e-06,
+      "loss": 0.28,
+      "step": 4427
+    },
+    {
+      "epoch": 1.0272590186753277,
+      "grad_norm": 24.15482898661228,
+      "learning_rate": 2e-06,
+      "loss": 0.2167,
+      "step": 4428
+    },
+    {
+      "epoch": 1.0274910103236283,
+      "grad_norm": 12.582332412328576,
+      "learning_rate": 2e-06,
+      "loss": 0.1751,
+      "step": 4429
+    },
+    {
+      "epoch": 1.027723001971929,
+      "grad_norm": 21.538185705261743,
+      "learning_rate": 2e-06,
+      "loss": 0.3409,
+      "step": 4430
+    },
+    {
+      "epoch": 1.0279549936202297,
+      "grad_norm": 8.712311921794157,
+      "learning_rate": 2e-06,
+      "loss": 0.1466,
+      "step": 4431
+    },
+    {
+      "epoch": 1.0281869852685304,
+      "grad_norm": 16.378555006915374,
+      "learning_rate": 2e-06,
+      "loss": 0.3026,
+      "step": 4432
+    },
+    {
+      "epoch": 1.028418976916831,
+      "grad_norm": 16.769818815480654,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 4433
+    },
+    {
+      "epoch": 1.0286509685651317,
+      "grad_norm": 7.228007866660501,
+      "learning_rate": 2e-06,
+      "loss": 0.2106,
+      "step": 4434
+    },
+    {
+      "epoch": 1.0288829602134324,
+      "grad_norm": 8.775747363226587,
+      "learning_rate": 2e-06,
+      "loss": 0.2117,
+      "step": 4435
+    },
+    {
+      "epoch": 1.029114951861733,
+      "grad_norm": 9.435564557301497,
+      "learning_rate": 2e-06,
+      "loss": 0.2511,
+      "step": 4436
+    },
+    {
+      "epoch": 1.0293469435100335,
+      "grad_norm": 9.178260857291992,
+      "learning_rate": 2e-06,
+      "loss": 0.2053,
+      "step": 4437
+    },
+    {
+      "epoch": 1.0295789351583342,
+      "grad_norm": 15.10714105935708,
+      "learning_rate": 2e-06,
+      "loss": 0.312,
+      "step": 4438
+    },
+    {
+      "epoch": 1.0298109268066349,
+      "grad_norm": 19.037191167720213,
+      "learning_rate": 2e-06,
+      "loss": 0.2424,
+      "step": 4439
+    },
+    {
+      "epoch": 1.0300429184549356,
+      "grad_norm": 13.052108536860008,
+      "learning_rate": 2e-06,
+      "loss": 0.2326,
+      "step": 4440
+    },
+    {
+      "epoch": 1.0302749101032362,
+      "grad_norm": 7.880553741749966,
+      "learning_rate": 2e-06,
+      "loss": 0.1743,
+      "step": 4441
+    },
+    {
+      "epoch": 1.030506901751537,
+      "grad_norm": 18.99204868125294,
+      "learning_rate": 2e-06,
+      "loss": 0.2647,
+      "step": 4442
+    },
+    {
+      "epoch": 1.0307388933998376,
+      "grad_norm": 41.273427411034746,
+      "learning_rate": 2e-06,
+      "loss": 0.2691,
+      "step": 4443
+    },
+    {
+      "epoch": 1.0309708850481383,
+      "grad_norm": 11.052697660150518,
+      "learning_rate": 2e-06,
+      "loss": 0.1936,
+      "step": 4444
+    },
+    {
+      "epoch": 1.031202876696439,
+      "grad_norm": 13.15387981143392,
+      "learning_rate": 2e-06,
+      "loss": 0.3272,
+      "step": 4445
+    },
+    {
+      "epoch": 1.0314348683447396,
+      "grad_norm": 16.851744488684126,
+      "learning_rate": 2e-06,
+      "loss": 0.3239,
+      "step": 4446
+    },
+    {
+      "epoch": 1.0316668599930403,
+      "grad_norm": 24.65220325923243,
+      "learning_rate": 2e-06,
+      "loss": 0.3975,
+      "step": 4447
+    },
+    {
+      "epoch": 1.031898851641341,
+      "grad_norm": 10.587443256446537,
+      "learning_rate": 2e-06,
+      "loss": 0.2074,
+      "step": 4448
+    },
+    {
+      "epoch": 1.0321308432896417,
+      "grad_norm": 9.47223283101745,
+      "learning_rate": 2e-06,
+      "loss": 0.1383,
+      "step": 4449
+    },
+    {
+      "epoch": 1.0323628349379423,
+      "grad_norm": 19.69363174664466,
+      "learning_rate": 2e-06,
+      "loss": 0.348,
+      "step": 4450
+    },
+    {
+      "epoch": 1.0325948265862428,
+      "grad_norm": 17.418878994857508,
+      "learning_rate": 2e-06,
+      "loss": 0.3025,
+      "step": 4451
+    },
+    {
+      "epoch": 1.0328268182345435,
+      "grad_norm": 17.068050882970642,
+      "learning_rate": 2e-06,
+      "loss": 0.1959,
+      "step": 4452
+    },
+    {
+      "epoch": 1.0330588098828442,
+      "grad_norm": 14.348268274620187,
+      "learning_rate": 2e-06,
+      "loss": 0.2605,
+      "step": 4453
+    },
+    {
+      "epoch": 1.0332908015311448,
+      "grad_norm": 9.200927968182237,
+      "learning_rate": 2e-06,
+      "loss": 0.3117,
+      "step": 4454
+    },
+    {
+      "epoch": 1.0335227931794455,
+      "grad_norm": 11.880208685210507,
+      "learning_rate": 2e-06,
+      "loss": 0.3102,
+      "step": 4455
+    },
+    {
+      "epoch": 1.0337547848277462,
+      "grad_norm": 13.195463192921046,
+      "learning_rate": 2e-06,
+      "loss": 0.3442,
+      "step": 4456
+    },
+    {
+      "epoch": 1.0339867764760469,
+      "grad_norm": 16.39948374883616,
+      "learning_rate": 2e-06,
+      "loss": 0.2593,
+      "step": 4457
+    },
+    {
+      "epoch": 1.0342187681243475,
+      "grad_norm": 21.695075440702873,
+      "learning_rate": 2e-06,
+      "loss": 0.2773,
+      "step": 4458
+    },
+    {
+      "epoch": 1.0344507597726482,
+      "grad_norm": 23.236654747422598,
+      "learning_rate": 2e-06,
+      "loss": 0.3134,
+      "step": 4459
+    },
+    {
+      "epoch": 1.034682751420949,
+      "grad_norm": 12.113159851456805,
+      "learning_rate": 2e-06,
+      "loss": 0.2446,
+      "step": 4460
+    },
+    {
+      "epoch": 1.0349147430692496,
+      "grad_norm": 16.68370613230402,
+      "learning_rate": 2e-06,
+      "loss": 0.2297,
+      "step": 4461
+    },
+    {
+      "epoch": 1.0351467347175503,
+      "grad_norm": 14.844254905487963,
+      "learning_rate": 2e-06,
+      "loss": 0.2473,
+      "step": 4462
+    },
+    {
+      "epoch": 1.035378726365851,
+      "grad_norm": 20.220611429267375,
+      "learning_rate": 2e-06,
+      "loss": 0.1801,
+      "step": 4463
+    },
+    {
+      "epoch": 1.0356107180141514,
+      "grad_norm": 14.251304825494103,
+      "learning_rate": 2e-06,
+      "loss": 0.263,
+      "step": 4464
+    },
+    {
+      "epoch": 1.035842709662452,
+      "grad_norm": 22.99719982100096,
+      "learning_rate": 2e-06,
+      "loss": 0.2855,
+      "step": 4465
+    },
+    {
+      "epoch": 1.0360747013107527,
+      "grad_norm": 8.585584935855628,
+      "learning_rate": 2e-06,
+      "loss": 0.2178,
+      "step": 4466
+    },
+    {
+      "epoch": 1.0363066929590534,
+      "grad_norm": 18.387469782533767,
+      "learning_rate": 2e-06,
+      "loss": 0.21,
+      "step": 4467
+    },
+    {
+      "epoch": 1.036538684607354,
+      "grad_norm": 11.401769727716564,
+      "learning_rate": 2e-06,
+      "loss": 0.2416,
+      "step": 4468
+    },
+    {
+      "epoch": 1.0367706762556548,
+      "grad_norm": 15.94523695286945,
+      "learning_rate": 2e-06,
+      "loss": 0.311,
+      "step": 4469
+    },
+    {
+      "epoch": 1.0370026679039555,
+      "grad_norm": 11.196828766891255,
+      "learning_rate": 2e-06,
+      "loss": 0.2191,
+      "step": 4470
+    },
+    {
+      "epoch": 1.0372346595522561,
+      "grad_norm": 11.540823035696267,
+      "learning_rate": 2e-06,
+      "loss": 0.1927,
+      "step": 4471
+    },
+    {
+      "epoch": 1.0374666512005568,
+      "grad_norm": 14.536796413435079,
+      "learning_rate": 2e-06,
+      "loss": 0.3304,
+      "step": 4472
+    },
+    {
+      "epoch": 1.0376986428488575,
+      "grad_norm": 14.343576899291822,
+      "learning_rate": 2e-06,
+      "loss": 0.2313,
+      "step": 4473
+    },
+    {
+      "epoch": 1.0379306344971582,
+      "grad_norm": 16.922838282153773,
+      "learning_rate": 2e-06,
+      "loss": 0.3408,
+      "step": 4474
+    },
+    {
+      "epoch": 1.0381626261454588,
+      "grad_norm": 11.707108143377337,
+      "learning_rate": 2e-06,
+      "loss": 0.2202,
+      "step": 4475
+    },
+    {
+      "epoch": 1.0383946177937595,
+      "grad_norm": 8.517094184401891,
+      "learning_rate": 2e-06,
+      "loss": 0.2071,
+      "step": 4476
+    },
+    {
+      "epoch": 1.0386266094420602,
+      "grad_norm": 12.389320998949996,
+      "learning_rate": 2e-06,
+      "loss": 0.2455,
+      "step": 4477
+    },
+    {
+      "epoch": 1.0388586010903607,
+      "grad_norm": 13.645512621397236,
+      "learning_rate": 2e-06,
+      "loss": 0.2399,
+      "step": 4478
+    },
+    {
+      "epoch": 1.0390905927386613,
+      "grad_norm": 20.161252864133314,
+      "learning_rate": 2e-06,
+      "loss": 0.2882,
+      "step": 4479
+    },
+    {
+      "epoch": 1.039322584386962,
+      "grad_norm": 19.924490724699734,
+      "learning_rate": 2e-06,
+      "loss": 0.2941,
+      "step": 4480
+    },
+    {
+      "epoch": 1.0395545760352627,
+      "grad_norm": 17.642672690575644,
+      "learning_rate": 2e-06,
+      "loss": 0.2028,
+      "step": 4481
+    },
+    {
+      "epoch": 1.0397865676835634,
+      "grad_norm": 11.183339327538803,
+      "learning_rate": 2e-06,
+      "loss": 0.216,
+      "step": 4482
+    },
+    {
+      "epoch": 1.040018559331864,
+      "grad_norm": 16.449886451041284,
+      "learning_rate": 2e-06,
+      "loss": 0.2215,
+      "step": 4483
+    },
+    {
+      "epoch": 1.0402505509801647,
+      "grad_norm": 23.901687440717772,
+      "learning_rate": 2e-06,
+      "loss": 0.3176,
+      "step": 4484
+    },
+    {
+      "epoch": 1.0404825426284654,
+      "grad_norm": 10.718689618673862,
+      "learning_rate": 2e-06,
+      "loss": 0.2306,
+      "step": 4485
+    },
+    {
+      "epoch": 1.040714534276766,
+      "grad_norm": 13.623712550846152,
+      "learning_rate": 2e-06,
+      "loss": 0.2108,
+      "step": 4486
+    },
+    {
+      "epoch": 1.0409465259250668,
+      "grad_norm": 16.330073437541824,
+      "learning_rate": 2e-06,
+      "loss": 0.3593,
+      "step": 4487
+    },
+    {
+      "epoch": 1.0411785175733674,
+      "grad_norm": 12.825110261820797,
+      "learning_rate": 2e-06,
+      "loss": 0.2464,
+      "step": 4488
+    },
+    {
+      "epoch": 1.0414105092216681,
+      "grad_norm": 7.216701052005836,
+      "learning_rate": 2e-06,
+      "loss": 0.1575,
+      "step": 4489
+    },
+    {
+      "epoch": 1.0416425008699686,
+      "grad_norm": 22.15548889662668,
+      "learning_rate": 2e-06,
+      "loss": 0.1809,
+      "step": 4490
+    },
+    {
+      "epoch": 1.0418744925182692,
+      "grad_norm": 9.994987316571109,
+      "learning_rate": 2e-06,
+      "loss": 0.1943,
+      "step": 4491
+    },
+    {
+      "epoch": 1.04210648416657,
+      "grad_norm": 8.941695765183466,
+      "learning_rate": 2e-06,
+      "loss": 0.2319,
+      "step": 4492
+    },
+    {
+      "epoch": 1.0423384758148706,
+      "grad_norm": 15.189764942546732,
+      "learning_rate": 2e-06,
+      "loss": 0.2488,
+      "step": 4493
+    },
+    {
+      "epoch": 1.0425704674631713,
+      "grad_norm": 13.35000274072905,
+      "learning_rate": 2e-06,
+      "loss": 0.2118,
+      "step": 4494
+    },
+    {
+      "epoch": 1.042802459111472,
+      "grad_norm": 13.057315814446655,
+      "learning_rate": 2e-06,
+      "loss": 0.2257,
+      "step": 4495
+    },
+    {
+      "epoch": 1.0430344507597726,
+      "grad_norm": 12.011177476914142,
+      "learning_rate": 2e-06,
+      "loss": 0.2612,
+      "step": 4496
+    },
+    {
+      "epoch": 1.0432664424080733,
+      "grad_norm": 18.472444083720685,
+      "learning_rate": 2e-06,
+      "loss": 0.2278,
+      "step": 4497
+    },
+    {
+      "epoch": 1.043498434056374,
+      "grad_norm": 11.947161039265346,
+      "learning_rate": 2e-06,
+      "loss": 0.2028,
+      "step": 4498
+    },
+    {
+      "epoch": 1.0437304257046747,
+      "grad_norm": 11.683468527357304,
+      "learning_rate": 2e-06,
+      "loss": 0.2638,
+      "step": 4499
+    },
+    {
+      "epoch": 1.0439624173529753,
+      "grad_norm": 14.343236326278143,
+      "learning_rate": 2e-06,
+      "loss": 0.302,
+      "step": 4500
+    },
+    {
+      "epoch": 1.044194409001276,
+      "grad_norm": 10.710904819129617,
+      "learning_rate": 2e-06,
+      "loss": 0.2985,
+      "step": 4501
+    },
+    {
+      "epoch": 1.0444264006495767,
+      "grad_norm": 12.13213015738718,
+      "learning_rate": 2e-06,
+      "loss": 0.242,
+      "step": 4502
+    },
+    {
+      "epoch": 1.0446583922978774,
+      "grad_norm": 18.94819133437833,
+      "learning_rate": 2e-06,
+      "loss": 0.3138,
+      "step": 4503
+    },
+    {
+      "epoch": 1.0448903839461778,
+      "grad_norm": 12.539901478278948,
+      "learning_rate": 2e-06,
+      "loss": 0.2392,
+      "step": 4504
+    },
+    {
+      "epoch": 1.0451223755944785,
+      "grad_norm": 24.18437820339448,
+      "learning_rate": 2e-06,
+      "loss": 0.2755,
+      "step": 4505
+    },
+    {
+      "epoch": 1.0453543672427792,
+      "grad_norm": 23.383318703382532,
+      "learning_rate": 2e-06,
+      "loss": 0.3919,
+      "step": 4506
+    },
+    {
+      "epoch": 1.0455863588910799,
+      "grad_norm": 9.67869670349899,
+      "learning_rate": 2e-06,
+      "loss": 0.2123,
+      "step": 4507
+    },
+    {
+      "epoch": 1.0458183505393805,
+      "grad_norm": 14.588716396765122,
+      "learning_rate": 2e-06,
+      "loss": 0.2996,
+      "step": 4508
+    },
+    {
+      "epoch": 1.0460503421876812,
+      "grad_norm": 12.840467825576411,
+      "learning_rate": 2e-06,
+      "loss": 0.2757,
+      "step": 4509
+    },
+    {
+      "epoch": 1.046282333835982,
+      "grad_norm": 14.272494811690157,
+      "learning_rate": 2e-06,
+      "loss": 0.2401,
+      "step": 4510
+    },
+    {
+      "epoch": 1.0465143254842826,
+      "grad_norm": 20.01918249140347,
+      "learning_rate": 2e-06,
+      "loss": 0.3383,
+      "step": 4511
+    },
+    {
+      "epoch": 1.0467463171325833,
+      "grad_norm": 7.556423609497603,
+      "learning_rate": 2e-06,
+      "loss": 0.1243,
+      "step": 4512
+    },
+    {
+      "epoch": 1.046978308780884,
+      "grad_norm": 8.6970267586175,
+      "learning_rate": 2e-06,
+      "loss": 0.1976,
+      "step": 4513
+    },
+    {
+      "epoch": 1.0472103004291846,
+      "grad_norm": 24.096318328090813,
+      "learning_rate": 2e-06,
+      "loss": 0.2707,
+      "step": 4514
+    },
+    {
+      "epoch": 1.0474422920774853,
+      "grad_norm": 7.8467329542157565,
+      "learning_rate": 2e-06,
+      "loss": 0.1836,
+      "step": 4515
+    },
+    {
+      "epoch": 1.047674283725786,
+      "grad_norm": 29.928182855435207,
+      "learning_rate": 2e-06,
+      "loss": 0.3649,
+      "step": 4516
+    },
+    {
+      "epoch": 1.0479062753740864,
+      "grad_norm": 13.927439113839654,
+      "learning_rate": 2e-06,
+      "loss": 0.3017,
+      "step": 4517
+    },
+    {
+      "epoch": 1.048138267022387,
+      "grad_norm": 18.88795094602496,
+      "learning_rate": 2e-06,
+      "loss": 0.225,
+      "step": 4518
+    },
+    {
+      "epoch": 1.0483702586706878,
+      "grad_norm": 9.620669829192538,
+      "learning_rate": 2e-06,
+      "loss": 0.2321,
+      "step": 4519
+    },
+    {
+      "epoch": 1.0486022503189885,
+      "grad_norm": 17.223106187454317,
+      "learning_rate": 2e-06,
+      "loss": 0.2882,
+      "step": 4520
+    },
+    {
+      "epoch": 1.0488342419672891,
+      "grad_norm": 8.646155394897107,
+      "learning_rate": 2e-06,
+      "loss": 0.3076,
+      "step": 4521
+    },
+    {
+      "epoch": 1.0490662336155898,
+      "grad_norm": 8.778649493758728,
+      "learning_rate": 2e-06,
+      "loss": 0.2392,
+      "step": 4522
+    },
+    {
+      "epoch": 1.0492982252638905,
+      "grad_norm": 7.918122268225447,
+      "learning_rate": 2e-06,
+      "loss": 0.1816,
+      "step": 4523
+    },
+    {
+      "epoch": 1.0495302169121912,
+      "grad_norm": 11.01843669150523,
+      "learning_rate": 2e-06,
+      "loss": 0.2619,
+      "step": 4524
+    },
+    {
+      "epoch": 1.0497622085604918,
+      "grad_norm": 13.875284633960788,
+      "learning_rate": 2e-06,
+      "loss": 0.2297,
+      "step": 4525
+    },
+    {
+      "epoch": 1.0499942002087925,
+      "grad_norm": 20.84523992516674,
+      "learning_rate": 2e-06,
+      "loss": 0.2333,
+      "step": 4526
+    },
+    {
+      "epoch": 1.0502261918570932,
+      "grad_norm": 7.755997981352427,
+      "learning_rate": 2e-06,
+      "loss": 0.1873,
+      "step": 4527
+    },
+    {
+      "epoch": 1.0504581835053939,
+      "grad_norm": 9.333137819737173,
+      "learning_rate": 2e-06,
+      "loss": 0.2753,
+      "step": 4528
+    },
+    {
+      "epoch": 1.0506901751536946,
+      "grad_norm": 39.160017340477,
+      "learning_rate": 2e-06,
+      "loss": 0.2399,
+      "step": 4529
+    },
+    {
+      "epoch": 1.0509221668019952,
+      "grad_norm": 25.282757342372953,
+      "learning_rate": 2e-06,
+      "loss": 0.2774,
+      "step": 4530
+    },
+    {
+      "epoch": 1.0511541584502957,
+      "grad_norm": 12.916061156883767,
+      "learning_rate": 2e-06,
+      "loss": 0.1881,
+      "step": 4531
+    },
+    {
+      "epoch": 1.0513861500985964,
+      "grad_norm": 17.357923519315666,
+      "learning_rate": 2e-06,
+      "loss": 0.3112,
+      "step": 4532
+    },
+    {
+      "epoch": 1.051618141746897,
+      "grad_norm": 8.057295108120414,
+      "learning_rate": 2e-06,
+      "loss": 0.1589,
+      "step": 4533
+    },
+    {
+      "epoch": 1.0518501333951977,
+      "grad_norm": 11.473880621998514,
+      "learning_rate": 2e-06,
+      "loss": 0.1511,
+      "step": 4534
+    },
+    {
+      "epoch": 1.0520821250434984,
+      "grad_norm": 7.7417751611696835,
+      "learning_rate": 2e-06,
+      "loss": 0.1299,
+      "step": 4535
+    },
+    {
+      "epoch": 1.052314116691799,
+      "grad_norm": 15.442864167699556,
+      "learning_rate": 2e-06,
+      "loss": 0.3142,
+      "step": 4536
+    },
+    {
+      "epoch": 1.0525461083400998,
+      "grad_norm": 17.60484150290361,
+      "learning_rate": 2e-06,
+      "loss": 0.2372,
+      "step": 4537
+    },
+    {
+      "epoch": 1.0527780999884004,
+      "grad_norm": 16.936105884989708,
+      "learning_rate": 2e-06,
+      "loss": 0.289,
+      "step": 4538
+    },
+    {
+      "epoch": 1.0530100916367011,
+      "grad_norm": 11.843293598367202,
+      "learning_rate": 2e-06,
+      "loss": 0.2139,
+      "step": 4539
+    },
+    {
+      "epoch": 1.0532420832850018,
+      "grad_norm": 23.64743039681951,
+      "learning_rate": 2e-06,
+      "loss": 0.3383,
+      "step": 4540
+    },
+    {
+      "epoch": 1.0534740749333025,
+      "grad_norm": 17.546494934902448,
+      "learning_rate": 2e-06,
+      "loss": 0.2541,
+      "step": 4541
+    },
+    {
+      "epoch": 1.0537060665816032,
+      "grad_norm": 15.785905089669802,
+      "learning_rate": 2e-06,
+      "loss": 0.2919,
+      "step": 4542
+    },
+    {
+      "epoch": 1.0539380582299038,
+      "grad_norm": 15.552562721537294,
+      "learning_rate": 2e-06,
+      "loss": 0.313,
+      "step": 4543
+    },
+    {
+      "epoch": 1.0541700498782043,
+      "grad_norm": 13.38274443962515,
+      "learning_rate": 2e-06,
+      "loss": 0.2333,
+      "step": 4544
+    },
+    {
+      "epoch": 1.054402041526505,
+      "grad_norm": 14.364837287011094,
+      "learning_rate": 2e-06,
+      "loss": 0.3176,
+      "step": 4545
+    },
+    {
+      "epoch": 1.0546340331748056,
+      "grad_norm": 10.675992961826456,
+      "learning_rate": 2e-06,
+      "loss": 0.1771,
+      "step": 4546
+    },
+    {
+      "epoch": 1.0548660248231063,
+      "grad_norm": 25.91576359257879,
+      "learning_rate": 2e-06,
+      "loss": 0.2628,
+      "step": 4547
+    },
+    {
+      "epoch": 1.055098016471407,
+      "grad_norm": 10.117754686145902,
+      "learning_rate": 2e-06,
+      "loss": 0.2334,
+      "step": 4548
+    },
+    {
+      "epoch": 1.0553300081197077,
+      "grad_norm": 10.059227091582052,
+      "learning_rate": 2e-06,
+      "loss": 0.1506,
+      "step": 4549
+    },
+    {
+      "epoch": 1.0555619997680084,
+      "grad_norm": 16.908727808786647,
+      "learning_rate": 2e-06,
+      "loss": 0.2702,
+      "step": 4550
+    },
+    {
+      "epoch": 1.055793991416309,
+      "grad_norm": 14.386589699765908,
+      "learning_rate": 2e-06,
+      "loss": 0.2534,
+      "step": 4551
+    },
+    {
+      "epoch": 1.0560259830646097,
+      "grad_norm": 10.248441451994353,
+      "learning_rate": 2e-06,
+      "loss": 0.2022,
+      "step": 4552
+    },
+    {
+      "epoch": 1.0562579747129104,
+      "grad_norm": 7.8765393725949,
+      "learning_rate": 2e-06,
+      "loss": 0.1391,
+      "step": 4553
+    },
+    {
+      "epoch": 1.056489966361211,
+      "grad_norm": 13.745768749508281,
+      "learning_rate": 2e-06,
+      "loss": 0.3445,
+      "step": 4554
+    },
+    {
+      "epoch": 1.0567219580095117,
+      "grad_norm": 20.73572027129585,
+      "learning_rate": 2e-06,
+      "loss": 0.2545,
+      "step": 4555
+    },
+    {
+      "epoch": 1.0569539496578124,
+      "grad_norm": 19.14940776893297,
+      "learning_rate": 2e-06,
+      "loss": 0.239,
+      "step": 4556
+    },
+    {
+      "epoch": 1.0571859413061129,
+      "grad_norm": 12.890430025290328,
+      "learning_rate": 2e-06,
+      "loss": 0.29,
+      "step": 4557
+    },
+    {
+      "epoch": 1.0574179329544136,
+      "grad_norm": 18.20419221939733,
+      "learning_rate": 2e-06,
+      "loss": 0.3279,
+      "step": 4558
+    },
+    {
+      "epoch": 1.0576499246027142,
+      "grad_norm": 12.980910847550414,
+      "learning_rate": 2e-06,
+      "loss": 0.2111,
+      "step": 4559
+    },
+    {
+      "epoch": 1.057881916251015,
+      "grad_norm": 8.458796386653953,
+      "learning_rate": 2e-06,
+      "loss": 0.1938,
+      "step": 4560
+    },
+    {
+      "epoch": 1.0581139078993156,
+      "grad_norm": 21.75891206290162,
+      "learning_rate": 2e-06,
+      "loss": 0.2288,
+      "step": 4561
+    },
+    {
+      "epoch": 1.0583458995476163,
+      "grad_norm": 13.117360758513612,
+      "learning_rate": 2e-06,
+      "loss": 0.2496,
+      "step": 4562
+    },
+    {
+      "epoch": 1.058577891195917,
+      "grad_norm": 18.092807169325763,
+      "learning_rate": 2e-06,
+      "loss": 0.2883,
+      "step": 4563
+    },
+    {
+      "epoch": 1.0588098828442176,
+      "grad_norm": 28.11071353474641,
+      "learning_rate": 2e-06,
+      "loss": 0.2404,
+      "step": 4564
+    },
+    {
+      "epoch": 1.0590418744925183,
+      "grad_norm": 11.093359793460808,
+      "learning_rate": 2e-06,
+      "loss": 0.2871,
+      "step": 4565
+    },
+    {
+      "epoch": 1.059273866140819,
+      "grad_norm": 5.708558446753598,
+      "learning_rate": 2e-06,
+      "loss": 0.1646,
+      "step": 4566
+    },
+    {
+      "epoch": 1.0595058577891197,
+      "grad_norm": 8.007983606895225,
+      "learning_rate": 2e-06,
+      "loss": 0.1235,
+      "step": 4567
+    },
+    {
+      "epoch": 1.0597378494374203,
+      "grad_norm": 23.691691410023708,
+      "learning_rate": 2e-06,
+      "loss": 0.4476,
+      "step": 4568
+    },
+    {
+      "epoch": 1.059969841085721,
+      "grad_norm": 14.342884710913854,
+      "learning_rate": 2e-06,
+      "loss": 0.2121,
+      "step": 4569
+    },
+    {
+      "epoch": 1.0602018327340215,
+      "grad_norm": 20.421082262288795,
+      "learning_rate": 2e-06,
+      "loss": 0.2337,
+      "step": 4570
+    },
+    {
+      "epoch": 1.0604338243823221,
+      "grad_norm": 14.283943578587204,
+      "learning_rate": 2e-06,
+      "loss": 0.3167,
+      "step": 4571
+    },
+    {
+      "epoch": 1.0606658160306228,
+      "grad_norm": 14.32307432911022,
+      "learning_rate": 2e-06,
+      "loss": 0.3219,
+      "step": 4572
+    },
+    {
+      "epoch": 1.0608978076789235,
+      "grad_norm": 23.713879420441952,
+      "learning_rate": 2e-06,
+      "loss": 0.42,
+      "step": 4573
+    },
+    {
+      "epoch": 1.0611297993272242,
+      "grad_norm": 14.388879223470608,
+      "learning_rate": 2e-06,
+      "loss": 0.2093,
+      "step": 4574
+    },
+    {
+      "epoch": 1.0613617909755249,
+      "grad_norm": 14.770522362129077,
+      "learning_rate": 2e-06,
+      "loss": 0.2532,
+      "step": 4575
+    },
+    {
+      "epoch": 1.0615937826238255,
+      "grad_norm": 9.670132297957418,
+      "learning_rate": 2e-06,
+      "loss": 0.2032,
+      "step": 4576
+    },
+    {
+      "epoch": 1.0618257742721262,
+      "grad_norm": 12.174822545988004,
+      "learning_rate": 2e-06,
+      "loss": 0.1677,
+      "step": 4577
+    },
+    {
+      "epoch": 1.0620577659204269,
+      "grad_norm": 16.409913639531826,
+      "learning_rate": 2e-06,
+      "loss": 0.305,
+      "step": 4578
+    },
+    {
+      "epoch": 1.0622897575687276,
+      "grad_norm": 13.679220521114875,
+      "learning_rate": 2e-06,
+      "loss": 0.3606,
+      "step": 4579
+    },
+    {
+      "epoch": 1.0625217492170282,
+      "grad_norm": 15.351767353163767,
+      "learning_rate": 2e-06,
+      "loss": 0.2492,
+      "step": 4580
+    },
+    {
+      "epoch": 1.062753740865329,
+      "grad_norm": 14.55264537341593,
+      "learning_rate": 2e-06,
+      "loss": 0.194,
+      "step": 4581
+    },
+    {
+      "epoch": 1.0629857325136296,
+      "grad_norm": 23.70433697778693,
+      "learning_rate": 2e-06,
+      "loss": 0.2432,
+      "step": 4582
+    },
+    {
+      "epoch": 1.0632177241619303,
+      "grad_norm": 13.494892360681849,
+      "learning_rate": 2e-06,
+      "loss": 0.2695,
+      "step": 4583
+    },
+    {
+      "epoch": 1.0634497158102307,
+      "grad_norm": 14.626691010991978,
+      "learning_rate": 2e-06,
+      "loss": 0.2345,
+      "step": 4584
+    },
+    {
+      "epoch": 1.0636817074585314,
+      "grad_norm": 13.807386499526016,
+      "learning_rate": 2e-06,
+      "loss": 0.1773,
+      "step": 4585
+    },
+    {
+      "epoch": 1.063913699106832,
+      "grad_norm": 14.377411897301746,
+      "learning_rate": 2e-06,
+      "loss": 0.206,
+      "step": 4586
+    },
+    {
+      "epoch": 1.0641456907551328,
+      "grad_norm": 10.717646496233396,
+      "learning_rate": 2e-06,
+      "loss": 0.1543,
+      "step": 4587
+    },
+    {
+      "epoch": 1.0643776824034334,
+      "grad_norm": 12.121576229930133,
+      "learning_rate": 2e-06,
+      "loss": 0.1937,
+      "step": 4588
+    },
+    {
+      "epoch": 1.0646096740517341,
+      "grad_norm": 10.377540319043618,
+      "learning_rate": 2e-06,
+      "loss": 0.2044,
+      "step": 4589
+    },
+    {
+      "epoch": 1.0648416657000348,
+      "grad_norm": 16.025433031691097,
+      "learning_rate": 2e-06,
+      "loss": 0.2279,
+      "step": 4590
+    },
+    {
+      "epoch": 1.0650736573483355,
+      "grad_norm": 28.09748111844974,
+      "learning_rate": 2e-06,
+      "loss": 0.271,
+      "step": 4591
+    },
+    {
+      "epoch": 1.0653056489966362,
+      "grad_norm": 15.233408719058405,
+      "learning_rate": 2e-06,
+      "loss": 0.2557,
+      "step": 4592
+    },
+    {
+      "epoch": 1.0655376406449368,
+      "grad_norm": 11.688531867610594,
+      "learning_rate": 2e-06,
+      "loss": 0.2213,
+      "step": 4593
+    },
+    {
+      "epoch": 1.0657696322932375,
+      "grad_norm": 12.478963044020029,
+      "learning_rate": 2e-06,
+      "loss": 0.3794,
+      "step": 4594
+    },
+    {
+      "epoch": 1.0660016239415382,
+      "grad_norm": 12.255769188520668,
+      "learning_rate": 2e-06,
+      "loss": 0.1958,
+      "step": 4595
+    },
+    {
+      "epoch": 1.0662336155898389,
+      "grad_norm": 14.479545376110917,
+      "learning_rate": 2e-06,
+      "loss": 0.2721,
+      "step": 4596
+    },
+    {
+      "epoch": 1.0664656072381393,
+      "grad_norm": 9.375689405742289,
+      "learning_rate": 2e-06,
+      "loss": 0.2932,
+      "step": 4597
+    },
+    {
+      "epoch": 1.06669759888644,
+      "grad_norm": 10.315314477724502,
+      "learning_rate": 2e-06,
+      "loss": 0.2008,
+      "step": 4598
+    },
+    {
+      "epoch": 1.0669295905347407,
+      "grad_norm": 21.631875638569465,
+      "learning_rate": 2e-06,
+      "loss": 0.3366,
+      "step": 4599
+    },
+    {
+      "epoch": 1.0671615821830414,
+      "grad_norm": 27.067384794430104,
+      "learning_rate": 2e-06,
+      "loss": 0.4603,
+      "step": 4600
+    },
+    {
+      "epoch": 1.067393573831342,
+      "grad_norm": 9.802424258760192,
+      "learning_rate": 2e-06,
+      "loss": 0.1782,
+      "step": 4601
+    },
+    {
+      "epoch": 1.0676255654796427,
+      "grad_norm": 10.320769127337003,
+      "learning_rate": 2e-06,
+      "loss": 0.1942,
+      "step": 4602
+    },
+    {
+      "epoch": 1.0678575571279434,
+      "grad_norm": 12.10330234840912,
+      "learning_rate": 2e-06,
+      "loss": 0.2389,
+      "step": 4603
+    },
+    {
+      "epoch": 1.068089548776244,
+      "grad_norm": 9.482829401573873,
+      "learning_rate": 2e-06,
+      "loss": 0.2014,
+      "step": 4604
+    },
+    {
+      "epoch": 1.0683215404245447,
+      "grad_norm": 13.2444842153338,
+      "learning_rate": 2e-06,
+      "loss": 0.275,
+      "step": 4605
+    },
+    {
+      "epoch": 1.0685535320728454,
+      "grad_norm": 9.059155667239775,
+      "learning_rate": 2e-06,
+      "loss": 0.1656,
+      "step": 4606
+    },
+    {
+      "epoch": 1.068785523721146,
+      "grad_norm": 22.107718337840467,
+      "learning_rate": 2e-06,
+      "loss": 0.3723,
+      "step": 4607
+    },
+    {
+      "epoch": 1.0690175153694468,
+      "grad_norm": 16.721630962015563,
+      "learning_rate": 2e-06,
+      "loss": 0.2691,
+      "step": 4608
+    },
+    {
+      "epoch": 1.0692495070177475,
+      "grad_norm": 14.552520243084661,
+      "learning_rate": 2e-06,
+      "loss": 0.2485,
+      "step": 4609
+    },
+    {
+      "epoch": 1.0694814986660481,
+      "grad_norm": 16.363779480701567,
+      "learning_rate": 2e-06,
+      "loss": 0.2091,
+      "step": 4610
+    },
+    {
+      "epoch": 1.0697134903143486,
+      "grad_norm": 6.404394903208128,
+      "learning_rate": 2e-06,
+      "loss": 0.2208,
+      "step": 4611
+    },
+    {
+      "epoch": 1.0699454819626493,
+      "grad_norm": 20.95385822012562,
+      "learning_rate": 2e-06,
+      "loss": 0.2749,
+      "step": 4612
+    },
+    {
+      "epoch": 1.07017747361095,
+      "grad_norm": 8.509211364750067,
+      "learning_rate": 2e-06,
+      "loss": 0.1986,
+      "step": 4613
+    },
+    {
+      "epoch": 1.0704094652592506,
+      "grad_norm": 9.683183528937615,
+      "learning_rate": 2e-06,
+      "loss": 0.1457,
+      "step": 4614
+    },
+    {
+      "epoch": 1.0706414569075513,
+      "grad_norm": 9.248345401803073,
+      "learning_rate": 2e-06,
+      "loss": 0.2679,
+      "step": 4615
+    },
+    {
+      "epoch": 1.070873448555852,
+      "grad_norm": 10.304306796998121,
+      "learning_rate": 2e-06,
+      "loss": 0.2017,
+      "step": 4616
+    },
+    {
+      "epoch": 1.0711054402041527,
+      "grad_norm": 12.830372489140107,
+      "learning_rate": 2e-06,
+      "loss": 0.2048,
+      "step": 4617
+    },
+    {
+      "epoch": 1.0713374318524533,
+      "grad_norm": 15.16407227138631,
+      "learning_rate": 2e-06,
+      "loss": 0.3194,
+      "step": 4618
+    },
+    {
+      "epoch": 1.071569423500754,
+      "grad_norm": 19.579768580627913,
+      "learning_rate": 2e-06,
+      "loss": 0.2052,
+      "step": 4619
+    },
+    {
+      "epoch": 1.0718014151490547,
+      "grad_norm": 18.4708758331504,
+      "learning_rate": 2e-06,
+      "loss": 0.2765,
+      "step": 4620
+    },
+    {
+      "epoch": 1.0720334067973554,
+      "grad_norm": 16.340610589492915,
+      "learning_rate": 2e-06,
+      "loss": 0.1827,
+      "step": 4621
+    },
+    {
+      "epoch": 1.072265398445656,
+      "grad_norm": 15.476040559442902,
+      "learning_rate": 2e-06,
+      "loss": 0.3229,
+      "step": 4622
+    },
+    {
+      "epoch": 1.0724973900939565,
+      "grad_norm": 11.731130456848682,
+      "learning_rate": 2e-06,
+      "loss": 0.2831,
+      "step": 4623
+    },
+    {
+      "epoch": 1.0727293817422572,
+      "grad_norm": 14.988365226355235,
+      "learning_rate": 2e-06,
+      "loss": 0.3421,
+      "step": 4624
+    },
+    {
+      "epoch": 1.0729613733905579,
+      "grad_norm": 16.246876392051842,
+      "learning_rate": 2e-06,
+      "loss": 0.2171,
+      "step": 4625
+    },
+    {
+      "epoch": 1.0731933650388585,
+      "grad_norm": 16.63092189706823,
+      "learning_rate": 2e-06,
+      "loss": 0.2903,
+      "step": 4626
+    },
+    {
+      "epoch": 1.0734253566871592,
+      "grad_norm": 9.829098076638097,
+      "learning_rate": 2e-06,
+      "loss": 0.1791,
+      "step": 4627
+    },
+    {
+      "epoch": 1.07365734833546,
+      "grad_norm": 12.44972854207715,
+      "learning_rate": 2e-06,
+      "loss": 0.1847,
+      "step": 4628
+    },
+    {
+      "epoch": 1.0738893399837606,
+      "grad_norm": 15.653560086600205,
+      "learning_rate": 2e-06,
+      "loss": 0.2672,
+      "step": 4629
+    },
+    {
+      "epoch": 1.0741213316320612,
+      "grad_norm": 6.496762545985574,
+      "learning_rate": 2e-06,
+      "loss": 0.1861,
+      "step": 4630
+    },
+    {
+      "epoch": 1.074353323280362,
+      "grad_norm": 8.684805225175854,
+      "learning_rate": 2e-06,
+      "loss": 0.1961,
+      "step": 4631
+    },
+    {
+      "epoch": 1.0745853149286626,
+      "grad_norm": 13.618572813623743,
+      "learning_rate": 2e-06,
+      "loss": 0.2212,
+      "step": 4632
+    },
+    {
+      "epoch": 1.0748173065769633,
+      "grad_norm": 11.440261368799872,
+      "learning_rate": 2e-06,
+      "loss": 0.2728,
+      "step": 4633
+    },
+    {
+      "epoch": 1.075049298225264,
+      "grad_norm": 8.376117985507744,
+      "learning_rate": 2e-06,
+      "loss": 0.2127,
+      "step": 4634
+    },
+    {
+      "epoch": 1.0752812898735646,
+      "grad_norm": 10.788704052608766,
+      "learning_rate": 2e-06,
+      "loss": 0.2423,
+      "step": 4635
+    },
+    {
+      "epoch": 1.0755132815218653,
+      "grad_norm": 12.356605240160762,
+      "learning_rate": 2e-06,
+      "loss": 0.2149,
+      "step": 4636
+    },
+    {
+      "epoch": 1.0757452731701658,
+      "grad_norm": 8.646560020577828,
+      "learning_rate": 2e-06,
+      "loss": 0.1408,
+      "step": 4637
+    },
+    {
+      "epoch": 1.0759772648184665,
+      "grad_norm": 22.873044231235614,
+      "learning_rate": 2e-06,
+      "loss": 0.3564,
+      "step": 4638
+    },
+    {
+      "epoch": 1.0762092564667671,
+      "grad_norm": 13.115236584747734,
+      "learning_rate": 2e-06,
+      "loss": 0.327,
+      "step": 4639
+    },
+    {
+      "epoch": 1.0764412481150678,
+      "grad_norm": 16.990624415839985,
+      "learning_rate": 2e-06,
+      "loss": 0.2237,
+      "step": 4640
+    },
+    {
+      "epoch": 1.0766732397633685,
+      "grad_norm": 24.284509006768378,
+      "learning_rate": 2e-06,
+      "loss": 0.3058,
+      "step": 4641
+    },
+    {
+      "epoch": 1.0769052314116692,
+      "grad_norm": 13.264059901096275,
+      "learning_rate": 2e-06,
+      "loss": 0.2127,
+      "step": 4642
+    },
+    {
+      "epoch": 1.0771372230599698,
+      "grad_norm": 17.962152197713916,
+      "learning_rate": 2e-06,
+      "loss": 0.4182,
+      "step": 4643
+    },
+    {
+      "epoch": 1.0773692147082705,
+      "grad_norm": 15.13566581385218,
+      "learning_rate": 2e-06,
+      "loss": 0.3004,
+      "step": 4644
+    },
+    {
+      "epoch": 1.0776012063565712,
+      "grad_norm": 9.107776851142189,
+      "learning_rate": 2e-06,
+      "loss": 0.1799,
+      "step": 4645
+    },
+    {
+      "epoch": 1.0778331980048719,
+      "grad_norm": 15.416609074726267,
+      "learning_rate": 2e-06,
+      "loss": 0.2687,
+      "step": 4646
+    },
+    {
+      "epoch": 1.0780651896531726,
+      "grad_norm": 60.35641994256627,
+      "learning_rate": 2e-06,
+      "loss": 0.2461,
+      "step": 4647
+    },
+    {
+      "epoch": 1.0782971813014732,
+      "grad_norm": 11.96075666327412,
+      "learning_rate": 2e-06,
+      "loss": 0.1971,
+      "step": 4648
+    },
+    {
+      "epoch": 1.078529172949774,
+      "grad_norm": 13.424063532979817,
+      "learning_rate": 2e-06,
+      "loss": 0.2106,
+      "step": 4649
+    },
+    {
+      "epoch": 1.0787611645980744,
+      "grad_norm": 12.773540979848594,
+      "learning_rate": 2e-06,
+      "loss": 0.1968,
+      "step": 4650
+    },
+    {
+      "epoch": 1.078993156246375,
+      "grad_norm": 19.417735202497628,
+      "learning_rate": 2e-06,
+      "loss": 0.2292,
+      "step": 4651
+    },
+    {
+      "epoch": 1.0792251478946757,
+      "grad_norm": 6.308682768305041,
+      "learning_rate": 2e-06,
+      "loss": 0.1197,
+      "step": 4652
+    },
+    {
+      "epoch": 1.0794571395429764,
+      "grad_norm": 7.2959765587506205,
+      "learning_rate": 2e-06,
+      "loss": 0.2321,
+      "step": 4653
+    },
+    {
+      "epoch": 1.079689131191277,
+      "grad_norm": 15.44589618064587,
+      "learning_rate": 2e-06,
+      "loss": 0.2507,
+      "step": 4654
+    },
+    {
+      "epoch": 1.0799211228395778,
+      "grad_norm": 9.29934262830167,
+      "learning_rate": 2e-06,
+      "loss": 0.2287,
+      "step": 4655
+    },
+    {
+      "epoch": 1.0801531144878784,
+      "grad_norm": 15.286437259258648,
+      "learning_rate": 2e-06,
+      "loss": 0.3175,
+      "step": 4656
+    },
+    {
+      "epoch": 1.080385106136179,
+      "grad_norm": 13.27664930644032,
+      "learning_rate": 2e-06,
+      "loss": 0.2383,
+      "step": 4657
+    },
+    {
+      "epoch": 1.0806170977844798,
+      "grad_norm": 12.263893963668028,
+      "learning_rate": 2e-06,
+      "loss": 0.2594,
+      "step": 4658
+    },
+    {
+      "epoch": 1.0808490894327805,
+      "grad_norm": 6.0357490100565006,
+      "learning_rate": 2e-06,
+      "loss": 0.1549,
+      "step": 4659
+    },
+    {
+      "epoch": 1.0810810810810811,
+      "grad_norm": 16.18608723157087,
+      "learning_rate": 2e-06,
+      "loss": 0.2984,
+      "step": 4660
+    },
+    {
+      "epoch": 1.0813130727293818,
+      "grad_norm": 9.181713104002375,
+      "learning_rate": 2e-06,
+      "loss": 0.1386,
+      "step": 4661
+    },
+    {
+      "epoch": 1.0815450643776825,
+      "grad_norm": 5.644022388995841,
+      "learning_rate": 2e-06,
+      "loss": 0.1688,
+      "step": 4662
+    },
+    {
+      "epoch": 1.0817770560259832,
+      "grad_norm": 14.739106617725085,
+      "learning_rate": 2e-06,
+      "loss": 0.2945,
+      "step": 4663
+    },
+    {
+      "epoch": 1.0820090476742836,
+      "grad_norm": 11.676188528472048,
+      "learning_rate": 2e-06,
+      "loss": 0.1579,
+      "step": 4664
+    },
+    {
+      "epoch": 1.0822410393225843,
+      "grad_norm": 9.657311316004925,
+      "learning_rate": 2e-06,
+      "loss": 0.1834,
+      "step": 4665
+    },
+    {
+      "epoch": 1.082473030970885,
+      "grad_norm": 31.149815682668027,
+      "learning_rate": 2e-06,
+      "loss": 0.5156,
+      "step": 4666
+    },
+    {
+      "epoch": 1.0827050226191857,
+      "grad_norm": 22.49755979397703,
+      "learning_rate": 2e-06,
+      "loss": 0.3144,
+      "step": 4667
+    },
+    {
+      "epoch": 1.0829370142674863,
+      "grad_norm": 16.400850906073345,
+      "learning_rate": 2e-06,
+      "loss": 0.24,
+      "step": 4668
+    },
+    {
+      "epoch": 1.083169005915787,
+      "grad_norm": 14.898844803167906,
+      "learning_rate": 2e-06,
+      "loss": 0.2425,
+      "step": 4669
+    },
+    {
+      "epoch": 1.0834009975640877,
+      "grad_norm": 15.375770635265186,
+      "learning_rate": 2e-06,
+      "loss": 0.235,
+      "step": 4670
+    },
+    {
+      "epoch": 1.0836329892123884,
+      "grad_norm": 8.903667496388419,
+      "learning_rate": 2e-06,
+      "loss": 0.1767,
+      "step": 4671
+    },
+    {
+      "epoch": 1.083864980860689,
+      "grad_norm": 20.36801862973905,
+      "learning_rate": 2e-06,
+      "loss": 0.34,
+      "step": 4672
+    },
+    {
+      "epoch": 1.0840969725089897,
+      "grad_norm": 16.623785198060162,
+      "learning_rate": 2e-06,
+      "loss": 0.3051,
+      "step": 4673
+    },
+    {
+      "epoch": 1.0843289641572904,
+      "grad_norm": 21.712004714291528,
+      "learning_rate": 2e-06,
+      "loss": 0.2214,
+      "step": 4674
+    },
+    {
+      "epoch": 1.084560955805591,
+      "grad_norm": 18.64164274149171,
+      "learning_rate": 2e-06,
+      "loss": 0.3313,
+      "step": 4675
+    },
+    {
+      "epoch": 1.0847929474538915,
+      "grad_norm": 15.021075091917751,
+      "learning_rate": 2e-06,
+      "loss": 0.2372,
+      "step": 4676
+    },
+    {
+      "epoch": 1.0850249391021922,
+      "grad_norm": 13.172486147361289,
+      "learning_rate": 2e-06,
+      "loss": 0.2282,
+      "step": 4677
+    },
+    {
+      "epoch": 1.085256930750493,
+      "grad_norm": 14.952373836405085,
+      "learning_rate": 2e-06,
+      "loss": 0.2649,
+      "step": 4678
+    },
+    {
+      "epoch": 1.0854889223987936,
+      "grad_norm": 12.057002528751237,
+      "learning_rate": 2e-06,
+      "loss": 0.2557,
+      "step": 4679
+    },
+    {
+      "epoch": 1.0857209140470943,
+      "grad_norm": 14.097720996300128,
+      "learning_rate": 2e-06,
+      "loss": 0.1673,
+      "step": 4680
+    },
+    {
+      "epoch": 1.085952905695395,
+      "grad_norm": 13.424131645799678,
+      "learning_rate": 2e-06,
+      "loss": 0.2547,
+      "step": 4681
+    },
+    {
+      "epoch": 1.0861848973436956,
+      "grad_norm": 26.119273464189042,
+      "learning_rate": 2e-06,
+      "loss": 0.2301,
+      "step": 4682
+    },
+    {
+      "epoch": 1.0864168889919963,
+      "grad_norm": 10.85903450705642,
+      "learning_rate": 2e-06,
+      "loss": 0.3203,
+      "step": 4683
+    },
+    {
+      "epoch": 1.086648880640297,
+      "grad_norm": 12.192959640288787,
+      "learning_rate": 2e-06,
+      "loss": 0.2221,
+      "step": 4684
+    },
+    {
+      "epoch": 1.0868808722885976,
+      "grad_norm": 12.991969981697782,
+      "learning_rate": 2e-06,
+      "loss": 0.2164,
+      "step": 4685
+    },
+    {
+      "epoch": 1.0871128639368983,
+      "grad_norm": 13.438897594700979,
+      "learning_rate": 2e-06,
+      "loss": 0.2253,
+      "step": 4686
+    },
+    {
+      "epoch": 1.087344855585199,
+      "grad_norm": 21.185991926389537,
+      "learning_rate": 2e-06,
+      "loss": 0.2621,
+      "step": 4687
+    },
+    {
+      "epoch": 1.0875768472334997,
+      "grad_norm": 7.49060315068765,
+      "learning_rate": 2e-06,
+      "loss": 0.152,
+      "step": 4688
+    },
+    {
+      "epoch": 1.0878088388818004,
+      "grad_norm": 10.692675079619084,
+      "learning_rate": 2e-06,
+      "loss": 0.206,
+      "step": 4689
+    },
+    {
+      "epoch": 1.088040830530101,
+      "grad_norm": 12.085593295785767,
+      "learning_rate": 2e-06,
+      "loss": 0.2123,
+      "step": 4690
+    },
+    {
+      "epoch": 1.0882728221784015,
+      "grad_norm": 15.658009069301887,
+      "learning_rate": 2e-06,
+      "loss": 0.2292,
+      "step": 4691
+    },
+    {
+      "epoch": 1.0885048138267022,
+      "grad_norm": 21.839791385577954,
+      "learning_rate": 2e-06,
+      "loss": 0.2165,
+      "step": 4692
+    },
+    {
+      "epoch": 1.0887368054750028,
+      "grad_norm": 20.168667031089573,
+      "learning_rate": 2e-06,
+      "loss": 0.4257,
+      "step": 4693
+    },
+    {
+      "epoch": 1.0889687971233035,
+      "grad_norm": 13.532395866046311,
+      "learning_rate": 2e-06,
+      "loss": 0.3163,
+      "step": 4694
+    },
+    {
+      "epoch": 1.0892007887716042,
+      "grad_norm": 7.037441919712227,
+      "learning_rate": 2e-06,
+      "loss": 0.1576,
+      "step": 4695
+    },
+    {
+      "epoch": 1.0894327804199049,
+      "grad_norm": 6.080678326111714,
+      "learning_rate": 2e-06,
+      "loss": 0.1554,
+      "step": 4696
+    },
+    {
+      "epoch": 1.0896647720682056,
+      "grad_norm": 11.62312585625033,
+      "learning_rate": 2e-06,
+      "loss": 0.2072,
+      "step": 4697
+    },
+    {
+      "epoch": 1.0898967637165062,
+      "grad_norm": 25.07140286851281,
+      "learning_rate": 2e-06,
+      "loss": 0.3817,
+      "step": 4698
+    },
+    {
+      "epoch": 1.090128755364807,
+      "grad_norm": 14.454005153102496,
+      "learning_rate": 2e-06,
+      "loss": 0.237,
+      "step": 4699
+    },
+    {
+      "epoch": 1.0903607470131076,
+      "grad_norm": 33.65593280593324,
+      "learning_rate": 2e-06,
+      "loss": 0.3603,
+      "step": 4700
+    },
+    {
+      "epoch": 1.0905927386614083,
+      "grad_norm": 24.21819465836278,
+      "learning_rate": 2e-06,
+      "loss": 0.3173,
+      "step": 4701
+    },
+    {
+      "epoch": 1.090824730309709,
+      "grad_norm": 6.002003352997946,
+      "learning_rate": 2e-06,
+      "loss": 0.1649,
+      "step": 4702
+    },
+    {
+      "epoch": 1.0910567219580094,
+      "grad_norm": 15.363931384625337,
+      "learning_rate": 2e-06,
+      "loss": 0.3317,
+      "step": 4703
+    },
+    {
+      "epoch": 1.09128871360631,
+      "grad_norm": 19.767472015606725,
+      "learning_rate": 2e-06,
+      "loss": 0.3107,
+      "step": 4704
+    },
+    {
+      "epoch": 1.0915207052546108,
+      "grad_norm": 15.38698131756278,
+      "learning_rate": 2e-06,
+      "loss": 0.2003,
+      "step": 4705
+    },
+    {
+      "epoch": 1.0917526969029114,
+      "grad_norm": 12.01919931645781,
+      "learning_rate": 2e-06,
+      "loss": 0.2028,
+      "step": 4706
+    },
+    {
+      "epoch": 1.0919846885512121,
+      "grad_norm": 15.942894539948433,
+      "learning_rate": 2e-06,
+      "loss": 0.2649,
+      "step": 4707
+    },
+    {
+      "epoch": 1.0922166801995128,
+      "grad_norm": 14.38917599038256,
+      "learning_rate": 2e-06,
+      "loss": 0.2061,
+      "step": 4708
+    },
+    {
+      "epoch": 1.0924486718478135,
+      "grad_norm": 11.833878380303931,
+      "learning_rate": 2e-06,
+      "loss": 0.2203,
+      "step": 4709
+    },
+    {
+      "epoch": 1.0926806634961141,
+      "grad_norm": 16.914912934828145,
+      "learning_rate": 2e-06,
+      "loss": 0.2783,
+      "step": 4710
+    },
+    {
+      "epoch": 1.0929126551444148,
+      "grad_norm": 17.266074010543782,
+      "learning_rate": 2e-06,
+      "loss": 0.3055,
+      "step": 4711
+    },
+    {
+      "epoch": 1.0931446467927155,
+      "grad_norm": 18.75284359060675,
+      "learning_rate": 2e-06,
+      "loss": 0.283,
+      "step": 4712
+    },
+    {
+      "epoch": 1.0933766384410162,
+      "grad_norm": 12.43982576036956,
+      "learning_rate": 2e-06,
+      "loss": 0.1862,
+      "step": 4713
+    },
+    {
+      "epoch": 1.0936086300893169,
+      "grad_norm": 14.80054201183015,
+      "learning_rate": 2e-06,
+      "loss": 0.2475,
+      "step": 4714
+    },
+    {
+      "epoch": 1.0938406217376175,
+      "grad_norm": 13.096084263417465,
+      "learning_rate": 2e-06,
+      "loss": 0.2297,
+      "step": 4715
+    },
+    {
+      "epoch": 1.0940726133859182,
+      "grad_norm": 23.74250380776062,
+      "learning_rate": 2e-06,
+      "loss": 0.2798,
+      "step": 4716
+    },
+    {
+      "epoch": 1.0943046050342187,
+      "grad_norm": 8.639370985491647,
+      "learning_rate": 2e-06,
+      "loss": 0.1993,
+      "step": 4717
+    },
+    {
+      "epoch": 1.0945365966825193,
+      "grad_norm": 8.52413219768701,
+      "learning_rate": 2e-06,
+      "loss": 0.1727,
+      "step": 4718
+    },
+    {
+      "epoch": 1.09476858833082,
+      "grad_norm": 11.6044749708807,
+      "learning_rate": 2e-06,
+      "loss": 0.2995,
+      "step": 4719
+    },
+    {
+      "epoch": 1.0950005799791207,
+      "grad_norm": 15.624858689431957,
+      "learning_rate": 2e-06,
+      "loss": 0.2698,
+      "step": 4720
+    },
+    {
+      "epoch": 1.0952325716274214,
+      "grad_norm": 13.945303085396459,
+      "learning_rate": 2e-06,
+      "loss": 0.2106,
+      "step": 4721
+    },
+    {
+      "epoch": 1.095464563275722,
+      "grad_norm": 10.769513824946431,
+      "learning_rate": 2e-06,
+      "loss": 0.1963,
+      "step": 4722
+    },
+    {
+      "epoch": 1.0956965549240227,
+      "grad_norm": 7.428153969726964,
+      "learning_rate": 2e-06,
+      "loss": 0.148,
+      "step": 4723
+    },
+    {
+      "epoch": 1.0959285465723234,
+      "grad_norm": 6.88998868347155,
+      "learning_rate": 2e-06,
+      "loss": 0.1802,
+      "step": 4724
+    },
+    {
+      "epoch": 1.096160538220624,
+      "grad_norm": 14.252947491435332,
+      "learning_rate": 2e-06,
+      "loss": 0.2731,
+      "step": 4725
+    },
+    {
+      "epoch": 1.0963925298689248,
+      "grad_norm": 18.767434337446932,
+      "learning_rate": 2e-06,
+      "loss": 0.1984,
+      "step": 4726
+    },
+    {
+      "epoch": 1.0966245215172254,
+      "grad_norm": 25.407062038419664,
+      "learning_rate": 2e-06,
+      "loss": 0.3428,
+      "step": 4727
+    },
+    {
+      "epoch": 1.0968565131655261,
+      "grad_norm": 19.46092075545016,
+      "learning_rate": 2e-06,
+      "loss": 0.2915,
+      "step": 4728
+    },
+    {
+      "epoch": 1.0970885048138268,
+      "grad_norm": 20.120073541123094,
+      "learning_rate": 2e-06,
+      "loss": 0.3822,
+      "step": 4729
+    },
+    {
+      "epoch": 1.0973204964621273,
+      "grad_norm": 22.039775692665884,
+      "learning_rate": 2e-06,
+      "loss": 0.3699,
+      "step": 4730
+    },
+    {
+      "epoch": 1.097552488110428,
+      "grad_norm": 19.658684197724533,
+      "learning_rate": 2e-06,
+      "loss": 0.2764,
+      "step": 4731
+    },
+    {
+      "epoch": 1.0977844797587286,
+      "grad_norm": 12.168061926618066,
+      "learning_rate": 2e-06,
+      "loss": 0.1944,
+      "step": 4732
+    },
+    {
+      "epoch": 1.0980164714070293,
+      "grad_norm": 24.83924510420318,
+      "learning_rate": 2e-06,
+      "loss": 0.3434,
+      "step": 4733
+    },
+    {
+      "epoch": 1.09824846305533,
+      "grad_norm": 12.746618260356046,
+      "learning_rate": 2e-06,
+      "loss": 0.2208,
+      "step": 4734
+    },
+    {
+      "epoch": 1.0984804547036306,
+      "grad_norm": 13.560829849978433,
+      "learning_rate": 2e-06,
+      "loss": 0.2237,
+      "step": 4735
+    },
+    {
+      "epoch": 1.0987124463519313,
+      "grad_norm": 16.147177241511702,
+      "learning_rate": 2e-06,
+      "loss": 0.2865,
+      "step": 4736
+    },
+    {
+      "epoch": 1.098944438000232,
+      "grad_norm": 15.593195676890463,
+      "learning_rate": 2e-06,
+      "loss": 0.1484,
+      "step": 4737
+    },
+    {
+      "epoch": 1.0991764296485327,
+      "grad_norm": 17.79741115414023,
+      "learning_rate": 2e-06,
+      "loss": 0.2794,
+      "step": 4738
+    },
+    {
+      "epoch": 1.0994084212968334,
+      "grad_norm": 12.267696363765928,
+      "learning_rate": 2e-06,
+      "loss": 0.213,
+      "step": 4739
+    },
+    {
+      "epoch": 1.099640412945134,
+      "grad_norm": 21.270542895126816,
+      "learning_rate": 2e-06,
+      "loss": 0.3346,
+      "step": 4740
+    },
+    {
+      "epoch": 1.0998724045934347,
+      "grad_norm": 13.268440838091314,
+      "learning_rate": 2e-06,
+      "loss": 0.2893,
+      "step": 4741
+    },
+    {
+      "epoch": 1.1001043962417354,
+      "grad_norm": 5.773451254069926,
+      "learning_rate": 2e-06,
+      "loss": 0.1248,
+      "step": 4742
+    },
+    {
+      "epoch": 1.100336387890036,
+      "grad_norm": 13.38291685039027,
+      "learning_rate": 2e-06,
+      "loss": 0.2305,
+      "step": 4743
+    },
+    {
+      "epoch": 1.1005683795383365,
+      "grad_norm": 17.201303310278664,
+      "learning_rate": 2e-06,
+      "loss": 0.3608,
+      "step": 4744
+    },
+    {
+      "epoch": 1.1008003711866372,
+      "grad_norm": 14.377630959205913,
+      "learning_rate": 2e-06,
+      "loss": 0.2518,
+      "step": 4745
+    },
+    {
+      "epoch": 1.1010323628349379,
+      "grad_norm": 14.809347676129226,
+      "learning_rate": 2e-06,
+      "loss": 0.2533,
+      "step": 4746
+    },
+    {
+      "epoch": 1.1012643544832386,
+      "grad_norm": 9.986337062899754,
+      "learning_rate": 2e-06,
+      "loss": 0.2014,
+      "step": 4747
+    },
+    {
+      "epoch": 1.1014963461315392,
+      "grad_norm": 15.922279096775062,
+      "learning_rate": 2e-06,
+      "loss": 0.2832,
+      "step": 4748
+    },
+    {
+      "epoch": 1.10172833777984,
+      "grad_norm": 12.216428482913866,
+      "learning_rate": 2e-06,
+      "loss": 0.2412,
+      "step": 4749
+    },
+    {
+      "epoch": 1.1019603294281406,
+      "grad_norm": 8.778545487508715,
+      "learning_rate": 2e-06,
+      "loss": 0.213,
+      "step": 4750
+    },
+    {
+      "epoch": 1.1021923210764413,
+      "grad_norm": 13.367201916070908,
+      "learning_rate": 2e-06,
+      "loss": 0.2322,
+      "step": 4751
+    },
+    {
+      "epoch": 1.102424312724742,
+      "grad_norm": 17.779014580733584,
+      "learning_rate": 2e-06,
+      "loss": 0.2878,
+      "step": 4752
+    },
+    {
+      "epoch": 1.1026563043730426,
+      "grad_norm": 15.27352344365719,
+      "learning_rate": 2e-06,
+      "loss": 0.3188,
+      "step": 4753
+    },
+    {
+      "epoch": 1.1028882960213433,
+      "grad_norm": 16.461117593032657,
+      "learning_rate": 2e-06,
+      "loss": 0.2435,
+      "step": 4754
+    },
+    {
+      "epoch": 1.103120287669644,
+      "grad_norm": 14.525277461670653,
+      "learning_rate": 2e-06,
+      "loss": 0.2675,
+      "step": 4755
+    },
+    {
+      "epoch": 1.1033522793179444,
+      "grad_norm": 16.976506099386146,
+      "learning_rate": 2e-06,
+      "loss": 0.1877,
+      "step": 4756
+    },
+    {
+      "epoch": 1.1035842709662451,
+      "grad_norm": 10.436412155903202,
+      "learning_rate": 2e-06,
+      "loss": 0.2611,
+      "step": 4757
+    },
+    {
+      "epoch": 1.1038162626145458,
+      "grad_norm": 17.45570996860586,
+      "learning_rate": 2e-06,
+      "loss": 0.2223,
+      "step": 4758
+    },
+    {
+      "epoch": 1.1040482542628465,
+      "grad_norm": 9.11640215787863,
+      "learning_rate": 2e-06,
+      "loss": 0.1965,
+      "step": 4759
+    },
+    {
+      "epoch": 1.1042802459111472,
+      "grad_norm": 18.53955266775229,
+      "learning_rate": 2e-06,
+      "loss": 0.2927,
+      "step": 4760
+    },
+    {
+      "epoch": 1.1045122375594478,
+      "grad_norm": 11.813660934324743,
+      "learning_rate": 2e-06,
+      "loss": 0.2288,
+      "step": 4761
+    },
+    {
+      "epoch": 1.1047442292077485,
+      "grad_norm": 28.113973632222756,
+      "learning_rate": 2e-06,
+      "loss": 0.4137,
+      "step": 4762
+    },
+    {
+      "epoch": 1.1049762208560492,
+      "grad_norm": 13.059036772244967,
+      "learning_rate": 2e-06,
+      "loss": 0.2059,
+      "step": 4763
+    },
+    {
+      "epoch": 1.1052082125043499,
+      "grad_norm": 26.718482609331993,
+      "learning_rate": 2e-06,
+      "loss": 0.5283,
+      "step": 4764
+    },
+    {
+      "epoch": 1.1054402041526505,
+      "grad_norm": 21.388850421340358,
+      "learning_rate": 2e-06,
+      "loss": 0.2686,
+      "step": 4765
+    },
+    {
+      "epoch": 1.1056721958009512,
+      "grad_norm": 13.831933938166646,
+      "learning_rate": 2e-06,
+      "loss": 0.2863,
+      "step": 4766
+    },
+    {
+      "epoch": 1.105904187449252,
+      "grad_norm": 14.716751435725632,
+      "learning_rate": 2e-06,
+      "loss": 0.2895,
+      "step": 4767
+    },
+    {
+      "epoch": 1.1061361790975526,
+      "grad_norm": 28.502863024905714,
+      "learning_rate": 2e-06,
+      "loss": 0.2986,
+      "step": 4768
+    },
+    {
+      "epoch": 1.1063681707458533,
+      "grad_norm": 11.712588324651502,
+      "learning_rate": 2e-06,
+      "loss": 0.2583,
+      "step": 4769
+    },
+    {
+      "epoch": 1.1066001623941537,
+      "grad_norm": 14.2455947813304,
+      "learning_rate": 2e-06,
+      "loss": 0.3269,
+      "step": 4770
+    },
+    {
+      "epoch": 1.1068321540424544,
+      "grad_norm": 20.12904263154396,
+      "learning_rate": 2e-06,
+      "loss": 0.3651,
+      "step": 4771
+    },
+    {
+      "epoch": 1.107064145690755,
+      "grad_norm": 10.495923075726504,
+      "learning_rate": 2e-06,
+      "loss": 0.2407,
+      "step": 4772
+    },
+    {
+      "epoch": 1.1072961373390557,
+      "grad_norm": 13.363417014459092,
+      "learning_rate": 2e-06,
+      "loss": 0.3103,
+      "step": 4773
+    },
+    {
+      "epoch": 1.1075281289873564,
+      "grad_norm": 9.752561848539205,
+      "learning_rate": 2e-06,
+      "loss": 0.1802,
+      "step": 4774
+    },
+    {
+      "epoch": 1.107760120635657,
+      "grad_norm": 10.545529316723636,
+      "learning_rate": 2e-06,
+      "loss": 0.2058,
+      "step": 4775
+    },
+    {
+      "epoch": 1.1079921122839578,
+      "grad_norm": 11.52069645016782,
+      "learning_rate": 2e-06,
+      "loss": 0.2783,
+      "step": 4776
+    },
+    {
+      "epoch": 1.1082241039322585,
+      "grad_norm": 11.148106046786955,
+      "learning_rate": 2e-06,
+      "loss": 0.1324,
+      "step": 4777
+    },
+    {
+      "epoch": 1.1084560955805591,
+      "grad_norm": 11.319629458424366,
+      "learning_rate": 2e-06,
+      "loss": 0.3652,
+      "step": 4778
+    },
+    {
+      "epoch": 1.1086880872288598,
+      "grad_norm": 13.690492530236053,
+      "learning_rate": 2e-06,
+      "loss": 0.2366,
+      "step": 4779
+    },
+    {
+      "epoch": 1.1089200788771605,
+      "grad_norm": 9.632568788720638,
+      "learning_rate": 2e-06,
+      "loss": 0.2421,
+      "step": 4780
+    },
+    {
+      "epoch": 1.1091520705254612,
+      "grad_norm": 12.249528407753822,
+      "learning_rate": 2e-06,
+      "loss": 0.2144,
+      "step": 4781
+    },
+    {
+      "epoch": 1.1093840621737618,
+      "grad_norm": 14.319966268970228,
+      "learning_rate": 2e-06,
+      "loss": 0.2755,
+      "step": 4782
+    },
+    {
+      "epoch": 1.1096160538220623,
+      "grad_norm": 25.04977436131108,
+      "learning_rate": 2e-06,
+      "loss": 0.2159,
+      "step": 4783
+    },
+    {
+      "epoch": 1.109848045470363,
+      "grad_norm": 13.118374923070153,
+      "learning_rate": 2e-06,
+      "loss": 0.2431,
+      "step": 4784
+    },
+    {
+      "epoch": 1.1100800371186637,
+      "grad_norm": 13.918332645947876,
+      "learning_rate": 2e-06,
+      "loss": 0.3029,
+      "step": 4785
+    },
+    {
+      "epoch": 1.1103120287669643,
+      "grad_norm": 17.479860616245222,
+      "learning_rate": 2e-06,
+      "loss": 0.3775,
+      "step": 4786
+    },
+    {
+      "epoch": 1.110544020415265,
+      "grad_norm": 14.610461992797951,
+      "learning_rate": 2e-06,
+      "loss": 0.2226,
+      "step": 4787
+    },
+    {
+      "epoch": 1.1107760120635657,
+      "grad_norm": 18.220638020090497,
+      "learning_rate": 2e-06,
+      "loss": 0.2144,
+      "step": 4788
+    },
+    {
+      "epoch": 1.1110080037118664,
+      "grad_norm": 11.609900531076313,
+      "learning_rate": 2e-06,
+      "loss": 0.2249,
+      "step": 4789
+    },
+    {
+      "epoch": 1.111239995360167,
+      "grad_norm": 11.198809304515796,
+      "learning_rate": 2e-06,
+      "loss": 0.2243,
+      "step": 4790
+    },
+    {
+      "epoch": 1.1114719870084677,
+      "grad_norm": 17.489913518711848,
+      "learning_rate": 2e-06,
+      "loss": 0.2504,
+      "step": 4791
+    },
+    {
+      "epoch": 1.1117039786567684,
+      "grad_norm": 12.78910159612698,
+      "learning_rate": 2e-06,
+      "loss": 0.3694,
+      "step": 4792
+    },
+    {
+      "epoch": 1.111935970305069,
+      "grad_norm": 15.772843023314104,
+      "learning_rate": 2e-06,
+      "loss": 0.3384,
+      "step": 4793
+    },
+    {
+      "epoch": 1.1121679619533698,
+      "grad_norm": 11.195851794747387,
+      "learning_rate": 2e-06,
+      "loss": 0.2107,
+      "step": 4794
+    },
+    {
+      "epoch": 1.1123999536016704,
+      "grad_norm": 9.780183781534575,
+      "learning_rate": 2e-06,
+      "loss": 0.1513,
+      "step": 4795
+    },
+    {
+      "epoch": 1.112631945249971,
+      "grad_norm": 11.903280414707714,
+      "learning_rate": 2e-06,
+      "loss": 0.2401,
+      "step": 4796
+    },
+    {
+      "epoch": 1.1128639368982716,
+      "grad_norm": 12.375824587748216,
+      "learning_rate": 2e-06,
+      "loss": 0.3704,
+      "step": 4797
+    },
+    {
+      "epoch": 1.1130959285465722,
+      "grad_norm": 15.841434537765656,
+      "learning_rate": 2e-06,
+      "loss": 0.267,
+      "step": 4798
+    },
+    {
+      "epoch": 1.113327920194873,
+      "grad_norm": 16.05445964383955,
+      "learning_rate": 2e-06,
+      "loss": 0.2876,
+      "step": 4799
+    },
+    {
+      "epoch": 1.1135599118431736,
+      "grad_norm": 16.170941162145347,
+      "learning_rate": 2e-06,
+      "loss": 0.4808,
+      "step": 4800
+    },
+    {
+      "epoch": 1.1137919034914743,
+      "grad_norm": 7.983958612287163,
+      "learning_rate": 2e-06,
+      "loss": 0.2322,
+      "step": 4801
+    },
+    {
+      "epoch": 1.114023895139775,
+      "grad_norm": 17.947298006751954,
+      "learning_rate": 2e-06,
+      "loss": 0.2394,
+      "step": 4802
+    },
+    {
+      "epoch": 1.1142558867880756,
+      "grad_norm": 14.904696965127243,
+      "learning_rate": 2e-06,
+      "loss": 0.2698,
+      "step": 4803
+    },
+    {
+      "epoch": 1.1144878784363763,
+      "grad_norm": 24.716054456606248,
+      "learning_rate": 2e-06,
+      "loss": 0.3391,
+      "step": 4804
+    },
+    {
+      "epoch": 1.114719870084677,
+      "grad_norm": 17.012778768945804,
+      "learning_rate": 2e-06,
+      "loss": 0.2561,
+      "step": 4805
+    },
+    {
+      "epoch": 1.1149518617329777,
+      "grad_norm": 11.77157842120241,
+      "learning_rate": 2e-06,
+      "loss": 0.2345,
+      "step": 4806
+    },
+    {
+      "epoch": 1.1151838533812783,
+      "grad_norm": 17.869538222418846,
+      "learning_rate": 2e-06,
+      "loss": 0.2186,
+      "step": 4807
+    },
+    {
+      "epoch": 1.115415845029579,
+      "grad_norm": 7.464176480789736,
+      "learning_rate": 2e-06,
+      "loss": 0.1725,
+      "step": 4808
+    },
+    {
+      "epoch": 1.1156478366778795,
+      "grad_norm": 5.421341474728665,
+      "learning_rate": 2e-06,
+      "loss": 0.209,
+      "step": 4809
+    },
+    {
+      "epoch": 1.1158798283261802,
+      "grad_norm": 10.89685000856278,
+      "learning_rate": 2e-06,
+      "loss": 0.1944,
+      "step": 4810
+    },
+    {
+      "epoch": 1.1161118199744808,
+      "grad_norm": 13.971761511853458,
+      "learning_rate": 2e-06,
+      "loss": 0.3304,
+      "step": 4811
+    },
+    {
+      "epoch": 1.1163438116227815,
+      "grad_norm": 34.40954485204369,
+      "learning_rate": 2e-06,
+      "loss": 0.394,
+      "step": 4812
+    },
+    {
+      "epoch": 1.1165758032710822,
+      "grad_norm": 15.488875948943251,
+      "learning_rate": 2e-06,
+      "loss": 0.2046,
+      "step": 4813
+    },
+    {
+      "epoch": 1.1168077949193829,
+      "grad_norm": 12.132601391775486,
+      "learning_rate": 2e-06,
+      "loss": 0.2219,
+      "step": 4814
+    },
+    {
+      "epoch": 1.1170397865676835,
+      "grad_norm": 13.424740585546829,
+      "learning_rate": 2e-06,
+      "loss": 0.221,
+      "step": 4815
+    },
+    {
+      "epoch": 1.1172717782159842,
+      "grad_norm": 8.659678873098684,
+      "learning_rate": 2e-06,
+      "loss": 0.2093,
+      "step": 4816
+    },
+    {
+      "epoch": 1.117503769864285,
+      "grad_norm": 11.50236687791197,
+      "learning_rate": 2e-06,
+      "loss": 0.2311,
+      "step": 4817
+    },
+    {
+      "epoch": 1.1177357615125856,
+      "grad_norm": 18.164120226051143,
+      "learning_rate": 2e-06,
+      "loss": 0.2499,
+      "step": 4818
+    },
+    {
+      "epoch": 1.1179677531608863,
+      "grad_norm": 17.800646370079132,
+      "learning_rate": 2e-06,
+      "loss": 0.223,
+      "step": 4819
+    },
+    {
+      "epoch": 1.118199744809187,
+      "grad_norm": 9.323484868669501,
+      "learning_rate": 2e-06,
+      "loss": 0.1765,
+      "step": 4820
+    },
+    {
+      "epoch": 1.1184317364574876,
+      "grad_norm": 16.824993656613074,
+      "learning_rate": 2e-06,
+      "loss": 0.312,
+      "step": 4821
+    },
+    {
+      "epoch": 1.1186637281057883,
+      "grad_norm": 12.203168704174818,
+      "learning_rate": 2e-06,
+      "loss": 0.2344,
+      "step": 4822
+    },
+    {
+      "epoch": 1.118895719754089,
+      "grad_norm": 31.56245938171965,
+      "learning_rate": 2e-06,
+      "loss": 0.3754,
+      "step": 4823
+    },
+    {
+      "epoch": 1.1191277114023894,
+      "grad_norm": 8.968222747142645,
+      "learning_rate": 2e-06,
+      "loss": 0.1932,
+      "step": 4824
+    },
+    {
+      "epoch": 1.11935970305069,
+      "grad_norm": 16.992866517419166,
+      "learning_rate": 2e-06,
+      "loss": 0.2995,
+      "step": 4825
+    },
+    {
+      "epoch": 1.1195916946989908,
+      "grad_norm": 10.459429594675886,
+      "learning_rate": 2e-06,
+      "loss": 0.352,
+      "step": 4826
+    },
+    {
+      "epoch": 1.1198236863472915,
+      "grad_norm": 13.449766556141963,
+      "learning_rate": 2e-06,
+      "loss": 0.2041,
+      "step": 4827
+    },
+    {
+      "epoch": 1.1200556779955921,
+      "grad_norm": 12.05557560578355,
+      "learning_rate": 2e-06,
+      "loss": 0.2017,
+      "step": 4828
+    },
+    {
+      "epoch": 1.1202876696438928,
+      "grad_norm": 9.367438870463518,
+      "learning_rate": 2e-06,
+      "loss": 0.2387,
+      "step": 4829
+    },
+    {
+      "epoch": 1.1205196612921935,
+      "grad_norm": 17.76365816605677,
+      "learning_rate": 2e-06,
+      "loss": 0.1864,
+      "step": 4830
+    },
+    {
+      "epoch": 1.1207516529404942,
+      "grad_norm": 15.404675931816326,
+      "learning_rate": 2e-06,
+      "loss": 0.2811,
+      "step": 4831
+    },
+    {
+      "epoch": 1.1209836445887948,
+      "grad_norm": 15.29386673652393,
+      "learning_rate": 2e-06,
+      "loss": 0.2481,
+      "step": 4832
+    },
+    {
+      "epoch": 1.1212156362370955,
+      "grad_norm": 8.263621968845886,
+      "learning_rate": 2e-06,
+      "loss": 0.2153,
+      "step": 4833
+    },
+    {
+      "epoch": 1.1214476278853962,
+      "grad_norm": 11.598973711992286,
+      "learning_rate": 2e-06,
+      "loss": 0.2205,
+      "step": 4834
+    },
+    {
+      "epoch": 1.1216796195336969,
+      "grad_norm": 12.121246868598893,
+      "learning_rate": 2e-06,
+      "loss": 0.2727,
+      "step": 4835
+    },
+    {
+      "epoch": 1.1219116111819973,
+      "grad_norm": 21.845578092591122,
+      "learning_rate": 2e-06,
+      "loss": 0.247,
+      "step": 4836
+    },
+    {
+      "epoch": 1.122143602830298,
+      "grad_norm": 26.244677083522305,
+      "learning_rate": 2e-06,
+      "loss": 0.4809,
+      "step": 4837
+    },
+    {
+      "epoch": 1.1223755944785987,
+      "grad_norm": 14.525950272406515,
+      "learning_rate": 2e-06,
+      "loss": 0.2579,
+      "step": 4838
+    },
+    {
+      "epoch": 1.1226075861268994,
+      "grad_norm": 8.921911193753202,
+      "learning_rate": 2e-06,
+      "loss": 0.2388,
+      "step": 4839
+    },
+    {
+      "epoch": 1.1228395777752,
+      "grad_norm": 18.573858193690356,
+      "learning_rate": 2e-06,
+      "loss": 0.2852,
+      "step": 4840
+    },
+    {
+      "epoch": 1.1230715694235007,
+      "grad_norm": 15.231148127881918,
+      "learning_rate": 2e-06,
+      "loss": 0.3048,
+      "step": 4841
+    },
+    {
+      "epoch": 1.1233035610718014,
+      "grad_norm": 14.311189724319316,
+      "learning_rate": 2e-06,
+      "loss": 0.174,
+      "step": 4842
+    },
+    {
+      "epoch": 1.123535552720102,
+      "grad_norm": 9.759725795906759,
+      "learning_rate": 2e-06,
+      "loss": 0.2344,
+      "step": 4843
+    },
+    {
+      "epoch": 1.1237675443684028,
+      "grad_norm": 16.32478232769143,
+      "learning_rate": 2e-06,
+      "loss": 0.3097,
+      "step": 4844
+    },
+    {
+      "epoch": 1.1239995360167034,
+      "grad_norm": 11.181963687751907,
+      "learning_rate": 2e-06,
+      "loss": 0.2216,
+      "step": 4845
+    },
+    {
+      "epoch": 1.1242315276650041,
+      "grad_norm": 10.353225268015038,
+      "learning_rate": 2e-06,
+      "loss": 0.2954,
+      "step": 4846
+    },
+    {
+      "epoch": 1.1244635193133048,
+      "grad_norm": 10.309147818411686,
+      "learning_rate": 2e-06,
+      "loss": 0.1935,
+      "step": 4847
+    },
+    {
+      "epoch": 1.1246955109616055,
+      "grad_norm": 14.911985063186178,
+      "learning_rate": 2e-06,
+      "loss": 0.228,
+      "step": 4848
+    },
+    {
+      "epoch": 1.1249275026099061,
+      "grad_norm": 21.648534920152432,
+      "learning_rate": 2e-06,
+      "loss": 0.3459,
+      "step": 4849
+    },
+    {
+      "epoch": 1.1251594942582068,
+      "grad_norm": 12.10124114368933,
+      "learning_rate": 2e-06,
+      "loss": 0.2225,
+      "step": 4850
+    },
+    {
+      "epoch": 1.1253914859065073,
+      "grad_norm": 15.01947620297207,
+      "learning_rate": 2e-06,
+      "loss": 0.2392,
+      "step": 4851
+    },
+    {
+      "epoch": 1.125623477554808,
+      "grad_norm": 22.524618954163767,
+      "learning_rate": 2e-06,
+      "loss": 0.2461,
+      "step": 4852
+    },
+    {
+      "epoch": 1.1258554692031086,
+      "grad_norm": 20.873088264065554,
+      "learning_rate": 2e-06,
+      "loss": 0.3608,
+      "step": 4853
+    },
+    {
+      "epoch": 1.1260874608514093,
+      "grad_norm": 13.904202955837375,
+      "learning_rate": 2e-06,
+      "loss": 0.2254,
+      "step": 4854
+    },
+    {
+      "epoch": 1.12631945249971,
+      "grad_norm": 26.509242877229937,
+      "learning_rate": 2e-06,
+      "loss": 0.3359,
+      "step": 4855
+    },
+    {
+      "epoch": 1.1265514441480107,
+      "grad_norm": 13.028136016976488,
+      "learning_rate": 2e-06,
+      "loss": 0.2455,
+      "step": 4856
+    },
+    {
+      "epoch": 1.1267834357963114,
+      "grad_norm": 7.319785950351981,
+      "learning_rate": 2e-06,
+      "loss": 0.1814,
+      "step": 4857
+    },
+    {
+      "epoch": 1.127015427444612,
+      "grad_norm": 20.918159000298974,
+      "learning_rate": 2e-06,
+      "loss": 0.3011,
+      "step": 4858
+    },
+    {
+      "epoch": 1.1272474190929127,
+      "grad_norm": 9.220278402493546,
+      "learning_rate": 2e-06,
+      "loss": 0.2005,
+      "step": 4859
+    },
+    {
+      "epoch": 1.1274794107412134,
+      "grad_norm": 7.8435657373328445,
+      "learning_rate": 2e-06,
+      "loss": 0.1529,
+      "step": 4860
+    },
+    {
+      "epoch": 1.127711402389514,
+      "grad_norm": 12.331950357133152,
+      "learning_rate": 2e-06,
+      "loss": 0.2644,
+      "step": 4861
+    },
+    {
+      "epoch": 1.1279433940378145,
+      "grad_norm": 11.468091056813932,
+      "learning_rate": 2e-06,
+      "loss": 0.2825,
+      "step": 4862
+    },
+    {
+      "epoch": 1.1281753856861152,
+      "grad_norm": 9.125446465633535,
+      "learning_rate": 2e-06,
+      "loss": 0.1779,
+      "step": 4863
+    },
+    {
+      "epoch": 1.1284073773344159,
+      "grad_norm": 16.561095932862465,
+      "learning_rate": 2e-06,
+      "loss": 0.3371,
+      "step": 4864
+    },
+    {
+      "epoch": 1.1286393689827166,
+      "grad_norm": 13.373293675830446,
+      "learning_rate": 2e-06,
+      "loss": 0.3443,
+      "step": 4865
+    },
+    {
+      "epoch": 1.1288713606310172,
+      "grad_norm": 20.618005722280156,
+      "learning_rate": 2e-06,
+      "loss": 0.302,
+      "step": 4866
+    },
+    {
+      "epoch": 1.129103352279318,
+      "grad_norm": 20.43437800180757,
+      "learning_rate": 2e-06,
+      "loss": 0.4071,
+      "step": 4867
+    },
+    {
+      "epoch": 1.1293353439276186,
+      "grad_norm": 23.09409807583735,
+      "learning_rate": 2e-06,
+      "loss": 0.3352,
+      "step": 4868
+    },
+    {
+      "epoch": 1.1295673355759193,
+      "grad_norm": 21.548952088035673,
+      "learning_rate": 2e-06,
+      "loss": 0.2033,
+      "step": 4869
+    },
+    {
+      "epoch": 1.12979932722422,
+      "grad_norm": 13.179588125487964,
+      "learning_rate": 2e-06,
+      "loss": 0.1926,
+      "step": 4870
+    },
+    {
+      "epoch": 1.1300313188725206,
+      "grad_norm": 17.560998374400327,
+      "learning_rate": 2e-06,
+      "loss": 0.3046,
+      "step": 4871
+    },
+    {
+      "epoch": 1.1302633105208213,
+      "grad_norm": 15.359237858726829,
+      "learning_rate": 2e-06,
+      "loss": 0.2663,
+      "step": 4872
+    },
+    {
+      "epoch": 1.130495302169122,
+      "grad_norm": 19.264672448088973,
+      "learning_rate": 2e-06,
+      "loss": 0.311,
+      "step": 4873
+    },
+    {
+      "epoch": 1.1307272938174227,
+      "grad_norm": 18.49971199105671,
+      "learning_rate": 2e-06,
+      "loss": 0.2726,
+      "step": 4874
+    },
+    {
+      "epoch": 1.1309592854657233,
+      "grad_norm": 16.811942294775953,
+      "learning_rate": 2e-06,
+      "loss": 0.3512,
+      "step": 4875
+    },
+    {
+      "epoch": 1.131191277114024,
+      "grad_norm": 10.463711626604983,
+      "learning_rate": 2e-06,
+      "loss": 0.2304,
+      "step": 4876
+    },
+    {
+      "epoch": 1.1314232687623245,
+      "grad_norm": 12.918641341385287,
+      "learning_rate": 2e-06,
+      "loss": 0.2987,
+      "step": 4877
+    },
+    {
+      "epoch": 1.1316552604106251,
+      "grad_norm": 7.4286140248356825,
+      "learning_rate": 2e-06,
+      "loss": 0.1508,
+      "step": 4878
+    },
+    {
+      "epoch": 1.1318872520589258,
+      "grad_norm": 20.665785145642293,
+      "learning_rate": 2e-06,
+      "loss": 0.365,
+      "step": 4879
+    },
+    {
+      "epoch": 1.1321192437072265,
+      "grad_norm": 12.166636770416769,
+      "learning_rate": 2e-06,
+      "loss": 0.2197,
+      "step": 4880
+    },
+    {
+      "epoch": 1.1323512353555272,
+      "grad_norm": 15.393645357245871,
+      "learning_rate": 2e-06,
+      "loss": 0.3478,
+      "step": 4881
+    },
+    {
+      "epoch": 1.1325832270038279,
+      "grad_norm": 9.480527037233777,
+      "learning_rate": 2e-06,
+      "loss": 0.1504,
+      "step": 4882
+    },
+    {
+      "epoch": 1.1328152186521285,
+      "grad_norm": 10.986611255263721,
+      "learning_rate": 2e-06,
+      "loss": 0.2744,
+      "step": 4883
+    },
+    {
+      "epoch": 1.1330472103004292,
+      "grad_norm": 8.984061788935852,
+      "learning_rate": 2e-06,
+      "loss": 0.2151,
+      "step": 4884
+    },
+    {
+      "epoch": 1.1332792019487299,
+      "grad_norm": 19.65763052847948,
+      "learning_rate": 2e-06,
+      "loss": 0.2741,
+      "step": 4885
+    },
+    {
+      "epoch": 1.1335111935970306,
+      "grad_norm": 10.736693671979038,
+      "learning_rate": 2e-06,
+      "loss": 0.1959,
+      "step": 4886
+    },
+    {
+      "epoch": 1.1337431852453312,
+      "grad_norm": 14.183338903446597,
+      "learning_rate": 2e-06,
+      "loss": 0.281,
+      "step": 4887
+    },
+    {
+      "epoch": 1.133975176893632,
+      "grad_norm": 9.740995781048545,
+      "learning_rate": 2e-06,
+      "loss": 0.1822,
+      "step": 4888
+    },
+    {
+      "epoch": 1.1342071685419324,
+      "grad_norm": 21.632445425240174,
+      "learning_rate": 2e-06,
+      "loss": 0.3031,
+      "step": 4889
+    },
+    {
+      "epoch": 1.134439160190233,
+      "grad_norm": 24.816556530453955,
+      "learning_rate": 2e-06,
+      "loss": 0.3508,
+      "step": 4890
+    },
+    {
+      "epoch": 1.1346711518385337,
+      "grad_norm": 13.454497821313662,
+      "learning_rate": 2e-06,
+      "loss": 0.196,
+      "step": 4891
+    },
+    {
+      "epoch": 1.1349031434868344,
+      "grad_norm": 6.912301352121711,
+      "learning_rate": 2e-06,
+      "loss": 0.18,
+      "step": 4892
+    },
+    {
+      "epoch": 1.135135135135135,
+      "grad_norm": 20.439880621178403,
+      "learning_rate": 2e-06,
+      "loss": 0.2886,
+      "step": 4893
+    },
+    {
+      "epoch": 1.1353671267834358,
+      "grad_norm": 19.2432781790515,
+      "learning_rate": 2e-06,
+      "loss": 0.3268,
+      "step": 4894
+    },
+    {
+      "epoch": 1.1355991184317364,
+      "grad_norm": 12.769986602500271,
+      "learning_rate": 2e-06,
+      "loss": 0.2096,
+      "step": 4895
+    },
+    {
+      "epoch": 1.1358311100800371,
+      "grad_norm": 11.600290293189488,
+      "learning_rate": 2e-06,
+      "loss": 0.1359,
+      "step": 4896
+    },
+    {
+      "epoch": 1.1360631017283378,
+      "grad_norm": 15.612197475543613,
+      "learning_rate": 2e-06,
+      "loss": 0.2858,
+      "step": 4897
+    },
+    {
+      "epoch": 1.1362950933766385,
+      "grad_norm": 8.466076316561889,
+      "learning_rate": 2e-06,
+      "loss": 0.1672,
+      "step": 4898
+    },
+    {
+      "epoch": 1.1365270850249392,
+      "grad_norm": 14.876673188857252,
+      "learning_rate": 2e-06,
+      "loss": 0.3044,
+      "step": 4899
+    },
+    {
+      "epoch": 1.1367590766732398,
+      "grad_norm": 11.133115228172553,
+      "learning_rate": 2e-06,
+      "loss": 0.1983,
+      "step": 4900
+    },
+    {
+      "epoch": 1.1369910683215405,
+      "grad_norm": 8.143066059932629,
+      "learning_rate": 2e-06,
+      "loss": 0.1884,
+      "step": 4901
+    },
+    {
+      "epoch": 1.1372230599698412,
+      "grad_norm": 17.890975380490254,
+      "learning_rate": 2e-06,
+      "loss": 0.2398,
+      "step": 4902
+    },
+    {
+      "epoch": 1.1374550516181419,
+      "grad_norm": 15.559934780180642,
+      "learning_rate": 2e-06,
+      "loss": 0.2069,
+      "step": 4903
+    },
+    {
+      "epoch": 1.1376870432664423,
+      "grad_norm": 21.19586601560053,
+      "learning_rate": 2e-06,
+      "loss": 0.3452,
+      "step": 4904
+    },
+    {
+      "epoch": 1.137919034914743,
+      "grad_norm": 18.280108012771613,
+      "learning_rate": 2e-06,
+      "loss": 0.3406,
+      "step": 4905
+    },
+    {
+      "epoch": 1.1381510265630437,
+      "grad_norm": 13.982547036255285,
+      "learning_rate": 2e-06,
+      "loss": 0.2576,
+      "step": 4906
+    },
+    {
+      "epoch": 1.1383830182113444,
+      "grad_norm": 11.804514631444446,
+      "learning_rate": 2e-06,
+      "loss": 0.229,
+      "step": 4907
+    },
+    {
+      "epoch": 1.138615009859645,
+      "grad_norm": 19.541278639122623,
+      "learning_rate": 2e-06,
+      "loss": 0.2622,
+      "step": 4908
+    },
+    {
+      "epoch": 1.1388470015079457,
+      "grad_norm": 14.760808289312461,
+      "learning_rate": 2e-06,
+      "loss": 0.2904,
+      "step": 4909
+    },
+    {
+      "epoch": 1.1390789931562464,
+      "grad_norm": 19.011030729462316,
+      "learning_rate": 2e-06,
+      "loss": 0.2777,
+      "step": 4910
+    },
+    {
+      "epoch": 1.139310984804547,
+      "grad_norm": 6.137233877763751,
+      "learning_rate": 2e-06,
+      "loss": 0.1433,
+      "step": 4911
+    },
+    {
+      "epoch": 1.1395429764528477,
+      "grad_norm": 14.918947462083826,
+      "learning_rate": 2e-06,
+      "loss": 0.2392,
+      "step": 4912
+    },
+    {
+      "epoch": 1.1397749681011484,
+      "grad_norm": 14.772137907592853,
+      "learning_rate": 2e-06,
+      "loss": 0.1439,
+      "step": 4913
+    },
+    {
+      "epoch": 1.140006959749449,
+      "grad_norm": 14.608881128373772,
+      "learning_rate": 2e-06,
+      "loss": 0.2408,
+      "step": 4914
+    },
+    {
+      "epoch": 1.1402389513977498,
+      "grad_norm": 9.58689032170824,
+      "learning_rate": 2e-06,
+      "loss": 0.1699,
+      "step": 4915
+    },
+    {
+      "epoch": 1.1404709430460502,
+      "grad_norm": 11.368371089686674,
+      "learning_rate": 2e-06,
+      "loss": 0.1777,
+      "step": 4916
+    },
+    {
+      "epoch": 1.140702934694351,
+      "grad_norm": 14.5766053931017,
+      "learning_rate": 2e-06,
+      "loss": 0.1859,
+      "step": 4917
+    },
+    {
+      "epoch": 1.1409349263426516,
+      "grad_norm": 9.122800332401326,
+      "learning_rate": 2e-06,
+      "loss": 0.1833,
+      "step": 4918
+    },
+    {
+      "epoch": 1.1411669179909523,
+      "grad_norm": 9.61504687863536,
+      "learning_rate": 2e-06,
+      "loss": 0.1986,
+      "step": 4919
+    },
+    {
+      "epoch": 1.141398909639253,
+      "grad_norm": 21.523746206148463,
+      "learning_rate": 2e-06,
+      "loss": 0.3443,
+      "step": 4920
+    },
+    {
+      "epoch": 1.1416309012875536,
+      "grad_norm": 14.444470032653033,
+      "learning_rate": 2e-06,
+      "loss": 0.1923,
+      "step": 4921
+    },
+    {
+      "epoch": 1.1418628929358543,
+      "grad_norm": 18.236309948457443,
+      "learning_rate": 2e-06,
+      "loss": 0.2983,
+      "step": 4922
+    },
+    {
+      "epoch": 1.142094884584155,
+      "grad_norm": 13.92540097571774,
+      "learning_rate": 2e-06,
+      "loss": 0.193,
+      "step": 4923
+    },
+    {
+      "epoch": 1.1423268762324557,
+      "grad_norm": 24.180937619439668,
+      "learning_rate": 2e-06,
+      "loss": 0.3859,
+      "step": 4924
+    },
+    {
+      "epoch": 1.1425588678807563,
+      "grad_norm": 15.437820412405143,
+      "learning_rate": 2e-06,
+      "loss": 0.268,
+      "step": 4925
+    },
+    {
+      "epoch": 1.142790859529057,
+      "grad_norm": 13.582142530612812,
+      "learning_rate": 2e-06,
+      "loss": 0.2587,
+      "step": 4926
+    },
+    {
+      "epoch": 1.1430228511773577,
+      "grad_norm": 15.043145299085626,
+      "learning_rate": 2e-06,
+      "loss": 0.2439,
+      "step": 4927
+    },
+    {
+      "epoch": 1.1432548428256584,
+      "grad_norm": 16.86067192142225,
+      "learning_rate": 2e-06,
+      "loss": 0.2741,
+      "step": 4928
+    },
+    {
+      "epoch": 1.143486834473959,
+      "grad_norm": 18.408278721917398,
+      "learning_rate": 2e-06,
+      "loss": 0.2853,
+      "step": 4929
+    },
+    {
+      "epoch": 1.1437188261222595,
+      "grad_norm": 11.034525892499884,
+      "learning_rate": 2e-06,
+      "loss": 0.2126,
+      "step": 4930
+    },
+    {
+      "epoch": 1.1439508177705602,
+      "grad_norm": 8.690261559663353,
+      "learning_rate": 2e-06,
+      "loss": 0.1887,
+      "step": 4931
+    },
+    {
+      "epoch": 1.1441828094188609,
+      "grad_norm": 10.915617366321268,
+      "learning_rate": 2e-06,
+      "loss": 0.2147,
+      "step": 4932
+    },
+    {
+      "epoch": 1.1444148010671615,
+      "grad_norm": 6.999836918162795,
+      "learning_rate": 2e-06,
+      "loss": 0.2213,
+      "step": 4933
+    },
+    {
+      "epoch": 1.1446467927154622,
+      "grad_norm": 12.861398225532644,
+      "learning_rate": 2e-06,
+      "loss": 0.2748,
+      "step": 4934
+    },
+    {
+      "epoch": 1.144878784363763,
+      "grad_norm": 21.73627815965974,
+      "learning_rate": 2e-06,
+      "loss": 0.2468,
+      "step": 4935
+    },
+    {
+      "epoch": 1.1451107760120636,
+      "grad_norm": 5.840407410752199,
+      "learning_rate": 2e-06,
+      "loss": 0.1333,
+      "step": 4936
+    },
+    {
+      "epoch": 1.1453427676603642,
+      "grad_norm": 13.66647412745888,
+      "learning_rate": 2e-06,
+      "loss": 0.254,
+      "step": 4937
+    },
+    {
+      "epoch": 1.145574759308665,
+      "grad_norm": 10.955800205327387,
+      "learning_rate": 2e-06,
+      "loss": 0.2701,
+      "step": 4938
+    },
+    {
+      "epoch": 1.1458067509569656,
+      "grad_norm": 18.279641821552023,
+      "learning_rate": 2e-06,
+      "loss": 0.3168,
+      "step": 4939
+    },
+    {
+      "epoch": 1.1460387426052663,
+      "grad_norm": 13.678123638896912,
+      "learning_rate": 2e-06,
+      "loss": 0.1945,
+      "step": 4940
+    },
+    {
+      "epoch": 1.146270734253567,
+      "grad_norm": 20.690435466350603,
+      "learning_rate": 2e-06,
+      "loss": 0.3321,
+      "step": 4941
+    },
+    {
+      "epoch": 1.1465027259018674,
+      "grad_norm": 11.352089339007339,
+      "learning_rate": 2e-06,
+      "loss": 0.2042,
+      "step": 4942
+    },
+    {
+      "epoch": 1.146734717550168,
+      "grad_norm": 15.102901486629841,
+      "learning_rate": 2e-06,
+      "loss": 0.1913,
+      "step": 4943
+    },
+    {
+      "epoch": 1.1469667091984688,
+      "grad_norm": 16.465122437217744,
+      "learning_rate": 2e-06,
+      "loss": 0.2494,
+      "step": 4944
+    },
+    {
+      "epoch": 1.1471987008467694,
+      "grad_norm": 7.706423255130726,
+      "learning_rate": 2e-06,
+      "loss": 0.1918,
+      "step": 4945
+    },
+    {
+      "epoch": 1.1474306924950701,
+      "grad_norm": 13.249898609853025,
+      "learning_rate": 2e-06,
+      "loss": 0.2462,
+      "step": 4946
+    },
+    {
+      "epoch": 1.1476626841433708,
+      "grad_norm": 21.2608644464412,
+      "learning_rate": 2e-06,
+      "loss": 0.3209,
+      "step": 4947
+    },
+    {
+      "epoch": 1.1478946757916715,
+      "grad_norm": 8.674472393729499,
+      "learning_rate": 2e-06,
+      "loss": 0.1809,
+      "step": 4948
+    },
+    {
+      "epoch": 1.1481266674399722,
+      "grad_norm": 12.503974947717918,
+      "learning_rate": 2e-06,
+      "loss": 0.3836,
+      "step": 4949
+    },
+    {
+      "epoch": 1.1483586590882728,
+      "grad_norm": 20.12930606562257,
+      "learning_rate": 2e-06,
+      "loss": 0.1928,
+      "step": 4950
+    },
+    {
+      "epoch": 1.1485906507365735,
+      "grad_norm": 20.94522005014584,
+      "learning_rate": 2e-06,
+      "loss": 0.2479,
+      "step": 4951
+    },
+    {
+      "epoch": 1.1488226423848742,
+      "grad_norm": 24.26873660368573,
+      "learning_rate": 2e-06,
+      "loss": 0.2137,
+      "step": 4952
+    },
+    {
+      "epoch": 1.1490546340331749,
+      "grad_norm": 10.406117471367917,
+      "learning_rate": 2e-06,
+      "loss": 0.2668,
+      "step": 4953
+    },
+    {
+      "epoch": 1.1492866256814755,
+      "grad_norm": 11.010379737891737,
+      "learning_rate": 2e-06,
+      "loss": 0.1913,
+      "step": 4954
+    },
+    {
+      "epoch": 1.1495186173297762,
+      "grad_norm": 12.299171427398806,
+      "learning_rate": 2e-06,
+      "loss": 0.1321,
+      "step": 4955
+    },
+    {
+      "epoch": 1.149750608978077,
+      "grad_norm": 17.37082274280718,
+      "learning_rate": 2e-06,
+      "loss": 0.3217,
+      "step": 4956
+    },
+    {
+      "epoch": 1.1499826006263774,
+      "grad_norm": 14.921465025425027,
+      "learning_rate": 2e-06,
+      "loss": 0.3425,
+      "step": 4957
+    },
+    {
+      "epoch": 1.150214592274678,
+      "grad_norm": 17.521596392801534,
+      "learning_rate": 2e-06,
+      "loss": 0.3958,
+      "step": 4958
+    },
+    {
+      "epoch": 1.1504465839229787,
+      "grad_norm": 15.05574840704271,
+      "learning_rate": 2e-06,
+      "loss": 0.2477,
+      "step": 4959
+    },
+    {
+      "epoch": 1.1506785755712794,
+      "grad_norm": 9.486800586659157,
+      "learning_rate": 2e-06,
+      "loss": 0.2234,
+      "step": 4960
+    },
+    {
+      "epoch": 1.15091056721958,
+      "grad_norm": 10.712302390953361,
+      "learning_rate": 2e-06,
+      "loss": 0.2072,
+      "step": 4961
+    },
+    {
+      "epoch": 1.1511425588678807,
+      "grad_norm": 10.90126140152789,
+      "learning_rate": 2e-06,
+      "loss": 0.3454,
+      "step": 4962
+    },
+    {
+      "epoch": 1.1513745505161814,
+      "grad_norm": 20.62003729182546,
+      "learning_rate": 2e-06,
+      "loss": 0.3847,
+      "step": 4963
+    },
+    {
+      "epoch": 1.151606542164482,
+      "grad_norm": 5.952093748397747,
+      "learning_rate": 2e-06,
+      "loss": 0.1854,
+      "step": 4964
+    },
+    {
+      "epoch": 1.1518385338127828,
+      "grad_norm": 21.558466694925713,
+      "learning_rate": 2e-06,
+      "loss": 0.3083,
+      "step": 4965
+    },
+    {
+      "epoch": 1.1520705254610835,
+      "grad_norm": 16.098757853215655,
+      "learning_rate": 2e-06,
+      "loss": 0.2644,
+      "step": 4966
+    },
+    {
+      "epoch": 1.1523025171093841,
+      "grad_norm": 10.083478843015879,
+      "learning_rate": 2e-06,
+      "loss": 0.293,
+      "step": 4967
+    },
+    {
+      "epoch": 1.1525345087576848,
+      "grad_norm": 14.425534089566499,
+      "learning_rate": 2e-06,
+      "loss": 0.2881,
+      "step": 4968
+    },
+    {
+      "epoch": 1.1527665004059853,
+      "grad_norm": 18.679608904856032,
+      "learning_rate": 2e-06,
+      "loss": 0.3362,
+      "step": 4969
+    },
+    {
+      "epoch": 1.152998492054286,
+      "grad_norm": 21.405848462538046,
+      "learning_rate": 2e-06,
+      "loss": 0.2952,
+      "step": 4970
+    },
+    {
+      "epoch": 1.1532304837025866,
+      "grad_norm": 18.534218012077417,
+      "learning_rate": 2e-06,
+      "loss": 0.2789,
+      "step": 4971
+    },
+    {
+      "epoch": 1.1534624753508873,
+      "grad_norm": 16.90363676355196,
+      "learning_rate": 2e-06,
+      "loss": 0.2342,
+      "step": 4972
+    },
+    {
+      "epoch": 1.153694466999188,
+      "grad_norm": 20.269954202409547,
+      "learning_rate": 2e-06,
+      "loss": 0.2938,
+      "step": 4973
+    },
+    {
+      "epoch": 1.1539264586474887,
+      "grad_norm": 14.655589208363546,
+      "learning_rate": 2e-06,
+      "loss": 0.2959,
+      "step": 4974
+    },
+    {
+      "epoch": 1.1541584502957893,
+      "grad_norm": 16.179437731609507,
+      "learning_rate": 2e-06,
+      "loss": 0.3986,
+      "step": 4975
+    },
+    {
+      "epoch": 1.15439044194409,
+      "grad_norm": 17.471580867845255,
+      "learning_rate": 2e-06,
+      "loss": 0.3414,
+      "step": 4976
+    },
+    {
+      "epoch": 1.1546224335923907,
+      "grad_norm": 15.168744371031757,
+      "learning_rate": 2e-06,
+      "loss": 0.2526,
+      "step": 4977
+    },
+    {
+      "epoch": 1.1548544252406914,
+      "grad_norm": 14.814605235511708,
+      "learning_rate": 2e-06,
+      "loss": 0.297,
+      "step": 4978
+    },
+    {
+      "epoch": 1.155086416888992,
+      "grad_norm": 5.064559732869771,
+      "learning_rate": 2e-06,
+      "loss": 0.1687,
+      "step": 4979
+    },
+    {
+      "epoch": 1.1553184085372927,
+      "grad_norm": 22.91198371044645,
+      "learning_rate": 2e-06,
+      "loss": 0.3917,
+      "step": 4980
+    },
+    {
+      "epoch": 1.1555504001855934,
+      "grad_norm": 10.770969937868816,
+      "learning_rate": 2e-06,
+      "loss": 0.2501,
+      "step": 4981
+    },
+    {
+      "epoch": 1.155782391833894,
+      "grad_norm": 15.982737134996356,
+      "learning_rate": 2e-06,
+      "loss": 0.2392,
+      "step": 4982
+    },
+    {
+      "epoch": 1.1560143834821948,
+      "grad_norm": 16.99469144176511,
+      "learning_rate": 2e-06,
+      "loss": 0.3181,
+      "step": 4983
+    },
+    {
+      "epoch": 1.1562463751304952,
+      "grad_norm": 25.361483749720282,
+      "learning_rate": 2e-06,
+      "loss": 0.2793,
+      "step": 4984
+    },
+    {
+      "epoch": 1.156478366778796,
+      "grad_norm": 11.43453219225867,
+      "learning_rate": 2e-06,
+      "loss": 0.2394,
+      "step": 4985
+    },
+    {
+      "epoch": 1.1567103584270966,
+      "grad_norm": 17.527726479771758,
+      "learning_rate": 2e-06,
+      "loss": 0.3313,
+      "step": 4986
+    },
+    {
+      "epoch": 1.1569423500753973,
+      "grad_norm": 23.029281340142415,
+      "learning_rate": 2e-06,
+      "loss": 0.2906,
+      "step": 4987
+    },
+    {
+      "epoch": 1.157174341723698,
+      "grad_norm": 15.859396445352443,
+      "learning_rate": 2e-06,
+      "loss": 0.1847,
+      "step": 4988
+    },
+    {
+      "epoch": 1.1574063333719986,
+      "grad_norm": 18.52207039945924,
+      "learning_rate": 2e-06,
+      "loss": 0.232,
+      "step": 4989
+    },
+    {
+      "epoch": 1.1576383250202993,
+      "grad_norm": 18.5323290413136,
+      "learning_rate": 2e-06,
+      "loss": 0.2638,
+      "step": 4990
+    },
+    {
+      "epoch": 1.1578703166686,
+      "grad_norm": 11.726869560068051,
+      "learning_rate": 2e-06,
+      "loss": 0.2262,
+      "step": 4991
+    },
+    {
+      "epoch": 1.1581023083169006,
+      "grad_norm": 10.325741206233513,
+      "learning_rate": 2e-06,
+      "loss": 0.2112,
+      "step": 4992
+    },
+    {
+      "epoch": 1.1583342999652013,
+      "grad_norm": 18.185472349818276,
+      "learning_rate": 2e-06,
+      "loss": 0.2729,
+      "step": 4993
+    },
+    {
+      "epoch": 1.158566291613502,
+      "grad_norm": 16.719998948739807,
+      "learning_rate": 2e-06,
+      "loss": 0.2897,
+      "step": 4994
+    },
+    {
+      "epoch": 1.1587982832618025,
+      "grad_norm": 16.404411179286353,
+      "learning_rate": 2e-06,
+      "loss": 0.2828,
+      "step": 4995
+    },
+    {
+      "epoch": 1.1590302749101031,
+      "grad_norm": 15.826079148596243,
+      "learning_rate": 2e-06,
+      "loss": 0.1987,
+      "step": 4996
+    },
+    {
+      "epoch": 1.1592622665584038,
+      "grad_norm": 23.52084532680064,
+      "learning_rate": 2e-06,
+      "loss": 0.3933,
+      "step": 4997
+    },
+    {
+      "epoch": 1.1594942582067045,
+      "grad_norm": 19.727342270029716,
+      "learning_rate": 2e-06,
+      "loss": 0.3705,
+      "step": 4998
+    },
+    {
+      "epoch": 1.1597262498550052,
+      "grad_norm": 8.532466128418308,
+      "learning_rate": 2e-06,
+      "loss": 0.2316,
+      "step": 4999
+    },
+    {
+      "epoch": 1.1599582415033058,
+      "grad_norm": 18.545541497081892,
+      "learning_rate": 2e-06,
+      "loss": 0.2471,
+      "step": 5000
+    },
+    {
+      "epoch": 1.1601902331516065,
+      "grad_norm": 7.39760627147255,
+      "learning_rate": 2e-06,
+      "loss": 0.1612,
+      "step": 5001
+    },
+    {
+      "epoch": 1.1604222247999072,
+      "grad_norm": 10.978935659260085,
+      "learning_rate": 2e-06,
+      "loss": 0.2346,
+      "step": 5002
+    },
+    {
+      "epoch": 1.1606542164482079,
+      "grad_norm": 11.515475457268554,
+      "learning_rate": 2e-06,
+      "loss": 0.2053,
+      "step": 5003
+    },
+    {
+      "epoch": 1.1608862080965086,
+      "grad_norm": 6.859243508484112,
+      "learning_rate": 2e-06,
+      "loss": 0.1754,
+      "step": 5004
+    },
+    {
+      "epoch": 1.1611181997448092,
+      "grad_norm": 22.30127623099719,
+      "learning_rate": 2e-06,
+      "loss": 0.2655,
+      "step": 5005
+    },
+    {
+      "epoch": 1.16135019139311,
+      "grad_norm": 18.018850253614218,
+      "learning_rate": 2e-06,
+      "loss": 0.2342,
+      "step": 5006
+    },
+    {
+      "epoch": 1.1615821830414106,
+      "grad_norm": 14.185858142945714,
+      "learning_rate": 2e-06,
+      "loss": 0.2357,
+      "step": 5007
+    },
+    {
+      "epoch": 1.1618141746897113,
+      "grad_norm": 13.850770859563564,
+      "learning_rate": 2e-06,
+      "loss": 0.2282,
+      "step": 5008
+    },
+    {
+      "epoch": 1.162046166338012,
+      "grad_norm": 20.168032983436802,
+      "learning_rate": 2e-06,
+      "loss": 0.3488,
+      "step": 5009
+    },
+    {
+      "epoch": 1.1622781579863124,
+      "grad_norm": 14.213162878346074,
+      "learning_rate": 2e-06,
+      "loss": 0.1906,
+      "step": 5010
+    },
+    {
+      "epoch": 1.162510149634613,
+      "grad_norm": 18.527199878180998,
+      "learning_rate": 2e-06,
+      "loss": 0.2876,
+      "step": 5011
+    },
+    {
+      "epoch": 1.1627421412829138,
+      "grad_norm": 21.948041045089763,
+      "learning_rate": 2e-06,
+      "loss": 0.2023,
+      "step": 5012
+    },
+    {
+      "epoch": 1.1629741329312144,
+      "grad_norm": 18.20419217348716,
+      "learning_rate": 2e-06,
+      "loss": 0.2678,
+      "step": 5013
+    },
+    {
+      "epoch": 1.163206124579515,
+      "grad_norm": 7.939425692141144,
+      "learning_rate": 2e-06,
+      "loss": 0.2372,
+      "step": 5014
+    },
+    {
+      "epoch": 1.1634381162278158,
+      "grad_norm": 12.672603130303628,
+      "learning_rate": 2e-06,
+      "loss": 0.2832,
+      "step": 5015
+    },
+    {
+      "epoch": 1.1636701078761165,
+      "grad_norm": 10.981135126370553,
+      "learning_rate": 2e-06,
+      "loss": 0.3249,
+      "step": 5016
+    },
+    {
+      "epoch": 1.1639020995244171,
+      "grad_norm": 15.315053555166035,
+      "learning_rate": 2e-06,
+      "loss": 0.2706,
+      "step": 5017
+    },
+    {
+      "epoch": 1.1641340911727178,
+      "grad_norm": 7.395346913250289,
+      "learning_rate": 2e-06,
+      "loss": 0.1825,
+      "step": 5018
+    },
+    {
+      "epoch": 1.1643660828210185,
+      "grad_norm": 25.789706744104237,
+      "learning_rate": 2e-06,
+      "loss": 0.2779,
+      "step": 5019
+    },
+    {
+      "epoch": 1.1645980744693192,
+      "grad_norm": 13.929896223441505,
+      "learning_rate": 2e-06,
+      "loss": 0.3325,
+      "step": 5020
+    },
+    {
+      "epoch": 1.1648300661176199,
+      "grad_norm": 20.150725102614853,
+      "learning_rate": 2e-06,
+      "loss": 0.2426,
+      "step": 5021
+    },
+    {
+      "epoch": 1.1650620577659203,
+      "grad_norm": 16.45818500299967,
+      "learning_rate": 2e-06,
+      "loss": 0.1771,
+      "step": 5022
+    },
+    {
+      "epoch": 1.165294049414221,
+      "grad_norm": 15.739444570190075,
+      "learning_rate": 2e-06,
+      "loss": 0.2365,
+      "step": 5023
+    },
+    {
+      "epoch": 1.1655260410625217,
+      "grad_norm": 10.570517956343688,
+      "learning_rate": 2e-06,
+      "loss": 0.2052,
+      "step": 5024
+    },
+    {
+      "epoch": 1.1657580327108223,
+      "grad_norm": 13.3629168062596,
+      "learning_rate": 2e-06,
+      "loss": 0.2241,
+      "step": 5025
+    },
+    {
+      "epoch": 1.165990024359123,
+      "grad_norm": 18.754419599758915,
+      "learning_rate": 2e-06,
+      "loss": 0.2735,
+      "step": 5026
+    },
+    {
+      "epoch": 1.1662220160074237,
+      "grad_norm": 19.664815170672313,
+      "learning_rate": 2e-06,
+      "loss": 0.2899,
+      "step": 5027
+    },
+    {
+      "epoch": 1.1664540076557244,
+      "grad_norm": 9.366801142256714,
+      "learning_rate": 2e-06,
+      "loss": 0.23,
+      "step": 5028
+    },
+    {
+      "epoch": 1.166685999304025,
+      "grad_norm": 14.935637719397329,
+      "learning_rate": 2e-06,
+      "loss": 0.2281,
+      "step": 5029
+    },
+    {
+      "epoch": 1.1669179909523257,
+      "grad_norm": 24.75933529551093,
+      "learning_rate": 2e-06,
+      "loss": 0.2572,
+      "step": 5030
+    },
+    {
+      "epoch": 1.1671499826006264,
+      "grad_norm": 14.730901640082049,
+      "learning_rate": 2e-06,
+      "loss": 0.3131,
+      "step": 5031
+    },
+    {
+      "epoch": 1.167381974248927,
+      "grad_norm": 5.845621968031191,
+      "learning_rate": 2e-06,
+      "loss": 0.103,
+      "step": 5032
+    },
+    {
+      "epoch": 1.1676139658972278,
+      "grad_norm": 15.167261756957625,
+      "learning_rate": 2e-06,
+      "loss": 0.2014,
+      "step": 5033
+    },
+    {
+      "epoch": 1.1678459575455284,
+      "grad_norm": 16.816500981299157,
+      "learning_rate": 2e-06,
+      "loss": 0.3123,
+      "step": 5034
+    },
+    {
+      "epoch": 1.1680779491938291,
+      "grad_norm": 12.555567932873625,
+      "learning_rate": 2e-06,
+      "loss": 0.209,
+      "step": 5035
+    },
+    {
+      "epoch": 1.1683099408421298,
+      "grad_norm": 15.661823316978516,
+      "learning_rate": 2e-06,
+      "loss": 0.3003,
+      "step": 5036
+    },
+    {
+      "epoch": 1.1685419324904303,
+      "grad_norm": 7.488072913478131,
+      "learning_rate": 2e-06,
+      "loss": 0.1905,
+      "step": 5037
+    },
+    {
+      "epoch": 1.168773924138731,
+      "grad_norm": 15.02359869467387,
+      "learning_rate": 2e-06,
+      "loss": 0.2309,
+      "step": 5038
+    },
+    {
+      "epoch": 1.1690059157870316,
+      "grad_norm": 14.78971080478935,
+      "learning_rate": 2e-06,
+      "loss": 0.2297,
+      "step": 5039
+    },
+    {
+      "epoch": 1.1692379074353323,
+      "grad_norm": 20.143819338315392,
+      "learning_rate": 2e-06,
+      "loss": 0.3366,
+      "step": 5040
+    },
+    {
+      "epoch": 1.169469899083633,
+      "grad_norm": 20.31800677163891,
+      "learning_rate": 2e-06,
+      "loss": 0.3499,
+      "step": 5041
+    },
+    {
+      "epoch": 1.1697018907319336,
+      "grad_norm": 18.584130400204625,
+      "learning_rate": 2e-06,
+      "loss": 0.2986,
+      "step": 5042
+    },
+    {
+      "epoch": 1.1699338823802343,
+      "grad_norm": 9.014805971449672,
+      "learning_rate": 2e-06,
+      "loss": 0.2106,
+      "step": 5043
+    },
+    {
+      "epoch": 1.170165874028535,
+      "grad_norm": 15.855667851271633,
+      "learning_rate": 2e-06,
+      "loss": 0.3955,
+      "step": 5044
+    },
+    {
+      "epoch": 1.1703978656768357,
+      "grad_norm": 16.573583773977177,
+      "learning_rate": 2e-06,
+      "loss": 0.3397,
+      "step": 5045
+    },
+    {
+      "epoch": 1.1706298573251364,
+      "grad_norm": 38.35167617624872,
+      "learning_rate": 2e-06,
+      "loss": 0.2736,
+      "step": 5046
+    },
+    {
+      "epoch": 1.170861848973437,
+      "grad_norm": 22.72179482763607,
+      "learning_rate": 2e-06,
+      "loss": 0.3405,
+      "step": 5047
+    },
+    {
+      "epoch": 1.1710938406217377,
+      "grad_norm": 12.249486716090715,
+      "learning_rate": 2e-06,
+      "loss": 0.1978,
+      "step": 5048
+    },
+    {
+      "epoch": 1.1713258322700382,
+      "grad_norm": 9.266429827972644,
+      "learning_rate": 2e-06,
+      "loss": 0.2376,
+      "step": 5049
+    },
+    {
+      "epoch": 1.1715578239183388,
+      "grad_norm": 18.193981352814255,
+      "learning_rate": 2e-06,
+      "loss": 0.3892,
+      "step": 5050
+    },
+    {
+      "epoch": 1.1717898155666395,
+      "grad_norm": 15.126893755477953,
+      "learning_rate": 2e-06,
+      "loss": 0.2697,
+      "step": 5051
+    },
+    {
+      "epoch": 1.1720218072149402,
+      "grad_norm": 11.528059200038399,
+      "learning_rate": 2e-06,
+      "loss": 0.2112,
+      "step": 5052
+    },
+    {
+      "epoch": 1.1722537988632409,
+      "grad_norm": 5.696819551783051,
+      "learning_rate": 2e-06,
+      "loss": 0.1733,
+      "step": 5053
+    },
+    {
+      "epoch": 1.1724857905115416,
+      "grad_norm": 21.68939285127513,
+      "learning_rate": 2e-06,
+      "loss": 0.238,
+      "step": 5054
+    },
+    {
+      "epoch": 1.1727177821598422,
+      "grad_norm": 11.54426577532738,
+      "learning_rate": 2e-06,
+      "loss": 0.2835,
+      "step": 5055
+    },
+    {
+      "epoch": 1.172949773808143,
+      "grad_norm": 16.36253134447032,
+      "learning_rate": 2e-06,
+      "loss": 0.3236,
+      "step": 5056
+    },
+    {
+      "epoch": 1.1731817654564436,
+      "grad_norm": 8.142562005217773,
+      "learning_rate": 2e-06,
+      "loss": 0.212,
+      "step": 5057
+    },
+    {
+      "epoch": 1.1734137571047443,
+      "grad_norm": 12.615320983816575,
+      "learning_rate": 2e-06,
+      "loss": 0.2282,
+      "step": 5058
+    },
+    {
+      "epoch": 1.173645748753045,
+      "grad_norm": 12.514326523956044,
+      "learning_rate": 2e-06,
+      "loss": 0.1682,
+      "step": 5059
+    },
+    {
+      "epoch": 1.1738777404013456,
+      "grad_norm": 12.633760376909049,
+      "learning_rate": 2e-06,
+      "loss": 0.2375,
+      "step": 5060
+    },
+    {
+      "epoch": 1.1741097320496463,
+      "grad_norm": 18.721158363810936,
+      "learning_rate": 2e-06,
+      "loss": 0.3598,
+      "step": 5061
+    },
+    {
+      "epoch": 1.174341723697947,
+      "grad_norm": 19.54568283304831,
+      "learning_rate": 2e-06,
+      "loss": 0.3662,
+      "step": 5062
+    },
+    {
+      "epoch": 1.1745737153462477,
+      "grad_norm": 56.5685419514945,
+      "learning_rate": 2e-06,
+      "loss": 0.1762,
+      "step": 5063
+    },
+    {
+      "epoch": 1.1748057069945481,
+      "grad_norm": 20.023831440292504,
+      "learning_rate": 2e-06,
+      "loss": 0.2206,
+      "step": 5064
+    },
+    {
+      "epoch": 1.1750376986428488,
+      "grad_norm": 11.468361701197614,
+      "learning_rate": 2e-06,
+      "loss": 0.2195,
+      "step": 5065
+    },
+    {
+      "epoch": 1.1752696902911495,
+      "grad_norm": 12.883223148041303,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 5066
+    },
+    {
+      "epoch": 1.1755016819394501,
+      "grad_norm": 12.738172382017016,
+      "learning_rate": 2e-06,
+      "loss": 0.3331,
+      "step": 5067
+    },
+    {
+      "epoch": 1.1757336735877508,
+      "grad_norm": 15.362041684758266,
+      "learning_rate": 2e-06,
+      "loss": 0.3555,
+      "step": 5068
+    },
+    {
+      "epoch": 1.1759656652360515,
+      "grad_norm": 12.414739283145341,
+      "learning_rate": 2e-06,
+      "loss": 0.1942,
+      "step": 5069
+    },
+    {
+      "epoch": 1.1761976568843522,
+      "grad_norm": 17.904270630927922,
+      "learning_rate": 2e-06,
+      "loss": 0.2559,
+      "step": 5070
+    },
+    {
+      "epoch": 1.1764296485326529,
+      "grad_norm": 13.538195283396922,
+      "learning_rate": 2e-06,
+      "loss": 0.2464,
+      "step": 5071
+    },
+    {
+      "epoch": 1.1766616401809535,
+      "grad_norm": 13.422021738421487,
+      "learning_rate": 2e-06,
+      "loss": 0.2749,
+      "step": 5072
+    },
+    {
+      "epoch": 1.1768936318292542,
+      "grad_norm": 22.53580094618496,
+      "learning_rate": 2e-06,
+      "loss": 0.243,
+      "step": 5073
+    },
+    {
+      "epoch": 1.177125623477555,
+      "grad_norm": 11.55881963636086,
+      "learning_rate": 2e-06,
+      "loss": 0.2796,
+      "step": 5074
+    },
+    {
+      "epoch": 1.1773576151258554,
+      "grad_norm": 10.049045185397313,
+      "learning_rate": 2e-06,
+      "loss": 0.1981,
+      "step": 5075
+    },
+    {
+      "epoch": 1.177589606774156,
+      "grad_norm": 22.65546336238449,
+      "learning_rate": 2e-06,
+      "loss": 0.2988,
+      "step": 5076
+    },
+    {
+      "epoch": 1.1778215984224567,
+      "grad_norm": 13.048630873201166,
+      "learning_rate": 2e-06,
+      "loss": 0.2681,
+      "step": 5077
+    },
+    {
+      "epoch": 1.1780535900707574,
+      "grad_norm": 13.900332711139294,
+      "learning_rate": 2e-06,
+      "loss": 0.2521,
+      "step": 5078
+    },
+    {
+      "epoch": 1.178285581719058,
+      "grad_norm": 11.32114169357351,
+      "learning_rate": 2e-06,
+      "loss": 0.2488,
+      "step": 5079
+    },
+    {
+      "epoch": 1.1785175733673587,
+      "grad_norm": 24.06534770278926,
+      "learning_rate": 2e-06,
+      "loss": 0.4219,
+      "step": 5080
+    },
+    {
+      "epoch": 1.1787495650156594,
+      "grad_norm": 20.324166681833166,
+      "learning_rate": 2e-06,
+      "loss": 0.249,
+      "step": 5081
+    },
+    {
+      "epoch": 1.17898155666396,
+      "grad_norm": 10.903753798413435,
+      "learning_rate": 2e-06,
+      "loss": 0.2103,
+      "step": 5082
+    },
+    {
+      "epoch": 1.1792135483122608,
+      "grad_norm": 29.388568864614136,
+      "learning_rate": 2e-06,
+      "loss": 0.4449,
+      "step": 5083
+    },
+    {
+      "epoch": 1.1794455399605615,
+      "grad_norm": 12.80463042132413,
+      "learning_rate": 2e-06,
+      "loss": 0.2395,
+      "step": 5084
+    },
+    {
+      "epoch": 1.1796775316088621,
+      "grad_norm": 21.144388392722167,
+      "learning_rate": 2e-06,
+      "loss": 0.3401,
+      "step": 5085
+    },
+    {
+      "epoch": 1.1799095232571628,
+      "grad_norm": 12.57877912456603,
+      "learning_rate": 2e-06,
+      "loss": 0.2803,
+      "step": 5086
+    },
+    {
+      "epoch": 1.1801415149054635,
+      "grad_norm": 12.110198711063557,
+      "learning_rate": 2e-06,
+      "loss": 0.192,
+      "step": 5087
+    },
+    {
+      "epoch": 1.1803735065537642,
+      "grad_norm": 9.479556009422616,
+      "learning_rate": 2e-06,
+      "loss": 0.2175,
+      "step": 5088
+    },
+    {
+      "epoch": 1.1806054982020648,
+      "grad_norm": 9.087753336527165,
+      "learning_rate": 2e-06,
+      "loss": 0.1777,
+      "step": 5089
+    },
+    {
+      "epoch": 1.1808374898503653,
+      "grad_norm": 16.533214839801566,
+      "learning_rate": 2e-06,
+      "loss": 0.2729,
+      "step": 5090
+    },
+    {
+      "epoch": 1.181069481498666,
+      "grad_norm": 20.021190380675282,
+      "learning_rate": 2e-06,
+      "loss": 0.3022,
+      "step": 5091
+    },
+    {
+      "epoch": 1.1813014731469667,
+      "grad_norm": 19.108612020028357,
+      "learning_rate": 2e-06,
+      "loss": 0.3042,
+      "step": 5092
+    },
+    {
+      "epoch": 1.1815334647952673,
+      "grad_norm": 13.03954712661226,
+      "learning_rate": 2e-06,
+      "loss": 0.1961,
+      "step": 5093
+    },
+    {
+      "epoch": 1.181765456443568,
+      "grad_norm": 13.956540532981451,
+      "learning_rate": 2e-06,
+      "loss": 0.277,
+      "step": 5094
+    },
+    {
+      "epoch": 1.1819974480918687,
+      "grad_norm": 16.498525718064315,
+      "learning_rate": 2e-06,
+      "loss": 0.1667,
+      "step": 5095
+    },
+    {
+      "epoch": 1.1822294397401694,
+      "grad_norm": 24.182252375905538,
+      "learning_rate": 2e-06,
+      "loss": 0.3496,
+      "step": 5096
+    },
+    {
+      "epoch": 1.18246143138847,
+      "grad_norm": 15.170383802626947,
+      "learning_rate": 2e-06,
+      "loss": 0.308,
+      "step": 5097
+    },
+    {
+      "epoch": 1.1826934230367707,
+      "grad_norm": 12.57381961517264,
+      "learning_rate": 2e-06,
+      "loss": 0.2234,
+      "step": 5098
+    },
+    {
+      "epoch": 1.1829254146850714,
+      "grad_norm": 17.774575808972216,
+      "learning_rate": 2e-06,
+      "loss": 0.3411,
+      "step": 5099
+    },
+    {
+      "epoch": 1.183157406333372,
+      "grad_norm": 12.241022832388117,
+      "learning_rate": 2e-06,
+      "loss": 0.3423,
+      "step": 5100
+    },
+    {
+      "epoch": 1.1833893979816728,
+      "grad_norm": 13.921320685454633,
+      "learning_rate": 2e-06,
+      "loss": 0.1748,
+      "step": 5101
+    },
+    {
+      "epoch": 1.1836213896299732,
+      "grad_norm": 11.853467367859825,
+      "learning_rate": 2e-06,
+      "loss": 0.17,
+      "step": 5102
+    },
+    {
+      "epoch": 1.1838533812782739,
+      "grad_norm": 14.718381089627021,
+      "learning_rate": 2e-06,
+      "loss": 0.2199,
+      "step": 5103
+    },
+    {
+      "epoch": 1.1840853729265746,
+      "grad_norm": 14.749166426120494,
+      "learning_rate": 2e-06,
+      "loss": 0.277,
+      "step": 5104
+    },
+    {
+      "epoch": 1.1843173645748752,
+      "grad_norm": 11.982207117697797,
+      "learning_rate": 2e-06,
+      "loss": 0.2446,
+      "step": 5105
+    },
+    {
+      "epoch": 1.184549356223176,
+      "grad_norm": 10.943136557718498,
+      "learning_rate": 2e-06,
+      "loss": 0.284,
+      "step": 5106
+    },
+    {
+      "epoch": 1.1847813478714766,
+      "grad_norm": 12.564714406437187,
+      "learning_rate": 2e-06,
+      "loss": 0.1975,
+      "step": 5107
+    },
+    {
+      "epoch": 1.1850133395197773,
+      "grad_norm": 16.674037184874972,
+      "learning_rate": 2e-06,
+      "loss": 0.2302,
+      "step": 5108
+    },
+    {
+      "epoch": 1.185245331168078,
+      "grad_norm": 21.112893619492684,
+      "learning_rate": 2e-06,
+      "loss": 0.4166,
+      "step": 5109
+    },
+    {
+      "epoch": 1.1854773228163786,
+      "grad_norm": 16.847327105118747,
+      "learning_rate": 2e-06,
+      "loss": 0.259,
+      "step": 5110
+    },
+    {
+      "epoch": 1.1857093144646793,
+      "grad_norm": 15.396429450961502,
+      "learning_rate": 2e-06,
+      "loss": 0.2732,
+      "step": 5111
+    },
+    {
+      "epoch": 1.18594130611298,
+      "grad_norm": 15.041230182302556,
+      "learning_rate": 2e-06,
+      "loss": 0.2603,
+      "step": 5112
+    },
+    {
+      "epoch": 1.1861732977612807,
+      "grad_norm": 11.679276575944929,
+      "learning_rate": 2e-06,
+      "loss": 0.2414,
+      "step": 5113
+    },
+    {
+      "epoch": 1.1864052894095813,
+      "grad_norm": 17.23626301583967,
+      "learning_rate": 2e-06,
+      "loss": 0.2465,
+      "step": 5114
+    },
+    {
+      "epoch": 1.186637281057882,
+      "grad_norm": 10.387629871060861,
+      "learning_rate": 2e-06,
+      "loss": 0.2685,
+      "step": 5115
+    },
+    {
+      "epoch": 1.1868692727061827,
+      "grad_norm": 13.583743595609374,
+      "learning_rate": 2e-06,
+      "loss": 0.2684,
+      "step": 5116
+    },
+    {
+      "epoch": 1.1871012643544832,
+      "grad_norm": 18.0518869085287,
+      "learning_rate": 2e-06,
+      "loss": 0.1706,
+      "step": 5117
+    },
+    {
+      "epoch": 1.1873332560027838,
+      "grad_norm": 8.292611403006696,
+      "learning_rate": 2e-06,
+      "loss": 0.1672,
+      "step": 5118
+    },
+    {
+      "epoch": 1.1875652476510845,
+      "grad_norm": 9.030629924098347,
+      "learning_rate": 2e-06,
+      "loss": 0.2098,
+      "step": 5119
+    },
+    {
+      "epoch": 1.1877972392993852,
+      "grad_norm": 22.190945386452597,
+      "learning_rate": 2e-06,
+      "loss": 0.4378,
+      "step": 5120
+    },
+    {
+      "epoch": 1.1880292309476859,
+      "grad_norm": 14.062678153543576,
+      "learning_rate": 2e-06,
+      "loss": 0.2778,
+      "step": 5121
+    },
+    {
+      "epoch": 1.1882612225959865,
+      "grad_norm": 9.700629503739886,
+      "learning_rate": 2e-06,
+      "loss": 0.3183,
+      "step": 5122
+    },
+    {
+      "epoch": 1.1884932142442872,
+      "grad_norm": 8.273146258477006,
+      "learning_rate": 2e-06,
+      "loss": 0.2306,
+      "step": 5123
+    },
+    {
+      "epoch": 1.188725205892588,
+      "grad_norm": 11.615305784614497,
+      "learning_rate": 2e-06,
+      "loss": 0.2241,
+      "step": 5124
+    },
+    {
+      "epoch": 1.1889571975408886,
+      "grad_norm": 9.89918448079035,
+      "learning_rate": 2e-06,
+      "loss": 0.1862,
+      "step": 5125
+    },
+    {
+      "epoch": 1.1891891891891893,
+      "grad_norm": 16.405532140936124,
+      "learning_rate": 2e-06,
+      "loss": 0.2831,
+      "step": 5126
+    },
+    {
+      "epoch": 1.18942118083749,
+      "grad_norm": 25.228975723271233,
+      "learning_rate": 2e-06,
+      "loss": 0.3914,
+      "step": 5127
+    },
+    {
+      "epoch": 1.1896531724857904,
+      "grad_norm": 18.17180925808904,
+      "learning_rate": 2e-06,
+      "loss": 0.2368,
+      "step": 5128
+    },
+    {
+      "epoch": 1.189885164134091,
+      "grad_norm": 8.534850939568784,
+      "learning_rate": 2e-06,
+      "loss": 0.2791,
+      "step": 5129
+    },
+    {
+      "epoch": 1.1901171557823917,
+      "grad_norm": 20.27568382310745,
+      "learning_rate": 2e-06,
+      "loss": 0.3322,
+      "step": 5130
+    },
+    {
+      "epoch": 1.1903491474306924,
+      "grad_norm": 14.323277855853348,
+      "learning_rate": 2e-06,
+      "loss": 0.3306,
+      "step": 5131
+    },
+    {
+      "epoch": 1.190581139078993,
+      "grad_norm": 7.469000808934431,
+      "learning_rate": 2e-06,
+      "loss": 0.1893,
+      "step": 5132
+    },
+    {
+      "epoch": 1.1908131307272938,
+      "grad_norm": 12.340451919803156,
+      "learning_rate": 2e-06,
+      "loss": 0.2664,
+      "step": 5133
+    },
+    {
+      "epoch": 1.1910451223755945,
+      "grad_norm": 15.508648743070854,
+      "learning_rate": 2e-06,
+      "loss": 0.2753,
+      "step": 5134
+    },
+    {
+      "epoch": 1.1912771140238951,
+      "grad_norm": 23.80622709464657,
+      "learning_rate": 2e-06,
+      "loss": 0.3063,
+      "step": 5135
+    },
+    {
+      "epoch": 1.1915091056721958,
+      "grad_norm": 16.937254153485966,
+      "learning_rate": 2e-06,
+      "loss": 0.2666,
+      "step": 5136
+    },
+    {
+      "epoch": 1.1917410973204965,
+      "grad_norm": 9.110017246897536,
+      "learning_rate": 2e-06,
+      "loss": 0.17,
+      "step": 5137
+    },
+    {
+      "epoch": 1.1919730889687972,
+      "grad_norm": 13.043804133984896,
+      "learning_rate": 2e-06,
+      "loss": 0.3281,
+      "step": 5138
+    },
+    {
+      "epoch": 1.1922050806170978,
+      "grad_norm": 8.90342892280623,
+      "learning_rate": 2e-06,
+      "loss": 0.1846,
+      "step": 5139
+    },
+    {
+      "epoch": 1.1924370722653985,
+      "grad_norm": 7.969223964850742,
+      "learning_rate": 2e-06,
+      "loss": 0.1504,
+      "step": 5140
+    },
+    {
+      "epoch": 1.1926690639136992,
+      "grad_norm": 13.442988145448872,
+      "learning_rate": 2e-06,
+      "loss": 0.2254,
+      "step": 5141
+    },
+    {
+      "epoch": 1.1929010555619999,
+      "grad_norm": 18.190046733000564,
+      "learning_rate": 2e-06,
+      "loss": 0.148,
+      "step": 5142
+    },
+    {
+      "epoch": 1.1931330472103003,
+      "grad_norm": 11.28922782319458,
+      "learning_rate": 2e-06,
+      "loss": 0.2086,
+      "step": 5143
+    },
+    {
+      "epoch": 1.193365038858601,
+      "grad_norm": 22.17112011918985,
+      "learning_rate": 2e-06,
+      "loss": 0.3036,
+      "step": 5144
+    },
+    {
+      "epoch": 1.1935970305069017,
+      "grad_norm": 14.839295424960572,
+      "learning_rate": 2e-06,
+      "loss": 0.2228,
+      "step": 5145
+    },
+    {
+      "epoch": 1.1938290221552024,
+      "grad_norm": 12.869616829651251,
+      "learning_rate": 2e-06,
+      "loss": 0.189,
+      "step": 5146
+    },
+    {
+      "epoch": 1.194061013803503,
+      "grad_norm": 15.443605579905961,
+      "learning_rate": 2e-06,
+      "loss": 0.2418,
+      "step": 5147
+    },
+    {
+      "epoch": 1.1942930054518037,
+      "grad_norm": 21.87299521722901,
+      "learning_rate": 2e-06,
+      "loss": 0.2444,
+      "step": 5148
+    },
+    {
+      "epoch": 1.1945249971001044,
+      "grad_norm": 9.004184331024451,
+      "learning_rate": 2e-06,
+      "loss": 0.1728,
+      "step": 5149
+    },
+    {
+      "epoch": 1.194756988748405,
+      "grad_norm": 12.551599524000297,
+      "learning_rate": 2e-06,
+      "loss": 0.2258,
+      "step": 5150
+    },
+    {
+      "epoch": 1.1949889803967058,
+      "grad_norm": 29.857523497550893,
+      "learning_rate": 2e-06,
+      "loss": 0.3028,
+      "step": 5151
+    },
+    {
+      "epoch": 1.1952209720450064,
+      "grad_norm": 12.085422933856577,
+      "learning_rate": 2e-06,
+      "loss": 0.2088,
+      "step": 5152
+    },
+    {
+      "epoch": 1.1954529636933071,
+      "grad_norm": 17.68757686122845,
+      "learning_rate": 2e-06,
+      "loss": 0.3159,
+      "step": 5153
+    },
+    {
+      "epoch": 1.1956849553416078,
+      "grad_norm": 18.730269953915307,
+      "learning_rate": 2e-06,
+      "loss": 0.2206,
+      "step": 5154
+    },
+    {
+      "epoch": 1.1959169469899082,
+      "grad_norm": 21.30782403888487,
+      "learning_rate": 2e-06,
+      "loss": 0.2301,
+      "step": 5155
+    },
+    {
+      "epoch": 1.196148938638209,
+      "grad_norm": 22.47097003616981,
+      "learning_rate": 2e-06,
+      "loss": 0.3451,
+      "step": 5156
+    },
+    {
+      "epoch": 1.1963809302865096,
+      "grad_norm": 15.510565862769823,
+      "learning_rate": 2e-06,
+      "loss": 0.3132,
+      "step": 5157
+    },
+    {
+      "epoch": 1.1966129219348103,
+      "grad_norm": 15.18217102573386,
+      "learning_rate": 2e-06,
+      "loss": 0.2763,
+      "step": 5158
+    },
+    {
+      "epoch": 1.196844913583111,
+      "grad_norm": 8.803567230647891,
+      "learning_rate": 2e-06,
+      "loss": 0.1541,
+      "step": 5159
+    },
+    {
+      "epoch": 1.1970769052314116,
+      "grad_norm": 13.327744557640425,
+      "learning_rate": 2e-06,
+      "loss": 0.1522,
+      "step": 5160
+    },
+    {
+      "epoch": 1.1973088968797123,
+      "grad_norm": 18.90690605913973,
+      "learning_rate": 2e-06,
+      "loss": 0.2968,
+      "step": 5161
+    },
+    {
+      "epoch": 1.197540888528013,
+      "grad_norm": 12.32255655411487,
+      "learning_rate": 2e-06,
+      "loss": 0.2587,
+      "step": 5162
+    },
+    {
+      "epoch": 1.1977728801763137,
+      "grad_norm": 11.8652432225982,
+      "learning_rate": 2e-06,
+      "loss": 0.1902,
+      "step": 5163
+    },
+    {
+      "epoch": 1.1980048718246143,
+      "grad_norm": 16.73378327149659,
+      "learning_rate": 2e-06,
+      "loss": 0.2478,
+      "step": 5164
+    },
+    {
+      "epoch": 1.198236863472915,
+      "grad_norm": 11.562678414791206,
+      "learning_rate": 2e-06,
+      "loss": 0.1989,
+      "step": 5165
+    },
+    {
+      "epoch": 1.1984688551212157,
+      "grad_norm": 10.397222114298854,
+      "learning_rate": 2e-06,
+      "loss": 0.2023,
+      "step": 5166
+    },
+    {
+      "epoch": 1.1987008467695164,
+      "grad_norm": 42.61136099362814,
+      "learning_rate": 2e-06,
+      "loss": 0.3812,
+      "step": 5167
+    },
+    {
+      "epoch": 1.198932838417817,
+      "grad_norm": 7.762564277411292,
+      "learning_rate": 2e-06,
+      "loss": 0.1743,
+      "step": 5168
+    },
+    {
+      "epoch": 1.1991648300661177,
+      "grad_norm": 19.320174208023595,
+      "learning_rate": 2e-06,
+      "loss": 0.2735,
+      "step": 5169
+    },
+    {
+      "epoch": 1.1993968217144182,
+      "grad_norm": 9.33982545876157,
+      "learning_rate": 2e-06,
+      "loss": 0.1582,
+      "step": 5170
+    },
+    {
+      "epoch": 1.1996288133627189,
+      "grad_norm": 24.12096917597885,
+      "learning_rate": 2e-06,
+      "loss": 0.269,
+      "step": 5171
+    },
+    {
+      "epoch": 1.1998608050110195,
+      "grad_norm": 14.189468767579216,
+      "learning_rate": 2e-06,
+      "loss": 0.2206,
+      "step": 5172
+    },
+    {
+      "epoch": 1.2000927966593202,
+      "grad_norm": 15.64630259236317,
+      "learning_rate": 2e-06,
+      "loss": 0.2856,
+      "step": 5173
+    },
+    {
+      "epoch": 1.200324788307621,
+      "grad_norm": 11.43412902266011,
+      "learning_rate": 2e-06,
+      "loss": 0.1606,
+      "step": 5174
+    },
+    {
+      "epoch": 1.2005567799559216,
+      "grad_norm": 17.885516797052404,
+      "learning_rate": 2e-06,
+      "loss": 0.2469,
+      "step": 5175
+    },
+    {
+      "epoch": 1.2007887716042223,
+      "grad_norm": 13.97343455823514,
+      "learning_rate": 2e-06,
+      "loss": 0.2441,
+      "step": 5176
+    },
+    {
+      "epoch": 1.201020763252523,
+      "grad_norm": 21.132176964097027,
+      "learning_rate": 2e-06,
+      "loss": 0.1858,
+      "step": 5177
+    },
+    {
+      "epoch": 1.2012527549008236,
+      "grad_norm": 16.177779257561756,
+      "learning_rate": 2e-06,
+      "loss": 0.2804,
+      "step": 5178
+    },
+    {
+      "epoch": 1.2014847465491243,
+      "grad_norm": 15.237446241231124,
+      "learning_rate": 2e-06,
+      "loss": 0.3138,
+      "step": 5179
+    },
+    {
+      "epoch": 1.201716738197425,
+      "grad_norm": 15.410025662753815,
+      "learning_rate": 2e-06,
+      "loss": 0.272,
+      "step": 5180
+    },
+    {
+      "epoch": 1.2019487298457256,
+      "grad_norm": 24.874371348199457,
+      "learning_rate": 2e-06,
+      "loss": 0.2788,
+      "step": 5181
+    },
+    {
+      "epoch": 1.202180721494026,
+      "grad_norm": 18.610897292084417,
+      "learning_rate": 2e-06,
+      "loss": 0.2966,
+      "step": 5182
+    },
+    {
+      "epoch": 1.2024127131423268,
+      "grad_norm": 18.36018014980279,
+      "learning_rate": 2e-06,
+      "loss": 0.2466,
+      "step": 5183
+    },
+    {
+      "epoch": 1.2026447047906275,
+      "grad_norm": 20.29533914354666,
+      "learning_rate": 2e-06,
+      "loss": 0.279,
+      "step": 5184
+    },
+    {
+      "epoch": 1.2028766964389281,
+      "grad_norm": 15.975613795013066,
+      "learning_rate": 2e-06,
+      "loss": 0.2532,
+      "step": 5185
+    },
+    {
+      "epoch": 1.2031086880872288,
+      "grad_norm": 10.660371213800708,
+      "learning_rate": 2e-06,
+      "loss": 0.2077,
+      "step": 5186
+    },
+    {
+      "epoch": 1.2033406797355295,
+      "grad_norm": 12.96893264451673,
+      "learning_rate": 2e-06,
+      "loss": 0.2392,
+      "step": 5187
+    },
+    {
+      "epoch": 1.2035726713838302,
+      "grad_norm": 15.938185909799879,
+      "learning_rate": 2e-06,
+      "loss": 0.2569,
+      "step": 5188
+    },
+    {
+      "epoch": 1.2038046630321309,
+      "grad_norm": 15.88553069087415,
+      "learning_rate": 2e-06,
+      "loss": 0.2742,
+      "step": 5189
+    },
+    {
+      "epoch": 1.2040366546804315,
+      "grad_norm": 17.130342950677043,
+      "learning_rate": 2e-06,
+      "loss": 0.2267,
+      "step": 5190
+    },
+    {
+      "epoch": 1.2042686463287322,
+      "grad_norm": 14.692193200428905,
+      "learning_rate": 2e-06,
+      "loss": 0.2488,
+      "step": 5191
+    },
+    {
+      "epoch": 1.2045006379770329,
+      "grad_norm": 20.5376558559371,
+      "learning_rate": 2e-06,
+      "loss": 0.2569,
+      "step": 5192
+    },
+    {
+      "epoch": 1.2047326296253336,
+      "grad_norm": 12.547370979534392,
+      "learning_rate": 2e-06,
+      "loss": 0.1669,
+      "step": 5193
+    },
+    {
+      "epoch": 1.2049646212736342,
+      "grad_norm": 20.732742991471042,
+      "learning_rate": 2e-06,
+      "loss": 0.4357,
+      "step": 5194
+    },
+    {
+      "epoch": 1.205196612921935,
+      "grad_norm": 16.429927193275017,
+      "learning_rate": 2e-06,
+      "loss": 0.2503,
+      "step": 5195
+    },
+    {
+      "epoch": 1.2054286045702356,
+      "grad_norm": 11.049573146865242,
+      "learning_rate": 2e-06,
+      "loss": 0.1848,
+      "step": 5196
+    },
+    {
+      "epoch": 1.205660596218536,
+      "grad_norm": 14.675234242172602,
+      "learning_rate": 2e-06,
+      "loss": 0.2755,
+      "step": 5197
+    },
+    {
+      "epoch": 1.2058925878668367,
+      "grad_norm": 15.300604270336887,
+      "learning_rate": 2e-06,
+      "loss": 0.317,
+      "step": 5198
+    },
+    {
+      "epoch": 1.2061245795151374,
+      "grad_norm": 9.036597914764565,
+      "learning_rate": 2e-06,
+      "loss": 0.1548,
+      "step": 5199
+    },
+    {
+      "epoch": 1.206356571163438,
+      "grad_norm": 12.011136214657867,
+      "learning_rate": 2e-06,
+      "loss": 0.1626,
+      "step": 5200
+    },
+    {
+      "epoch": 1.2065885628117388,
+      "grad_norm": 16.154875873107986,
+      "learning_rate": 2e-06,
+      "loss": 0.2903,
+      "step": 5201
+    },
+    {
+      "epoch": 1.2068205544600394,
+      "grad_norm": 17.69607289312409,
+      "learning_rate": 2e-06,
+      "loss": 0.2915,
+      "step": 5202
+    },
+    {
+      "epoch": 1.2070525461083401,
+      "grad_norm": 17.132895546179085,
+      "learning_rate": 2e-06,
+      "loss": 0.2851,
+      "step": 5203
+    },
+    {
+      "epoch": 1.2072845377566408,
+      "grad_norm": 17.23651791207007,
+      "learning_rate": 2e-06,
+      "loss": 0.2503,
+      "step": 5204
+    },
+    {
+      "epoch": 1.2075165294049415,
+      "grad_norm": 13.061251148229335,
+      "learning_rate": 2e-06,
+      "loss": 0.2728,
+      "step": 5205
+    },
+    {
+      "epoch": 1.2077485210532422,
+      "grad_norm": 9.275794904296339,
+      "learning_rate": 2e-06,
+      "loss": 0.1567,
+      "step": 5206
+    },
+    {
+      "epoch": 1.2079805127015428,
+      "grad_norm": 10.008621925169576,
+      "learning_rate": 2e-06,
+      "loss": 0.2422,
+      "step": 5207
+    },
+    {
+      "epoch": 1.2082125043498433,
+      "grad_norm": 7.5606010156297,
+      "learning_rate": 2e-06,
+      "loss": 0.1861,
+      "step": 5208
+    },
+    {
+      "epoch": 1.208444495998144,
+      "grad_norm": 13.35580662177829,
+      "learning_rate": 2e-06,
+      "loss": 0.3143,
+      "step": 5209
+    },
+    {
+      "epoch": 1.2086764876464446,
+      "grad_norm": 13.10733565592071,
+      "learning_rate": 2e-06,
+      "loss": 0.1783,
+      "step": 5210
+    },
+    {
+      "epoch": 1.2089084792947453,
+      "grad_norm": 7.70315944281134,
+      "learning_rate": 2e-06,
+      "loss": 0.131,
+      "step": 5211
+    },
+    {
+      "epoch": 1.209140470943046,
+      "grad_norm": 11.866743411131484,
+      "learning_rate": 2e-06,
+      "loss": 0.2679,
+      "step": 5212
+    },
+    {
+      "epoch": 1.2093724625913467,
+      "grad_norm": 22.510614560943893,
+      "learning_rate": 2e-06,
+      "loss": 0.2429,
+      "step": 5213
+    },
+    {
+      "epoch": 1.2096044542396474,
+      "grad_norm": 10.310466738956148,
+      "learning_rate": 2e-06,
+      "loss": 0.2073,
+      "step": 5214
+    },
+    {
+      "epoch": 1.209836445887948,
+      "grad_norm": 20.320785351962254,
+      "learning_rate": 2e-06,
+      "loss": 0.3057,
+      "step": 5215
+    },
+    {
+      "epoch": 1.2100684375362487,
+      "grad_norm": 9.40601990023123,
+      "learning_rate": 2e-06,
+      "loss": 0.182,
+      "step": 5216
+    },
+    {
+      "epoch": 1.2103004291845494,
+      "grad_norm": 13.741165780691722,
+      "learning_rate": 2e-06,
+      "loss": 0.2182,
+      "step": 5217
+    },
+    {
+      "epoch": 1.21053242083285,
+      "grad_norm": 7.508991657284485,
+      "learning_rate": 2e-06,
+      "loss": 0.1946,
+      "step": 5218
+    },
+    {
+      "epoch": 1.2107644124811507,
+      "grad_norm": 15.514854424391363,
+      "learning_rate": 2e-06,
+      "loss": 0.2248,
+      "step": 5219
+    },
+    {
+      "epoch": 1.2109964041294514,
+      "grad_norm": 10.895204473041584,
+      "learning_rate": 2e-06,
+      "loss": 0.1865,
+      "step": 5220
+    },
+    {
+      "epoch": 1.211228395777752,
+      "grad_norm": 19.450039982901867,
+      "learning_rate": 2e-06,
+      "loss": 0.219,
+      "step": 5221
+    },
+    {
+      "epoch": 1.2114603874260528,
+      "grad_norm": 13.302386848635217,
+      "learning_rate": 2e-06,
+      "loss": 0.1976,
+      "step": 5222
+    },
+    {
+      "epoch": 1.2116923790743532,
+      "grad_norm": 22.35231992069119,
+      "learning_rate": 2e-06,
+      "loss": 0.4383,
+      "step": 5223
+    },
+    {
+      "epoch": 1.211924370722654,
+      "grad_norm": 16.850637258242692,
+      "learning_rate": 2e-06,
+      "loss": 0.1745,
+      "step": 5224
+    },
+    {
+      "epoch": 1.2121563623709546,
+      "grad_norm": 7.775618491069661,
+      "learning_rate": 2e-06,
+      "loss": 0.1796,
+      "step": 5225
+    },
+    {
+      "epoch": 1.2123883540192553,
+      "grad_norm": 26.089677514709958,
+      "learning_rate": 2e-06,
+      "loss": 0.3559,
+      "step": 5226
+    },
+    {
+      "epoch": 1.212620345667556,
+      "grad_norm": 21.937358674413783,
+      "learning_rate": 2e-06,
+      "loss": 0.2087,
+      "step": 5227
+    },
+    {
+      "epoch": 1.2128523373158566,
+      "grad_norm": 8.702241396578975,
+      "learning_rate": 2e-06,
+      "loss": 0.2015,
+      "step": 5228
+    },
+    {
+      "epoch": 1.2130843289641573,
+      "grad_norm": 7.409840180631241,
+      "learning_rate": 2e-06,
+      "loss": 0.1414,
+      "step": 5229
+    },
+    {
+      "epoch": 1.213316320612458,
+      "grad_norm": 19.95628813984114,
+      "learning_rate": 2e-06,
+      "loss": 0.3464,
+      "step": 5230
+    },
+    {
+      "epoch": 1.2135483122607587,
+      "grad_norm": 17.23062502380466,
+      "learning_rate": 2e-06,
+      "loss": 0.2711,
+      "step": 5231
+    },
+    {
+      "epoch": 1.2137803039090593,
+      "grad_norm": 20.833991635692122,
+      "learning_rate": 2e-06,
+      "loss": 0.4338,
+      "step": 5232
+    },
+    {
+      "epoch": 1.21401229555736,
+      "grad_norm": 15.283066556761923,
+      "learning_rate": 2e-06,
+      "loss": 0.3238,
+      "step": 5233
+    },
+    {
+      "epoch": 1.2142442872056607,
+      "grad_norm": 14.901162875963417,
+      "learning_rate": 2e-06,
+      "loss": 0.1931,
+      "step": 5234
+    },
+    {
+      "epoch": 1.2144762788539611,
+      "grad_norm": 21.469509336173594,
+      "learning_rate": 2e-06,
+      "loss": 0.2316,
+      "step": 5235
+    },
+    {
+      "epoch": 1.2147082705022618,
+      "grad_norm": 17.957005529775234,
+      "learning_rate": 2e-06,
+      "loss": 0.3315,
+      "step": 5236
+    },
+    {
+      "epoch": 1.2149402621505625,
+      "grad_norm": 18.140735392836362,
+      "learning_rate": 2e-06,
+      "loss": 0.2248,
+      "step": 5237
+    },
+    {
+      "epoch": 1.2151722537988632,
+      "grad_norm": 26.17420284310708,
+      "learning_rate": 2e-06,
+      "loss": 0.3931,
+      "step": 5238
+    },
+    {
+      "epoch": 1.2154042454471639,
+      "grad_norm": 7.8289304917804055,
+      "learning_rate": 2e-06,
+      "loss": 0.1446,
+      "step": 5239
+    },
+    {
+      "epoch": 1.2156362370954645,
+      "grad_norm": 18.23081021770153,
+      "learning_rate": 2e-06,
+      "loss": 0.1974,
+      "step": 5240
+    },
+    {
+      "epoch": 1.2158682287437652,
+      "grad_norm": 18.589950781673117,
+      "learning_rate": 2e-06,
+      "loss": 0.2152,
+      "step": 5241
+    },
+    {
+      "epoch": 1.216100220392066,
+      "grad_norm": 15.53438228175382,
+      "learning_rate": 2e-06,
+      "loss": 0.2284,
+      "step": 5242
+    },
+    {
+      "epoch": 1.2163322120403666,
+      "grad_norm": 17.21491317910429,
+      "learning_rate": 2e-06,
+      "loss": 0.2799,
+      "step": 5243
+    },
+    {
+      "epoch": 1.2165642036886672,
+      "grad_norm": 17.67007580606666,
+      "learning_rate": 2e-06,
+      "loss": 0.4263,
+      "step": 5244
+    },
+    {
+      "epoch": 1.216796195336968,
+      "grad_norm": 16.006138323506924,
+      "learning_rate": 2e-06,
+      "loss": 0.2022,
+      "step": 5245
+    },
+    {
+      "epoch": 1.2170281869852686,
+      "grad_norm": 12.50912584709491,
+      "learning_rate": 2e-06,
+      "loss": 0.2468,
+      "step": 5246
+    },
+    {
+      "epoch": 1.2172601786335693,
+      "grad_norm": 19.46136469945236,
+      "learning_rate": 2e-06,
+      "loss": 0.3507,
+      "step": 5247
+    },
+    {
+      "epoch": 1.21749217028187,
+      "grad_norm": 14.635269609314967,
+      "learning_rate": 2e-06,
+      "loss": 0.2371,
+      "step": 5248
+    },
+    {
+      "epoch": 1.2177241619301706,
+      "grad_norm": 13.798822892634078,
+      "learning_rate": 2e-06,
+      "loss": 0.218,
+      "step": 5249
+    },
+    {
+      "epoch": 1.217956153578471,
+      "grad_norm": 27.779776858612273,
+      "learning_rate": 2e-06,
+      "loss": 0.5181,
+      "step": 5250
+    },
+    {
+      "epoch": 1.2181881452267718,
+      "grad_norm": 15.34299875845018,
+      "learning_rate": 2e-06,
+      "loss": 0.3343,
+      "step": 5251
+    },
+    {
+      "epoch": 1.2184201368750724,
+      "grad_norm": 18.417075663930053,
+      "learning_rate": 2e-06,
+      "loss": 0.2751,
+      "step": 5252
+    },
+    {
+      "epoch": 1.2186521285233731,
+      "grad_norm": 18.78961744953273,
+      "learning_rate": 2e-06,
+      "loss": 0.3231,
+      "step": 5253
+    },
+    {
+      "epoch": 1.2188841201716738,
+      "grad_norm": 13.46130141911124,
+      "learning_rate": 2e-06,
+      "loss": 0.2129,
+      "step": 5254
+    },
+    {
+      "epoch": 1.2191161118199745,
+      "grad_norm": 11.833100567305175,
+      "learning_rate": 2e-06,
+      "loss": 0.3017,
+      "step": 5255
+    },
+    {
+      "epoch": 1.2193481034682752,
+      "grad_norm": 7.404813410538326,
+      "learning_rate": 2e-06,
+      "loss": 0.2085,
+      "step": 5256
+    },
+    {
+      "epoch": 1.2195800951165758,
+      "grad_norm": 27.538389014789562,
+      "learning_rate": 2e-06,
+      "loss": 0.3444,
+      "step": 5257
+    },
+    {
+      "epoch": 1.2198120867648765,
+      "grad_norm": 14.48501764302651,
+      "learning_rate": 2e-06,
+      "loss": 0.265,
+      "step": 5258
+    },
+    {
+      "epoch": 1.2200440784131772,
+      "grad_norm": 15.734888066600165,
+      "learning_rate": 2e-06,
+      "loss": 0.2839,
+      "step": 5259
+    },
+    {
+      "epoch": 1.2202760700614779,
+      "grad_norm": 14.022442780422809,
+      "learning_rate": 2e-06,
+      "loss": 0.3132,
+      "step": 5260
+    },
+    {
+      "epoch": 1.2205080617097783,
+      "grad_norm": 15.731696974707647,
+      "learning_rate": 2e-06,
+      "loss": 0.1934,
+      "step": 5261
+    },
+    {
+      "epoch": 1.220740053358079,
+      "grad_norm": 12.422595153554099,
+      "learning_rate": 2e-06,
+      "loss": 0.2521,
+      "step": 5262
+    },
+    {
+      "epoch": 1.2209720450063797,
+      "grad_norm": 15.45177514885588,
+      "learning_rate": 2e-06,
+      "loss": 0.2583,
+      "step": 5263
+    },
+    {
+      "epoch": 1.2212040366546804,
+      "grad_norm": 14.29565801586227,
+      "learning_rate": 2e-06,
+      "loss": 0.2793,
+      "step": 5264
+    },
+    {
+      "epoch": 1.221436028302981,
+      "grad_norm": 15.599363130704816,
+      "learning_rate": 2e-06,
+      "loss": 0.2823,
+      "step": 5265
+    },
+    {
+      "epoch": 1.2216680199512817,
+      "grad_norm": 15.440296781675615,
+      "learning_rate": 2e-06,
+      "loss": 0.3655,
+      "step": 5266
+    },
+    {
+      "epoch": 1.2219000115995824,
+      "grad_norm": 13.407628310415342,
+      "learning_rate": 2e-06,
+      "loss": 0.2767,
+      "step": 5267
+    },
+    {
+      "epoch": 1.222132003247883,
+      "grad_norm": 10.727707106075549,
+      "learning_rate": 2e-06,
+      "loss": 0.1716,
+      "step": 5268
+    },
+    {
+      "epoch": 1.2223639948961837,
+      "grad_norm": 10.130188590250924,
+      "learning_rate": 2e-06,
+      "loss": 0.2332,
+      "step": 5269
+    },
+    {
+      "epoch": 1.2225959865444844,
+      "grad_norm": 14.498875997948055,
+      "learning_rate": 2e-06,
+      "loss": 0.278,
+      "step": 5270
+    },
+    {
+      "epoch": 1.222827978192785,
+      "grad_norm": 8.002068878752675,
+      "learning_rate": 2e-06,
+      "loss": 0.2471,
+      "step": 5271
+    },
+    {
+      "epoch": 1.2230599698410858,
+      "grad_norm": 11.603943189818388,
+      "learning_rate": 2e-06,
+      "loss": 0.2822,
+      "step": 5272
+    },
+    {
+      "epoch": 1.2232919614893865,
+      "grad_norm": 8.812245906502385,
+      "learning_rate": 2e-06,
+      "loss": 0.173,
+      "step": 5273
+    },
+    {
+      "epoch": 1.2235239531376871,
+      "grad_norm": 21.6674746111459,
+      "learning_rate": 2e-06,
+      "loss": 0.2749,
+      "step": 5274
+    },
+    {
+      "epoch": 1.2237559447859878,
+      "grad_norm": 21.08493131129624,
+      "learning_rate": 2e-06,
+      "loss": 0.3471,
+      "step": 5275
+    },
+    {
+      "epoch": 1.2239879364342883,
+      "grad_norm": 16.69822828663652,
+      "learning_rate": 2e-06,
+      "loss": 0.2287,
+      "step": 5276
+    },
+    {
+      "epoch": 1.224219928082589,
+      "grad_norm": 10.561775512795705,
+      "learning_rate": 2e-06,
+      "loss": 0.1804,
+      "step": 5277
+    },
+    {
+      "epoch": 1.2244519197308896,
+      "grad_norm": 28.678018135887456,
+      "learning_rate": 2e-06,
+      "loss": 0.3214,
+      "step": 5278
+    },
+    {
+      "epoch": 1.2246839113791903,
+      "grad_norm": 28.824433534080182,
+      "learning_rate": 2e-06,
+      "loss": 0.2695,
+      "step": 5279
+    },
+    {
+      "epoch": 1.224915903027491,
+      "grad_norm": 14.031124142089942,
+      "learning_rate": 2e-06,
+      "loss": 0.2114,
+      "step": 5280
+    },
+    {
+      "epoch": 1.2251478946757917,
+      "grad_norm": 17.859482295756116,
+      "learning_rate": 2e-06,
+      "loss": 0.226,
+      "step": 5281
+    },
+    {
+      "epoch": 1.2253798863240923,
+      "grad_norm": 16.643741572523084,
+      "learning_rate": 2e-06,
+      "loss": 0.1829,
+      "step": 5282
+    },
+    {
+      "epoch": 1.225611877972393,
+      "grad_norm": 37.07114976670886,
+      "learning_rate": 2e-06,
+      "loss": 0.232,
+      "step": 5283
+    },
+    {
+      "epoch": 1.2258438696206937,
+      "grad_norm": 20.04040061838349,
+      "learning_rate": 2e-06,
+      "loss": 0.4911,
+      "step": 5284
+    },
+    {
+      "epoch": 1.2260758612689944,
+      "grad_norm": 11.54246641036162,
+      "learning_rate": 2e-06,
+      "loss": 0.2101,
+      "step": 5285
+    },
+    {
+      "epoch": 1.226307852917295,
+      "grad_norm": 16.36165634216338,
+      "learning_rate": 2e-06,
+      "loss": 0.2945,
+      "step": 5286
+    },
+    {
+      "epoch": 1.2265398445655957,
+      "grad_norm": 19.920970194021628,
+      "learning_rate": 2e-06,
+      "loss": 0.2737,
+      "step": 5287
+    },
+    {
+      "epoch": 1.2267718362138962,
+      "grad_norm": 17.9952410931923,
+      "learning_rate": 2e-06,
+      "loss": 0.3436,
+      "step": 5288
+    },
+    {
+      "epoch": 1.2270038278621969,
+      "grad_norm": 11.289988956344413,
+      "learning_rate": 2e-06,
+      "loss": 0.2206,
+      "step": 5289
+    },
+    {
+      "epoch": 1.2272358195104975,
+      "grad_norm": 13.383557975275709,
+      "learning_rate": 2e-06,
+      "loss": 0.2419,
+      "step": 5290
+    },
+    {
+      "epoch": 1.2274678111587982,
+      "grad_norm": 22.369959716146592,
+      "learning_rate": 2e-06,
+      "loss": 0.2851,
+      "step": 5291
+    },
+    {
+      "epoch": 1.227699802807099,
+      "grad_norm": 12.843101286087245,
+      "learning_rate": 2e-06,
+      "loss": 0.2112,
+      "step": 5292
+    },
+    {
+      "epoch": 1.2279317944553996,
+      "grad_norm": 9.344779922844397,
+      "learning_rate": 2e-06,
+      "loss": 0.267,
+      "step": 5293
+    },
+    {
+      "epoch": 1.2281637861037003,
+      "grad_norm": 12.59966525810847,
+      "learning_rate": 2e-06,
+      "loss": 0.295,
+      "step": 5294
+    },
+    {
+      "epoch": 1.228395777752001,
+      "grad_norm": 14.793820877515458,
+      "learning_rate": 2e-06,
+      "loss": 0.2337,
+      "step": 5295
+    },
+    {
+      "epoch": 1.2286277694003016,
+      "grad_norm": 9.697869595422336,
+      "learning_rate": 2e-06,
+      "loss": 0.1847,
+      "step": 5296
+    },
+    {
+      "epoch": 1.2288597610486023,
+      "grad_norm": 12.580748125443968,
+      "learning_rate": 2e-06,
+      "loss": 0.211,
+      "step": 5297
+    },
+    {
+      "epoch": 1.229091752696903,
+      "grad_norm": 8.40386513019791,
+      "learning_rate": 2e-06,
+      "loss": 0.1698,
+      "step": 5298
+    },
+    {
+      "epoch": 1.2293237443452036,
+      "grad_norm": 7.596042516068372,
+      "learning_rate": 2e-06,
+      "loss": 0.2155,
+      "step": 5299
+    },
+    {
+      "epoch": 1.2295557359935043,
+      "grad_norm": 17.352822181947214,
+      "learning_rate": 2e-06,
+      "loss": 0.3248,
+      "step": 5300
+    },
+    {
+      "epoch": 1.229787727641805,
+      "grad_norm": 14.722065240431393,
+      "learning_rate": 2e-06,
+      "loss": 0.2228,
+      "step": 5301
+    },
+    {
+      "epoch": 1.2300197192901057,
+      "grad_norm": 8.723387036606349,
+      "learning_rate": 2e-06,
+      "loss": 0.1781,
+      "step": 5302
+    },
+    {
+      "epoch": 1.2302517109384061,
+      "grad_norm": 15.040416722840469,
+      "learning_rate": 2e-06,
+      "loss": 0.2241,
+      "step": 5303
+    },
+    {
+      "epoch": 1.2304837025867068,
+      "grad_norm": 13.198481853039846,
+      "learning_rate": 2e-06,
+      "loss": 0.234,
+      "step": 5304
+    },
+    {
+      "epoch": 1.2307156942350075,
+      "grad_norm": 9.054704316573181,
+      "learning_rate": 2e-06,
+      "loss": 0.2446,
+      "step": 5305
+    },
+    {
+      "epoch": 1.2309476858833082,
+      "grad_norm": 16.080836154342443,
+      "learning_rate": 2e-06,
+      "loss": 0.2978,
+      "step": 5306
+    },
+    {
+      "epoch": 1.2311796775316088,
+      "grad_norm": 9.419831444878099,
+      "learning_rate": 2e-06,
+      "loss": 0.1617,
+      "step": 5307
+    },
+    {
+      "epoch": 1.2314116691799095,
+      "grad_norm": 13.487378036543648,
+      "learning_rate": 2e-06,
+      "loss": 0.1953,
+      "step": 5308
+    },
+    {
+      "epoch": 1.2316436608282102,
+      "grad_norm": 13.178197224965563,
+      "learning_rate": 2e-06,
+      "loss": 0.3204,
+      "step": 5309
+    },
+    {
+      "epoch": 1.2318756524765109,
+      "grad_norm": 15.681660473782781,
+      "learning_rate": 2e-06,
+      "loss": 0.3323,
+      "step": 5310
+    },
+    {
+      "epoch": 1.2321076441248116,
+      "grad_norm": 25.239516257597224,
+      "learning_rate": 2e-06,
+      "loss": 0.3799,
+      "step": 5311
+    },
+    {
+      "epoch": 1.2323396357731122,
+      "grad_norm": 8.244385037922948,
+      "learning_rate": 2e-06,
+      "loss": 0.1961,
+      "step": 5312
+    },
+    {
+      "epoch": 1.232571627421413,
+      "grad_norm": 12.190565719041365,
+      "learning_rate": 2e-06,
+      "loss": 0.311,
+      "step": 5313
+    },
+    {
+      "epoch": 1.2328036190697136,
+      "grad_norm": 9.850539308026018,
+      "learning_rate": 2e-06,
+      "loss": 0.2206,
+      "step": 5314
+    },
+    {
+      "epoch": 1.233035610718014,
+      "grad_norm": 14.153920435151814,
+      "learning_rate": 2e-06,
+      "loss": 0.3262,
+      "step": 5315
+    },
+    {
+      "epoch": 1.2332676023663147,
+      "grad_norm": 13.520522643805842,
+      "learning_rate": 2e-06,
+      "loss": 0.1809,
+      "step": 5316
+    },
+    {
+      "epoch": 1.2334995940146154,
+      "grad_norm": 19.47076432215645,
+      "learning_rate": 2e-06,
+      "loss": 0.2154,
+      "step": 5317
+    },
+    {
+      "epoch": 1.233731585662916,
+      "grad_norm": 11.485832864111378,
+      "learning_rate": 2e-06,
+      "loss": 0.2006,
+      "step": 5318
+    },
+    {
+      "epoch": 1.2339635773112168,
+      "grad_norm": 20.080611120412676,
+      "learning_rate": 2e-06,
+      "loss": 0.316,
+      "step": 5319
+    },
+    {
+      "epoch": 1.2341955689595174,
+      "grad_norm": 8.779411883202771,
+      "learning_rate": 2e-06,
+      "loss": 0.1902,
+      "step": 5320
+    },
+    {
+      "epoch": 1.234427560607818,
+      "grad_norm": 16.488788887028306,
+      "learning_rate": 2e-06,
+      "loss": 0.2165,
+      "step": 5321
+    },
+    {
+      "epoch": 1.2346595522561188,
+      "grad_norm": 11.51692899731653,
+      "learning_rate": 2e-06,
+      "loss": 0.258,
+      "step": 5322
+    },
+    {
+      "epoch": 1.2348915439044195,
+      "grad_norm": 17.678905519432654,
+      "learning_rate": 2e-06,
+      "loss": 0.3973,
+      "step": 5323
+    },
+    {
+      "epoch": 1.2351235355527201,
+      "grad_norm": 11.581841331604895,
+      "learning_rate": 2e-06,
+      "loss": 0.2456,
+      "step": 5324
+    },
+    {
+      "epoch": 1.2353555272010208,
+      "grad_norm": 18.37405604865033,
+      "learning_rate": 2e-06,
+      "loss": 0.3861,
+      "step": 5325
+    },
+    {
+      "epoch": 1.2355875188493215,
+      "grad_norm": 15.130367359648647,
+      "learning_rate": 2e-06,
+      "loss": 0.4564,
+      "step": 5326
+    },
+    {
+      "epoch": 1.2358195104976222,
+      "grad_norm": 14.957660846372246,
+      "learning_rate": 2e-06,
+      "loss": 0.2367,
+      "step": 5327
+    },
+    {
+      "epoch": 1.2360515021459229,
+      "grad_norm": 10.56472444321113,
+      "learning_rate": 2e-06,
+      "loss": 0.197,
+      "step": 5328
+    },
+    {
+      "epoch": 1.2362834937942235,
+      "grad_norm": 13.940692025178311,
+      "learning_rate": 2e-06,
+      "loss": 0.1999,
+      "step": 5329
+    },
+    {
+      "epoch": 1.236515485442524,
+      "grad_norm": 11.270627604067421,
+      "learning_rate": 2e-06,
+      "loss": 0.278,
+      "step": 5330
+    },
+    {
+      "epoch": 1.2367474770908247,
+      "grad_norm": 16.92028277449169,
+      "learning_rate": 2e-06,
+      "loss": 0.2766,
+      "step": 5331
+    },
+    {
+      "epoch": 1.2369794687391253,
+      "grad_norm": 26.307067518093277,
+      "learning_rate": 2e-06,
+      "loss": 0.3855,
+      "step": 5332
+    },
+    {
+      "epoch": 1.237211460387426,
+      "grad_norm": 21.92979348271037,
+      "learning_rate": 2e-06,
+      "loss": 0.3515,
+      "step": 5333
+    },
+    {
+      "epoch": 1.2374434520357267,
+      "grad_norm": 11.581362836366731,
+      "learning_rate": 2e-06,
+      "loss": 0.2065,
+      "step": 5334
+    },
+    {
+      "epoch": 1.2376754436840274,
+      "grad_norm": 11.930440029646041,
+      "learning_rate": 2e-06,
+      "loss": 0.1782,
+      "step": 5335
+    },
+    {
+      "epoch": 1.237907435332328,
+      "grad_norm": 17.013879424895844,
+      "learning_rate": 2e-06,
+      "loss": 0.2203,
+      "step": 5336
+    },
+    {
+      "epoch": 1.2381394269806287,
+      "grad_norm": 7.1496544046508195,
+      "learning_rate": 2e-06,
+      "loss": 0.1477,
+      "step": 5337
+    },
+    {
+      "epoch": 1.2383714186289294,
+      "grad_norm": 15.173156812522874,
+      "learning_rate": 2e-06,
+      "loss": 0.253,
+      "step": 5338
+    },
+    {
+      "epoch": 1.23860341027723,
+      "grad_norm": 13.382184341323773,
+      "learning_rate": 2e-06,
+      "loss": 0.2531,
+      "step": 5339
+    },
+    {
+      "epoch": 1.2388354019255308,
+      "grad_norm": 10.946003794488881,
+      "learning_rate": 2e-06,
+      "loss": 0.2282,
+      "step": 5340
+    },
+    {
+      "epoch": 1.2390673935738312,
+      "grad_norm": 10.11429496594036,
+      "learning_rate": 2e-06,
+      "loss": 0.2194,
+      "step": 5341
+    },
+    {
+      "epoch": 1.239299385222132,
+      "grad_norm": 12.266409181966331,
+      "learning_rate": 2e-06,
+      "loss": 0.1854,
+      "step": 5342
+    },
+    {
+      "epoch": 1.2395313768704326,
+      "grad_norm": 14.34239606199122,
+      "learning_rate": 2e-06,
+      "loss": 0.2947,
+      "step": 5343
+    },
+    {
+      "epoch": 1.2397633685187333,
+      "grad_norm": 17.562564624110667,
+      "learning_rate": 2e-06,
+      "loss": 0.1894,
+      "step": 5344
+    },
+    {
+      "epoch": 1.239995360167034,
+      "grad_norm": 13.432435067887818,
+      "learning_rate": 2e-06,
+      "loss": 0.2164,
+      "step": 5345
+    },
+    {
+      "epoch": 1.2402273518153346,
+      "grad_norm": 22.792988260322883,
+      "learning_rate": 2e-06,
+      "loss": 0.3237,
+      "step": 5346
+    },
+    {
+      "epoch": 1.2404593434636353,
+      "grad_norm": 11.917086324715964,
+      "learning_rate": 2e-06,
+      "loss": 0.2028,
+      "step": 5347
+    },
+    {
+      "epoch": 1.240691335111936,
+      "grad_norm": 18.645434561808916,
+      "learning_rate": 2e-06,
+      "loss": 0.3638,
+      "step": 5348
+    },
+    {
+      "epoch": 1.2409233267602366,
+      "grad_norm": 14.115886033627724,
+      "learning_rate": 2e-06,
+      "loss": 0.1957,
+      "step": 5349
+    },
+    {
+      "epoch": 1.2411553184085373,
+      "grad_norm": 11.39626873072166,
+      "learning_rate": 2e-06,
+      "loss": 0.2367,
+      "step": 5350
+    },
+    {
+      "epoch": 1.241387310056838,
+      "grad_norm": 18.343971576177992,
+      "learning_rate": 2e-06,
+      "loss": 0.301,
+      "step": 5351
+    },
+    {
+      "epoch": 1.2416193017051387,
+      "grad_norm": 14.328369898087471,
+      "learning_rate": 2e-06,
+      "loss": 0.196,
+      "step": 5352
+    },
+    {
+      "epoch": 1.2418512933534394,
+      "grad_norm": 23.31079139215547,
+      "learning_rate": 2e-06,
+      "loss": 0.3009,
+      "step": 5353
+    },
+    {
+      "epoch": 1.24208328500174,
+      "grad_norm": 15.595543149206021,
+      "learning_rate": 2e-06,
+      "loss": 0.2458,
+      "step": 5354
+    },
+    {
+      "epoch": 1.2423152766500407,
+      "grad_norm": 12.699706007266634,
+      "learning_rate": 2e-06,
+      "loss": 0.2639,
+      "step": 5355
+    },
+    {
+      "epoch": 1.2425472682983412,
+      "grad_norm": 17.694766975284093,
+      "learning_rate": 2e-06,
+      "loss": 0.3025,
+      "step": 5356
+    },
+    {
+      "epoch": 1.2427792599466418,
+      "grad_norm": 19.481656997033788,
+      "learning_rate": 2e-06,
+      "loss": 0.2393,
+      "step": 5357
+    },
+    {
+      "epoch": 1.2430112515949425,
+      "grad_norm": 26.897648558409387,
+      "learning_rate": 2e-06,
+      "loss": 0.2656,
+      "step": 5358
+    },
+    {
+      "epoch": 1.2432432432432432,
+      "grad_norm": 17.579246825669056,
+      "learning_rate": 2e-06,
+      "loss": 0.2459,
+      "step": 5359
+    },
+    {
+      "epoch": 1.2434752348915439,
+      "grad_norm": 15.492054232320067,
+      "learning_rate": 2e-06,
+      "loss": 0.194,
+      "step": 5360
+    },
+    {
+      "epoch": 1.2437072265398446,
+      "grad_norm": 10.026590173467095,
+      "learning_rate": 2e-06,
+      "loss": 0.1695,
+      "step": 5361
+    },
+    {
+      "epoch": 1.2439392181881452,
+      "grad_norm": 16.692618658838594,
+      "learning_rate": 2e-06,
+      "loss": 0.2551,
+      "step": 5362
+    },
+    {
+      "epoch": 1.244171209836446,
+      "grad_norm": 15.049609606356615,
+      "learning_rate": 2e-06,
+      "loss": 0.2791,
+      "step": 5363
+    },
+    {
+      "epoch": 1.2444032014847466,
+      "grad_norm": 11.389033854712501,
+      "learning_rate": 2e-06,
+      "loss": 0.3525,
+      "step": 5364
+    },
+    {
+      "epoch": 1.2446351931330473,
+      "grad_norm": 6.8271310257402815,
+      "learning_rate": 2e-06,
+      "loss": 0.1401,
+      "step": 5365
+    },
+    {
+      "epoch": 1.244867184781348,
+      "grad_norm": 24.358447016584893,
+      "learning_rate": 2e-06,
+      "loss": 0.2946,
+      "step": 5366
+    },
+    {
+      "epoch": 1.2450991764296486,
+      "grad_norm": 15.490606812265609,
+      "learning_rate": 2e-06,
+      "loss": 0.3321,
+      "step": 5367
+    },
+    {
+      "epoch": 1.245331168077949,
+      "grad_norm": 17.53430950849937,
+      "learning_rate": 2e-06,
+      "loss": 0.2033,
+      "step": 5368
+    },
+    {
+      "epoch": 1.2455631597262498,
+      "grad_norm": 6.766701685141638,
+      "learning_rate": 2e-06,
+      "loss": 0.1528,
+      "step": 5369
+    },
+    {
+      "epoch": 1.2457951513745504,
+      "grad_norm": 21.917208890331153,
+      "learning_rate": 2e-06,
+      "loss": 0.3338,
+      "step": 5370
+    },
+    {
+      "epoch": 1.2460271430228511,
+      "grad_norm": 9.526048106428112,
+      "learning_rate": 2e-06,
+      "loss": 0.1639,
+      "step": 5371
+    },
+    {
+      "epoch": 1.2462591346711518,
+      "grad_norm": 11.603815223248493,
+      "learning_rate": 2e-06,
+      "loss": 0.2868,
+      "step": 5372
+    },
+    {
+      "epoch": 1.2464911263194525,
+      "grad_norm": 13.988597939735435,
+      "learning_rate": 2e-06,
+      "loss": 0.259,
+      "step": 5373
+    },
+    {
+      "epoch": 1.2467231179677531,
+      "grad_norm": 8.844100357062398,
+      "learning_rate": 2e-06,
+      "loss": 0.1971,
+      "step": 5374
+    },
+    {
+      "epoch": 1.2469551096160538,
+      "grad_norm": 15.231375274391446,
+      "learning_rate": 2e-06,
+      "loss": 0.2439,
+      "step": 5375
+    },
+    {
+      "epoch": 1.2471871012643545,
+      "grad_norm": 13.556775542243084,
+      "learning_rate": 2e-06,
+      "loss": 0.2513,
+      "step": 5376
+    },
+    {
+      "epoch": 1.2474190929126552,
+      "grad_norm": 11.837907914182448,
+      "learning_rate": 2e-06,
+      "loss": 0.1514,
+      "step": 5377
+    },
+    {
+      "epoch": 1.2476510845609559,
+      "grad_norm": 11.071337144240893,
+      "learning_rate": 2e-06,
+      "loss": 0.1871,
+      "step": 5378
+    },
+    {
+      "epoch": 1.2478830762092565,
+      "grad_norm": 9.181267980204264,
+      "learning_rate": 2e-06,
+      "loss": 0.1558,
+      "step": 5379
+    },
+    {
+      "epoch": 1.2481150678575572,
+      "grad_norm": 27.662846838543977,
+      "learning_rate": 2e-06,
+      "loss": 0.4109,
+      "step": 5380
+    },
+    {
+      "epoch": 1.248347059505858,
+      "grad_norm": 7.471747957628051,
+      "learning_rate": 2e-06,
+      "loss": 0.1225,
+      "step": 5381
+    },
+    {
+      "epoch": 1.2485790511541586,
+      "grad_norm": 11.818739316598782,
+      "learning_rate": 2e-06,
+      "loss": 0.2175,
+      "step": 5382
+    },
+    {
+      "epoch": 1.248811042802459,
+      "grad_norm": 13.116379005884413,
+      "learning_rate": 2e-06,
+      "loss": 0.2719,
+      "step": 5383
+    },
+    {
+      "epoch": 1.2490430344507597,
+      "grad_norm": 15.287798792169049,
+      "learning_rate": 2e-06,
+      "loss": 0.1912,
+      "step": 5384
+    },
+    {
+      "epoch": 1.2492750260990604,
+      "grad_norm": 21.297823361697418,
+      "learning_rate": 2e-06,
+      "loss": 0.3731,
+      "step": 5385
+    },
+    {
+      "epoch": 1.249507017747361,
+      "grad_norm": 14.772441145934714,
+      "learning_rate": 2e-06,
+      "loss": 0.2237,
+      "step": 5386
+    },
+    {
+      "epoch": 1.2497390093956617,
+      "grad_norm": 22.743960581731905,
+      "learning_rate": 2e-06,
+      "loss": 0.2809,
+      "step": 5387
+    },
+    {
+      "epoch": 1.2499710010439624,
+      "grad_norm": 8.798316896561857,
+      "learning_rate": 2e-06,
+      "loss": 0.1979,
+      "step": 5388
+    },
+    {
+      "epoch": 1.250202992692263,
+      "grad_norm": 17.322801821705692,
+      "learning_rate": 2e-06,
+      "loss": 0.2744,
+      "step": 5389
+    },
+    {
+      "epoch": 1.2504349843405638,
+      "grad_norm": 22.401139532469738,
+      "learning_rate": 2e-06,
+      "loss": 0.2902,
+      "step": 5390
+    },
+    {
+      "epoch": 1.2506669759888644,
+      "grad_norm": 12.33349075041886,
+      "learning_rate": 2e-06,
+      "loss": 0.2465,
+      "step": 5391
+    },
+    {
+      "epoch": 1.2508989676371651,
+      "grad_norm": 12.554184765560295,
+      "learning_rate": 2e-06,
+      "loss": 0.2667,
+      "step": 5392
+    },
+    {
+      "epoch": 1.2511309592854658,
+      "grad_norm": 19.246451335744148,
+      "learning_rate": 2e-06,
+      "loss": 0.268,
+      "step": 5393
+    },
+    {
+      "epoch": 1.2513629509337663,
+      "grad_norm": 15.757899753835336,
+      "learning_rate": 2e-06,
+      "loss": 0.3866,
+      "step": 5394
+    },
+    {
+      "epoch": 1.251594942582067,
+      "grad_norm": 11.682845962868761,
+      "learning_rate": 2e-06,
+      "loss": 0.1962,
+      "step": 5395
+    },
+    {
+      "epoch": 1.2518269342303676,
+      "grad_norm": 8.234530902056399,
+      "learning_rate": 2e-06,
+      "loss": 0.1765,
+      "step": 5396
+    },
+    {
+      "epoch": 1.2520589258786683,
+      "grad_norm": 9.16550352575417,
+      "learning_rate": 2e-06,
+      "loss": 0.2197,
+      "step": 5397
+    },
+    {
+      "epoch": 1.252290917526969,
+      "grad_norm": 13.833352896694109,
+      "learning_rate": 2e-06,
+      "loss": 0.2307,
+      "step": 5398
+    },
+    {
+      "epoch": 1.2525229091752696,
+      "grad_norm": 13.330435337641362,
+      "learning_rate": 2e-06,
+      "loss": 0.3172,
+      "step": 5399
+    },
+    {
+      "epoch": 1.2527549008235703,
+      "grad_norm": 13.575020941419025,
+      "learning_rate": 2e-06,
+      "loss": 0.3236,
+      "step": 5400
+    },
+    {
+      "epoch": 1.252986892471871,
+      "grad_norm": 15.24233503470825,
+      "learning_rate": 2e-06,
+      "loss": 0.1454,
+      "step": 5401
+    },
+    {
+      "epoch": 1.2532188841201717,
+      "grad_norm": 7.776532329048835,
+      "learning_rate": 2e-06,
+      "loss": 0.1922,
+      "step": 5402
+    },
+    {
+      "epoch": 1.2534508757684724,
+      "grad_norm": 10.196711226595042,
+      "learning_rate": 2e-06,
+      "loss": 0.2424,
+      "step": 5403
+    },
+    {
+      "epoch": 1.253682867416773,
+      "grad_norm": 14.61397139989457,
+      "learning_rate": 2e-06,
+      "loss": 0.205,
+      "step": 5404
+    },
+    {
+      "epoch": 1.2539148590650737,
+      "grad_norm": 10.510717889247282,
+      "learning_rate": 2e-06,
+      "loss": 0.2287,
+      "step": 5405
+    },
+    {
+      "epoch": 1.2541468507133744,
+      "grad_norm": 6.766161353966618,
+      "learning_rate": 2e-06,
+      "loss": 0.2261,
+      "step": 5406
+    },
+    {
+      "epoch": 1.254378842361675,
+      "grad_norm": 11.411526589754951,
+      "learning_rate": 2e-06,
+      "loss": 0.2431,
+      "step": 5407
+    },
+    {
+      "epoch": 1.2546108340099758,
+      "grad_norm": 12.28838509585074,
+      "learning_rate": 2e-06,
+      "loss": 0.1855,
+      "step": 5408
+    },
+    {
+      "epoch": 1.2548428256582764,
+      "grad_norm": 7.946790336662799,
+      "learning_rate": 2e-06,
+      "loss": 0.1554,
+      "step": 5409
+    },
+    {
+      "epoch": 1.2550748173065769,
+      "grad_norm": 12.619103129116512,
+      "learning_rate": 2e-06,
+      "loss": 0.1923,
+      "step": 5410
+    },
+    {
+      "epoch": 1.2553068089548776,
+      "grad_norm": 27.705684365921115,
+      "learning_rate": 2e-06,
+      "loss": 0.5062,
+      "step": 5411
+    },
+    {
+      "epoch": 1.2555388006031782,
+      "grad_norm": 4.933290693333517,
+      "learning_rate": 2e-06,
+      "loss": 0.1248,
+      "step": 5412
+    },
+    {
+      "epoch": 1.255770792251479,
+      "grad_norm": 24.902814613229886,
+      "learning_rate": 2e-06,
+      "loss": 0.337,
+      "step": 5413
+    },
+    {
+      "epoch": 1.2560027838997796,
+      "grad_norm": 10.21067892420391,
+      "learning_rate": 2e-06,
+      "loss": 0.2146,
+      "step": 5414
+    },
+    {
+      "epoch": 1.2562347755480803,
+      "grad_norm": 6.780520081375767,
+      "learning_rate": 2e-06,
+      "loss": 0.2351,
+      "step": 5415
+    },
+    {
+      "epoch": 1.256466767196381,
+      "grad_norm": 13.71098978847903,
+      "learning_rate": 2e-06,
+      "loss": 0.2876,
+      "step": 5416
+    },
+    {
+      "epoch": 1.2566987588446816,
+      "grad_norm": 10.855468980185774,
+      "learning_rate": 2e-06,
+      "loss": 0.1825,
+      "step": 5417
+    },
+    {
+      "epoch": 1.2569307504929823,
+      "grad_norm": 21.315008502857644,
+      "learning_rate": 2e-06,
+      "loss": 0.3071,
+      "step": 5418
+    },
+    {
+      "epoch": 1.257162742141283,
+      "grad_norm": 14.134423743002685,
+      "learning_rate": 2e-06,
+      "loss": 0.1695,
+      "step": 5419
+    },
+    {
+      "epoch": 1.2573947337895834,
+      "grad_norm": 15.135285920332297,
+      "learning_rate": 2e-06,
+      "loss": 0.1712,
+      "step": 5420
+    },
+    {
+      "epoch": 1.2576267254378841,
+      "grad_norm": 13.025679352989053,
+      "learning_rate": 2e-06,
+      "loss": 0.1629,
+      "step": 5421
+    },
+    {
+      "epoch": 1.2578587170861848,
+      "grad_norm": 15.418028878233873,
+      "learning_rate": 2e-06,
+      "loss": 0.1973,
+      "step": 5422
+    },
+    {
+      "epoch": 1.2580907087344855,
+      "grad_norm": 14.90544663133779,
+      "learning_rate": 2e-06,
+      "loss": 0.2366,
+      "step": 5423
+    },
+    {
+      "epoch": 1.2583227003827862,
+      "grad_norm": 17.13177952043397,
+      "learning_rate": 2e-06,
+      "loss": 0.2224,
+      "step": 5424
+    },
+    {
+      "epoch": 1.2585546920310868,
+      "grad_norm": 15.768336904910315,
+      "learning_rate": 2e-06,
+      "loss": 0.1723,
+      "step": 5425
+    },
+    {
+      "epoch": 1.2587866836793875,
+      "grad_norm": 16.784910745172407,
+      "learning_rate": 2e-06,
+      "loss": 0.2809,
+      "step": 5426
+    },
+    {
+      "epoch": 1.2590186753276882,
+      "grad_norm": 14.877686438432477,
+      "learning_rate": 2e-06,
+      "loss": 0.2759,
+      "step": 5427
+    },
+    {
+      "epoch": 1.2592506669759889,
+      "grad_norm": 9.201621848621404,
+      "learning_rate": 2e-06,
+      "loss": 0.1827,
+      "step": 5428
+    },
+    {
+      "epoch": 1.2594826586242895,
+      "grad_norm": 14.700888196853745,
+      "learning_rate": 2e-06,
+      "loss": 0.1968,
+      "step": 5429
+    },
+    {
+      "epoch": 1.2597146502725902,
+      "grad_norm": 18.81228873623378,
+      "learning_rate": 2e-06,
+      "loss": 0.3326,
+      "step": 5430
+    },
+    {
+      "epoch": 1.259946641920891,
+      "grad_norm": 12.9065843099499,
+      "learning_rate": 2e-06,
+      "loss": 0.2946,
+      "step": 5431
+    },
+    {
+      "epoch": 1.2601786335691916,
+      "grad_norm": 9.30669535994372,
+      "learning_rate": 2e-06,
+      "loss": 0.2333,
+      "step": 5432
+    },
+    {
+      "epoch": 1.2604106252174923,
+      "grad_norm": 9.267711029265648,
+      "learning_rate": 2e-06,
+      "loss": 0.2008,
+      "step": 5433
+    },
+    {
+      "epoch": 1.260642616865793,
+      "grad_norm": 26.074534924266477,
+      "learning_rate": 2e-06,
+      "loss": 0.3929,
+      "step": 5434
+    },
+    {
+      "epoch": 1.2608746085140936,
+      "grad_norm": 13.773153795926959,
+      "learning_rate": 2e-06,
+      "loss": 0.1561,
+      "step": 5435
+    },
+    {
+      "epoch": 1.2611066001623943,
+      "grad_norm": 20.51695528132908,
+      "learning_rate": 2e-06,
+      "loss": 0.1994,
+      "step": 5436
+    },
+    {
+      "epoch": 1.2613385918106947,
+      "grad_norm": 13.866096902646714,
+      "learning_rate": 2e-06,
+      "loss": 0.25,
+      "step": 5437
+    },
+    {
+      "epoch": 1.2615705834589954,
+      "grad_norm": 15.601652672968376,
+      "learning_rate": 2e-06,
+      "loss": 0.1956,
+      "step": 5438
+    },
+    {
+      "epoch": 1.261802575107296,
+      "grad_norm": 9.407943419133732,
+      "learning_rate": 2e-06,
+      "loss": 0.2155,
+      "step": 5439
+    },
+    {
+      "epoch": 1.2620345667555968,
+      "grad_norm": 21.304630727887947,
+      "learning_rate": 2e-06,
+      "loss": 0.2215,
+      "step": 5440
+    },
+    {
+      "epoch": 1.2622665584038975,
+      "grad_norm": 18.56505166670855,
+      "learning_rate": 2e-06,
+      "loss": 0.3319,
+      "step": 5441
+    },
+    {
+      "epoch": 1.2624985500521981,
+      "grad_norm": 11.90138217449211,
+      "learning_rate": 2e-06,
+      "loss": 0.2573,
+      "step": 5442
+    },
+    {
+      "epoch": 1.2627305417004988,
+      "grad_norm": 15.059252412285858,
+      "learning_rate": 2e-06,
+      "loss": 0.2218,
+      "step": 5443
+    },
+    {
+      "epoch": 1.2629625333487995,
+      "grad_norm": 8.775996304124295,
+      "learning_rate": 2e-06,
+      "loss": 0.1364,
+      "step": 5444
+    },
+    {
+      "epoch": 1.2631945249971002,
+      "grad_norm": 12.641643012310846,
+      "learning_rate": 2e-06,
+      "loss": 0.2702,
+      "step": 5445
+    },
+    {
+      "epoch": 1.2634265166454008,
+      "grad_norm": 20.388734064223414,
+      "learning_rate": 2e-06,
+      "loss": 0.3179,
+      "step": 5446
+    },
+    {
+      "epoch": 1.2636585082937013,
+      "grad_norm": 17.263471978052475,
+      "learning_rate": 2e-06,
+      "loss": 0.3227,
+      "step": 5447
+    },
+    {
+      "epoch": 1.263890499942002,
+      "grad_norm": 13.912072866064792,
+      "learning_rate": 2e-06,
+      "loss": 0.2209,
+      "step": 5448
+    },
+    {
+      "epoch": 1.2641224915903027,
+      "grad_norm": 17.87351186270327,
+      "learning_rate": 2e-06,
+      "loss": 0.347,
+      "step": 5449
+    },
+    {
+      "epoch": 1.2643544832386033,
+      "grad_norm": 20.927462227177482,
+      "learning_rate": 2e-06,
+      "loss": 0.3443,
+      "step": 5450
+    },
+    {
+      "epoch": 1.264586474886904,
+      "grad_norm": 14.92230797263229,
+      "learning_rate": 2e-06,
+      "loss": 0.2529,
+      "step": 5451
+    },
+    {
+      "epoch": 1.2648184665352047,
+      "grad_norm": 11.658954372633668,
+      "learning_rate": 2e-06,
+      "loss": 0.1356,
+      "step": 5452
+    },
+    {
+      "epoch": 1.2650504581835054,
+      "grad_norm": 16.889816040298218,
+      "learning_rate": 2e-06,
+      "loss": 0.2544,
+      "step": 5453
+    },
+    {
+      "epoch": 1.265282449831806,
+      "grad_norm": 17.986590759970184,
+      "learning_rate": 2e-06,
+      "loss": 0.2652,
+      "step": 5454
+    },
+    {
+      "epoch": 1.2655144414801067,
+      "grad_norm": 15.375217352482847,
+      "learning_rate": 2e-06,
+      "loss": 0.2162,
+      "step": 5455
+    },
+    {
+      "epoch": 1.2657464331284074,
+      "grad_norm": 13.563201322950821,
+      "learning_rate": 2e-06,
+      "loss": 0.2364,
+      "step": 5456
+    },
+    {
+      "epoch": 1.265978424776708,
+      "grad_norm": 14.413978406555218,
+      "learning_rate": 2e-06,
+      "loss": 0.2424,
+      "step": 5457
+    },
+    {
+      "epoch": 1.2662104164250088,
+      "grad_norm": 16.31505711897065,
+      "learning_rate": 2e-06,
+      "loss": 0.2526,
+      "step": 5458
+    },
+    {
+      "epoch": 1.2664424080733094,
+      "grad_norm": 8.752232702803143,
+      "learning_rate": 2e-06,
+      "loss": 0.1601,
+      "step": 5459
+    },
+    {
+      "epoch": 1.2666743997216101,
+      "grad_norm": 9.794706568555943,
+      "learning_rate": 2e-06,
+      "loss": 0.1836,
+      "step": 5460
+    },
+    {
+      "epoch": 1.2669063913699108,
+      "grad_norm": 13.647159948468177,
+      "learning_rate": 2e-06,
+      "loss": 0.2018,
+      "step": 5461
+    },
+    {
+      "epoch": 1.2671383830182115,
+      "grad_norm": 14.812271457713734,
+      "learning_rate": 2e-06,
+      "loss": 0.2389,
+      "step": 5462
+    },
+    {
+      "epoch": 1.267370374666512,
+      "grad_norm": 8.905656894633575,
+      "learning_rate": 2e-06,
+      "loss": 0.1496,
+      "step": 5463
+    },
+    {
+      "epoch": 1.2676023663148126,
+      "grad_norm": 9.256200576462955,
+      "learning_rate": 2e-06,
+      "loss": 0.1749,
+      "step": 5464
+    },
+    {
+      "epoch": 1.2678343579631133,
+      "grad_norm": 12.138821016330132,
+      "learning_rate": 2e-06,
+      "loss": 0.3476,
+      "step": 5465
+    },
+    {
+      "epoch": 1.268066349611414,
+      "grad_norm": 14.589430451668205,
+      "learning_rate": 2e-06,
+      "loss": 0.1956,
+      "step": 5466
+    },
+    {
+      "epoch": 1.2682983412597146,
+      "grad_norm": 18.385243783916064,
+      "learning_rate": 2e-06,
+      "loss": 0.2236,
+      "step": 5467
+    },
+    {
+      "epoch": 1.2685303329080153,
+      "grad_norm": 15.912545380757068,
+      "learning_rate": 2e-06,
+      "loss": 0.2797,
+      "step": 5468
+    },
+    {
+      "epoch": 1.268762324556316,
+      "grad_norm": 24.290486722780003,
+      "learning_rate": 2e-06,
+      "loss": 0.336,
+      "step": 5469
+    },
+    {
+      "epoch": 1.2689943162046167,
+      "grad_norm": 17.68733338274814,
+      "learning_rate": 2e-06,
+      "loss": 0.3079,
+      "step": 5470
+    },
+    {
+      "epoch": 1.2692263078529173,
+      "grad_norm": 29.58373573181717,
+      "learning_rate": 2e-06,
+      "loss": 0.2578,
+      "step": 5471
+    },
+    {
+      "epoch": 1.269458299501218,
+      "grad_norm": 21.49698383474083,
+      "learning_rate": 2e-06,
+      "loss": 0.2823,
+      "step": 5472
+    },
+    {
+      "epoch": 1.2696902911495185,
+      "grad_norm": 20.99655313594657,
+      "learning_rate": 2e-06,
+      "loss": 0.3464,
+      "step": 5473
+    },
+    {
+      "epoch": 1.2699222827978192,
+      "grad_norm": 12.60122524250526,
+      "learning_rate": 2e-06,
+      "loss": 0.404,
+      "step": 5474
+    },
+    {
+      "epoch": 1.2701542744461198,
+      "grad_norm": 11.14334254645051,
+      "learning_rate": 2e-06,
+      "loss": 0.259,
+      "step": 5475
+    },
+    {
+      "epoch": 1.2703862660944205,
+      "grad_norm": 6.256460629933859,
+      "learning_rate": 2e-06,
+      "loss": 0.1994,
+      "step": 5476
+    },
+    {
+      "epoch": 1.2706182577427212,
+      "grad_norm": 15.74119024883826,
+      "learning_rate": 2e-06,
+      "loss": 0.2372,
+      "step": 5477
+    },
+    {
+      "epoch": 1.2708502493910219,
+      "grad_norm": 22.71832340845438,
+      "learning_rate": 2e-06,
+      "loss": 0.3303,
+      "step": 5478
+    },
+    {
+      "epoch": 1.2710822410393225,
+      "grad_norm": 9.136667866436637,
+      "learning_rate": 2e-06,
+      "loss": 0.214,
+      "step": 5479
+    },
+    {
+      "epoch": 1.2713142326876232,
+      "grad_norm": 22.00409773731631,
+      "learning_rate": 2e-06,
+      "loss": 0.2791,
+      "step": 5480
+    },
+    {
+      "epoch": 1.271546224335924,
+      "grad_norm": 14.514457099733832,
+      "learning_rate": 2e-06,
+      "loss": 0.1993,
+      "step": 5481
+    },
+    {
+      "epoch": 1.2717782159842246,
+      "grad_norm": 19.24414215082918,
+      "learning_rate": 2e-06,
+      "loss": 0.239,
+      "step": 5482
+    },
+    {
+      "epoch": 1.2720102076325253,
+      "grad_norm": 10.895891454072098,
+      "learning_rate": 2e-06,
+      "loss": 0.1715,
+      "step": 5483
+    },
+    {
+      "epoch": 1.272242199280826,
+      "grad_norm": 9.632069525604589,
+      "learning_rate": 2e-06,
+      "loss": 0.328,
+      "step": 5484
+    },
+    {
+      "epoch": 1.2724741909291266,
+      "grad_norm": 9.230036573343929,
+      "learning_rate": 2e-06,
+      "loss": 0.1802,
+      "step": 5485
+    },
+    {
+      "epoch": 1.2727061825774273,
+      "grad_norm": 8.919987934719751,
+      "learning_rate": 2e-06,
+      "loss": 0.1919,
+      "step": 5486
+    },
+    {
+      "epoch": 1.272938174225728,
+      "grad_norm": 21.874171401796,
+      "learning_rate": 2e-06,
+      "loss": 0.4171,
+      "step": 5487
+    },
+    {
+      "epoch": 1.2731701658740286,
+      "grad_norm": 21.073806899512824,
+      "learning_rate": 2e-06,
+      "loss": 0.2203,
+      "step": 5488
+    },
+    {
+      "epoch": 1.2734021575223293,
+      "grad_norm": 19.802684449859967,
+      "learning_rate": 2e-06,
+      "loss": 0.2651,
+      "step": 5489
+    },
+    {
+      "epoch": 1.2736341491706298,
+      "grad_norm": 8.130642986566825,
+      "learning_rate": 2e-06,
+      "loss": 0.2091,
+      "step": 5490
+    },
+    {
+      "epoch": 1.2738661408189305,
+      "grad_norm": 16.65568102886136,
+      "learning_rate": 2e-06,
+      "loss": 0.195,
+      "step": 5491
+    },
+    {
+      "epoch": 1.2740981324672311,
+      "grad_norm": 46.93369863196454,
+      "learning_rate": 2e-06,
+      "loss": 0.2481,
+      "step": 5492
+    },
+    {
+      "epoch": 1.2743301241155318,
+      "grad_norm": 19.882996928850734,
+      "learning_rate": 2e-06,
+      "loss": 0.3797,
+      "step": 5493
+    },
+    {
+      "epoch": 1.2745621157638325,
+      "grad_norm": 6.786034128081566,
+      "learning_rate": 2e-06,
+      "loss": 0.1929,
+      "step": 5494
+    },
+    {
+      "epoch": 1.2747941074121332,
+      "grad_norm": 10.954183622735066,
+      "learning_rate": 2e-06,
+      "loss": 0.1846,
+      "step": 5495
+    },
+    {
+      "epoch": 1.2750260990604338,
+      "grad_norm": 17.289913825248284,
+      "learning_rate": 2e-06,
+      "loss": 0.2575,
+      "step": 5496
+    },
+    {
+      "epoch": 1.2752580907087345,
+      "grad_norm": 9.900332056828477,
+      "learning_rate": 2e-06,
+      "loss": 0.1919,
+      "step": 5497
+    },
+    {
+      "epoch": 1.2754900823570352,
+      "grad_norm": 11.216172941552085,
+      "learning_rate": 2e-06,
+      "loss": 0.2311,
+      "step": 5498
+    },
+    {
+      "epoch": 1.2757220740053359,
+      "grad_norm": 18.067500250176654,
+      "learning_rate": 2e-06,
+      "loss": 0.2861,
+      "step": 5499
+    },
+    {
+      "epoch": 1.2759540656536363,
+      "grad_norm": 9.544040849535888,
+      "learning_rate": 2e-06,
+      "loss": 0.2313,
+      "step": 5500
+    },
+    {
+      "epoch": 1.276186057301937,
+      "grad_norm": 19.109610262028383,
+      "learning_rate": 2e-06,
+      "loss": 0.2947,
+      "step": 5501
+    },
+    {
+      "epoch": 1.2764180489502377,
+      "grad_norm": 8.074254510862858,
+      "learning_rate": 2e-06,
+      "loss": 0.1712,
+      "step": 5502
+    },
+    {
+      "epoch": 1.2766500405985384,
+      "grad_norm": 16.010980414431298,
+      "learning_rate": 2e-06,
+      "loss": 0.3747,
+      "step": 5503
+    },
+    {
+      "epoch": 1.276882032246839,
+      "grad_norm": 13.51093579441853,
+      "learning_rate": 2e-06,
+      "loss": 0.2592,
+      "step": 5504
+    },
+    {
+      "epoch": 1.2771140238951397,
+      "grad_norm": 14.669785165177656,
+      "learning_rate": 2e-06,
+      "loss": 0.2628,
+      "step": 5505
+    },
+    {
+      "epoch": 1.2773460155434404,
+      "grad_norm": 7.660559763630664,
+      "learning_rate": 2e-06,
+      "loss": 0.1481,
+      "step": 5506
+    },
+    {
+      "epoch": 1.277578007191741,
+      "grad_norm": 20.24682997014625,
+      "learning_rate": 2e-06,
+      "loss": 0.3538,
+      "step": 5507
+    },
+    {
+      "epoch": 1.2778099988400418,
+      "grad_norm": 13.362958359090564,
+      "learning_rate": 2e-06,
+      "loss": 0.2271,
+      "step": 5508
+    },
+    {
+      "epoch": 1.2780419904883424,
+      "grad_norm": 8.164473390361263,
+      "learning_rate": 2e-06,
+      "loss": 0.1568,
+      "step": 5509
+    },
+    {
+      "epoch": 1.2782739821366431,
+      "grad_norm": 6.75118446837882,
+      "learning_rate": 2e-06,
+      "loss": 0.1275,
+      "step": 5510
+    },
+    {
+      "epoch": 1.2785059737849438,
+      "grad_norm": 8.801135729909866,
+      "learning_rate": 2e-06,
+      "loss": 0.1721,
+      "step": 5511
+    },
+    {
+      "epoch": 1.2787379654332445,
+      "grad_norm": 19.006232049977932,
+      "learning_rate": 2e-06,
+      "loss": 0.3433,
+      "step": 5512
+    },
+    {
+      "epoch": 1.2789699570815452,
+      "grad_norm": 8.830534900711758,
+      "learning_rate": 2e-06,
+      "loss": 0.1636,
+      "step": 5513
+    },
+    {
+      "epoch": 1.2792019487298458,
+      "grad_norm": 3.593403289783236,
+      "learning_rate": 2e-06,
+      "loss": 0.1181,
+      "step": 5514
+    },
+    {
+      "epoch": 1.2794339403781465,
+      "grad_norm": 20.464133931736466,
+      "learning_rate": 2e-06,
+      "loss": 0.3697,
+      "step": 5515
+    },
+    {
+      "epoch": 1.2796659320264472,
+      "grad_norm": 22.020806574162382,
+      "learning_rate": 2e-06,
+      "loss": 0.2337,
+      "step": 5516
+    },
+    {
+      "epoch": 1.2798979236747476,
+      "grad_norm": 11.288169129492667,
+      "learning_rate": 2e-06,
+      "loss": 0.2503,
+      "step": 5517
+    },
+    {
+      "epoch": 1.2801299153230483,
+      "grad_norm": 12.874188117641705,
+      "learning_rate": 2e-06,
+      "loss": 0.3134,
+      "step": 5518
+    },
+    {
+      "epoch": 1.280361906971349,
+      "grad_norm": 12.527139165639522,
+      "learning_rate": 2e-06,
+      "loss": 0.3227,
+      "step": 5519
+    },
+    {
+      "epoch": 1.2805938986196497,
+      "grad_norm": 14.050860916922632,
+      "learning_rate": 2e-06,
+      "loss": 0.4138,
+      "step": 5520
+    },
+    {
+      "epoch": 1.2808258902679504,
+      "grad_norm": 17.575926637021528,
+      "learning_rate": 2e-06,
+      "loss": 0.4182,
+      "step": 5521
+    },
+    {
+      "epoch": 1.281057881916251,
+      "grad_norm": 38.30070940939618,
+      "learning_rate": 2e-06,
+      "loss": 0.2344,
+      "step": 5522
+    },
+    {
+      "epoch": 1.2812898735645517,
+      "grad_norm": 15.721424393613805,
+      "learning_rate": 2e-06,
+      "loss": 0.2039,
+      "step": 5523
+    },
+    {
+      "epoch": 1.2815218652128524,
+      "grad_norm": 14.438274765990126,
+      "learning_rate": 2e-06,
+      "loss": 0.2811,
+      "step": 5524
+    },
+    {
+      "epoch": 1.281753856861153,
+      "grad_norm": 11.964371077643447,
+      "learning_rate": 2e-06,
+      "loss": 0.1871,
+      "step": 5525
+    },
+    {
+      "epoch": 1.2819858485094537,
+      "grad_norm": 11.192127859492148,
+      "learning_rate": 2e-06,
+      "loss": 0.2232,
+      "step": 5526
+    },
+    {
+      "epoch": 1.2822178401577542,
+      "grad_norm": 16.55720382812037,
+      "learning_rate": 2e-06,
+      "loss": 0.3419,
+      "step": 5527
+    },
+    {
+      "epoch": 1.2824498318060549,
+      "grad_norm": 11.859714499129243,
+      "learning_rate": 2e-06,
+      "loss": 0.3109,
+      "step": 5528
+    },
+    {
+      "epoch": 1.2826818234543556,
+      "grad_norm": 12.773674037445563,
+      "learning_rate": 2e-06,
+      "loss": 0.2546,
+      "step": 5529
+    },
+    {
+      "epoch": 1.2829138151026562,
+      "grad_norm": 20.39897546810617,
+      "learning_rate": 2e-06,
+      "loss": 0.3194,
+      "step": 5530
+    },
+    {
+      "epoch": 1.283145806750957,
+      "grad_norm": 16.604649378816013,
+      "learning_rate": 2e-06,
+      "loss": 0.2979,
+      "step": 5531
+    },
+    {
+      "epoch": 1.2833777983992576,
+      "grad_norm": 10.145216773558865,
+      "learning_rate": 2e-06,
+      "loss": 0.1958,
+      "step": 5532
+    },
+    {
+      "epoch": 1.2836097900475583,
+      "grad_norm": 12.156091156673625,
+      "learning_rate": 2e-06,
+      "loss": 0.1559,
+      "step": 5533
+    },
+    {
+      "epoch": 1.283841781695859,
+      "grad_norm": 14.025709421254515,
+      "learning_rate": 2e-06,
+      "loss": 0.1721,
+      "step": 5534
+    },
+    {
+      "epoch": 1.2840737733441596,
+      "grad_norm": 11.107260139493441,
+      "learning_rate": 2e-06,
+      "loss": 0.266,
+      "step": 5535
+    },
+    {
+      "epoch": 1.2843057649924603,
+      "grad_norm": 11.576453611507793,
+      "learning_rate": 2e-06,
+      "loss": 0.2343,
+      "step": 5536
+    },
+    {
+      "epoch": 1.284537756640761,
+      "grad_norm": 8.007601321516523,
+      "learning_rate": 2e-06,
+      "loss": 0.1732,
+      "step": 5537
+    },
+    {
+      "epoch": 1.2847697482890617,
+      "grad_norm": 11.928698418916918,
+      "learning_rate": 2e-06,
+      "loss": 0.1612,
+      "step": 5538
+    },
+    {
+      "epoch": 1.2850017399373623,
+      "grad_norm": 17.465166018146665,
+      "learning_rate": 2e-06,
+      "loss": 0.3162,
+      "step": 5539
+    },
+    {
+      "epoch": 1.285233731585663,
+      "grad_norm": 5.783970569015845,
+      "learning_rate": 2e-06,
+      "loss": 0.1845,
+      "step": 5540
+    },
+    {
+      "epoch": 1.2854657232339637,
+      "grad_norm": 9.157428800986871,
+      "learning_rate": 2e-06,
+      "loss": 0.2112,
+      "step": 5541
+    },
+    {
+      "epoch": 1.2856977148822644,
+      "grad_norm": 4.1076278120894,
+      "learning_rate": 2e-06,
+      "loss": 0.1123,
+      "step": 5542
+    },
+    {
+      "epoch": 1.2859297065305648,
+      "grad_norm": 10.11312397969375,
+      "learning_rate": 2e-06,
+      "loss": 0.1765,
+      "step": 5543
+    },
+    {
+      "epoch": 1.2861616981788655,
+      "grad_norm": 11.274897323273564,
+      "learning_rate": 2e-06,
+      "loss": 0.1913,
+      "step": 5544
+    },
+    {
+      "epoch": 1.2863936898271662,
+      "grad_norm": 19.606507289753136,
+      "learning_rate": 2e-06,
+      "loss": 0.3079,
+      "step": 5545
+    },
+    {
+      "epoch": 1.2866256814754669,
+      "grad_norm": 9.684645461882754,
+      "learning_rate": 2e-06,
+      "loss": 0.1835,
+      "step": 5546
+    },
+    {
+      "epoch": 1.2868576731237675,
+      "grad_norm": 9.737675214521724,
+      "learning_rate": 2e-06,
+      "loss": 0.2299,
+      "step": 5547
+    },
+    {
+      "epoch": 1.2870896647720682,
+      "grad_norm": 10.599090403823574,
+      "learning_rate": 2e-06,
+      "loss": 0.2373,
+      "step": 5548
+    },
+    {
+      "epoch": 1.2873216564203689,
+      "grad_norm": 16.630432627012777,
+      "learning_rate": 2e-06,
+      "loss": 0.3984,
+      "step": 5549
+    },
+    {
+      "epoch": 1.2875536480686696,
+      "grad_norm": 17.729293437700733,
+      "learning_rate": 2e-06,
+      "loss": 0.2671,
+      "step": 5550
+    },
+    {
+      "epoch": 1.2877856397169702,
+      "grad_norm": 12.05835176586732,
+      "learning_rate": 2e-06,
+      "loss": 0.2906,
+      "step": 5551
+    },
+    {
+      "epoch": 1.288017631365271,
+      "grad_norm": 14.440985815174287,
+      "learning_rate": 2e-06,
+      "loss": 0.2058,
+      "step": 5552
+    },
+    {
+      "epoch": 1.2882496230135714,
+      "grad_norm": 16.84003641847929,
+      "learning_rate": 2e-06,
+      "loss": 0.32,
+      "step": 5553
+    },
+    {
+      "epoch": 1.288481614661872,
+      "grad_norm": 20.495041375308226,
+      "learning_rate": 2e-06,
+      "loss": 0.1994,
+      "step": 5554
+    },
+    {
+      "epoch": 1.2887136063101727,
+      "grad_norm": 21.208082266804475,
+      "learning_rate": 2e-06,
+      "loss": 0.4303,
+      "step": 5555
+    },
+    {
+      "epoch": 1.2889455979584734,
+      "grad_norm": 10.462591620428805,
+      "learning_rate": 2e-06,
+      "loss": 0.1862,
+      "step": 5556
+    },
+    {
+      "epoch": 1.289177589606774,
+      "grad_norm": 33.57346108492076,
+      "learning_rate": 2e-06,
+      "loss": 0.4521,
+      "step": 5557
+    },
+    {
+      "epoch": 1.2894095812550748,
+      "grad_norm": 9.690303304788598,
+      "learning_rate": 2e-06,
+      "loss": 0.1659,
+      "step": 5558
+    },
+    {
+      "epoch": 1.2896415729033754,
+      "grad_norm": 26.12477597482978,
+      "learning_rate": 2e-06,
+      "loss": 0.3697,
+      "step": 5559
+    },
+    {
+      "epoch": 1.2898735645516761,
+      "grad_norm": 10.753843000487956,
+      "learning_rate": 2e-06,
+      "loss": 0.1845,
+      "step": 5560
+    },
+    {
+      "epoch": 1.2901055561999768,
+      "grad_norm": 11.94320212794658,
+      "learning_rate": 2e-06,
+      "loss": 0.3165,
+      "step": 5561
+    },
+    {
+      "epoch": 1.2903375478482775,
+      "grad_norm": 8.074090848780601,
+      "learning_rate": 2e-06,
+      "loss": 0.1782,
+      "step": 5562
+    },
+    {
+      "epoch": 1.2905695394965782,
+      "grad_norm": 9.986874563274212,
+      "learning_rate": 2e-06,
+      "loss": 0.2196,
+      "step": 5563
+    },
+    {
+      "epoch": 1.2908015311448788,
+      "grad_norm": 13.88506300781901,
+      "learning_rate": 2e-06,
+      "loss": 0.2352,
+      "step": 5564
+    },
+    {
+      "epoch": 1.2910335227931795,
+      "grad_norm": 12.780056004774648,
+      "learning_rate": 2e-06,
+      "loss": 0.206,
+      "step": 5565
+    },
+    {
+      "epoch": 1.2912655144414802,
+      "grad_norm": 17.948578222912435,
+      "learning_rate": 2e-06,
+      "loss": 0.3228,
+      "step": 5566
+    },
+    {
+      "epoch": 1.2914975060897809,
+      "grad_norm": 8.911005690791237,
+      "learning_rate": 2e-06,
+      "loss": 0.2329,
+      "step": 5567
+    },
+    {
+      "epoch": 1.2917294977380815,
+      "grad_norm": 9.354454545488181,
+      "learning_rate": 2e-06,
+      "loss": 0.2418,
+      "step": 5568
+    },
+    {
+      "epoch": 1.2919614893863822,
+      "grad_norm": 21.323098522700068,
+      "learning_rate": 2e-06,
+      "loss": 0.2735,
+      "step": 5569
+    },
+    {
+      "epoch": 1.2921934810346827,
+      "grad_norm": 15.446059925915666,
+      "learning_rate": 2e-06,
+      "loss": 0.2856,
+      "step": 5570
+    },
+    {
+      "epoch": 1.2924254726829834,
+      "grad_norm": 8.774450735495995,
+      "learning_rate": 2e-06,
+      "loss": 0.2065,
+      "step": 5571
+    },
+    {
+      "epoch": 1.292657464331284,
+      "grad_norm": 8.875266415022931,
+      "learning_rate": 2e-06,
+      "loss": 0.1516,
+      "step": 5572
+    },
+    {
+      "epoch": 1.2928894559795847,
+      "grad_norm": 25.699395396889404,
+      "learning_rate": 2e-06,
+      "loss": 0.1989,
+      "step": 5573
+    },
+    {
+      "epoch": 1.2931214476278854,
+      "grad_norm": 13.780637111179585,
+      "learning_rate": 2e-06,
+      "loss": 0.1974,
+      "step": 5574
+    },
+    {
+      "epoch": 1.293353439276186,
+      "grad_norm": 10.596894116088386,
+      "learning_rate": 2e-06,
+      "loss": 0.1873,
+      "step": 5575
+    },
+    {
+      "epoch": 1.2935854309244867,
+      "grad_norm": 15.882696742741262,
+      "learning_rate": 2e-06,
+      "loss": 0.275,
+      "step": 5576
+    },
+    {
+      "epoch": 1.2938174225727874,
+      "grad_norm": 16.749738561496294,
+      "learning_rate": 2e-06,
+      "loss": 0.2766,
+      "step": 5577
+    },
+    {
+      "epoch": 1.294049414221088,
+      "grad_norm": 22.970839761942415,
+      "learning_rate": 2e-06,
+      "loss": 0.303,
+      "step": 5578
+    },
+    {
+      "epoch": 1.2942814058693888,
+      "grad_norm": 10.956263232344353,
+      "learning_rate": 2e-06,
+      "loss": 0.2567,
+      "step": 5579
+    },
+    {
+      "epoch": 1.2945133975176892,
+      "grad_norm": 12.000426780774758,
+      "learning_rate": 2e-06,
+      "loss": 0.2221,
+      "step": 5580
+    },
+    {
+      "epoch": 1.29474538916599,
+      "grad_norm": 11.959888501070827,
+      "learning_rate": 2e-06,
+      "loss": 0.2105,
+      "step": 5581
+    },
+    {
+      "epoch": 1.2949773808142906,
+      "grad_norm": 13.004029224978572,
+      "learning_rate": 2e-06,
+      "loss": 0.2816,
+      "step": 5582
+    },
+    {
+      "epoch": 1.2952093724625913,
+      "grad_norm": 16.552747214762597,
+      "learning_rate": 2e-06,
+      "loss": 0.2822,
+      "step": 5583
+    },
+    {
+      "epoch": 1.295441364110892,
+      "grad_norm": 7.828758165100487,
+      "learning_rate": 2e-06,
+      "loss": 0.1445,
+      "step": 5584
+    },
+    {
+      "epoch": 1.2956733557591926,
+      "grad_norm": 9.69840028310193,
+      "learning_rate": 2e-06,
+      "loss": 0.192,
+      "step": 5585
+    },
+    {
+      "epoch": 1.2959053474074933,
+      "grad_norm": 8.594676831269815,
+      "learning_rate": 2e-06,
+      "loss": 0.2646,
+      "step": 5586
+    },
+    {
+      "epoch": 1.296137339055794,
+      "grad_norm": 19.72568385506288,
+      "learning_rate": 2e-06,
+      "loss": 0.4113,
+      "step": 5587
+    },
+    {
+      "epoch": 1.2963693307040947,
+      "grad_norm": 11.60141821509818,
+      "learning_rate": 2e-06,
+      "loss": 0.2368,
+      "step": 5588
+    },
+    {
+      "epoch": 1.2966013223523953,
+      "grad_norm": 11.360073503368694,
+      "learning_rate": 2e-06,
+      "loss": 0.1508,
+      "step": 5589
+    },
+    {
+      "epoch": 1.296833314000696,
+      "grad_norm": 11.466245032578335,
+      "learning_rate": 2e-06,
+      "loss": 0.1911,
+      "step": 5590
+    },
+    {
+      "epoch": 1.2970653056489967,
+      "grad_norm": 11.352394051290077,
+      "learning_rate": 2e-06,
+      "loss": 0.1763,
+      "step": 5591
+    },
+    {
+      "epoch": 1.2972972972972974,
+      "grad_norm": 16.11533083404958,
+      "learning_rate": 2e-06,
+      "loss": 0.2902,
+      "step": 5592
+    },
+    {
+      "epoch": 1.297529288945598,
+      "grad_norm": 12.36723660219495,
+      "learning_rate": 2e-06,
+      "loss": 0.2096,
+      "step": 5593
+    },
+    {
+      "epoch": 1.2977612805938987,
+      "grad_norm": 30.89625173241812,
+      "learning_rate": 2e-06,
+      "loss": 0.2666,
+      "step": 5594
+    },
+    {
+      "epoch": 1.2979932722421994,
+      "grad_norm": 19.514344172121138,
+      "learning_rate": 2e-06,
+      "loss": 0.2418,
+      "step": 5595
+    },
+    {
+      "epoch": 1.2982252638905,
+      "grad_norm": 10.788689229972393,
+      "learning_rate": 2e-06,
+      "loss": 0.1786,
+      "step": 5596
+    },
+    {
+      "epoch": 1.2984572555388005,
+      "grad_norm": 15.45400603168222,
+      "learning_rate": 2e-06,
+      "loss": 0.2733,
+      "step": 5597
+    },
+    {
+      "epoch": 1.2986892471871012,
+      "grad_norm": 13.722921586738387,
+      "learning_rate": 2e-06,
+      "loss": 0.2962,
+      "step": 5598
+    },
+    {
+      "epoch": 1.298921238835402,
+      "grad_norm": 19.39647911275564,
+      "learning_rate": 2e-06,
+      "loss": 0.3514,
+      "step": 5599
+    },
+    {
+      "epoch": 1.2991532304837026,
+      "grad_norm": 15.274235619440297,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 5600
+    },
+    {
+      "epoch": 1.2993852221320032,
+      "grad_norm": 16.11752315764069,
+      "learning_rate": 2e-06,
+      "loss": 0.2394,
+      "step": 5601
+    },
+    {
+      "epoch": 1.299617213780304,
+      "grad_norm": 24.618352423068448,
+      "learning_rate": 2e-06,
+      "loss": 0.3625,
+      "step": 5602
+    },
+    {
+      "epoch": 1.2998492054286046,
+      "grad_norm": 24.593954311058177,
+      "learning_rate": 2e-06,
+      "loss": 0.2486,
+      "step": 5603
+    },
+    {
+      "epoch": 1.3000811970769053,
+      "grad_norm": 23.962619857204736,
+      "learning_rate": 2e-06,
+      "loss": 0.2788,
+      "step": 5604
+    },
+    {
+      "epoch": 1.300313188725206,
+      "grad_norm": 13.745233766397064,
+      "learning_rate": 2e-06,
+      "loss": 0.292,
+      "step": 5605
+    },
+    {
+      "epoch": 1.3005451803735064,
+      "grad_norm": 184.99953181781865,
+      "learning_rate": 2e-06,
+      "loss": 0.2061,
+      "step": 5606
+    },
+    {
+      "epoch": 1.300777172021807,
+      "grad_norm": 11.009695789207703,
+      "learning_rate": 2e-06,
+      "loss": 0.2157,
+      "step": 5607
+    },
+    {
+      "epoch": 1.3010091636701078,
+      "grad_norm": 15.03713330803401,
+      "learning_rate": 2e-06,
+      "loss": 0.2397,
+      "step": 5608
+    },
+    {
+      "epoch": 1.3012411553184084,
+      "grad_norm": 17.794040596623752,
+      "learning_rate": 2e-06,
+      "loss": 0.2294,
+      "step": 5609
+    },
+    {
+      "epoch": 1.3014731469667091,
+      "grad_norm": 12.91998866491081,
+      "learning_rate": 2e-06,
+      "loss": 0.2401,
+      "step": 5610
+    },
+    {
+      "epoch": 1.3017051386150098,
+      "grad_norm": 15.712347452714706,
+      "learning_rate": 2e-06,
+      "loss": 0.1369,
+      "step": 5611
+    },
+    {
+      "epoch": 1.3019371302633105,
+      "grad_norm": 18.194023250326826,
+      "learning_rate": 2e-06,
+      "loss": 0.2016,
+      "step": 5612
+    },
+    {
+      "epoch": 1.3021691219116112,
+      "grad_norm": 10.73759855968218,
+      "learning_rate": 2e-06,
+      "loss": 0.1686,
+      "step": 5613
+    },
+    {
+      "epoch": 1.3024011135599118,
+      "grad_norm": 12.442952329564145,
+      "learning_rate": 2e-06,
+      "loss": 0.2391,
+      "step": 5614
+    },
+    {
+      "epoch": 1.3026331052082125,
+      "grad_norm": 23.27162006463831,
+      "learning_rate": 2e-06,
+      "loss": 0.2731,
+      "step": 5615
+    },
+    {
+      "epoch": 1.3028650968565132,
+      "grad_norm": 15.22373633972883,
+      "learning_rate": 2e-06,
+      "loss": 0.3134,
+      "step": 5616
+    },
+    {
+      "epoch": 1.3030970885048139,
+      "grad_norm": 8.903182393289645,
+      "learning_rate": 2e-06,
+      "loss": 0.1805,
+      "step": 5617
+    },
+    {
+      "epoch": 1.3033290801531145,
+      "grad_norm": 25.937526008653386,
+      "learning_rate": 2e-06,
+      "loss": 0.2313,
+      "step": 5618
+    },
+    {
+      "epoch": 1.3035610718014152,
+      "grad_norm": 17.577990312401784,
+      "learning_rate": 2e-06,
+      "loss": 0.3776,
+      "step": 5619
+    },
+    {
+      "epoch": 1.303793063449716,
+      "grad_norm": 22.427564565601266,
+      "learning_rate": 2e-06,
+      "loss": 0.256,
+      "step": 5620
+    },
+    {
+      "epoch": 1.3040250550980166,
+      "grad_norm": 15.660251500511174,
+      "learning_rate": 2e-06,
+      "loss": 0.2918,
+      "step": 5621
+    },
+    {
+      "epoch": 1.3042570467463173,
+      "grad_norm": 28.10834557005505,
+      "learning_rate": 2e-06,
+      "loss": 0.3871,
+      "step": 5622
+    },
+    {
+      "epoch": 1.3044890383946177,
+      "grad_norm": 8.673493743204125,
+      "learning_rate": 2e-06,
+      "loss": 0.2583,
+      "step": 5623
+    },
+    {
+      "epoch": 1.3047210300429184,
+      "grad_norm": 21.417289777883838,
+      "learning_rate": 2e-06,
+      "loss": 0.3396,
+      "step": 5624
+    },
+    {
+      "epoch": 1.304953021691219,
+      "grad_norm": 16.89284063484694,
+      "learning_rate": 2e-06,
+      "loss": 0.3283,
+      "step": 5625
+    },
+    {
+      "epoch": 1.3051850133395198,
+      "grad_norm": 29.120207682173884,
+      "learning_rate": 2e-06,
+      "loss": 0.3296,
+      "step": 5626
+    },
+    {
+      "epoch": 1.3054170049878204,
+      "grad_norm": 8.919708053307183,
+      "learning_rate": 2e-06,
+      "loss": 0.1461,
+      "step": 5627
+    },
+    {
+      "epoch": 1.305648996636121,
+      "grad_norm": 19.78253962837575,
+      "learning_rate": 2e-06,
+      "loss": 0.2993,
+      "step": 5628
+    },
+    {
+      "epoch": 1.3058809882844218,
+      "grad_norm": 17.402054742095746,
+      "learning_rate": 2e-06,
+      "loss": 0.3022,
+      "step": 5629
+    },
+    {
+      "epoch": 1.3061129799327225,
+      "grad_norm": 11.4823499978865,
+      "learning_rate": 2e-06,
+      "loss": 0.2225,
+      "step": 5630
+    },
+    {
+      "epoch": 1.3063449715810231,
+      "grad_norm": 7.3361505909982005,
+      "learning_rate": 2e-06,
+      "loss": 0.1902,
+      "step": 5631
+    },
+    {
+      "epoch": 1.3065769632293238,
+      "grad_norm": 27.559188119699623,
+      "learning_rate": 2e-06,
+      "loss": 0.2572,
+      "step": 5632
+    },
+    {
+      "epoch": 1.3068089548776243,
+      "grad_norm": 15.273929151441568,
+      "learning_rate": 2e-06,
+      "loss": 0.2725,
+      "step": 5633
+    },
+    {
+      "epoch": 1.307040946525925,
+      "grad_norm": 22.424000462796233,
+      "learning_rate": 2e-06,
+      "loss": 0.3106,
+      "step": 5634
+    },
+    {
+      "epoch": 1.3072729381742256,
+      "grad_norm": 14.713171143247271,
+      "learning_rate": 2e-06,
+      "loss": 0.1986,
+      "step": 5635
+    },
+    {
+      "epoch": 1.3075049298225263,
+      "grad_norm": 19.197749791692836,
+      "learning_rate": 2e-06,
+      "loss": 0.2414,
+      "step": 5636
+    },
+    {
+      "epoch": 1.307736921470827,
+      "grad_norm": 9.70171890144793,
+      "learning_rate": 2e-06,
+      "loss": 0.2229,
+      "step": 5637
+    },
+    {
+      "epoch": 1.3079689131191277,
+      "grad_norm": 11.136763234840572,
+      "learning_rate": 2e-06,
+      "loss": 0.1674,
+      "step": 5638
+    },
+    {
+      "epoch": 1.3082009047674283,
+      "grad_norm": 26.03212305103045,
+      "learning_rate": 2e-06,
+      "loss": 0.3436,
+      "step": 5639
+    },
+    {
+      "epoch": 1.308432896415729,
+      "grad_norm": 21.30713547042178,
+      "learning_rate": 2e-06,
+      "loss": 0.2795,
+      "step": 5640
+    },
+    {
+      "epoch": 1.3086648880640297,
+      "grad_norm": 10.993965951343093,
+      "learning_rate": 2e-06,
+      "loss": 0.1959,
+      "step": 5641
+    },
+    {
+      "epoch": 1.3088968797123304,
+      "grad_norm": 17.769077676138195,
+      "learning_rate": 2e-06,
+      "loss": 0.2873,
+      "step": 5642
+    },
+    {
+      "epoch": 1.309128871360631,
+      "grad_norm": 18.506312434168304,
+      "learning_rate": 2e-06,
+      "loss": 0.255,
+      "step": 5643
+    },
+    {
+      "epoch": 1.3093608630089317,
+      "grad_norm": 6.793232387244837,
+      "learning_rate": 2e-06,
+      "loss": 0.174,
+      "step": 5644
+    },
+    {
+      "epoch": 1.3095928546572324,
+      "grad_norm": 32.49566834679004,
+      "learning_rate": 2e-06,
+      "loss": 0.2574,
+      "step": 5645
+    },
+    {
+      "epoch": 1.309824846305533,
+      "grad_norm": 17.02251550801001,
+      "learning_rate": 2e-06,
+      "loss": 0.3174,
+      "step": 5646
+    },
+    {
+      "epoch": 1.3100568379538338,
+      "grad_norm": 12.841461665684056,
+      "learning_rate": 2e-06,
+      "loss": 0.3524,
+      "step": 5647
+    },
+    {
+      "epoch": 1.3102888296021344,
+      "grad_norm": 19.076546400110157,
+      "learning_rate": 2e-06,
+      "loss": 0.2798,
+      "step": 5648
+    },
+    {
+      "epoch": 1.3105208212504351,
+      "grad_norm": 13.18433741810789,
+      "learning_rate": 2e-06,
+      "loss": 0.3233,
+      "step": 5649
+    },
+    {
+      "epoch": 1.3107528128987356,
+      "grad_norm": 11.135069470099385,
+      "learning_rate": 2e-06,
+      "loss": 0.1642,
+      "step": 5650
+    },
+    {
+      "epoch": 1.3109848045470363,
+      "grad_norm": 15.315677523471457,
+      "learning_rate": 2e-06,
+      "loss": 0.2652,
+      "step": 5651
+    },
+    {
+      "epoch": 1.311216796195337,
+      "grad_norm": 10.62905135472249,
+      "learning_rate": 2e-06,
+      "loss": 0.2165,
+      "step": 5652
+    },
+    {
+      "epoch": 1.3114487878436376,
+      "grad_norm": 11.231634605332838,
+      "learning_rate": 2e-06,
+      "loss": 0.2613,
+      "step": 5653
+    },
+    {
+      "epoch": 1.3116807794919383,
+      "grad_norm": 9.078649052887393,
+      "learning_rate": 2e-06,
+      "loss": 0.2143,
+      "step": 5654
+    },
+    {
+      "epoch": 1.311912771140239,
+      "grad_norm": 21.39401032753167,
+      "learning_rate": 2e-06,
+      "loss": 0.3495,
+      "step": 5655
+    },
+    {
+      "epoch": 1.3121447627885396,
+      "grad_norm": 17.904213558896597,
+      "learning_rate": 2e-06,
+      "loss": 0.2527,
+      "step": 5656
+    },
+    {
+      "epoch": 1.3123767544368403,
+      "grad_norm": 15.558548274464785,
+      "learning_rate": 2e-06,
+      "loss": 0.1924,
+      "step": 5657
+    },
+    {
+      "epoch": 1.312608746085141,
+      "grad_norm": 12.029762240958295,
+      "learning_rate": 2e-06,
+      "loss": 0.2173,
+      "step": 5658
+    },
+    {
+      "epoch": 1.3128407377334417,
+      "grad_norm": 25.30959589435329,
+      "learning_rate": 2e-06,
+      "loss": 0.2702,
+      "step": 5659
+    },
+    {
+      "epoch": 1.3130727293817421,
+      "grad_norm": 10.092299250446088,
+      "learning_rate": 2e-06,
+      "loss": 0.1949,
+      "step": 5660
+    },
+    {
+      "epoch": 1.3133047210300428,
+      "grad_norm": 13.306741014180796,
+      "learning_rate": 2e-06,
+      "loss": 0.2432,
+      "step": 5661
+    },
+    {
+      "epoch": 1.3135367126783435,
+      "grad_norm": 17.398002407527894,
+      "learning_rate": 2e-06,
+      "loss": 0.2507,
+      "step": 5662
+    },
+    {
+      "epoch": 1.3137687043266442,
+      "grad_norm": 13.958106857172252,
+      "learning_rate": 2e-06,
+      "loss": 0.2637,
+      "step": 5663
+    },
+    {
+      "epoch": 1.3140006959749448,
+      "grad_norm": 11.373230060748023,
+      "learning_rate": 2e-06,
+      "loss": 0.2399,
+      "step": 5664
+    },
+    {
+      "epoch": 1.3142326876232455,
+      "grad_norm": 15.86847179090274,
+      "learning_rate": 2e-06,
+      "loss": 0.2329,
+      "step": 5665
+    },
+    {
+      "epoch": 1.3144646792715462,
+      "grad_norm": 15.508532456260358,
+      "learning_rate": 2e-06,
+      "loss": 0.3161,
+      "step": 5666
+    },
+    {
+      "epoch": 1.3146966709198469,
+      "grad_norm": 15.532683378784451,
+      "learning_rate": 2e-06,
+      "loss": 0.3558,
+      "step": 5667
+    },
+    {
+      "epoch": 1.3149286625681476,
+      "grad_norm": 19.714299042898194,
+      "learning_rate": 2e-06,
+      "loss": 0.3513,
+      "step": 5668
+    },
+    {
+      "epoch": 1.3151606542164482,
+      "grad_norm": 13.958754944833625,
+      "learning_rate": 2e-06,
+      "loss": 0.2168,
+      "step": 5669
+    },
+    {
+      "epoch": 1.315392645864749,
+      "grad_norm": 30.644144100112875,
+      "learning_rate": 2e-06,
+      "loss": 0.3604,
+      "step": 5670
+    },
+    {
+      "epoch": 1.3156246375130496,
+      "grad_norm": 16.352621053156582,
+      "learning_rate": 2e-06,
+      "loss": 0.2741,
+      "step": 5671
+    },
+    {
+      "epoch": 1.3158566291613503,
+      "grad_norm": 19.73007160153037,
+      "learning_rate": 2e-06,
+      "loss": 0.261,
+      "step": 5672
+    },
+    {
+      "epoch": 1.316088620809651,
+      "grad_norm": 14.995894794886874,
+      "learning_rate": 2e-06,
+      "loss": 0.2534,
+      "step": 5673
+    },
+    {
+      "epoch": 1.3163206124579516,
+      "grad_norm": 15.945047875864542,
+      "learning_rate": 2e-06,
+      "loss": 0.3808,
+      "step": 5674
+    },
+    {
+      "epoch": 1.3165526041062523,
+      "grad_norm": 14.018774776451082,
+      "learning_rate": 2e-06,
+      "loss": 0.2702,
+      "step": 5675
+    },
+    {
+      "epoch": 1.3167845957545528,
+      "grad_norm": 12.70859105222736,
+      "learning_rate": 2e-06,
+      "loss": 0.24,
+      "step": 5676
+    },
+    {
+      "epoch": 1.3170165874028534,
+      "grad_norm": 9.678998173041654,
+      "learning_rate": 2e-06,
+      "loss": 0.1932,
+      "step": 5677
+    },
+    {
+      "epoch": 1.3172485790511541,
+      "grad_norm": 15.736690719275137,
+      "learning_rate": 2e-06,
+      "loss": 0.2398,
+      "step": 5678
+    },
+    {
+      "epoch": 1.3174805706994548,
+      "grad_norm": 13.60977667580227,
+      "learning_rate": 2e-06,
+      "loss": 0.2472,
+      "step": 5679
+    },
+    {
+      "epoch": 1.3177125623477555,
+      "grad_norm": 11.102773871126534,
+      "learning_rate": 2e-06,
+      "loss": 0.2326,
+      "step": 5680
+    },
+    {
+      "epoch": 1.3179445539960561,
+      "grad_norm": 17.291796040037262,
+      "learning_rate": 2e-06,
+      "loss": 0.346,
+      "step": 5681
+    },
+    {
+      "epoch": 1.3181765456443568,
+      "grad_norm": 18.166601354990185,
+      "learning_rate": 2e-06,
+      "loss": 0.2617,
+      "step": 5682
+    },
+    {
+      "epoch": 1.3184085372926575,
+      "grad_norm": 10.95806449961031,
+      "learning_rate": 2e-06,
+      "loss": 0.237,
+      "step": 5683
+    },
+    {
+      "epoch": 1.3186405289409582,
+      "grad_norm": 160.91351095779942,
+      "learning_rate": 2e-06,
+      "loss": 0.2951,
+      "step": 5684
+    },
+    {
+      "epoch": 1.3188725205892589,
+      "grad_norm": 10.603170771079542,
+      "learning_rate": 2e-06,
+      "loss": 0.1857,
+      "step": 5685
+    },
+    {
+      "epoch": 1.3191045122375593,
+      "grad_norm": 6.870903833321501,
+      "learning_rate": 2e-06,
+      "loss": 0.1642,
+      "step": 5686
+    },
+    {
+      "epoch": 1.31933650388586,
+      "grad_norm": 13.412277495806073,
+      "learning_rate": 2e-06,
+      "loss": 0.2572,
+      "step": 5687
+    },
+    {
+      "epoch": 1.3195684955341607,
+      "grad_norm": 16.00296720744652,
+      "learning_rate": 2e-06,
+      "loss": 0.2616,
+      "step": 5688
+    },
+    {
+      "epoch": 1.3198004871824613,
+      "grad_norm": 14.523893821884103,
+      "learning_rate": 2e-06,
+      "loss": 0.2931,
+      "step": 5689
+    },
+    {
+      "epoch": 1.320032478830762,
+      "grad_norm": 8.344921740838892,
+      "learning_rate": 2e-06,
+      "loss": 0.1381,
+      "step": 5690
+    },
+    {
+      "epoch": 1.3202644704790627,
+      "grad_norm": 18.55719833842272,
+      "learning_rate": 2e-06,
+      "loss": 0.2493,
+      "step": 5691
+    },
+    {
+      "epoch": 1.3204964621273634,
+      "grad_norm": 14.12605618900697,
+      "learning_rate": 2e-06,
+      "loss": 0.1944,
+      "step": 5692
+    },
+    {
+      "epoch": 1.320728453775664,
+      "grad_norm": 9.998184058061966,
+      "learning_rate": 2e-06,
+      "loss": 0.1542,
+      "step": 5693
+    },
+    {
+      "epoch": 1.3209604454239647,
+      "grad_norm": 15.911168575838987,
+      "learning_rate": 2e-06,
+      "loss": 0.4191,
+      "step": 5694
+    },
+    {
+      "epoch": 1.3211924370722654,
+      "grad_norm": 24.85883116892657,
+      "learning_rate": 2e-06,
+      "loss": 0.2957,
+      "step": 5695
+    },
+    {
+      "epoch": 1.321424428720566,
+      "grad_norm": 17.623611354513486,
+      "learning_rate": 2e-06,
+      "loss": 0.3902,
+      "step": 5696
+    },
+    {
+      "epoch": 1.3216564203688668,
+      "grad_norm": 16.48576591994104,
+      "learning_rate": 2e-06,
+      "loss": 0.3015,
+      "step": 5697
+    },
+    {
+      "epoch": 1.3218884120171674,
+      "grad_norm": 15.575788365937843,
+      "learning_rate": 2e-06,
+      "loss": 0.2635,
+      "step": 5698
+    },
+    {
+      "epoch": 1.3221204036654681,
+      "grad_norm": 20.85661933589363,
+      "learning_rate": 2e-06,
+      "loss": 0.2316,
+      "step": 5699
+    },
+    {
+      "epoch": 1.3223523953137688,
+      "grad_norm": 16.061174354807644,
+      "learning_rate": 2e-06,
+      "loss": 0.1946,
+      "step": 5700
+    },
+    {
+      "epoch": 1.3225843869620695,
+      "grad_norm": 17.84776593809947,
+      "learning_rate": 2e-06,
+      "loss": 0.2291,
+      "step": 5701
+    },
+    {
+      "epoch": 1.3228163786103702,
+      "grad_norm": 18.804374385719715,
+      "learning_rate": 2e-06,
+      "loss": 0.3436,
+      "step": 5702
+    },
+    {
+      "epoch": 1.3230483702586706,
+      "grad_norm": 13.939505997269132,
+      "learning_rate": 2e-06,
+      "loss": 0.2215,
+      "step": 5703
+    },
+    {
+      "epoch": 1.3232803619069713,
+      "grad_norm": 22.54661147857366,
+      "learning_rate": 2e-06,
+      "loss": 0.4171,
+      "step": 5704
+    },
+    {
+      "epoch": 1.323512353555272,
+      "grad_norm": 14.117182775507976,
+      "learning_rate": 2e-06,
+      "loss": 0.2152,
+      "step": 5705
+    },
+    {
+      "epoch": 1.3237443452035726,
+      "grad_norm": 18.175557503429257,
+      "learning_rate": 2e-06,
+      "loss": 0.2274,
+      "step": 5706
+    },
+    {
+      "epoch": 1.3239763368518733,
+      "grad_norm": 11.06563729243645,
+      "learning_rate": 2e-06,
+      "loss": 0.1603,
+      "step": 5707
+    },
+    {
+      "epoch": 1.324208328500174,
+      "grad_norm": 11.94708026064263,
+      "learning_rate": 2e-06,
+      "loss": 0.2145,
+      "step": 5708
+    },
+    {
+      "epoch": 1.3244403201484747,
+      "grad_norm": 19.632307362212533,
+      "learning_rate": 2e-06,
+      "loss": 0.3409,
+      "step": 5709
+    },
+    {
+      "epoch": 1.3246723117967754,
+      "grad_norm": 12.854868599897623,
+      "learning_rate": 2e-06,
+      "loss": 0.2245,
+      "step": 5710
+    },
+    {
+      "epoch": 1.324904303445076,
+      "grad_norm": 23.40658671310776,
+      "learning_rate": 2e-06,
+      "loss": 0.2682,
+      "step": 5711
+    },
+    {
+      "epoch": 1.3251362950933767,
+      "grad_norm": 16.268358378842592,
+      "learning_rate": 2e-06,
+      "loss": 0.3602,
+      "step": 5712
+    },
+    {
+      "epoch": 1.3253682867416772,
+      "grad_norm": 11.210694208149162,
+      "learning_rate": 2e-06,
+      "loss": 0.2451,
+      "step": 5713
+    },
+    {
+      "epoch": 1.3256002783899778,
+      "grad_norm": 13.093366884959643,
+      "learning_rate": 2e-06,
+      "loss": 0.1851,
+      "step": 5714
+    },
+    {
+      "epoch": 1.3258322700382785,
+      "grad_norm": 13.9470039807361,
+      "learning_rate": 2e-06,
+      "loss": 0.2528,
+      "step": 5715
+    },
+    {
+      "epoch": 1.3260642616865792,
+      "grad_norm": 13.826574290698455,
+      "learning_rate": 2e-06,
+      "loss": 0.2275,
+      "step": 5716
+    },
+    {
+      "epoch": 1.3262962533348799,
+      "grad_norm": 19.0989336579408,
+      "learning_rate": 2e-06,
+      "loss": 0.4041,
+      "step": 5717
+    },
+    {
+      "epoch": 1.3265282449831806,
+      "grad_norm": 10.94421028576376,
+      "learning_rate": 2e-06,
+      "loss": 0.1777,
+      "step": 5718
+    },
+    {
+      "epoch": 1.3267602366314812,
+      "grad_norm": 11.10383820485182,
+      "learning_rate": 2e-06,
+      "loss": 0.1856,
+      "step": 5719
+    },
+    {
+      "epoch": 1.326992228279782,
+      "grad_norm": 11.421442448851714,
+      "learning_rate": 2e-06,
+      "loss": 0.215,
+      "step": 5720
+    },
+    {
+      "epoch": 1.3272242199280826,
+      "grad_norm": 16.064869933616933,
+      "learning_rate": 2e-06,
+      "loss": 0.311,
+      "step": 5721
+    },
+    {
+      "epoch": 1.3274562115763833,
+      "grad_norm": 18.75785660722465,
+      "learning_rate": 2e-06,
+      "loss": 0.4559,
+      "step": 5722
+    },
+    {
+      "epoch": 1.327688203224684,
+      "grad_norm": 11.390146013990016,
+      "learning_rate": 2e-06,
+      "loss": 0.2577,
+      "step": 5723
+    },
+    {
+      "epoch": 1.3279201948729846,
+      "grad_norm": 16.85163301480196,
+      "learning_rate": 2e-06,
+      "loss": 0.2651,
+      "step": 5724
+    },
+    {
+      "epoch": 1.3281521865212853,
+      "grad_norm": 13.0317462251221,
+      "learning_rate": 2e-06,
+      "loss": 0.2271,
+      "step": 5725
+    },
+    {
+      "epoch": 1.328384178169586,
+      "grad_norm": 14.351959885689283,
+      "learning_rate": 2e-06,
+      "loss": 0.226,
+      "step": 5726
+    },
+    {
+      "epoch": 1.3286161698178867,
+      "grad_norm": 9.984793704893075,
+      "learning_rate": 2e-06,
+      "loss": 0.1668,
+      "step": 5727
+    },
+    {
+      "epoch": 1.3288481614661873,
+      "grad_norm": 17.625833578442002,
+      "learning_rate": 2e-06,
+      "loss": 0.343,
+      "step": 5728
+    },
+    {
+      "epoch": 1.329080153114488,
+      "grad_norm": 13.273409325079097,
+      "learning_rate": 2e-06,
+      "loss": 0.2235,
+      "step": 5729
+    },
+    {
+      "epoch": 1.3293121447627885,
+      "grad_norm": 8.8052389353841,
+      "learning_rate": 2e-06,
+      "loss": 0.223,
+      "step": 5730
+    },
+    {
+      "epoch": 1.3295441364110892,
+      "grad_norm": 15.961048639991073,
+      "learning_rate": 2e-06,
+      "loss": 0.1882,
+      "step": 5731
+    },
+    {
+      "epoch": 1.3297761280593898,
+      "grad_norm": 15.980771784027198,
+      "learning_rate": 2e-06,
+      "loss": 0.2133,
+      "step": 5732
+    },
+    {
+      "epoch": 1.3300081197076905,
+      "grad_norm": 34.87453938575482,
+      "learning_rate": 2e-06,
+      "loss": 0.3107,
+      "step": 5733
+    },
+    {
+      "epoch": 1.3302401113559912,
+      "grad_norm": 7.971302638747583,
+      "learning_rate": 2e-06,
+      "loss": 0.1623,
+      "step": 5734
+    },
+    {
+      "epoch": 1.3304721030042919,
+      "grad_norm": 13.732619994934568,
+      "learning_rate": 2e-06,
+      "loss": 0.3478,
+      "step": 5735
+    },
+    {
+      "epoch": 1.3307040946525925,
+      "grad_norm": 13.81265807913985,
+      "learning_rate": 2e-06,
+      "loss": 0.2507,
+      "step": 5736
+    },
+    {
+      "epoch": 1.3309360863008932,
+      "grad_norm": 17.587535840998314,
+      "learning_rate": 2e-06,
+      "loss": 0.299,
+      "step": 5737
+    },
+    {
+      "epoch": 1.331168077949194,
+      "grad_norm": 18.227607371744853,
+      "learning_rate": 2e-06,
+      "loss": 0.3449,
+      "step": 5738
+    },
+    {
+      "epoch": 1.3314000695974946,
+      "grad_norm": 12.651673644096768,
+      "learning_rate": 2e-06,
+      "loss": 0.2299,
+      "step": 5739
+    },
+    {
+      "epoch": 1.331632061245795,
+      "grad_norm": 13.107582649881053,
+      "learning_rate": 2e-06,
+      "loss": 0.2587,
+      "step": 5740
+    },
+    {
+      "epoch": 1.3318640528940957,
+      "grad_norm": 19.926768899159622,
+      "learning_rate": 2e-06,
+      "loss": 0.2526,
+      "step": 5741
+    },
+    {
+      "epoch": 1.3320960445423964,
+      "grad_norm": 9.882153373560206,
+      "learning_rate": 2e-06,
+      "loss": 0.2351,
+      "step": 5742
+    },
+    {
+      "epoch": 1.332328036190697,
+      "grad_norm": 15.401606571846399,
+      "learning_rate": 2e-06,
+      "loss": 0.2877,
+      "step": 5743
+    },
+    {
+      "epoch": 1.3325600278389977,
+      "grad_norm": 8.051494076240969,
+      "learning_rate": 2e-06,
+      "loss": 0.1099,
+      "step": 5744
+    },
+    {
+      "epoch": 1.3327920194872984,
+      "grad_norm": 12.485400896606112,
+      "learning_rate": 2e-06,
+      "loss": 0.2127,
+      "step": 5745
+    },
+    {
+      "epoch": 1.333024011135599,
+      "grad_norm": 19.37818023128224,
+      "learning_rate": 2e-06,
+      "loss": 0.3585,
+      "step": 5746
+    },
+    {
+      "epoch": 1.3332560027838998,
+      "grad_norm": 11.50542375491202,
+      "learning_rate": 2e-06,
+      "loss": 0.1986,
+      "step": 5747
+    },
+    {
+      "epoch": 1.3334879944322005,
+      "grad_norm": 9.535853097525685,
+      "learning_rate": 2e-06,
+      "loss": 0.1785,
+      "step": 5748
+    },
+    {
+      "epoch": 1.3337199860805011,
+      "grad_norm": 13.825585578950808,
+      "learning_rate": 2e-06,
+      "loss": 0.2251,
+      "step": 5749
+    },
+    {
+      "epoch": 1.3339519777288018,
+      "grad_norm": 20.489808046635204,
+      "learning_rate": 2e-06,
+      "loss": 0.3068,
+      "step": 5750
+    },
+    {
+      "epoch": 1.3341839693771025,
+      "grad_norm": 7.979473578310454,
+      "learning_rate": 2e-06,
+      "loss": 0.219,
+      "step": 5751
+    },
+    {
+      "epoch": 1.3344159610254032,
+      "grad_norm": 20.741850635447804,
+      "learning_rate": 2e-06,
+      "loss": 0.2966,
+      "step": 5752
+    },
+    {
+      "epoch": 1.3346479526737038,
+      "grad_norm": 14.487809285683062,
+      "learning_rate": 2e-06,
+      "loss": 0.2511,
+      "step": 5753
+    },
+    {
+      "epoch": 1.3348799443220045,
+      "grad_norm": 13.953602449611058,
+      "learning_rate": 2e-06,
+      "loss": 0.2804,
+      "step": 5754
+    },
+    {
+      "epoch": 1.3351119359703052,
+      "grad_norm": 17.240216254771312,
+      "learning_rate": 2e-06,
+      "loss": 0.1892,
+      "step": 5755
+    },
+    {
+      "epoch": 1.3353439276186057,
+      "grad_norm": 10.47617287821396,
+      "learning_rate": 2e-06,
+      "loss": 0.1832,
+      "step": 5756
+    },
+    {
+      "epoch": 1.3355759192669063,
+      "grad_norm": 17.52330908616922,
+      "learning_rate": 2e-06,
+      "loss": 0.2934,
+      "step": 5757
+    },
+    {
+      "epoch": 1.335807910915207,
+      "grad_norm": 13.566061758419231,
+      "learning_rate": 2e-06,
+      "loss": 0.2273,
+      "step": 5758
+    },
+    {
+      "epoch": 1.3360399025635077,
+      "grad_norm": 12.742318765067107,
+      "learning_rate": 2e-06,
+      "loss": 0.1997,
+      "step": 5759
+    },
+    {
+      "epoch": 1.3362718942118084,
+      "grad_norm": 18.812041464643745,
+      "learning_rate": 2e-06,
+      "loss": 0.2981,
+      "step": 5760
+    },
+    {
+      "epoch": 1.336503885860109,
+      "grad_norm": 7.438888234983939,
+      "learning_rate": 2e-06,
+      "loss": 0.1601,
+      "step": 5761
+    },
+    {
+      "epoch": 1.3367358775084097,
+      "grad_norm": 16.70120584326876,
+      "learning_rate": 2e-06,
+      "loss": 0.2006,
+      "step": 5762
+    },
+    {
+      "epoch": 1.3369678691567104,
+      "grad_norm": 18.450037872236187,
+      "learning_rate": 2e-06,
+      "loss": 0.2465,
+      "step": 5763
+    },
+    {
+      "epoch": 1.337199860805011,
+      "grad_norm": 15.59582496370636,
+      "learning_rate": 2e-06,
+      "loss": 0.3092,
+      "step": 5764
+    },
+    {
+      "epoch": 1.3374318524533118,
+      "grad_norm": 24.090622475541675,
+      "learning_rate": 2e-06,
+      "loss": 0.3752,
+      "step": 5765
+    },
+    {
+      "epoch": 1.3376638441016122,
+      "grad_norm": 11.557989090391212,
+      "learning_rate": 2e-06,
+      "loss": 0.2061,
+      "step": 5766
+    },
+    {
+      "epoch": 1.3378958357499129,
+      "grad_norm": 7.013857459223439,
+      "learning_rate": 2e-06,
+      "loss": 0.218,
+      "step": 5767
+    },
+    {
+      "epoch": 1.3381278273982136,
+      "grad_norm": 10.23666384150274,
+      "learning_rate": 2e-06,
+      "loss": 0.1466,
+      "step": 5768
+    },
+    {
+      "epoch": 1.3383598190465142,
+      "grad_norm": 11.148655360805588,
+      "learning_rate": 2e-06,
+      "loss": 0.1825,
+      "step": 5769
+    },
+    {
+      "epoch": 1.338591810694815,
+      "grad_norm": 16.186729561676344,
+      "learning_rate": 2e-06,
+      "loss": 0.3968,
+      "step": 5770
+    },
+    {
+      "epoch": 1.3388238023431156,
+      "grad_norm": 6.943969848029348,
+      "learning_rate": 2e-06,
+      "loss": 0.1478,
+      "step": 5771
+    },
+    {
+      "epoch": 1.3390557939914163,
+      "grad_norm": 17.341373830915476,
+      "learning_rate": 2e-06,
+      "loss": 0.1947,
+      "step": 5772
+    },
+    {
+      "epoch": 1.339287785639717,
+      "grad_norm": 24.482447956744732,
+      "learning_rate": 2e-06,
+      "loss": 0.3202,
+      "step": 5773
+    },
+    {
+      "epoch": 1.3395197772880176,
+      "grad_norm": 21.39557609795782,
+      "learning_rate": 2e-06,
+      "loss": 0.3008,
+      "step": 5774
+    },
+    {
+      "epoch": 1.3397517689363183,
+      "grad_norm": 6.321646081575424,
+      "learning_rate": 2e-06,
+      "loss": 0.18,
+      "step": 5775
+    },
+    {
+      "epoch": 1.339983760584619,
+      "grad_norm": 14.757152685348238,
+      "learning_rate": 2e-06,
+      "loss": 0.3238,
+      "step": 5776
+    },
+    {
+      "epoch": 1.3402157522329197,
+      "grad_norm": 12.278175383917478,
+      "learning_rate": 2e-06,
+      "loss": 0.2181,
+      "step": 5777
+    },
+    {
+      "epoch": 1.3404477438812203,
+      "grad_norm": 12.23761349802958,
+      "learning_rate": 2e-06,
+      "loss": 0.3068,
+      "step": 5778
+    },
+    {
+      "epoch": 1.340679735529521,
+      "grad_norm": 11.403333857518392,
+      "learning_rate": 2e-06,
+      "loss": 0.1405,
+      "step": 5779
+    },
+    {
+      "epoch": 1.3409117271778217,
+      "grad_norm": 13.05117023683827,
+      "learning_rate": 2e-06,
+      "loss": 0.2465,
+      "step": 5780
+    },
+    {
+      "epoch": 1.3411437188261224,
+      "grad_norm": 22.224875632817582,
+      "learning_rate": 2e-06,
+      "loss": 0.2594,
+      "step": 5781
+    },
+    {
+      "epoch": 1.341375710474423,
+      "grad_norm": 20.353481270647716,
+      "learning_rate": 2e-06,
+      "loss": 0.3515,
+      "step": 5782
+    },
+    {
+      "epoch": 1.3416077021227235,
+      "grad_norm": 12.682528250567298,
+      "learning_rate": 2e-06,
+      "loss": 0.1926,
+      "step": 5783
+    },
+    {
+      "epoch": 1.3418396937710242,
+      "grad_norm": 28.57266479023652,
+      "learning_rate": 2e-06,
+      "loss": 0.2512,
+      "step": 5784
+    },
+    {
+      "epoch": 1.3420716854193249,
+      "grad_norm": 16.909366137872915,
+      "learning_rate": 2e-06,
+      "loss": 0.2867,
+      "step": 5785
+    },
+    {
+      "epoch": 1.3423036770676255,
+      "grad_norm": 13.171274976616166,
+      "learning_rate": 2e-06,
+      "loss": 0.2623,
+      "step": 5786
+    },
+    {
+      "epoch": 1.3425356687159262,
+      "grad_norm": 8.080789224669203,
+      "learning_rate": 2e-06,
+      "loss": 0.1932,
+      "step": 5787
+    },
+    {
+      "epoch": 1.342767660364227,
+      "grad_norm": 20.2668518242124,
+      "learning_rate": 2e-06,
+      "loss": 0.3434,
+      "step": 5788
+    },
+    {
+      "epoch": 1.3429996520125276,
+      "grad_norm": 15.16901875665676,
+      "learning_rate": 2e-06,
+      "loss": 0.2457,
+      "step": 5789
+    },
+    {
+      "epoch": 1.3432316436608283,
+      "grad_norm": 11.042566276296274,
+      "learning_rate": 2e-06,
+      "loss": 0.1578,
+      "step": 5790
+    },
+    {
+      "epoch": 1.343463635309129,
+      "grad_norm": 10.584691156901858,
+      "learning_rate": 2e-06,
+      "loss": 0.2686,
+      "step": 5791
+    },
+    {
+      "epoch": 1.3436956269574296,
+      "grad_norm": 13.434105792076489,
+      "learning_rate": 2e-06,
+      "loss": 0.3587,
+      "step": 5792
+    },
+    {
+      "epoch": 1.34392761860573,
+      "grad_norm": 17.34269805708881,
+      "learning_rate": 2e-06,
+      "loss": 0.2988,
+      "step": 5793
+    },
+    {
+      "epoch": 1.3441596102540307,
+      "grad_norm": 6.675450785331753,
+      "learning_rate": 2e-06,
+      "loss": 0.1739,
+      "step": 5794
+    },
+    {
+      "epoch": 1.3443916019023314,
+      "grad_norm": 13.984086856770102,
+      "learning_rate": 2e-06,
+      "loss": 0.3053,
+      "step": 5795
+    },
+    {
+      "epoch": 1.344623593550632,
+      "grad_norm": 19.1634289852142,
+      "learning_rate": 2e-06,
+      "loss": 0.3829,
+      "step": 5796
+    },
+    {
+      "epoch": 1.3448555851989328,
+      "grad_norm": 11.212872422045528,
+      "learning_rate": 2e-06,
+      "loss": 0.2311,
+      "step": 5797
+    },
+    {
+      "epoch": 1.3450875768472335,
+      "grad_norm": 8.339351003547979,
+      "learning_rate": 2e-06,
+      "loss": 0.186,
+      "step": 5798
+    },
+    {
+      "epoch": 1.3453195684955341,
+      "grad_norm": 11.20007845000388,
+      "learning_rate": 2e-06,
+      "loss": 0.2126,
+      "step": 5799
+    },
+    {
+      "epoch": 1.3455515601438348,
+      "grad_norm": 9.150632215768846,
+      "learning_rate": 2e-06,
+      "loss": 0.1566,
+      "step": 5800
+    },
+    {
+      "epoch": 1.3457835517921355,
+      "grad_norm": 10.867733909146118,
+      "learning_rate": 2e-06,
+      "loss": 0.2175,
+      "step": 5801
+    },
+    {
+      "epoch": 1.3460155434404362,
+      "grad_norm": 44.450382387130546,
+      "learning_rate": 2e-06,
+      "loss": 0.3324,
+      "step": 5802
+    },
+    {
+      "epoch": 1.3462475350887368,
+      "grad_norm": 19.048516268670234,
+      "learning_rate": 2e-06,
+      "loss": 0.2678,
+      "step": 5803
+    },
+    {
+      "epoch": 1.3464795267370375,
+      "grad_norm": 12.280476212517785,
+      "learning_rate": 2e-06,
+      "loss": 0.2765,
+      "step": 5804
+    },
+    {
+      "epoch": 1.3467115183853382,
+      "grad_norm": 14.774793325255686,
+      "learning_rate": 2e-06,
+      "loss": 0.4255,
+      "step": 5805
+    },
+    {
+      "epoch": 1.3469435100336389,
+      "grad_norm": 11.714254562758581,
+      "learning_rate": 2e-06,
+      "loss": 0.2319,
+      "step": 5806
+    },
+    {
+      "epoch": 1.3471755016819396,
+      "grad_norm": 12.16881579986135,
+      "learning_rate": 2e-06,
+      "loss": 0.2334,
+      "step": 5807
+    },
+    {
+      "epoch": 1.3474074933302402,
+      "grad_norm": 15.737049960037895,
+      "learning_rate": 2e-06,
+      "loss": 0.2389,
+      "step": 5808
+    },
+    {
+      "epoch": 1.3476394849785407,
+      "grad_norm": 14.653130929599717,
+      "learning_rate": 2e-06,
+      "loss": 0.2888,
+      "step": 5809
+    },
+    {
+      "epoch": 1.3478714766268414,
+      "grad_norm": 16.50927703121179,
+      "learning_rate": 2e-06,
+      "loss": 0.2738,
+      "step": 5810
+    },
+    {
+      "epoch": 1.348103468275142,
+      "grad_norm": 13.745080652083896,
+      "learning_rate": 2e-06,
+      "loss": 0.2365,
+      "step": 5811
+    },
+    {
+      "epoch": 1.3483354599234427,
+      "grad_norm": 18.49258799469895,
+      "learning_rate": 2e-06,
+      "loss": 0.3036,
+      "step": 5812
+    },
+    {
+      "epoch": 1.3485674515717434,
+      "grad_norm": 18.51938083068741,
+      "learning_rate": 2e-06,
+      "loss": 0.2746,
+      "step": 5813
+    },
+    {
+      "epoch": 1.348799443220044,
+      "grad_norm": 12.325727775873908,
+      "learning_rate": 2e-06,
+      "loss": 0.2631,
+      "step": 5814
+    },
+    {
+      "epoch": 1.3490314348683448,
+      "grad_norm": 17.84369574907302,
+      "learning_rate": 2e-06,
+      "loss": 0.3388,
+      "step": 5815
+    },
+    {
+      "epoch": 1.3492634265166454,
+      "grad_norm": 13.687209491713253,
+      "learning_rate": 2e-06,
+      "loss": 0.1935,
+      "step": 5816
+    },
+    {
+      "epoch": 1.3494954181649461,
+      "grad_norm": 16.68203773438771,
+      "learning_rate": 2e-06,
+      "loss": 0.2844,
+      "step": 5817
+    },
+    {
+      "epoch": 1.3497274098132468,
+      "grad_norm": 9.734625692979442,
+      "learning_rate": 2e-06,
+      "loss": 0.2316,
+      "step": 5818
+    },
+    {
+      "epoch": 1.3499594014615472,
+      "grad_norm": 17.018315278088533,
+      "learning_rate": 2e-06,
+      "loss": 0.2562,
+      "step": 5819
+    },
+    {
+      "epoch": 1.350191393109848,
+      "grad_norm": 5.075900093156938,
+      "learning_rate": 2e-06,
+      "loss": 0.1322,
+      "step": 5820
+    },
+    {
+      "epoch": 1.3504233847581486,
+      "grad_norm": 12.15002342495086,
+      "learning_rate": 2e-06,
+      "loss": 0.1723,
+      "step": 5821
+    },
+    {
+      "epoch": 1.3506553764064493,
+      "grad_norm": 13.438205754472456,
+      "learning_rate": 2e-06,
+      "loss": 0.2599,
+      "step": 5822
+    },
+    {
+      "epoch": 1.35088736805475,
+      "grad_norm": 14.590596311024154,
+      "learning_rate": 2e-06,
+      "loss": 0.2037,
+      "step": 5823
+    },
+    {
+      "epoch": 1.3511193597030506,
+      "grad_norm": 11.914815783448612,
+      "learning_rate": 2e-06,
+      "loss": 0.1507,
+      "step": 5824
+    },
+    {
+      "epoch": 1.3513513513513513,
+      "grad_norm": 11.406277237176791,
+      "learning_rate": 2e-06,
+      "loss": 0.1968,
+      "step": 5825
+    },
+    {
+      "epoch": 1.351583342999652,
+      "grad_norm": 23.409525749652285,
+      "learning_rate": 2e-06,
+      "loss": 0.2491,
+      "step": 5826
+    },
+    {
+      "epoch": 1.3518153346479527,
+      "grad_norm": 10.095249207393666,
+      "learning_rate": 2e-06,
+      "loss": 0.1637,
+      "step": 5827
+    },
+    {
+      "epoch": 1.3520473262962533,
+      "grad_norm": 11.506175779660916,
+      "learning_rate": 2e-06,
+      "loss": 0.2528,
+      "step": 5828
+    },
+    {
+      "epoch": 1.352279317944554,
+      "grad_norm": 15.188116430508073,
+      "learning_rate": 2e-06,
+      "loss": 0.3157,
+      "step": 5829
+    },
+    {
+      "epoch": 1.3525113095928547,
+      "grad_norm": 10.428369062323641,
+      "learning_rate": 2e-06,
+      "loss": 0.1897,
+      "step": 5830
+    },
+    {
+      "epoch": 1.3527433012411554,
+      "grad_norm": 11.703161185958193,
+      "learning_rate": 2e-06,
+      "loss": 0.2059,
+      "step": 5831
+    },
+    {
+      "epoch": 1.352975292889456,
+      "grad_norm": 18.495688046319522,
+      "learning_rate": 2e-06,
+      "loss": 0.2096,
+      "step": 5832
+    },
+    {
+      "epoch": 1.3532072845377567,
+      "grad_norm": 15.077754673818836,
+      "learning_rate": 2e-06,
+      "loss": 0.1848,
+      "step": 5833
+    },
+    {
+      "epoch": 1.3534392761860574,
+      "grad_norm": 8.695499011423294,
+      "learning_rate": 2e-06,
+      "loss": 0.1427,
+      "step": 5834
+    },
+    {
+      "epoch": 1.353671267834358,
+      "grad_norm": 10.649967194948475,
+      "learning_rate": 2e-06,
+      "loss": 0.2084,
+      "step": 5835
+    },
+    {
+      "epoch": 1.3539032594826585,
+      "grad_norm": 9.647015626245473,
+      "learning_rate": 2e-06,
+      "loss": 0.164,
+      "step": 5836
+    },
+    {
+      "epoch": 1.3541352511309592,
+      "grad_norm": 18.321975887396846,
+      "learning_rate": 2e-06,
+      "loss": 0.2208,
+      "step": 5837
+    },
+    {
+      "epoch": 1.35436724277926,
+      "grad_norm": 12.6264369995146,
+      "learning_rate": 2e-06,
+      "loss": 0.2561,
+      "step": 5838
+    },
+    {
+      "epoch": 1.3545992344275606,
+      "grad_norm": 13.3140753106527,
+      "learning_rate": 2e-06,
+      "loss": 0.271,
+      "step": 5839
+    },
+    {
+      "epoch": 1.3548312260758613,
+      "grad_norm": 18.567865143723125,
+      "learning_rate": 2e-06,
+      "loss": 0.3481,
+      "step": 5840
+    },
+    {
+      "epoch": 1.355063217724162,
+      "grad_norm": 12.737075327613965,
+      "learning_rate": 2e-06,
+      "loss": 0.4176,
+      "step": 5841
+    },
+    {
+      "epoch": 1.3552952093724626,
+      "grad_norm": 11.288649162226255,
+      "learning_rate": 2e-06,
+      "loss": 0.2044,
+      "step": 5842
+    },
+    {
+      "epoch": 1.3555272010207633,
+      "grad_norm": 18.55176280025957,
+      "learning_rate": 2e-06,
+      "loss": 0.2101,
+      "step": 5843
+    },
+    {
+      "epoch": 1.355759192669064,
+      "grad_norm": 17.54094405996478,
+      "learning_rate": 2e-06,
+      "loss": 0.2408,
+      "step": 5844
+    },
+    {
+      "epoch": 1.3559911843173647,
+      "grad_norm": 16.9841331544675,
+      "learning_rate": 2e-06,
+      "loss": 0.3381,
+      "step": 5845
+    },
+    {
+      "epoch": 1.356223175965665,
+      "grad_norm": 19.49676791705497,
+      "learning_rate": 2e-06,
+      "loss": 0.2199,
+      "step": 5846
+    },
+    {
+      "epoch": 1.3564551676139658,
+      "grad_norm": 14.364415926692745,
+      "learning_rate": 2e-06,
+      "loss": 0.1998,
+      "step": 5847
+    },
+    {
+      "epoch": 1.3566871592622665,
+      "grad_norm": 13.325930038440983,
+      "learning_rate": 2e-06,
+      "loss": 0.1628,
+      "step": 5848
+    },
+    {
+      "epoch": 1.3569191509105671,
+      "grad_norm": 23.459217434270393,
+      "learning_rate": 2e-06,
+      "loss": 0.2282,
+      "step": 5849
+    },
+    {
+      "epoch": 1.3571511425588678,
+      "grad_norm": 8.778681303386438,
+      "learning_rate": 2e-06,
+      "loss": 0.153,
+      "step": 5850
+    },
+    {
+      "epoch": 1.3573831342071685,
+      "grad_norm": 18.54260424822745,
+      "learning_rate": 2e-06,
+      "loss": 0.309,
+      "step": 5851
+    },
+    {
+      "epoch": 1.3576151258554692,
+      "grad_norm": 15.49191454628838,
+      "learning_rate": 2e-06,
+      "loss": 0.3286,
+      "step": 5852
+    },
+    {
+      "epoch": 1.3578471175037699,
+      "grad_norm": 12.041053783427783,
+      "learning_rate": 2e-06,
+      "loss": 0.2258,
+      "step": 5853
+    },
+    {
+      "epoch": 1.3580791091520705,
+      "grad_norm": 22.61898634773694,
+      "learning_rate": 2e-06,
+      "loss": 0.298,
+      "step": 5854
+    },
+    {
+      "epoch": 1.3583111008003712,
+      "grad_norm": 4.671569771383828,
+      "learning_rate": 2e-06,
+      "loss": 0.1256,
+      "step": 5855
+    },
+    {
+      "epoch": 1.3585430924486719,
+      "grad_norm": 28.314641736645186,
+      "learning_rate": 2e-06,
+      "loss": 0.2476,
+      "step": 5856
+    },
+    {
+      "epoch": 1.3587750840969726,
+      "grad_norm": 17.632781361738424,
+      "learning_rate": 2e-06,
+      "loss": 0.2985,
+      "step": 5857
+    },
+    {
+      "epoch": 1.3590070757452732,
+      "grad_norm": 8.707190057399295,
+      "learning_rate": 2e-06,
+      "loss": 0.2273,
+      "step": 5858
+    },
+    {
+      "epoch": 1.359239067393574,
+      "grad_norm": 10.346660254268741,
+      "learning_rate": 2e-06,
+      "loss": 0.1795,
+      "step": 5859
+    },
+    {
+      "epoch": 1.3594710590418746,
+      "grad_norm": 23.020950583167213,
+      "learning_rate": 2e-06,
+      "loss": 0.3505,
+      "step": 5860
+    },
+    {
+      "epoch": 1.3597030506901753,
+      "grad_norm": 13.438837661679297,
+      "learning_rate": 2e-06,
+      "loss": 0.2115,
+      "step": 5861
+    },
+    {
+      "epoch": 1.359935042338476,
+      "grad_norm": 11.591111693578771,
+      "learning_rate": 2e-06,
+      "loss": 0.1497,
+      "step": 5862
+    },
+    {
+      "epoch": 1.3601670339867764,
+      "grad_norm": 13.92468760116834,
+      "learning_rate": 2e-06,
+      "loss": 0.3266,
+      "step": 5863
+    },
+    {
+      "epoch": 1.360399025635077,
+      "grad_norm": 11.622340864961435,
+      "learning_rate": 2e-06,
+      "loss": 0.2338,
+      "step": 5864
+    },
+    {
+      "epoch": 1.3606310172833778,
+      "grad_norm": 18.978722409368135,
+      "learning_rate": 2e-06,
+      "loss": 0.2449,
+      "step": 5865
+    },
+    {
+      "epoch": 1.3608630089316784,
+      "grad_norm": 9.71805799882756,
+      "learning_rate": 2e-06,
+      "loss": 0.1905,
+      "step": 5866
+    },
+    {
+      "epoch": 1.3610950005799791,
+      "grad_norm": 16.208738621497222,
+      "learning_rate": 2e-06,
+      "loss": 0.233,
+      "step": 5867
+    },
+    {
+      "epoch": 1.3613269922282798,
+      "grad_norm": 13.70124124125462,
+      "learning_rate": 2e-06,
+      "loss": 0.2076,
+      "step": 5868
+    },
+    {
+      "epoch": 1.3615589838765805,
+      "grad_norm": 13.063859874737823,
+      "learning_rate": 2e-06,
+      "loss": 0.2459,
+      "step": 5869
+    },
+    {
+      "epoch": 1.3617909755248812,
+      "grad_norm": 8.165669539795408,
+      "learning_rate": 2e-06,
+      "loss": 0.1647,
+      "step": 5870
+    },
+    {
+      "epoch": 1.3620229671731818,
+      "grad_norm": 11.781677616521952,
+      "learning_rate": 2e-06,
+      "loss": 0.274,
+      "step": 5871
+    },
+    {
+      "epoch": 1.3622549588214825,
+      "grad_norm": 9.778774987540954,
+      "learning_rate": 2e-06,
+      "loss": 0.1572,
+      "step": 5872
+    },
+    {
+      "epoch": 1.362486950469783,
+      "grad_norm": 20.766841055575046,
+      "learning_rate": 2e-06,
+      "loss": 0.2041,
+      "step": 5873
+    },
+    {
+      "epoch": 1.3627189421180836,
+      "grad_norm": 11.59339540161316,
+      "learning_rate": 2e-06,
+      "loss": 0.2524,
+      "step": 5874
+    },
+    {
+      "epoch": 1.3629509337663843,
+      "grad_norm": 12.47148805393732,
+      "learning_rate": 2e-06,
+      "loss": 0.242,
+      "step": 5875
+    },
+    {
+      "epoch": 1.363182925414685,
+      "grad_norm": 10.960900356092315,
+      "learning_rate": 2e-06,
+      "loss": 0.1566,
+      "step": 5876
+    },
+    {
+      "epoch": 1.3634149170629857,
+      "grad_norm": 14.772461356180653,
+      "learning_rate": 2e-06,
+      "loss": 0.2168,
+      "step": 5877
+    },
+    {
+      "epoch": 1.3636469087112864,
+      "grad_norm": 17.572173076188516,
+      "learning_rate": 2e-06,
+      "loss": 0.2988,
+      "step": 5878
+    },
+    {
+      "epoch": 1.363878900359587,
+      "grad_norm": 9.980374339519635,
+      "learning_rate": 2e-06,
+      "loss": 0.2,
+      "step": 5879
+    },
+    {
+      "epoch": 1.3641108920078877,
+      "grad_norm": 16.310899671151198,
+      "learning_rate": 2e-06,
+      "loss": 0.157,
+      "step": 5880
+    },
+    {
+      "epoch": 1.3643428836561884,
+      "grad_norm": 15.22384696178038,
+      "learning_rate": 2e-06,
+      "loss": 0.2674,
+      "step": 5881
+    },
+    {
+      "epoch": 1.364574875304489,
+      "grad_norm": 113.19316199239812,
+      "learning_rate": 2e-06,
+      "loss": 0.3356,
+      "step": 5882
+    },
+    {
+      "epoch": 1.3648068669527897,
+      "grad_norm": 14.031477531072245,
+      "learning_rate": 2e-06,
+      "loss": 0.3136,
+      "step": 5883
+    },
+    {
+      "epoch": 1.3650388586010904,
+      "grad_norm": 22.121313316789323,
+      "learning_rate": 2e-06,
+      "loss": 0.1891,
+      "step": 5884
+    },
+    {
+      "epoch": 1.365270850249391,
+      "grad_norm": 17.825567248921306,
+      "learning_rate": 2e-06,
+      "loss": 0.3541,
+      "step": 5885
+    },
+    {
+      "epoch": 1.3655028418976918,
+      "grad_norm": 19.069180568099014,
+      "learning_rate": 2e-06,
+      "loss": 0.3158,
+      "step": 5886
+    },
+    {
+      "epoch": 1.3657348335459925,
+      "grad_norm": 25.075437472856258,
+      "learning_rate": 2e-06,
+      "loss": 0.4052,
+      "step": 5887
+    },
+    {
+      "epoch": 1.3659668251942931,
+      "grad_norm": 9.977701700428549,
+      "learning_rate": 2e-06,
+      "loss": 0.181,
+      "step": 5888
+    },
+    {
+      "epoch": 1.3661988168425936,
+      "grad_norm": 17.59684848990075,
+      "learning_rate": 2e-06,
+      "loss": 0.2857,
+      "step": 5889
+    },
+    {
+      "epoch": 1.3664308084908943,
+      "grad_norm": 15.09816100180224,
+      "learning_rate": 2e-06,
+      "loss": 0.3539,
+      "step": 5890
+    },
+    {
+      "epoch": 1.366662800139195,
+      "grad_norm": 10.943892970975487,
+      "learning_rate": 2e-06,
+      "loss": 0.1357,
+      "step": 5891
+    },
+    {
+      "epoch": 1.3668947917874956,
+      "grad_norm": 15.689614080268415,
+      "learning_rate": 2e-06,
+      "loss": 0.275,
+      "step": 5892
+    },
+    {
+      "epoch": 1.3671267834357963,
+      "grad_norm": 17.654429430044164,
+      "learning_rate": 2e-06,
+      "loss": 0.2506,
+      "step": 5893
+    },
+    {
+      "epoch": 1.367358775084097,
+      "grad_norm": 18.689276228543932,
+      "learning_rate": 2e-06,
+      "loss": 0.1976,
+      "step": 5894
+    },
+    {
+      "epoch": 1.3675907667323977,
+      "grad_norm": 18.938131307283825,
+      "learning_rate": 2e-06,
+      "loss": 0.3159,
+      "step": 5895
+    },
+    {
+      "epoch": 1.3678227583806983,
+      "grad_norm": 18.188342876250907,
+      "learning_rate": 2e-06,
+      "loss": 0.2705,
+      "step": 5896
+    },
+    {
+      "epoch": 1.368054750028999,
+      "grad_norm": 19.57321528336182,
+      "learning_rate": 2e-06,
+      "loss": 0.263,
+      "step": 5897
+    },
+    {
+      "epoch": 1.3682867416772997,
+      "grad_norm": 12.626901209943034,
+      "learning_rate": 2e-06,
+      "loss": 0.2686,
+      "step": 5898
+    },
+    {
+      "epoch": 1.3685187333256001,
+      "grad_norm": 16.377096327836597,
+      "learning_rate": 2e-06,
+      "loss": 0.2324,
+      "step": 5899
+    },
+    {
+      "epoch": 1.3687507249739008,
+      "grad_norm": 10.146682101900355,
+      "learning_rate": 2e-06,
+      "loss": 0.1951,
+      "step": 5900
+    },
+    {
+      "epoch": 1.3689827166222015,
+      "grad_norm": 9.085444222802984,
+      "learning_rate": 2e-06,
+      "loss": 0.2103,
+      "step": 5901
+    },
+    {
+      "epoch": 1.3692147082705022,
+      "grad_norm": 13.078879458322199,
+      "learning_rate": 2e-06,
+      "loss": 0.1458,
+      "step": 5902
+    },
+    {
+      "epoch": 1.3694466999188029,
+      "grad_norm": 22.771698063467927,
+      "learning_rate": 2e-06,
+      "loss": 0.3165,
+      "step": 5903
+    },
+    {
+      "epoch": 1.3696786915671035,
+      "grad_norm": 12.684698649225055,
+      "learning_rate": 2e-06,
+      "loss": 0.2419,
+      "step": 5904
+    },
+    {
+      "epoch": 1.3699106832154042,
+      "grad_norm": 18.780053747838103,
+      "learning_rate": 2e-06,
+      "loss": 0.3131,
+      "step": 5905
+    },
+    {
+      "epoch": 1.370142674863705,
+      "grad_norm": 13.218927833905115,
+      "learning_rate": 2e-06,
+      "loss": 0.1716,
+      "step": 5906
+    },
+    {
+      "epoch": 1.3703746665120056,
+      "grad_norm": 12.99323013996737,
+      "learning_rate": 2e-06,
+      "loss": 0.3394,
+      "step": 5907
+    },
+    {
+      "epoch": 1.3706066581603062,
+      "grad_norm": 14.975069838383702,
+      "learning_rate": 2e-06,
+      "loss": 0.2689,
+      "step": 5908
+    },
+    {
+      "epoch": 1.370838649808607,
+      "grad_norm": 13.194842859328135,
+      "learning_rate": 2e-06,
+      "loss": 0.255,
+      "step": 5909
+    },
+    {
+      "epoch": 1.3710706414569076,
+      "grad_norm": 13.572338888548854,
+      "learning_rate": 2e-06,
+      "loss": 0.2105,
+      "step": 5910
+    },
+    {
+      "epoch": 1.3713026331052083,
+      "grad_norm": 9.995146313684502,
+      "learning_rate": 2e-06,
+      "loss": 0.2161,
+      "step": 5911
+    },
+    {
+      "epoch": 1.371534624753509,
+      "grad_norm": 10.08923809195754,
+      "learning_rate": 2e-06,
+      "loss": 0.2171,
+      "step": 5912
+    },
+    {
+      "epoch": 1.3717666164018096,
+      "grad_norm": 26.709924443854867,
+      "learning_rate": 2e-06,
+      "loss": 0.3202,
+      "step": 5913
+    },
+    {
+      "epoch": 1.3719986080501103,
+      "grad_norm": 12.308014395977608,
+      "learning_rate": 2e-06,
+      "loss": 0.2215,
+      "step": 5914
+    },
+    {
+      "epoch": 1.372230599698411,
+      "grad_norm": 13.05156012354374,
+      "learning_rate": 2e-06,
+      "loss": 0.2219,
+      "step": 5915
+    },
+    {
+      "epoch": 1.3724625913467114,
+      "grad_norm": 15.342982310766626,
+      "learning_rate": 2e-06,
+      "loss": 0.2504,
+      "step": 5916
+    },
+    {
+      "epoch": 1.3726945829950121,
+      "grad_norm": 17.12977581566977,
+      "learning_rate": 2e-06,
+      "loss": 0.3918,
+      "step": 5917
+    },
+    {
+      "epoch": 1.3729265746433128,
+      "grad_norm": 10.564058055931197,
+      "learning_rate": 2e-06,
+      "loss": 0.2486,
+      "step": 5918
+    },
+    {
+      "epoch": 1.3731585662916135,
+      "grad_norm": 14.801827927834955,
+      "learning_rate": 2e-06,
+      "loss": 0.2071,
+      "step": 5919
+    },
+    {
+      "epoch": 1.3733905579399142,
+      "grad_norm": 19.601648310112644,
+      "learning_rate": 2e-06,
+      "loss": 0.3412,
+      "step": 5920
+    },
+    {
+      "epoch": 1.3736225495882148,
+      "grad_norm": 14.214701362197474,
+      "learning_rate": 2e-06,
+      "loss": 0.1959,
+      "step": 5921
+    },
+    {
+      "epoch": 1.3738545412365155,
+      "grad_norm": 12.208764144329555,
+      "learning_rate": 2e-06,
+      "loss": 0.3126,
+      "step": 5922
+    },
+    {
+      "epoch": 1.3740865328848162,
+      "grad_norm": 17.16152183037491,
+      "learning_rate": 2e-06,
+      "loss": 0.3282,
+      "step": 5923
+    },
+    {
+      "epoch": 1.3743185245331169,
+      "grad_norm": 15.279736880276335,
+      "learning_rate": 2e-06,
+      "loss": 0.2844,
+      "step": 5924
+    },
+    {
+      "epoch": 1.3745505161814175,
+      "grad_norm": 12.674344871267387,
+      "learning_rate": 2e-06,
+      "loss": 0.2018,
+      "step": 5925
+    },
+    {
+      "epoch": 1.374782507829718,
+      "grad_norm": 10.83384890181781,
+      "learning_rate": 2e-06,
+      "loss": 0.2479,
+      "step": 5926
+    },
+    {
+      "epoch": 1.3750144994780187,
+      "grad_norm": 12.041522272764032,
+      "learning_rate": 2e-06,
+      "loss": 0.2226,
+      "step": 5927
+    },
+    {
+      "epoch": 1.3752464911263194,
+      "grad_norm": 13.700646292901698,
+      "learning_rate": 2e-06,
+      "loss": 0.284,
+      "step": 5928
+    },
+    {
+      "epoch": 1.37547848277462,
+      "grad_norm": 16.9939098026401,
+      "learning_rate": 2e-06,
+      "loss": 0.2319,
+      "step": 5929
+    },
+    {
+      "epoch": 1.3757104744229207,
+      "grad_norm": 12.29850103313542,
+      "learning_rate": 2e-06,
+      "loss": 0.2211,
+      "step": 5930
+    },
+    {
+      "epoch": 1.3759424660712214,
+      "grad_norm": 22.822665379374467,
+      "learning_rate": 2e-06,
+      "loss": 0.1982,
+      "step": 5931
+    },
+    {
+      "epoch": 1.376174457719522,
+      "grad_norm": 13.545598501814286,
+      "learning_rate": 2e-06,
+      "loss": 0.187,
+      "step": 5932
+    },
+    {
+      "epoch": 1.3764064493678227,
+      "grad_norm": 12.914616833005294,
+      "learning_rate": 2e-06,
+      "loss": 0.2556,
+      "step": 5933
+    },
+    {
+      "epoch": 1.3766384410161234,
+      "grad_norm": 21.749768835773786,
+      "learning_rate": 2e-06,
+      "loss": 0.2858,
+      "step": 5934
+    },
+    {
+      "epoch": 1.376870432664424,
+      "grad_norm": 22.098796702958314,
+      "learning_rate": 2e-06,
+      "loss": 0.3402,
+      "step": 5935
+    },
+    {
+      "epoch": 1.3771024243127248,
+      "grad_norm": 21.733880085073434,
+      "learning_rate": 2e-06,
+      "loss": 0.3829,
+      "step": 5936
+    },
+    {
+      "epoch": 1.3773344159610255,
+      "grad_norm": 8.488122175170245,
+      "learning_rate": 2e-06,
+      "loss": 0.1605,
+      "step": 5937
+    },
+    {
+      "epoch": 1.3775664076093261,
+      "grad_norm": 12.911399378010156,
+      "learning_rate": 2e-06,
+      "loss": 0.2086,
+      "step": 5938
+    },
+    {
+      "epoch": 1.3777983992576268,
+      "grad_norm": 28.886112837195807,
+      "learning_rate": 2e-06,
+      "loss": 0.4906,
+      "step": 5939
+    },
+    {
+      "epoch": 1.3780303909059275,
+      "grad_norm": 21.59522435818442,
+      "learning_rate": 2e-06,
+      "loss": 0.2664,
+      "step": 5940
+    },
+    {
+      "epoch": 1.3782623825542282,
+      "grad_norm": 10.076643720998616,
+      "learning_rate": 2e-06,
+      "loss": 0.2323,
+      "step": 5941
+    },
+    {
+      "epoch": 1.3784943742025286,
+      "grad_norm": 13.516919113099314,
+      "learning_rate": 2e-06,
+      "loss": 0.3002,
+      "step": 5942
+    },
+    {
+      "epoch": 1.3787263658508293,
+      "grad_norm": 8.643098284175489,
+      "learning_rate": 2e-06,
+      "loss": 0.216,
+      "step": 5943
+    },
+    {
+      "epoch": 1.37895835749913,
+      "grad_norm": 14.300103934873006,
+      "learning_rate": 2e-06,
+      "loss": 0.2819,
+      "step": 5944
+    },
+    {
+      "epoch": 1.3791903491474307,
+      "grad_norm": 12.466558909522385,
+      "learning_rate": 2e-06,
+      "loss": 0.2555,
+      "step": 5945
+    },
+    {
+      "epoch": 1.3794223407957313,
+      "grad_norm": 8.138712893736352,
+      "learning_rate": 2e-06,
+      "loss": 0.2659,
+      "step": 5946
+    },
+    {
+      "epoch": 1.379654332444032,
+      "grad_norm": 12.649915403183721,
+      "learning_rate": 2e-06,
+      "loss": 0.2405,
+      "step": 5947
+    },
+    {
+      "epoch": 1.3798863240923327,
+      "grad_norm": 9.271014937152602,
+      "learning_rate": 2e-06,
+      "loss": 0.1644,
+      "step": 5948
+    },
+    {
+      "epoch": 1.3801183157406334,
+      "grad_norm": 12.106163458210577,
+      "learning_rate": 2e-06,
+      "loss": 0.2074,
+      "step": 5949
+    },
+    {
+      "epoch": 1.380350307388934,
+      "grad_norm": 14.840048188527057,
+      "learning_rate": 2e-06,
+      "loss": 0.227,
+      "step": 5950
+    },
+    {
+      "epoch": 1.3805822990372347,
+      "grad_norm": 15.092469306811585,
+      "learning_rate": 2e-06,
+      "loss": 0.3458,
+      "step": 5951
+    },
+    {
+      "epoch": 1.3808142906855352,
+      "grad_norm": 18.896819046822152,
+      "learning_rate": 2e-06,
+      "loss": 0.3011,
+      "step": 5952
+    },
+    {
+      "epoch": 1.3810462823338359,
+      "grad_norm": 12.53153675981379,
+      "learning_rate": 2e-06,
+      "loss": 0.3175,
+      "step": 5953
+    },
+    {
+      "epoch": 1.3812782739821365,
+      "grad_norm": 12.148576689161318,
+      "learning_rate": 2e-06,
+      "loss": 0.2515,
+      "step": 5954
+    },
+    {
+      "epoch": 1.3815102656304372,
+      "grad_norm": 17.49484792170981,
+      "learning_rate": 2e-06,
+      "loss": 0.2591,
+      "step": 5955
+    },
+    {
+      "epoch": 1.381742257278738,
+      "grad_norm": 10.394134652681377,
+      "learning_rate": 2e-06,
+      "loss": 0.2057,
+      "step": 5956
+    },
+    {
+      "epoch": 1.3819742489270386,
+      "grad_norm": 29.127601288530386,
+      "learning_rate": 2e-06,
+      "loss": 0.4277,
+      "step": 5957
+    },
+    {
+      "epoch": 1.3822062405753393,
+      "grad_norm": 22.843391511775554,
+      "learning_rate": 2e-06,
+      "loss": 0.3193,
+      "step": 5958
+    },
+    {
+      "epoch": 1.38243823222364,
+      "grad_norm": 5.510142521551222,
+      "learning_rate": 2e-06,
+      "loss": 0.1484,
+      "step": 5959
+    },
+    {
+      "epoch": 1.3826702238719406,
+      "grad_norm": 11.396703033934903,
+      "learning_rate": 2e-06,
+      "loss": 0.2945,
+      "step": 5960
+    },
+    {
+      "epoch": 1.3829022155202413,
+      "grad_norm": 9.600664042933897,
+      "learning_rate": 2e-06,
+      "loss": 0.1599,
+      "step": 5961
+    },
+    {
+      "epoch": 1.383134207168542,
+      "grad_norm": 5.892640207145712,
+      "learning_rate": 2e-06,
+      "loss": 0.1359,
+      "step": 5962
+    },
+    {
+      "epoch": 1.3833661988168426,
+      "grad_norm": 20.277757054717135,
+      "learning_rate": 2e-06,
+      "loss": 0.2424,
+      "step": 5963
+    },
+    {
+      "epoch": 1.3835981904651433,
+      "grad_norm": 10.898832193056212,
+      "learning_rate": 2e-06,
+      "loss": 0.1662,
+      "step": 5964
+    },
+    {
+      "epoch": 1.383830182113444,
+      "grad_norm": 19.88410341780845,
+      "learning_rate": 2e-06,
+      "loss": 0.2381,
+      "step": 5965
+    },
+    {
+      "epoch": 1.3840621737617447,
+      "grad_norm": 20.916354990874595,
+      "learning_rate": 2e-06,
+      "loss": 0.4176,
+      "step": 5966
+    },
+    {
+      "epoch": 1.3842941654100454,
+      "grad_norm": 12.718636226829762,
+      "learning_rate": 2e-06,
+      "loss": 0.1831,
+      "step": 5967
+    },
+    {
+      "epoch": 1.384526157058346,
+      "grad_norm": 15.40648006298018,
+      "learning_rate": 2e-06,
+      "loss": 0.2073,
+      "step": 5968
+    },
+    {
+      "epoch": 1.3847581487066465,
+      "grad_norm": 14.03245463978515,
+      "learning_rate": 2e-06,
+      "loss": 0.2025,
+      "step": 5969
+    },
+    {
+      "epoch": 1.3849901403549472,
+      "grad_norm": 23.08067780388722,
+      "learning_rate": 2e-06,
+      "loss": 0.3673,
+      "step": 5970
+    },
+    {
+      "epoch": 1.3852221320032478,
+      "grad_norm": 8.710463900190048,
+      "learning_rate": 2e-06,
+      "loss": 0.1394,
+      "step": 5971
+    },
+    {
+      "epoch": 1.3854541236515485,
+      "grad_norm": 19.634103965805803,
+      "learning_rate": 2e-06,
+      "loss": 0.3047,
+      "step": 5972
+    },
+    {
+      "epoch": 1.3856861152998492,
+      "grad_norm": 9.507533107547227,
+      "learning_rate": 2e-06,
+      "loss": 0.1999,
+      "step": 5973
+    },
+    {
+      "epoch": 1.3859181069481499,
+      "grad_norm": 17.26108100615538,
+      "learning_rate": 2e-06,
+      "loss": 0.3609,
+      "step": 5974
+    },
+    {
+      "epoch": 1.3861500985964506,
+      "grad_norm": 11.200525491341423,
+      "learning_rate": 2e-06,
+      "loss": 0.2384,
+      "step": 5975
+    },
+    {
+      "epoch": 1.3863820902447512,
+      "grad_norm": 26.638142154059025,
+      "learning_rate": 2e-06,
+      "loss": 0.4838,
+      "step": 5976
+    },
+    {
+      "epoch": 1.386614081893052,
+      "grad_norm": 9.526430529695023,
+      "learning_rate": 2e-06,
+      "loss": 0.3479,
+      "step": 5977
+    },
+    {
+      "epoch": 1.3868460735413526,
+      "grad_norm": 11.526083961069686,
+      "learning_rate": 2e-06,
+      "loss": 0.2981,
+      "step": 5978
+    },
+    {
+      "epoch": 1.387078065189653,
+      "grad_norm": 18.60022661873739,
+      "learning_rate": 2e-06,
+      "loss": 0.2312,
+      "step": 5979
+    },
+    {
+      "epoch": 1.3873100568379537,
+      "grad_norm": 12.776358544049172,
+      "learning_rate": 2e-06,
+      "loss": 0.2869,
+      "step": 5980
+    },
+    {
+      "epoch": 1.3875420484862544,
+      "grad_norm": 6.80341080091959,
+      "learning_rate": 2e-06,
+      "loss": 0.1541,
+      "step": 5981
+    },
+    {
+      "epoch": 1.387774040134555,
+      "grad_norm": 30.41245426392083,
+      "learning_rate": 2e-06,
+      "loss": 0.3814,
+      "step": 5982
+    },
+    {
+      "epoch": 1.3880060317828558,
+      "grad_norm": 8.694938302800548,
+      "learning_rate": 2e-06,
+      "loss": 0.2095,
+      "step": 5983
+    },
+    {
+      "epoch": 1.3882380234311564,
+      "grad_norm": 17.115266860722507,
+      "learning_rate": 2e-06,
+      "loss": 0.244,
+      "step": 5984
+    },
+    {
+      "epoch": 1.388470015079457,
+      "grad_norm": 11.428928951768523,
+      "learning_rate": 2e-06,
+      "loss": 0.2025,
+      "step": 5985
+    },
+    {
+      "epoch": 1.3887020067277578,
+      "grad_norm": 7.287599703203382,
+      "learning_rate": 2e-06,
+      "loss": 0.136,
+      "step": 5986
+    },
+    {
+      "epoch": 1.3889339983760585,
+      "grad_norm": 22.706310929356366,
+      "learning_rate": 2e-06,
+      "loss": 0.2294,
+      "step": 5987
+    },
+    {
+      "epoch": 1.3891659900243591,
+      "grad_norm": 13.099589997732332,
+      "learning_rate": 2e-06,
+      "loss": 0.2546,
+      "step": 5988
+    },
+    {
+      "epoch": 1.3893979816726598,
+      "grad_norm": 7.785644395983943,
+      "learning_rate": 2e-06,
+      "loss": 0.191,
+      "step": 5989
+    },
+    {
+      "epoch": 1.3896299733209605,
+      "grad_norm": 16.68842225821719,
+      "learning_rate": 2e-06,
+      "loss": 0.3427,
+      "step": 5990
+    },
+    {
+      "epoch": 1.3898619649692612,
+      "grad_norm": 14.016036584118243,
+      "learning_rate": 2e-06,
+      "loss": 0.2749,
+      "step": 5991
+    },
+    {
+      "epoch": 1.3900939566175619,
+      "grad_norm": 7.988120441251208,
+      "learning_rate": 2e-06,
+      "loss": 0.1942,
+      "step": 5992
+    },
+    {
+      "epoch": 1.3903259482658625,
+      "grad_norm": 112.7683572465281,
+      "learning_rate": 2e-06,
+      "loss": 0.3115,
+      "step": 5993
+    },
+    {
+      "epoch": 1.3905579399141632,
+      "grad_norm": 42.66441929587144,
+      "learning_rate": 2e-06,
+      "loss": 0.2085,
+      "step": 5994
+    },
+    {
+      "epoch": 1.3907899315624639,
+      "grad_norm": 10.947839024667601,
+      "learning_rate": 2e-06,
+      "loss": 0.2289,
+      "step": 5995
+    },
+    {
+      "epoch": 1.3910219232107643,
+      "grad_norm": 13.246445827515094,
+      "learning_rate": 2e-06,
+      "loss": 0.2489,
+      "step": 5996
+    },
+    {
+      "epoch": 1.391253914859065,
+      "grad_norm": 17.01408986194027,
+      "learning_rate": 2e-06,
+      "loss": 0.2695,
+      "step": 5997
+    },
+    {
+      "epoch": 1.3914859065073657,
+      "grad_norm": 9.757210001235624,
+      "learning_rate": 2e-06,
+      "loss": 0.215,
+      "step": 5998
+    },
+    {
+      "epoch": 1.3917178981556664,
+      "grad_norm": 14.177056956242843,
+      "learning_rate": 2e-06,
+      "loss": 0.2778,
+      "step": 5999
+    },
+    {
+      "epoch": 1.391949889803967,
+      "grad_norm": 5.241533037856061,
+      "learning_rate": 2e-06,
+      "loss": 0.149,
+      "step": 6000
+    },
+    {
+      "epoch": 1.3921818814522677,
+      "grad_norm": 11.535395075466417,
+      "learning_rate": 2e-06,
+      "loss": 0.2143,
+      "step": 6001
+    },
+    {
+      "epoch": 1.3924138731005684,
+      "grad_norm": 10.4332883563328,
+      "learning_rate": 2e-06,
+      "loss": 0.183,
+      "step": 6002
+    },
+    {
+      "epoch": 1.392645864748869,
+      "grad_norm": 6.525121061567562,
+      "learning_rate": 2e-06,
+      "loss": 0.1406,
+      "step": 6003
+    },
+    {
+      "epoch": 1.3928778563971698,
+      "grad_norm": 9.429201567332692,
+      "learning_rate": 2e-06,
+      "loss": 0.1821,
+      "step": 6004
+    },
+    {
+      "epoch": 1.3931098480454704,
+      "grad_norm": 12.268882730046338,
+      "learning_rate": 2e-06,
+      "loss": 0.2193,
+      "step": 6005
+    },
+    {
+      "epoch": 1.393341839693771,
+      "grad_norm": 21.003497758566507,
+      "learning_rate": 2e-06,
+      "loss": 0.3852,
+      "step": 6006
+    },
+    {
+      "epoch": 1.3935738313420716,
+      "grad_norm": 17.83956602871701,
+      "learning_rate": 2e-06,
+      "loss": 0.323,
+      "step": 6007
+    },
+    {
+      "epoch": 1.3938058229903723,
+      "grad_norm": 10.34881103448905,
+      "learning_rate": 2e-06,
+      "loss": 0.2825,
+      "step": 6008
+    },
+    {
+      "epoch": 1.394037814638673,
+      "grad_norm": 14.011766669465839,
+      "learning_rate": 2e-06,
+      "loss": 0.2148,
+      "step": 6009
+    },
+    {
+      "epoch": 1.3942698062869736,
+      "grad_norm": 19.869740415845733,
+      "learning_rate": 2e-06,
+      "loss": 0.394,
+      "step": 6010
+    },
+    {
+      "epoch": 1.3945017979352743,
+      "grad_norm": 18.403769506198863,
+      "learning_rate": 2e-06,
+      "loss": 0.3615,
+      "step": 6011
+    },
+    {
+      "epoch": 1.394733789583575,
+      "grad_norm": 7.7733595799105615,
+      "learning_rate": 2e-06,
+      "loss": 0.2009,
+      "step": 6012
+    },
+    {
+      "epoch": 1.3949657812318756,
+      "grad_norm": 7.321632491742111,
+      "learning_rate": 2e-06,
+      "loss": 0.1419,
+      "step": 6013
+    },
+    {
+      "epoch": 1.3951977728801763,
+      "grad_norm": 9.675069354228755,
+      "learning_rate": 2e-06,
+      "loss": 0.1768,
+      "step": 6014
+    },
+    {
+      "epoch": 1.395429764528477,
+      "grad_norm": 17.28073435681396,
+      "learning_rate": 2e-06,
+      "loss": 0.2887,
+      "step": 6015
+    },
+    {
+      "epoch": 1.3956617561767777,
+      "grad_norm": 9.356232891758404,
+      "learning_rate": 2e-06,
+      "loss": 0.2237,
+      "step": 6016
+    },
+    {
+      "epoch": 1.3958937478250784,
+      "grad_norm": 12.157659270560018,
+      "learning_rate": 2e-06,
+      "loss": 0.1606,
+      "step": 6017
+    },
+    {
+      "epoch": 1.396125739473379,
+      "grad_norm": 22.129004589054688,
+      "learning_rate": 2e-06,
+      "loss": 0.307,
+      "step": 6018
+    },
+    {
+      "epoch": 1.3963577311216797,
+      "grad_norm": 32.36328986876033,
+      "learning_rate": 2e-06,
+      "loss": 0.4857,
+      "step": 6019
+    },
+    {
+      "epoch": 1.3965897227699804,
+      "grad_norm": 19.052967549651004,
+      "learning_rate": 2e-06,
+      "loss": 0.3291,
+      "step": 6020
+    },
+    {
+      "epoch": 1.396821714418281,
+      "grad_norm": 23.30684610747512,
+      "learning_rate": 2e-06,
+      "loss": 0.2642,
+      "step": 6021
+    },
+    {
+      "epoch": 1.3970537060665815,
+      "grad_norm": 13.054309580640513,
+      "learning_rate": 2e-06,
+      "loss": 0.2567,
+      "step": 6022
+    },
+    {
+      "epoch": 1.3972856977148822,
+      "grad_norm": 11.28301974377214,
+      "learning_rate": 2e-06,
+      "loss": 0.2084,
+      "step": 6023
+    },
+    {
+      "epoch": 1.3975176893631829,
+      "grad_norm": 19.094865999095894,
+      "learning_rate": 2e-06,
+      "loss": 0.4077,
+      "step": 6024
+    },
+    {
+      "epoch": 1.3977496810114836,
+      "grad_norm": 20.856946893139874,
+      "learning_rate": 2e-06,
+      "loss": 0.2084,
+      "step": 6025
+    },
+    {
+      "epoch": 1.3979816726597842,
+      "grad_norm": 9.45025107913317,
+      "learning_rate": 2e-06,
+      "loss": 0.2486,
+      "step": 6026
+    },
+    {
+      "epoch": 1.398213664308085,
+      "grad_norm": 15.203941180895525,
+      "learning_rate": 2e-06,
+      "loss": 0.1621,
+      "step": 6027
+    },
+    {
+      "epoch": 1.3984456559563856,
+      "grad_norm": 6.03861772807484,
+      "learning_rate": 2e-06,
+      "loss": 0.1304,
+      "step": 6028
+    },
+    {
+      "epoch": 1.3986776476046863,
+      "grad_norm": 17.018859853107013,
+      "learning_rate": 2e-06,
+      "loss": 0.3151,
+      "step": 6029
+    },
+    {
+      "epoch": 1.398909639252987,
+      "grad_norm": 15.165568168648043,
+      "learning_rate": 2e-06,
+      "loss": 0.2561,
+      "step": 6030
+    },
+    {
+      "epoch": 1.3991416309012876,
+      "grad_norm": 13.53577123662282,
+      "learning_rate": 2e-06,
+      "loss": 0.2988,
+      "step": 6031
+    },
+    {
+      "epoch": 1.399373622549588,
+      "grad_norm": 14.900819174481283,
+      "learning_rate": 2e-06,
+      "loss": 0.2946,
+      "step": 6032
+    },
+    {
+      "epoch": 1.3996056141978888,
+      "grad_norm": 7.993752940333517,
+      "learning_rate": 2e-06,
+      "loss": 0.2056,
+      "step": 6033
+    },
+    {
+      "epoch": 1.3998376058461894,
+      "grad_norm": 22.356467342587003,
+      "learning_rate": 2e-06,
+      "loss": 0.1956,
+      "step": 6034
+    },
+    {
+      "epoch": 1.4000695974944901,
+      "grad_norm": 14.683958717312493,
+      "learning_rate": 2e-06,
+      "loss": 0.1932,
+      "step": 6035
+    },
+    {
+      "epoch": 1.4003015891427908,
+      "grad_norm": 12.143973408945179,
+      "learning_rate": 2e-06,
+      "loss": 0.2515,
+      "step": 6036
+    },
+    {
+      "epoch": 1.4005335807910915,
+      "grad_norm": 11.753610537279906,
+      "learning_rate": 2e-06,
+      "loss": 0.2235,
+      "step": 6037
+    },
+    {
+      "epoch": 1.4007655724393921,
+      "grad_norm": 22.482569340622312,
+      "learning_rate": 2e-06,
+      "loss": 0.2757,
+      "step": 6038
+    },
+    {
+      "epoch": 1.4009975640876928,
+      "grad_norm": 10.124105869305984,
+      "learning_rate": 2e-06,
+      "loss": 0.2144,
+      "step": 6039
+    },
+    {
+      "epoch": 1.4012295557359935,
+      "grad_norm": 9.48123463264121,
+      "learning_rate": 2e-06,
+      "loss": 0.1909,
+      "step": 6040
+    },
+    {
+      "epoch": 1.4014615473842942,
+      "grad_norm": 6.375181245128169,
+      "learning_rate": 2e-06,
+      "loss": 0.1122,
+      "step": 6041
+    },
+    {
+      "epoch": 1.4016935390325949,
+      "grad_norm": 25.529223668853717,
+      "learning_rate": 2e-06,
+      "loss": 0.3305,
+      "step": 6042
+    },
+    {
+      "epoch": 1.4019255306808955,
+      "grad_norm": 15.821112574407275,
+      "learning_rate": 2e-06,
+      "loss": 0.3286,
+      "step": 6043
+    },
+    {
+      "epoch": 1.4021575223291962,
+      "grad_norm": 9.7919902421017,
+      "learning_rate": 2e-06,
+      "loss": 0.1983,
+      "step": 6044
+    },
+    {
+      "epoch": 1.402389513977497,
+      "grad_norm": 47.05440000845356,
+      "learning_rate": 2e-06,
+      "loss": 0.2881,
+      "step": 6045
+    },
+    {
+      "epoch": 1.4026215056257976,
+      "grad_norm": 15.433693071590781,
+      "learning_rate": 2e-06,
+      "loss": 0.2866,
+      "step": 6046
+    },
+    {
+      "epoch": 1.4028534972740982,
+      "grad_norm": 19.2840330067713,
+      "learning_rate": 2e-06,
+      "loss": 0.413,
+      "step": 6047
+    },
+    {
+      "epoch": 1.403085488922399,
+      "grad_norm": 15.715282873660984,
+      "learning_rate": 2e-06,
+      "loss": 0.3135,
+      "step": 6048
+    },
+    {
+      "epoch": 1.4033174805706994,
+      "grad_norm": 12.690754735658464,
+      "learning_rate": 2e-06,
+      "loss": 0.1997,
+      "step": 6049
+    },
+    {
+      "epoch": 1.403549472219,
+      "grad_norm": 17.86836759439439,
+      "learning_rate": 2e-06,
+      "loss": 0.3227,
+      "step": 6050
+    },
+    {
+      "epoch": 1.4037814638673007,
+      "grad_norm": 20.685699119431618,
+      "learning_rate": 2e-06,
+      "loss": 0.2731,
+      "step": 6051
+    },
+    {
+      "epoch": 1.4040134555156014,
+      "grad_norm": 14.641062692101555,
+      "learning_rate": 2e-06,
+      "loss": 0.2976,
+      "step": 6052
+    },
+    {
+      "epoch": 1.404245447163902,
+      "grad_norm": 7.732580108107947,
+      "learning_rate": 2e-06,
+      "loss": 0.1796,
+      "step": 6053
+    },
+    {
+      "epoch": 1.4044774388122028,
+      "grad_norm": 18.032129481542793,
+      "learning_rate": 2e-06,
+      "loss": 0.3029,
+      "step": 6054
+    },
+    {
+      "epoch": 1.4047094304605034,
+      "grad_norm": 14.180763869770052,
+      "learning_rate": 2e-06,
+      "loss": 0.2168,
+      "step": 6055
+    },
+    {
+      "epoch": 1.4049414221088041,
+      "grad_norm": 13.529057984274884,
+      "learning_rate": 2e-06,
+      "loss": 0.2202,
+      "step": 6056
+    },
+    {
+      "epoch": 1.4051734137571048,
+      "grad_norm": 19.91666727951003,
+      "learning_rate": 2e-06,
+      "loss": 0.1565,
+      "step": 6057
+    },
+    {
+      "epoch": 1.4054054054054055,
+      "grad_norm": 20.439080791635305,
+      "learning_rate": 2e-06,
+      "loss": 0.3029,
+      "step": 6058
+    },
+    {
+      "epoch": 1.405637397053706,
+      "grad_norm": 8.094859396497109,
+      "learning_rate": 2e-06,
+      "loss": 0.159,
+      "step": 6059
+    },
+    {
+      "epoch": 1.4058693887020066,
+      "grad_norm": 17.211446044937162,
+      "learning_rate": 2e-06,
+      "loss": 0.2021,
+      "step": 6060
+    },
+    {
+      "epoch": 1.4061013803503073,
+      "grad_norm": 16.984237415332583,
+      "learning_rate": 2e-06,
+      "loss": 0.3073,
+      "step": 6061
+    },
+    {
+      "epoch": 1.406333371998608,
+      "grad_norm": 19.736131712976423,
+      "learning_rate": 2e-06,
+      "loss": 0.3034,
+      "step": 6062
+    },
+    {
+      "epoch": 1.4065653636469087,
+      "grad_norm": 8.803289909229532,
+      "learning_rate": 2e-06,
+      "loss": 0.2266,
+      "step": 6063
+    },
+    {
+      "epoch": 1.4067973552952093,
+      "grad_norm": 14.67567093094092,
+      "learning_rate": 2e-06,
+      "loss": 0.2873,
+      "step": 6064
+    },
+    {
+      "epoch": 1.40702934694351,
+      "grad_norm": 21.111374577296598,
+      "learning_rate": 2e-06,
+      "loss": 0.3436,
+      "step": 6065
+    },
+    {
+      "epoch": 1.4072613385918107,
+      "grad_norm": 23.285346279881995,
+      "learning_rate": 2e-06,
+      "loss": 0.353,
+      "step": 6066
+    },
+    {
+      "epoch": 1.4074933302401114,
+      "grad_norm": 30.91209969978155,
+      "learning_rate": 2e-06,
+      "loss": 0.3959,
+      "step": 6067
+    },
+    {
+      "epoch": 1.407725321888412,
+      "grad_norm": 23.50666321730142,
+      "learning_rate": 2e-06,
+      "loss": 0.3817,
+      "step": 6068
+    },
+    {
+      "epoch": 1.4079573135367127,
+      "grad_norm": 24.887261676979207,
+      "learning_rate": 2e-06,
+      "loss": 0.265,
+      "step": 6069
+    },
+    {
+      "epoch": 1.4081893051850134,
+      "grad_norm": 17.371120969303167,
+      "learning_rate": 2e-06,
+      "loss": 0.2875,
+      "step": 6070
+    },
+    {
+      "epoch": 1.408421296833314,
+      "grad_norm": 17.01279131314068,
+      "learning_rate": 2e-06,
+      "loss": 0.2949,
+      "step": 6071
+    },
+    {
+      "epoch": 1.4086532884816148,
+      "grad_norm": 8.158139020970319,
+      "learning_rate": 2e-06,
+      "loss": 0.2076,
+      "step": 6072
+    },
+    {
+      "epoch": 1.4088852801299154,
+      "grad_norm": 7.818186686590393,
+      "learning_rate": 2e-06,
+      "loss": 0.1549,
+      "step": 6073
+    },
+    {
+      "epoch": 1.409117271778216,
+      "grad_norm": 5.121188328909677,
+      "learning_rate": 2e-06,
+      "loss": 0.1714,
+      "step": 6074
+    },
+    {
+      "epoch": 1.4093492634265166,
+      "grad_norm": 11.775903756067791,
+      "learning_rate": 2e-06,
+      "loss": 0.2023,
+      "step": 6075
+    },
+    {
+      "epoch": 1.4095812550748172,
+      "grad_norm": 30.858707507154378,
+      "learning_rate": 2e-06,
+      "loss": 0.4582,
+      "step": 6076
+    },
+    {
+      "epoch": 1.409813246723118,
+      "grad_norm": 15.755543306378247,
+      "learning_rate": 2e-06,
+      "loss": 0.2169,
+      "step": 6077
+    },
+    {
+      "epoch": 1.4100452383714186,
+      "grad_norm": 18.69184492966843,
+      "learning_rate": 2e-06,
+      "loss": 0.2139,
+      "step": 6078
+    },
+    {
+      "epoch": 1.4102772300197193,
+      "grad_norm": 27.398404718078048,
+      "learning_rate": 2e-06,
+      "loss": 0.337,
+      "step": 6079
+    },
+    {
+      "epoch": 1.41050922166802,
+      "grad_norm": 25.75019073595999,
+      "learning_rate": 2e-06,
+      "loss": 0.3711,
+      "step": 6080
+    },
+    {
+      "epoch": 1.4107412133163206,
+      "grad_norm": 9.892592852299064,
+      "learning_rate": 2e-06,
+      "loss": 0.2051,
+      "step": 6081
+    },
+    {
+      "epoch": 1.4109732049646213,
+      "grad_norm": 7.408465030887885,
+      "learning_rate": 2e-06,
+      "loss": 0.1045,
+      "step": 6082
+    },
+    {
+      "epoch": 1.411205196612922,
+      "grad_norm": 14.06067060739889,
+      "learning_rate": 2e-06,
+      "loss": 0.1927,
+      "step": 6083
+    },
+    {
+      "epoch": 1.4114371882612227,
+      "grad_norm": 13.913663213295978,
+      "learning_rate": 2e-06,
+      "loss": 0.2588,
+      "step": 6084
+    },
+    {
+      "epoch": 1.4116691799095231,
+      "grad_norm": 10.085229486493613,
+      "learning_rate": 2e-06,
+      "loss": 0.2227,
+      "step": 6085
+    },
+    {
+      "epoch": 1.4119011715578238,
+      "grad_norm": 13.631124668880567,
+      "learning_rate": 2e-06,
+      "loss": 0.2588,
+      "step": 6086
+    },
+    {
+      "epoch": 1.4121331632061245,
+      "grad_norm": 10.88601808887599,
+      "learning_rate": 2e-06,
+      "loss": 0.2068,
+      "step": 6087
+    },
+    {
+      "epoch": 1.4123651548544252,
+      "grad_norm": 13.403020292657482,
+      "learning_rate": 2e-06,
+      "loss": 0.2976,
+      "step": 6088
+    },
+    {
+      "epoch": 1.4125971465027258,
+      "grad_norm": 10.344911101143142,
+      "learning_rate": 2e-06,
+      "loss": 0.2509,
+      "step": 6089
+    },
+    {
+      "epoch": 1.4128291381510265,
+      "grad_norm": 12.71640337739951,
+      "learning_rate": 2e-06,
+      "loss": 0.2407,
+      "step": 6090
+    },
+    {
+      "epoch": 1.4130611297993272,
+      "grad_norm": 15.261059190133613,
+      "learning_rate": 2e-06,
+      "loss": 0.3193,
+      "step": 6091
+    },
+    {
+      "epoch": 1.4132931214476279,
+      "grad_norm": 17.65399175249053,
+      "learning_rate": 2e-06,
+      "loss": 0.2981,
+      "step": 6092
+    },
+    {
+      "epoch": 1.4135251130959285,
+      "grad_norm": 18.415558010960087,
+      "learning_rate": 2e-06,
+      "loss": 0.3171,
+      "step": 6093
+    },
+    {
+      "epoch": 1.4137571047442292,
+      "grad_norm": 18.71760155254099,
+      "learning_rate": 2e-06,
+      "loss": 0.3432,
+      "step": 6094
+    },
+    {
+      "epoch": 1.41398909639253,
+      "grad_norm": 12.08233981768872,
+      "learning_rate": 2e-06,
+      "loss": 0.1896,
+      "step": 6095
+    },
+    {
+      "epoch": 1.4142210880408306,
+      "grad_norm": 12.924121884049468,
+      "learning_rate": 2e-06,
+      "loss": 0.2531,
+      "step": 6096
+    },
+    {
+      "epoch": 1.4144530796891313,
+      "grad_norm": 15.51268648637336,
+      "learning_rate": 2e-06,
+      "loss": 0.2829,
+      "step": 6097
+    },
+    {
+      "epoch": 1.414685071337432,
+      "grad_norm": 17.961260828460986,
+      "learning_rate": 2e-06,
+      "loss": 0.2108,
+      "step": 6098
+    },
+    {
+      "epoch": 1.4149170629857326,
+      "grad_norm": 17.365521433935196,
+      "learning_rate": 2e-06,
+      "loss": 0.2831,
+      "step": 6099
+    },
+    {
+      "epoch": 1.4151490546340333,
+      "grad_norm": 10.163031122556262,
+      "learning_rate": 2e-06,
+      "loss": 0.2289,
+      "step": 6100
+    },
+    {
+      "epoch": 1.415381046282334,
+      "grad_norm": 7.099404642130162,
+      "learning_rate": 2e-06,
+      "loss": 0.1514,
+      "step": 6101
+    },
+    {
+      "epoch": 1.4156130379306344,
+      "grad_norm": 11.052661536656478,
+      "learning_rate": 2e-06,
+      "loss": 0.1811,
+      "step": 6102
+    },
+    {
+      "epoch": 1.415845029578935,
+      "grad_norm": 12.598462452392125,
+      "learning_rate": 2e-06,
+      "loss": 0.1743,
+      "step": 6103
+    },
+    {
+      "epoch": 1.4160770212272358,
+      "grad_norm": 6.835223930186045,
+      "learning_rate": 2e-06,
+      "loss": 0.1597,
+      "step": 6104
+    },
+    {
+      "epoch": 1.4163090128755365,
+      "grad_norm": 26.35206340861612,
+      "learning_rate": 2e-06,
+      "loss": 0.3793,
+      "step": 6105
+    },
+    {
+      "epoch": 1.4165410045238371,
+      "grad_norm": 25.374028954103416,
+      "learning_rate": 2e-06,
+      "loss": 0.3094,
+      "step": 6106
+    },
+    {
+      "epoch": 1.4167729961721378,
+      "grad_norm": 11.833484379950999,
+      "learning_rate": 2e-06,
+      "loss": 0.244,
+      "step": 6107
+    },
+    {
+      "epoch": 1.4170049878204385,
+      "grad_norm": 20.596598339272045,
+      "learning_rate": 2e-06,
+      "loss": 0.3502,
+      "step": 6108
+    },
+    {
+      "epoch": 1.4172369794687392,
+      "grad_norm": 10.987960713511372,
+      "learning_rate": 2e-06,
+      "loss": 0.2556,
+      "step": 6109
+    },
+    {
+      "epoch": 1.4174689711170398,
+      "grad_norm": 7.404113399619784,
+      "learning_rate": 2e-06,
+      "loss": 0.1751,
+      "step": 6110
+    },
+    {
+      "epoch": 1.4177009627653405,
+      "grad_norm": 8.825514460580765,
+      "learning_rate": 2e-06,
+      "loss": 0.1497,
+      "step": 6111
+    },
+    {
+      "epoch": 1.417932954413641,
+      "grad_norm": 15.713391703876344,
+      "learning_rate": 2e-06,
+      "loss": 0.2371,
+      "step": 6112
+    },
+    {
+      "epoch": 1.4181649460619417,
+      "grad_norm": 9.618159708726328,
+      "learning_rate": 2e-06,
+      "loss": 0.2016,
+      "step": 6113
+    },
+    {
+      "epoch": 1.4183969377102423,
+      "grad_norm": 10.16052102551055,
+      "learning_rate": 2e-06,
+      "loss": 0.3071,
+      "step": 6114
+    },
+    {
+      "epoch": 1.418628929358543,
+      "grad_norm": 6.310866150220136,
+      "learning_rate": 2e-06,
+      "loss": 0.1548,
+      "step": 6115
+    },
+    {
+      "epoch": 1.4188609210068437,
+      "grad_norm": 19.80731411393623,
+      "learning_rate": 2e-06,
+      "loss": 0.3295,
+      "step": 6116
+    },
+    {
+      "epoch": 1.4190929126551444,
+      "grad_norm": 13.3606379253285,
+      "learning_rate": 2e-06,
+      "loss": 0.1826,
+      "step": 6117
+    },
+    {
+      "epoch": 1.419324904303445,
+      "grad_norm": 8.81503462660466,
+      "learning_rate": 2e-06,
+      "loss": 0.2251,
+      "step": 6118
+    },
+    {
+      "epoch": 1.4195568959517457,
+      "grad_norm": 15.069781449360283,
+      "learning_rate": 2e-06,
+      "loss": 0.2557,
+      "step": 6119
+    },
+    {
+      "epoch": 1.4197888876000464,
+      "grad_norm": 18.822472041359575,
+      "learning_rate": 2e-06,
+      "loss": 0.3187,
+      "step": 6120
+    },
+    {
+      "epoch": 1.420020879248347,
+      "grad_norm": 20.897490149361044,
+      "learning_rate": 2e-06,
+      "loss": 0.2403,
+      "step": 6121
+    },
+    {
+      "epoch": 1.4202528708966478,
+      "grad_norm": 6.836425013476632,
+      "learning_rate": 2e-06,
+      "loss": 0.1446,
+      "step": 6122
+    },
+    {
+      "epoch": 1.4204848625449484,
+      "grad_norm": 11.993374434727933,
+      "learning_rate": 2e-06,
+      "loss": 0.217,
+      "step": 6123
+    },
+    {
+      "epoch": 1.4207168541932491,
+      "grad_norm": 11.146419401766341,
+      "learning_rate": 2e-06,
+      "loss": 0.158,
+      "step": 6124
+    },
+    {
+      "epoch": 1.4209488458415498,
+      "grad_norm": 12.804609156327725,
+      "learning_rate": 2e-06,
+      "loss": 0.2305,
+      "step": 6125
+    },
+    {
+      "epoch": 1.4211808374898505,
+      "grad_norm": 20.536425822305404,
+      "learning_rate": 2e-06,
+      "loss": 0.2628,
+      "step": 6126
+    },
+    {
+      "epoch": 1.4214128291381511,
+      "grad_norm": 19.6182767891473,
+      "learning_rate": 2e-06,
+      "loss": 0.2312,
+      "step": 6127
+    },
+    {
+      "epoch": 1.4216448207864518,
+      "grad_norm": 8.67891356781281,
+      "learning_rate": 2e-06,
+      "loss": 0.1341,
+      "step": 6128
+    },
+    {
+      "epoch": 1.4218768124347523,
+      "grad_norm": 9.495972989677902,
+      "learning_rate": 2e-06,
+      "loss": 0.194,
+      "step": 6129
+    },
+    {
+      "epoch": 1.422108804083053,
+      "grad_norm": 16.465910741318215,
+      "learning_rate": 2e-06,
+      "loss": 0.2441,
+      "step": 6130
+    },
+    {
+      "epoch": 1.4223407957313536,
+      "grad_norm": 18.64237556218032,
+      "learning_rate": 2e-06,
+      "loss": 0.2429,
+      "step": 6131
+    },
+    {
+      "epoch": 1.4225727873796543,
+      "grad_norm": 15.131650364743992,
+      "learning_rate": 2e-06,
+      "loss": 0.2176,
+      "step": 6132
+    },
+    {
+      "epoch": 1.422804779027955,
+      "grad_norm": 16.521831690467256,
+      "learning_rate": 2e-06,
+      "loss": 0.1971,
+      "step": 6133
+    },
+    {
+      "epoch": 1.4230367706762557,
+      "grad_norm": 15.255936878428859,
+      "learning_rate": 2e-06,
+      "loss": 0.3409,
+      "step": 6134
+    },
+    {
+      "epoch": 1.4232687623245563,
+      "grad_norm": 19.319660936948484,
+      "learning_rate": 2e-06,
+      "loss": 0.1897,
+      "step": 6135
+    },
+    {
+      "epoch": 1.423500753972857,
+      "grad_norm": 19.973456561284625,
+      "learning_rate": 2e-06,
+      "loss": 0.3944,
+      "step": 6136
+    },
+    {
+      "epoch": 1.4237327456211577,
+      "grad_norm": 21.849948560087277,
+      "learning_rate": 2e-06,
+      "loss": 0.3276,
+      "step": 6137
+    },
+    {
+      "epoch": 1.4239647372694584,
+      "grad_norm": 16.2802313700536,
+      "learning_rate": 2e-06,
+      "loss": 0.2491,
+      "step": 6138
+    },
+    {
+      "epoch": 1.4241967289177588,
+      "grad_norm": 19.909440328810053,
+      "learning_rate": 2e-06,
+      "loss": 0.1586,
+      "step": 6139
+    },
+    {
+      "epoch": 1.4244287205660595,
+      "grad_norm": 14.748574132006137,
+      "learning_rate": 2e-06,
+      "loss": 0.2809,
+      "step": 6140
+    },
+    {
+      "epoch": 1.4246607122143602,
+      "grad_norm": 9.825467024454078,
+      "learning_rate": 2e-06,
+      "loss": 0.2456,
+      "step": 6141
+    },
+    {
+      "epoch": 1.4248927038626609,
+      "grad_norm": 15.769056569854035,
+      "learning_rate": 2e-06,
+      "loss": 0.2548,
+      "step": 6142
+    },
+    {
+      "epoch": 1.4251246955109615,
+      "grad_norm": 9.883953478158336,
+      "learning_rate": 2e-06,
+      "loss": 0.2044,
+      "step": 6143
+    },
+    {
+      "epoch": 1.4253566871592622,
+      "grad_norm": 17.016653772181083,
+      "learning_rate": 2e-06,
+      "loss": 0.406,
+      "step": 6144
+    },
+    {
+      "epoch": 1.425588678807563,
+      "grad_norm": 20.34748285243919,
+      "learning_rate": 2e-06,
+      "loss": 0.3685,
+      "step": 6145
+    },
+    {
+      "epoch": 1.4258206704558636,
+      "grad_norm": 11.292193060995796,
+      "learning_rate": 2e-06,
+      "loss": 0.2376,
+      "step": 6146
+    },
+    {
+      "epoch": 1.4260526621041643,
+      "grad_norm": 9.088756757460795,
+      "learning_rate": 2e-06,
+      "loss": 0.1936,
+      "step": 6147
+    },
+    {
+      "epoch": 1.426284653752465,
+      "grad_norm": 18.792375574742056,
+      "learning_rate": 2e-06,
+      "loss": 0.2258,
+      "step": 6148
+    },
+    {
+      "epoch": 1.4265166454007656,
+      "grad_norm": 25.108023742053295,
+      "learning_rate": 2e-06,
+      "loss": 0.2133,
+      "step": 6149
+    },
+    {
+      "epoch": 1.4267486370490663,
+      "grad_norm": 19.186927719163982,
+      "learning_rate": 2e-06,
+      "loss": 0.2004,
+      "step": 6150
+    },
+    {
+      "epoch": 1.426980628697367,
+      "grad_norm": 17.71675497443564,
+      "learning_rate": 2e-06,
+      "loss": 0.2085,
+      "step": 6151
+    },
+    {
+      "epoch": 1.4272126203456676,
+      "grad_norm": 22.360834089520896,
+      "learning_rate": 2e-06,
+      "loss": 0.3317,
+      "step": 6152
+    },
+    {
+      "epoch": 1.4274446119939683,
+      "grad_norm": 19.971085580506056,
+      "learning_rate": 2e-06,
+      "loss": 0.2874,
+      "step": 6153
+    },
+    {
+      "epoch": 1.427676603642269,
+      "grad_norm": 23.875068987608802,
+      "learning_rate": 2e-06,
+      "loss": 0.3523,
+      "step": 6154
+    },
+    {
+      "epoch": 1.4279085952905695,
+      "grad_norm": 14.028916962582743,
+      "learning_rate": 2e-06,
+      "loss": 0.2884,
+      "step": 6155
+    },
+    {
+      "epoch": 1.4281405869388701,
+      "grad_norm": 18.488646503647313,
+      "learning_rate": 2e-06,
+      "loss": 0.2513,
+      "step": 6156
+    },
+    {
+      "epoch": 1.4283725785871708,
+      "grad_norm": 18.390315378298006,
+      "learning_rate": 2e-06,
+      "loss": 0.3414,
+      "step": 6157
+    },
+    {
+      "epoch": 1.4286045702354715,
+      "grad_norm": 25.069614253868096,
+      "learning_rate": 2e-06,
+      "loss": 0.3823,
+      "step": 6158
+    },
+    {
+      "epoch": 1.4288365618837722,
+      "grad_norm": 16.644875413619268,
+      "learning_rate": 2e-06,
+      "loss": 0.2242,
+      "step": 6159
+    },
+    {
+      "epoch": 1.4290685535320728,
+      "grad_norm": 16.553733055395217,
+      "learning_rate": 2e-06,
+      "loss": 0.314,
+      "step": 6160
+    },
+    {
+      "epoch": 1.4293005451803735,
+      "grad_norm": 11.93423897501811,
+      "learning_rate": 2e-06,
+      "loss": 0.2209,
+      "step": 6161
+    },
+    {
+      "epoch": 1.4295325368286742,
+      "grad_norm": 10.172506414261784,
+      "learning_rate": 2e-06,
+      "loss": 0.2577,
+      "step": 6162
+    },
+    {
+      "epoch": 1.4297645284769749,
+      "grad_norm": 13.74352601131,
+      "learning_rate": 2e-06,
+      "loss": 0.1991,
+      "step": 6163
+    },
+    {
+      "epoch": 1.4299965201252756,
+      "grad_norm": 20.315569158586207,
+      "learning_rate": 2e-06,
+      "loss": 0.2665,
+      "step": 6164
+    },
+    {
+      "epoch": 1.430228511773576,
+      "grad_norm": 9.354263225246887,
+      "learning_rate": 2e-06,
+      "loss": 0.2413,
+      "step": 6165
+    },
+    {
+      "epoch": 1.4304605034218767,
+      "grad_norm": 12.08739139621951,
+      "learning_rate": 2e-06,
+      "loss": 0.2746,
+      "step": 6166
+    },
+    {
+      "epoch": 1.4306924950701774,
+      "grad_norm": 9.65658511342512,
+      "learning_rate": 2e-06,
+      "loss": 0.1819,
+      "step": 6167
+    },
+    {
+      "epoch": 1.430924486718478,
+      "grad_norm": 7.610932909883998,
+      "learning_rate": 2e-06,
+      "loss": 0.172,
+      "step": 6168
+    },
+    {
+      "epoch": 1.4311564783667787,
+      "grad_norm": 17.58117935316524,
+      "learning_rate": 2e-06,
+      "loss": 0.278,
+      "step": 6169
+    },
+    {
+      "epoch": 1.4313884700150794,
+      "grad_norm": 16.990502623789205,
+      "learning_rate": 2e-06,
+      "loss": 0.255,
+      "step": 6170
+    },
+    {
+      "epoch": 1.43162046166338,
+      "grad_norm": 13.475962120240382,
+      "learning_rate": 2e-06,
+      "loss": 0.2718,
+      "step": 6171
+    },
+    {
+      "epoch": 1.4318524533116808,
+      "grad_norm": 24.175582308504648,
+      "learning_rate": 2e-06,
+      "loss": 0.2386,
+      "step": 6172
+    },
+    {
+      "epoch": 1.4320844449599814,
+      "grad_norm": 5.392419668157998,
+      "learning_rate": 2e-06,
+      "loss": 0.0997,
+      "step": 6173
+    },
+    {
+      "epoch": 1.4323164366082821,
+      "grad_norm": 18.51628849044846,
+      "learning_rate": 2e-06,
+      "loss": 0.2131,
+      "step": 6174
+    },
+    {
+      "epoch": 1.4325484282565828,
+      "grad_norm": 14.240146276064726,
+      "learning_rate": 2e-06,
+      "loss": 0.1993,
+      "step": 6175
+    },
+    {
+      "epoch": 1.4327804199048835,
+      "grad_norm": 11.586250671216098,
+      "learning_rate": 2e-06,
+      "loss": 0.2414,
+      "step": 6176
+    },
+    {
+      "epoch": 1.4330124115531842,
+      "grad_norm": 11.519426842901828,
+      "learning_rate": 2e-06,
+      "loss": 0.2363,
+      "step": 6177
+    },
+    {
+      "epoch": 1.4332444032014848,
+      "grad_norm": 14.449292115185402,
+      "learning_rate": 2e-06,
+      "loss": 0.2333,
+      "step": 6178
+    },
+    {
+      "epoch": 1.4334763948497855,
+      "grad_norm": 14.108943556459652,
+      "learning_rate": 2e-06,
+      "loss": 0.2261,
+      "step": 6179
+    },
+    {
+      "epoch": 1.4337083864980862,
+      "grad_norm": 16.064294267055452,
+      "learning_rate": 2e-06,
+      "loss": 0.2962,
+      "step": 6180
+    },
+    {
+      "epoch": 1.4339403781463869,
+      "grad_norm": 10.455267909022457,
+      "learning_rate": 2e-06,
+      "loss": 0.1511,
+      "step": 6181
+    },
+    {
+      "epoch": 1.4341723697946873,
+      "grad_norm": 11.967123031710038,
+      "learning_rate": 2e-06,
+      "loss": 0.2125,
+      "step": 6182
+    },
+    {
+      "epoch": 1.434404361442988,
+      "grad_norm": 15.112844010484544,
+      "learning_rate": 2e-06,
+      "loss": 0.1897,
+      "step": 6183
+    },
+    {
+      "epoch": 1.4346363530912887,
+      "grad_norm": 13.501040212475816,
+      "learning_rate": 2e-06,
+      "loss": 0.2221,
+      "step": 6184
+    },
+    {
+      "epoch": 1.4348683447395894,
+      "grad_norm": 11.403029548215033,
+      "learning_rate": 2e-06,
+      "loss": 0.2974,
+      "step": 6185
+    },
+    {
+      "epoch": 1.43510033638789,
+      "grad_norm": 10.528756579434749,
+      "learning_rate": 2e-06,
+      "loss": 0.2717,
+      "step": 6186
+    },
+    {
+      "epoch": 1.4353323280361907,
+      "grad_norm": 17.4785306549272,
+      "learning_rate": 2e-06,
+      "loss": 0.3093,
+      "step": 6187
+    },
+    {
+      "epoch": 1.4355643196844914,
+      "grad_norm": 16.520461776544106,
+      "learning_rate": 2e-06,
+      "loss": 0.2704,
+      "step": 6188
+    },
+    {
+      "epoch": 1.435796311332792,
+      "grad_norm": 18.529090232095495,
+      "learning_rate": 2e-06,
+      "loss": 0.2756,
+      "step": 6189
+    },
+    {
+      "epoch": 1.4360283029810927,
+      "grad_norm": 19.45665700411441,
+      "learning_rate": 2e-06,
+      "loss": 0.1903,
+      "step": 6190
+    },
+    {
+      "epoch": 1.4362602946293934,
+      "grad_norm": 23.243262921443165,
+      "learning_rate": 2e-06,
+      "loss": 0.3617,
+      "step": 6191
+    },
+    {
+      "epoch": 1.4364922862776939,
+      "grad_norm": 14.006968855607878,
+      "learning_rate": 2e-06,
+      "loss": 0.2091,
+      "step": 6192
+    },
+    {
+      "epoch": 1.4367242779259946,
+      "grad_norm": 21.504718595315833,
+      "learning_rate": 2e-06,
+      "loss": 0.2635,
+      "step": 6193
+    },
+    {
+      "epoch": 1.4369562695742952,
+      "grad_norm": 13.817998624581417,
+      "learning_rate": 2e-06,
+      "loss": 0.2324,
+      "step": 6194
+    },
+    {
+      "epoch": 1.437188261222596,
+      "grad_norm": 7.810015772765436,
+      "learning_rate": 2e-06,
+      "loss": 0.1785,
+      "step": 6195
+    },
+    {
+      "epoch": 1.4374202528708966,
+      "grad_norm": 9.365374008202055,
+      "learning_rate": 2e-06,
+      "loss": 0.1485,
+      "step": 6196
+    },
+    {
+      "epoch": 1.4376522445191973,
+      "grad_norm": 10.895895587927544,
+      "learning_rate": 2e-06,
+      "loss": 0.3335,
+      "step": 6197
+    },
+    {
+      "epoch": 1.437884236167498,
+      "grad_norm": 14.286288695476802,
+      "learning_rate": 2e-06,
+      "loss": 0.2615,
+      "step": 6198
+    },
+    {
+      "epoch": 1.4381162278157986,
+      "grad_norm": 10.872284344699505,
+      "learning_rate": 2e-06,
+      "loss": 0.2664,
+      "step": 6199
+    },
+    {
+      "epoch": 1.4383482194640993,
+      "grad_norm": 10.360378431259461,
+      "learning_rate": 2e-06,
+      "loss": 0.223,
+      "step": 6200
+    },
+    {
+      "epoch": 1.4385802111124,
+      "grad_norm": 13.574591955794803,
+      "learning_rate": 2e-06,
+      "loss": 0.2481,
+      "step": 6201
+    },
+    {
+      "epoch": 1.4388122027607007,
+      "grad_norm": 19.38468404639361,
+      "learning_rate": 2e-06,
+      "loss": 0.2538,
+      "step": 6202
+    },
+    {
+      "epoch": 1.4390441944090013,
+      "grad_norm": 17.673282877067486,
+      "learning_rate": 2e-06,
+      "loss": 0.1993,
+      "step": 6203
+    },
+    {
+      "epoch": 1.439276186057302,
+      "grad_norm": 19.05185072264639,
+      "learning_rate": 2e-06,
+      "loss": 0.1937,
+      "step": 6204
+    },
+    {
+      "epoch": 1.4395081777056027,
+      "grad_norm": 14.081400982151527,
+      "learning_rate": 2e-06,
+      "loss": 0.2435,
+      "step": 6205
+    },
+    {
+      "epoch": 1.4397401693539034,
+      "grad_norm": 15.081745367531159,
+      "learning_rate": 2e-06,
+      "loss": 0.187,
+      "step": 6206
+    },
+    {
+      "epoch": 1.439972161002204,
+      "grad_norm": 7.762727291413856,
+      "learning_rate": 2e-06,
+      "loss": 0.1486,
+      "step": 6207
+    },
+    {
+      "epoch": 1.4402041526505045,
+      "grad_norm": 18.9592087976754,
+      "learning_rate": 2e-06,
+      "loss": 0.2228,
+      "step": 6208
+    },
+    {
+      "epoch": 1.4404361442988052,
+      "grad_norm": 11.651180155359892,
+      "learning_rate": 2e-06,
+      "loss": 0.2049,
+      "step": 6209
+    },
+    {
+      "epoch": 1.4406681359471059,
+      "grad_norm": 21.85022053628526,
+      "learning_rate": 2e-06,
+      "loss": 0.3898,
+      "step": 6210
+    },
+    {
+      "epoch": 1.4409001275954065,
+      "grad_norm": 15.022376335743237,
+      "learning_rate": 2e-06,
+      "loss": 0.2378,
+      "step": 6211
+    },
+    {
+      "epoch": 1.4411321192437072,
+      "grad_norm": 14.013016604280779,
+      "learning_rate": 2e-06,
+      "loss": 0.3044,
+      "step": 6212
+    },
+    {
+      "epoch": 1.4413641108920079,
+      "grad_norm": 15.010439248112638,
+      "learning_rate": 2e-06,
+      "loss": 0.25,
+      "step": 6213
+    },
+    {
+      "epoch": 1.4415961025403086,
+      "grad_norm": 14.021784314225817,
+      "learning_rate": 2e-06,
+      "loss": 0.1858,
+      "step": 6214
+    },
+    {
+      "epoch": 1.4418280941886092,
+      "grad_norm": 7.823186542884315,
+      "learning_rate": 2e-06,
+      "loss": 0.1652,
+      "step": 6215
+    },
+    {
+      "epoch": 1.44206008583691,
+      "grad_norm": 15.567837669828194,
+      "learning_rate": 2e-06,
+      "loss": 0.2944,
+      "step": 6216
+    },
+    {
+      "epoch": 1.4422920774852106,
+      "grad_norm": 14.942780282170178,
+      "learning_rate": 2e-06,
+      "loss": 0.3001,
+      "step": 6217
+    },
+    {
+      "epoch": 1.442524069133511,
+      "grad_norm": 19.08348088709312,
+      "learning_rate": 2e-06,
+      "loss": 0.3644,
+      "step": 6218
+    },
+    {
+      "epoch": 1.4427560607818117,
+      "grad_norm": 11.287593844636518,
+      "learning_rate": 2e-06,
+      "loss": 0.2539,
+      "step": 6219
+    },
+    {
+      "epoch": 1.4429880524301124,
+      "grad_norm": 7.621639203002738,
+      "learning_rate": 2e-06,
+      "loss": 0.2016,
+      "step": 6220
+    },
+    {
+      "epoch": 1.443220044078413,
+      "grad_norm": 18.406599332910428,
+      "learning_rate": 2e-06,
+      "loss": 0.2905,
+      "step": 6221
+    },
+    {
+      "epoch": 1.4434520357267138,
+      "grad_norm": 23.286825030451947,
+      "learning_rate": 2e-06,
+      "loss": 0.2663,
+      "step": 6222
+    },
+    {
+      "epoch": 1.4436840273750144,
+      "grad_norm": 8.222848890709107,
+      "learning_rate": 2e-06,
+      "loss": 0.2088,
+      "step": 6223
+    },
+    {
+      "epoch": 1.4439160190233151,
+      "grad_norm": 15.558196318996108,
+      "learning_rate": 2e-06,
+      "loss": 0.2139,
+      "step": 6224
+    },
+    {
+      "epoch": 1.4441480106716158,
+      "grad_norm": 12.775547720889165,
+      "learning_rate": 2e-06,
+      "loss": 0.1868,
+      "step": 6225
+    },
+    {
+      "epoch": 1.4443800023199165,
+      "grad_norm": 7.512811432134344,
+      "learning_rate": 2e-06,
+      "loss": 0.1356,
+      "step": 6226
+    },
+    {
+      "epoch": 1.4446119939682172,
+      "grad_norm": 13.737685430600592,
+      "learning_rate": 2e-06,
+      "loss": 0.2513,
+      "step": 6227
+    },
+    {
+      "epoch": 1.4448439856165178,
+      "grad_norm": 14.156243226031668,
+      "learning_rate": 2e-06,
+      "loss": 0.2744,
+      "step": 6228
+    },
+    {
+      "epoch": 1.4450759772648185,
+      "grad_norm": 17.606808588301096,
+      "learning_rate": 2e-06,
+      "loss": 0.2185,
+      "step": 6229
+    },
+    {
+      "epoch": 1.4453079689131192,
+      "grad_norm": 11.62668769274412,
+      "learning_rate": 2e-06,
+      "loss": 0.2055,
+      "step": 6230
+    },
+    {
+      "epoch": 1.4455399605614199,
+      "grad_norm": 14.338056904667104,
+      "learning_rate": 2e-06,
+      "loss": 0.2705,
+      "step": 6231
+    },
+    {
+      "epoch": 1.4457719522097205,
+      "grad_norm": 8.648296098897648,
+      "learning_rate": 2e-06,
+      "loss": 0.1655,
+      "step": 6232
+    },
+    {
+      "epoch": 1.4460039438580212,
+      "grad_norm": 13.206173762170703,
+      "learning_rate": 2e-06,
+      "loss": 0.2529,
+      "step": 6233
+    },
+    {
+      "epoch": 1.446235935506322,
+      "grad_norm": 9.84978112659568,
+      "learning_rate": 2e-06,
+      "loss": 0.1463,
+      "step": 6234
+    },
+    {
+      "epoch": 1.4464679271546224,
+      "grad_norm": 5.811228390154432,
+      "learning_rate": 2e-06,
+      "loss": 0.1444,
+      "step": 6235
+    },
+    {
+      "epoch": 1.446699918802923,
+      "grad_norm": 8.78844000550045,
+      "learning_rate": 2e-06,
+      "loss": 0.1566,
+      "step": 6236
+    },
+    {
+      "epoch": 1.4469319104512237,
+      "grad_norm": 14.425942706111064,
+      "learning_rate": 2e-06,
+      "loss": 0.2102,
+      "step": 6237
+    },
+    {
+      "epoch": 1.4471639020995244,
+      "grad_norm": 30.388185582567207,
+      "learning_rate": 2e-06,
+      "loss": 0.2826,
+      "step": 6238
+    },
+    {
+      "epoch": 1.447395893747825,
+      "grad_norm": 10.26658546069454,
+      "learning_rate": 2e-06,
+      "loss": 0.2397,
+      "step": 6239
+    },
+    {
+      "epoch": 1.4476278853961257,
+      "grad_norm": 12.24282778864517,
+      "learning_rate": 2e-06,
+      "loss": 0.2435,
+      "step": 6240
+    },
+    {
+      "epoch": 1.4478598770444264,
+      "grad_norm": 32.10949393944506,
+      "learning_rate": 2e-06,
+      "loss": 0.4579,
+      "step": 6241
+    },
+    {
+      "epoch": 1.448091868692727,
+      "grad_norm": 12.488039092828613,
+      "learning_rate": 2e-06,
+      "loss": 0.1936,
+      "step": 6242
+    },
+    {
+      "epoch": 1.4483238603410278,
+      "grad_norm": 19.789272082035247,
+      "learning_rate": 2e-06,
+      "loss": 0.3089,
+      "step": 6243
+    },
+    {
+      "epoch": 1.4485558519893285,
+      "grad_norm": 10.038105505453792,
+      "learning_rate": 2e-06,
+      "loss": 0.2293,
+      "step": 6244
+    },
+    {
+      "epoch": 1.448787843637629,
+      "grad_norm": 14.92566839643602,
+      "learning_rate": 2e-06,
+      "loss": 0.2395,
+      "step": 6245
+    },
+    {
+      "epoch": 1.4490198352859296,
+      "grad_norm": 13.084583951620527,
+      "learning_rate": 2e-06,
+      "loss": 0.2754,
+      "step": 6246
+    },
+    {
+      "epoch": 1.4492518269342303,
+      "grad_norm": 16.02842816644472,
+      "learning_rate": 2e-06,
+      "loss": 0.2676,
+      "step": 6247
+    },
+    {
+      "epoch": 1.449483818582531,
+      "grad_norm": 19.128811378051246,
+      "learning_rate": 2e-06,
+      "loss": 0.2822,
+      "step": 6248
+    },
+    {
+      "epoch": 1.4497158102308316,
+      "grad_norm": 22.26464674913668,
+      "learning_rate": 2e-06,
+      "loss": 0.2977,
+      "step": 6249
+    },
+    {
+      "epoch": 1.4499478018791323,
+      "grad_norm": 18.209839699130395,
+      "learning_rate": 2e-06,
+      "loss": 0.2602,
+      "step": 6250
+    },
+    {
+      "epoch": 1.450179793527433,
+      "grad_norm": 15.616951580530294,
+      "learning_rate": 2e-06,
+      "loss": 0.253,
+      "step": 6251
+    },
+    {
+      "epoch": 1.4504117851757337,
+      "grad_norm": 11.246926978514672,
+      "learning_rate": 2e-06,
+      "loss": 0.1721,
+      "step": 6252
+    },
+    {
+      "epoch": 1.4506437768240343,
+      "grad_norm": 16.39083591876464,
+      "learning_rate": 2e-06,
+      "loss": 0.2644,
+      "step": 6253
+    },
+    {
+      "epoch": 1.450875768472335,
+      "grad_norm": 20.47143320303876,
+      "learning_rate": 2e-06,
+      "loss": 0.2169,
+      "step": 6254
+    },
+    {
+      "epoch": 1.4511077601206357,
+      "grad_norm": 19.1369428998323,
+      "learning_rate": 2e-06,
+      "loss": 0.2878,
+      "step": 6255
+    },
+    {
+      "epoch": 1.4513397517689364,
+      "grad_norm": 12.488871471556534,
+      "learning_rate": 2e-06,
+      "loss": 0.2151,
+      "step": 6256
+    },
+    {
+      "epoch": 1.451571743417237,
+      "grad_norm": 11.303663814651218,
+      "learning_rate": 2e-06,
+      "loss": 0.1897,
+      "step": 6257
+    },
+    {
+      "epoch": 1.4518037350655377,
+      "grad_norm": 14.441768775753307,
+      "learning_rate": 2e-06,
+      "loss": 0.3127,
+      "step": 6258
+    },
+    {
+      "epoch": 1.4520357267138384,
+      "grad_norm": 16.363244917076656,
+      "learning_rate": 2e-06,
+      "loss": 0.2342,
+      "step": 6259
+    },
+    {
+      "epoch": 1.452267718362139,
+      "grad_norm": 23.380606574314694,
+      "learning_rate": 2e-06,
+      "loss": 0.25,
+      "step": 6260
+    },
+    {
+      "epoch": 1.4524997100104398,
+      "grad_norm": 7.810441383500039,
+      "learning_rate": 2e-06,
+      "loss": 0.1686,
+      "step": 6261
+    },
+    {
+      "epoch": 1.4527317016587402,
+      "grad_norm": 17.524082584791838,
+      "learning_rate": 2e-06,
+      "loss": 0.2508,
+      "step": 6262
+    },
+    {
+      "epoch": 1.452963693307041,
+      "grad_norm": 15.673484460883444,
+      "learning_rate": 2e-06,
+      "loss": 0.3772,
+      "step": 6263
+    },
+    {
+      "epoch": 1.4531956849553416,
+      "grad_norm": 8.920955052046363,
+      "learning_rate": 2e-06,
+      "loss": 0.1725,
+      "step": 6264
+    },
+    {
+      "epoch": 1.4534276766036422,
+      "grad_norm": 17.182446994054157,
+      "learning_rate": 2e-06,
+      "loss": 0.2424,
+      "step": 6265
+    },
+    {
+      "epoch": 1.453659668251943,
+      "grad_norm": 17.68772899689125,
+      "learning_rate": 2e-06,
+      "loss": 0.2875,
+      "step": 6266
+    },
+    {
+      "epoch": 1.4538916599002436,
+      "grad_norm": 19.9172602061717,
+      "learning_rate": 2e-06,
+      "loss": 0.3221,
+      "step": 6267
+    },
+    {
+      "epoch": 1.4541236515485443,
+      "grad_norm": 10.481349481722452,
+      "learning_rate": 2e-06,
+      "loss": 0.1947,
+      "step": 6268
+    },
+    {
+      "epoch": 1.454355643196845,
+      "grad_norm": 10.944112076996793,
+      "learning_rate": 2e-06,
+      "loss": 0.2254,
+      "step": 6269
+    },
+    {
+      "epoch": 1.4545876348451456,
+      "grad_norm": 15.685246629354507,
+      "learning_rate": 2e-06,
+      "loss": 0.3094,
+      "step": 6270
+    },
+    {
+      "epoch": 1.4548196264934463,
+      "grad_norm": 8.918649947722725,
+      "learning_rate": 2e-06,
+      "loss": 0.2104,
+      "step": 6271
+    },
+    {
+      "epoch": 1.4550516181417468,
+      "grad_norm": 11.16163245508632,
+      "learning_rate": 2e-06,
+      "loss": 0.1538,
+      "step": 6272
+    },
+    {
+      "epoch": 1.4552836097900474,
+      "grad_norm": 19.679477959880053,
+      "learning_rate": 2e-06,
+      "loss": 0.2507,
+      "step": 6273
+    },
+    {
+      "epoch": 1.4555156014383481,
+      "grad_norm": 12.987750627077375,
+      "learning_rate": 2e-06,
+      "loss": 0.2041,
+      "step": 6274
+    },
+    {
+      "epoch": 1.4557475930866488,
+      "grad_norm": 17.114922892888984,
+      "learning_rate": 2e-06,
+      "loss": 0.3117,
+      "step": 6275
+    },
+    {
+      "epoch": 1.4559795847349495,
+      "grad_norm": 10.067639029819325,
+      "learning_rate": 2e-06,
+      "loss": 0.1696,
+      "step": 6276
+    },
+    {
+      "epoch": 1.4562115763832502,
+      "grad_norm": 17.685941678809247,
+      "learning_rate": 2e-06,
+      "loss": 0.2471,
+      "step": 6277
+    },
+    {
+      "epoch": 1.4564435680315508,
+      "grad_norm": 20.106708525227468,
+      "learning_rate": 2e-06,
+      "loss": 0.2805,
+      "step": 6278
+    },
+    {
+      "epoch": 1.4566755596798515,
+      "grad_norm": 29.450049625966425,
+      "learning_rate": 2e-06,
+      "loss": 0.3727,
+      "step": 6279
+    },
+    {
+      "epoch": 1.4569075513281522,
+      "grad_norm": 12.063322046943563,
+      "learning_rate": 2e-06,
+      "loss": 0.1852,
+      "step": 6280
+    },
+    {
+      "epoch": 1.4571395429764529,
+      "grad_norm": 23.251493560396554,
+      "learning_rate": 2e-06,
+      "loss": 0.2404,
+      "step": 6281
+    },
+    {
+      "epoch": 1.4573715346247536,
+      "grad_norm": 16.424722310407798,
+      "learning_rate": 2e-06,
+      "loss": 0.2657,
+      "step": 6282
+    },
+    {
+      "epoch": 1.4576035262730542,
+      "grad_norm": 48.52568098535008,
+      "learning_rate": 2e-06,
+      "loss": 0.3805,
+      "step": 6283
+    },
+    {
+      "epoch": 1.457835517921355,
+      "grad_norm": 11.824165675935632,
+      "learning_rate": 2e-06,
+      "loss": 0.0905,
+      "step": 6284
+    },
+    {
+      "epoch": 1.4580675095696556,
+      "grad_norm": 12.360254817926524,
+      "learning_rate": 2e-06,
+      "loss": 0.3002,
+      "step": 6285
+    },
+    {
+      "epoch": 1.4582995012179563,
+      "grad_norm": 11.468732781327246,
+      "learning_rate": 2e-06,
+      "loss": 0.2397,
+      "step": 6286
+    },
+    {
+      "epoch": 1.458531492866257,
+      "grad_norm": 17.519255642282676,
+      "learning_rate": 2e-06,
+      "loss": 0.2276,
+      "step": 6287
+    },
+    {
+      "epoch": 1.4587634845145574,
+      "grad_norm": 16.463553573068534,
+      "learning_rate": 2e-06,
+      "loss": 0.2647,
+      "step": 6288
+    },
+    {
+      "epoch": 1.458995476162858,
+      "grad_norm": 10.964272322350954,
+      "learning_rate": 2e-06,
+      "loss": 0.1639,
+      "step": 6289
+    },
+    {
+      "epoch": 1.4592274678111588,
+      "grad_norm": 22.362764075921604,
+      "learning_rate": 2e-06,
+      "loss": 0.2763,
+      "step": 6290
+    },
+    {
+      "epoch": 1.4594594594594594,
+      "grad_norm": 12.523391116774407,
+      "learning_rate": 2e-06,
+      "loss": 0.2957,
+      "step": 6291
+    },
+    {
+      "epoch": 1.45969145110776,
+      "grad_norm": 11.906953940165494,
+      "learning_rate": 2e-06,
+      "loss": 0.2841,
+      "step": 6292
+    },
+    {
+      "epoch": 1.4599234427560608,
+      "grad_norm": 10.510627828123306,
+      "learning_rate": 2e-06,
+      "loss": 0.1849,
+      "step": 6293
+    },
+    {
+      "epoch": 1.4601554344043615,
+      "grad_norm": 11.924520737964247,
+      "learning_rate": 2e-06,
+      "loss": 0.2385,
+      "step": 6294
+    },
+    {
+      "epoch": 1.4603874260526621,
+      "grad_norm": 17.689641377689316,
+      "learning_rate": 2e-06,
+      "loss": 0.4174,
+      "step": 6295
+    },
+    {
+      "epoch": 1.4606194177009628,
+      "grad_norm": 11.519765885822281,
+      "learning_rate": 2e-06,
+      "loss": 0.2632,
+      "step": 6296
+    },
+    {
+      "epoch": 1.4608514093492635,
+      "grad_norm": 12.189681236199014,
+      "learning_rate": 2e-06,
+      "loss": 0.3487,
+      "step": 6297
+    },
+    {
+      "epoch": 1.461083400997564,
+      "grad_norm": 20.49920570696744,
+      "learning_rate": 2e-06,
+      "loss": 0.2388,
+      "step": 6298
+    },
+    {
+      "epoch": 1.4613153926458646,
+      "grad_norm": 12.667179012508328,
+      "learning_rate": 2e-06,
+      "loss": 0.282,
+      "step": 6299
+    },
+    {
+      "epoch": 1.4615473842941653,
+      "grad_norm": 12.849410460911564,
+      "learning_rate": 2e-06,
+      "loss": 0.227,
+      "step": 6300
+    },
+    {
+      "epoch": 1.461779375942466,
+      "grad_norm": 12.650327888469596,
+      "learning_rate": 2e-06,
+      "loss": 0.2,
+      "step": 6301
+    },
+    {
+      "epoch": 1.4620113675907667,
+      "grad_norm": 10.508182667607313,
+      "learning_rate": 2e-06,
+      "loss": 0.227,
+      "step": 6302
+    },
+    {
+      "epoch": 1.4622433592390673,
+      "grad_norm": 11.683754689306165,
+      "learning_rate": 2e-06,
+      "loss": 0.1806,
+      "step": 6303
+    },
+    {
+      "epoch": 1.462475350887368,
+      "grad_norm": 14.253812881316286,
+      "learning_rate": 2e-06,
+      "loss": 0.2678,
+      "step": 6304
+    },
+    {
+      "epoch": 1.4627073425356687,
+      "grad_norm": 12.658627134992473,
+      "learning_rate": 2e-06,
+      "loss": 0.2552,
+      "step": 6305
+    },
+    {
+      "epoch": 1.4629393341839694,
+      "grad_norm": 31.209899902519314,
+      "learning_rate": 2e-06,
+      "loss": 0.3512,
+      "step": 6306
+    },
+    {
+      "epoch": 1.46317132583227,
+      "grad_norm": 6.515647786907523,
+      "learning_rate": 2e-06,
+      "loss": 0.1727,
+      "step": 6307
+    },
+    {
+      "epoch": 1.4634033174805707,
+      "grad_norm": 14.165600444230142,
+      "learning_rate": 2e-06,
+      "loss": 0.2658,
+      "step": 6308
+    },
+    {
+      "epoch": 1.4636353091288714,
+      "grad_norm": 20.574582915051842,
+      "learning_rate": 2e-06,
+      "loss": 0.3432,
+      "step": 6309
+    },
+    {
+      "epoch": 1.463867300777172,
+      "grad_norm": 9.196302550572256,
+      "learning_rate": 2e-06,
+      "loss": 0.2192,
+      "step": 6310
+    },
+    {
+      "epoch": 1.4640992924254728,
+      "grad_norm": 12.910670650009056,
+      "learning_rate": 2e-06,
+      "loss": 0.2553,
+      "step": 6311
+    },
+    {
+      "epoch": 1.4643312840737734,
+      "grad_norm": 16.611732586870954,
+      "learning_rate": 2e-06,
+      "loss": 0.2743,
+      "step": 6312
+    },
+    {
+      "epoch": 1.4645632757220741,
+      "grad_norm": 11.80635259402948,
+      "learning_rate": 2e-06,
+      "loss": 0.2631,
+      "step": 6313
+    },
+    {
+      "epoch": 1.4647952673703748,
+      "grad_norm": 11.714774501709245,
+      "learning_rate": 2e-06,
+      "loss": 0.1788,
+      "step": 6314
+    },
+    {
+      "epoch": 1.4650272590186753,
+      "grad_norm": 14.071179173593196,
+      "learning_rate": 2e-06,
+      "loss": 0.2477,
+      "step": 6315
+    },
+    {
+      "epoch": 1.465259250666976,
+      "grad_norm": 14.841448731705155,
+      "learning_rate": 2e-06,
+      "loss": 0.2876,
+      "step": 6316
+    },
+    {
+      "epoch": 1.4654912423152766,
+      "grad_norm": 16.19049964702802,
+      "learning_rate": 2e-06,
+      "loss": 0.3139,
+      "step": 6317
+    },
+    {
+      "epoch": 1.4657232339635773,
+      "grad_norm": 11.0637061092165,
+      "learning_rate": 2e-06,
+      "loss": 0.2283,
+      "step": 6318
+    },
+    {
+      "epoch": 1.465955225611878,
+      "grad_norm": 11.873817866467068,
+      "learning_rate": 2e-06,
+      "loss": 0.1563,
+      "step": 6319
+    },
+    {
+      "epoch": 1.4661872172601786,
+      "grad_norm": 9.005565476151371,
+      "learning_rate": 2e-06,
+      "loss": 0.1996,
+      "step": 6320
+    },
+    {
+      "epoch": 1.4664192089084793,
+      "grad_norm": 13.384400838128574,
+      "learning_rate": 2e-06,
+      "loss": 0.3289,
+      "step": 6321
+    },
+    {
+      "epoch": 1.46665120055678,
+      "grad_norm": 20.766319695050672,
+      "learning_rate": 2e-06,
+      "loss": 0.2601,
+      "step": 6322
+    },
+    {
+      "epoch": 1.4668831922050807,
+      "grad_norm": 6.532545552029801,
+      "learning_rate": 2e-06,
+      "loss": 0.1765,
+      "step": 6323
+    },
+    {
+      "epoch": 1.4671151838533814,
+      "grad_norm": 7.998792218137757,
+      "learning_rate": 2e-06,
+      "loss": 0.1722,
+      "step": 6324
+    },
+    {
+      "epoch": 1.4673471755016818,
+      "grad_norm": 12.834191100976502,
+      "learning_rate": 2e-06,
+      "loss": 0.2775,
+      "step": 6325
+    },
+    {
+      "epoch": 1.4675791671499825,
+      "grad_norm": 9.317612115484362,
+      "learning_rate": 2e-06,
+      "loss": 0.2043,
+      "step": 6326
+    },
+    {
+      "epoch": 1.4678111587982832,
+      "grad_norm": 5.091989747213939,
+      "learning_rate": 2e-06,
+      "loss": 0.201,
+      "step": 6327
+    },
+    {
+      "epoch": 1.4680431504465838,
+      "grad_norm": 11.99678074994932,
+      "learning_rate": 2e-06,
+      "loss": 0.1815,
+      "step": 6328
+    },
+    {
+      "epoch": 1.4682751420948845,
+      "grad_norm": 21.94462944847759,
+      "learning_rate": 2e-06,
+      "loss": 0.3566,
+      "step": 6329
+    },
+    {
+      "epoch": 1.4685071337431852,
+      "grad_norm": 12.180969031739162,
+      "learning_rate": 2e-06,
+      "loss": 0.245,
+      "step": 6330
+    },
+    {
+      "epoch": 1.4687391253914859,
+      "grad_norm": 9.778283626208468,
+      "learning_rate": 2e-06,
+      "loss": 0.2044,
+      "step": 6331
+    },
+    {
+      "epoch": 1.4689711170397866,
+      "grad_norm": 16.4203032986916,
+      "learning_rate": 2e-06,
+      "loss": 0.2448,
+      "step": 6332
+    },
+    {
+      "epoch": 1.4692031086880872,
+      "grad_norm": 16.900025226867008,
+      "learning_rate": 2e-06,
+      "loss": 0.1976,
+      "step": 6333
+    },
+    {
+      "epoch": 1.469435100336388,
+      "grad_norm": 10.021115257040865,
+      "learning_rate": 2e-06,
+      "loss": 0.1496,
+      "step": 6334
+    },
+    {
+      "epoch": 1.4696670919846886,
+      "grad_norm": 13.389193715874105,
+      "learning_rate": 2e-06,
+      "loss": 0.2155,
+      "step": 6335
+    },
+    {
+      "epoch": 1.4698990836329893,
+      "grad_norm": 8.596070239371144,
+      "learning_rate": 2e-06,
+      "loss": 0.1977,
+      "step": 6336
+    },
+    {
+      "epoch": 1.47013107528129,
+      "grad_norm": 12.127040615020997,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 6337
+    },
+    {
+      "epoch": 1.4703630669295906,
+      "grad_norm": 16.134833289034738,
+      "learning_rate": 2e-06,
+      "loss": 0.215,
+      "step": 6338
+    },
+    {
+      "epoch": 1.4705950585778913,
+      "grad_norm": 10.771203339462213,
+      "learning_rate": 2e-06,
+      "loss": 0.2874,
+      "step": 6339
+    },
+    {
+      "epoch": 1.470827050226192,
+      "grad_norm": 16.109559074612886,
+      "learning_rate": 2e-06,
+      "loss": 0.1802,
+      "step": 6340
+    },
+    {
+      "epoch": 1.4710590418744927,
+      "grad_norm": 24.667718004035784,
+      "learning_rate": 2e-06,
+      "loss": 0.2801,
+      "step": 6341
+    },
+    {
+      "epoch": 1.4712910335227931,
+      "grad_norm": 11.749541340994458,
+      "learning_rate": 2e-06,
+      "loss": 0.1959,
+      "step": 6342
+    },
+    {
+      "epoch": 1.4715230251710938,
+      "grad_norm": 14.16110476245822,
+      "learning_rate": 2e-06,
+      "loss": 0.1827,
+      "step": 6343
+    },
+    {
+      "epoch": 1.4717550168193945,
+      "grad_norm": 10.151478075496506,
+      "learning_rate": 2e-06,
+      "loss": 0.127,
+      "step": 6344
+    },
+    {
+      "epoch": 1.4719870084676951,
+      "grad_norm": 16.90380252331379,
+      "learning_rate": 2e-06,
+      "loss": 0.1837,
+      "step": 6345
+    },
+    {
+      "epoch": 1.4722190001159958,
+      "grad_norm": 8.733030126409302,
+      "learning_rate": 2e-06,
+      "loss": 0.138,
+      "step": 6346
+    },
+    {
+      "epoch": 1.4724509917642965,
+      "grad_norm": 6.654863266480442,
+      "learning_rate": 2e-06,
+      "loss": 0.1539,
+      "step": 6347
+    },
+    {
+      "epoch": 1.4726829834125972,
+      "grad_norm": 23.177115613442897,
+      "learning_rate": 2e-06,
+      "loss": 0.1599,
+      "step": 6348
+    },
+    {
+      "epoch": 1.4729149750608979,
+      "grad_norm": 14.13503325100472,
+      "learning_rate": 2e-06,
+      "loss": 0.3002,
+      "step": 6349
+    },
+    {
+      "epoch": 1.4731469667091985,
+      "grad_norm": 19.238936368187463,
+      "learning_rate": 2e-06,
+      "loss": 0.3622,
+      "step": 6350
+    },
+    {
+      "epoch": 1.4733789583574992,
+      "grad_norm": 20.46860314621833,
+      "learning_rate": 2e-06,
+      "loss": 0.2823,
+      "step": 6351
+    },
+    {
+      "epoch": 1.4736109500057997,
+      "grad_norm": 10.96071359838017,
+      "learning_rate": 2e-06,
+      "loss": 0.178,
+      "step": 6352
+    },
+    {
+      "epoch": 1.4738429416541003,
+      "grad_norm": 17.243761699595712,
+      "learning_rate": 2e-06,
+      "loss": 0.3201,
+      "step": 6353
+    },
+    {
+      "epoch": 1.474074933302401,
+      "grad_norm": 11.761870335399129,
+      "learning_rate": 2e-06,
+      "loss": 0.1588,
+      "step": 6354
+    },
+    {
+      "epoch": 1.4743069249507017,
+      "grad_norm": 15.063788634244156,
+      "learning_rate": 2e-06,
+      "loss": 0.1985,
+      "step": 6355
+    },
+    {
+      "epoch": 1.4745389165990024,
+      "grad_norm": 10.205439964919952,
+      "learning_rate": 2e-06,
+      "loss": 0.1951,
+      "step": 6356
+    },
+    {
+      "epoch": 1.474770908247303,
+      "grad_norm": 19.755369133580384,
+      "learning_rate": 2e-06,
+      "loss": 0.314,
+      "step": 6357
+    },
+    {
+      "epoch": 1.4750028998956037,
+      "grad_norm": 18.02092442331726,
+      "learning_rate": 2e-06,
+      "loss": 0.2532,
+      "step": 6358
+    },
+    {
+      "epoch": 1.4752348915439044,
+      "grad_norm": 20.776127614591175,
+      "learning_rate": 2e-06,
+      "loss": 0.3867,
+      "step": 6359
+    },
+    {
+      "epoch": 1.475466883192205,
+      "grad_norm": 18.8490296827277,
+      "learning_rate": 2e-06,
+      "loss": 0.2271,
+      "step": 6360
+    },
+    {
+      "epoch": 1.4756988748405058,
+      "grad_norm": 16.243421950387734,
+      "learning_rate": 2e-06,
+      "loss": 0.3152,
+      "step": 6361
+    },
+    {
+      "epoch": 1.4759308664888064,
+      "grad_norm": 12.912182042718188,
+      "learning_rate": 2e-06,
+      "loss": 0.2522,
+      "step": 6362
+    },
+    {
+      "epoch": 1.4761628581371071,
+      "grad_norm": 13.434169048365156,
+      "learning_rate": 2e-06,
+      "loss": 0.1501,
+      "step": 6363
+    },
+    {
+      "epoch": 1.4763948497854078,
+      "grad_norm": 23.64559945466988,
+      "learning_rate": 2e-06,
+      "loss": 0.322,
+      "step": 6364
+    },
+    {
+      "epoch": 1.4766268414337085,
+      "grad_norm": 15.840938473928546,
+      "learning_rate": 2e-06,
+      "loss": 0.398,
+      "step": 6365
+    },
+    {
+      "epoch": 1.4768588330820092,
+      "grad_norm": 13.593669173006043,
+      "learning_rate": 2e-06,
+      "loss": 0.2298,
+      "step": 6366
+    },
+    {
+      "epoch": 1.4770908247303098,
+      "grad_norm": 15.484086423496137,
+      "learning_rate": 2e-06,
+      "loss": 0.2102,
+      "step": 6367
+    },
+    {
+      "epoch": 1.4773228163786103,
+      "grad_norm": 20.566877565115348,
+      "learning_rate": 2e-06,
+      "loss": 0.2452,
+      "step": 6368
+    },
+    {
+      "epoch": 1.477554808026911,
+      "grad_norm": 26.017129239071807,
+      "learning_rate": 2e-06,
+      "loss": 0.3534,
+      "step": 6369
+    },
+    {
+      "epoch": 1.4777867996752116,
+      "grad_norm": 22.452637281185886,
+      "learning_rate": 2e-06,
+      "loss": 0.3603,
+      "step": 6370
+    },
+    {
+      "epoch": 1.4780187913235123,
+      "grad_norm": 16.962058059153215,
+      "learning_rate": 2e-06,
+      "loss": 0.2599,
+      "step": 6371
+    },
+    {
+      "epoch": 1.478250782971813,
+      "grad_norm": 31.44988018042093,
+      "learning_rate": 2e-06,
+      "loss": 0.3149,
+      "step": 6372
+    },
+    {
+      "epoch": 1.4784827746201137,
+      "grad_norm": 17.97818075763738,
+      "learning_rate": 2e-06,
+      "loss": 0.2712,
+      "step": 6373
+    },
+    {
+      "epoch": 1.4787147662684144,
+      "grad_norm": 16.298172657415773,
+      "learning_rate": 2e-06,
+      "loss": 0.2514,
+      "step": 6374
+    },
+    {
+      "epoch": 1.478946757916715,
+      "grad_norm": 18.243457940810273,
+      "learning_rate": 2e-06,
+      "loss": 0.2485,
+      "step": 6375
+    },
+    {
+      "epoch": 1.4791787495650157,
+      "grad_norm": 11.290826885683321,
+      "learning_rate": 2e-06,
+      "loss": 0.2569,
+      "step": 6376
+    },
+    {
+      "epoch": 1.4794107412133164,
+      "grad_norm": 8.751435320787085,
+      "learning_rate": 2e-06,
+      "loss": 0.1369,
+      "step": 6377
+    },
+    {
+      "epoch": 1.4796427328616168,
+      "grad_norm": 16.818687395472264,
+      "learning_rate": 2e-06,
+      "loss": 0.209,
+      "step": 6378
+    },
+    {
+      "epoch": 1.4798747245099175,
+      "grad_norm": 15.913409104761076,
+      "learning_rate": 2e-06,
+      "loss": 0.2155,
+      "step": 6379
+    },
+    {
+      "epoch": 1.4801067161582182,
+      "grad_norm": 17.282055857289045,
+      "learning_rate": 2e-06,
+      "loss": 0.3269,
+      "step": 6380
+    },
+    {
+      "epoch": 1.4803387078065189,
+      "grad_norm": 17.084959173190963,
+      "learning_rate": 2e-06,
+      "loss": 0.3279,
+      "step": 6381
+    },
+    {
+      "epoch": 1.4805706994548196,
+      "grad_norm": 10.158009644371294,
+      "learning_rate": 2e-06,
+      "loss": 0.3108,
+      "step": 6382
+    },
+    {
+      "epoch": 1.4808026911031202,
+      "grad_norm": 13.161002799442894,
+      "learning_rate": 2e-06,
+      "loss": 0.2121,
+      "step": 6383
+    },
+    {
+      "epoch": 1.481034682751421,
+      "grad_norm": 10.16665338585957,
+      "learning_rate": 2e-06,
+      "loss": 0.1704,
+      "step": 6384
+    },
+    {
+      "epoch": 1.4812666743997216,
+      "grad_norm": 11.521240503832262,
+      "learning_rate": 2e-06,
+      "loss": 0.2038,
+      "step": 6385
+    },
+    {
+      "epoch": 1.4814986660480223,
+      "grad_norm": 15.73741550757018,
+      "learning_rate": 2e-06,
+      "loss": 0.3129,
+      "step": 6386
+    },
+    {
+      "epoch": 1.481730657696323,
+      "grad_norm": 12.864819746200281,
+      "learning_rate": 2e-06,
+      "loss": 0.2185,
+      "step": 6387
+    },
+    {
+      "epoch": 1.4819626493446236,
+      "grad_norm": 13.343572206351373,
+      "learning_rate": 2e-06,
+      "loss": 0.2572,
+      "step": 6388
+    },
+    {
+      "epoch": 1.4821946409929243,
+      "grad_norm": 11.726772864515828,
+      "learning_rate": 2e-06,
+      "loss": 0.214,
+      "step": 6389
+    },
+    {
+      "epoch": 1.482426632641225,
+      "grad_norm": 12.530670113616015,
+      "learning_rate": 2e-06,
+      "loss": 0.2401,
+      "step": 6390
+    },
+    {
+      "epoch": 1.4826586242895257,
+      "grad_norm": 19.523033448539547,
+      "learning_rate": 2e-06,
+      "loss": 0.3102,
+      "step": 6391
+    },
+    {
+      "epoch": 1.4828906159378263,
+      "grad_norm": 14.813811087640675,
+      "learning_rate": 2e-06,
+      "loss": 0.2472,
+      "step": 6392
+    },
+    {
+      "epoch": 1.483122607586127,
+      "grad_norm": 11.124681524188954,
+      "learning_rate": 2e-06,
+      "loss": 0.1997,
+      "step": 6393
+    },
+    {
+      "epoch": 1.4833545992344277,
+      "grad_norm": 13.281192962940535,
+      "learning_rate": 2e-06,
+      "loss": 0.1957,
+      "step": 6394
+    },
+    {
+      "epoch": 1.4835865908827282,
+      "grad_norm": 9.593242911666561,
+      "learning_rate": 2e-06,
+      "loss": 0.1657,
+      "step": 6395
+    },
+    {
+      "epoch": 1.4838185825310288,
+      "grad_norm": 14.775396233103777,
+      "learning_rate": 2e-06,
+      "loss": 0.2757,
+      "step": 6396
+    },
+    {
+      "epoch": 1.4840505741793295,
+      "grad_norm": 10.907066557516638,
+      "learning_rate": 2e-06,
+      "loss": 0.1952,
+      "step": 6397
+    },
+    {
+      "epoch": 1.4842825658276302,
+      "grad_norm": 12.904636573491848,
+      "learning_rate": 2e-06,
+      "loss": 0.2891,
+      "step": 6398
+    },
+    {
+      "epoch": 1.4845145574759309,
+      "grad_norm": 16.960841103912564,
+      "learning_rate": 2e-06,
+      "loss": 0.3691,
+      "step": 6399
+    },
+    {
+      "epoch": 1.4847465491242315,
+      "grad_norm": 10.291992699295115,
+      "learning_rate": 2e-06,
+      "loss": 0.2032,
+      "step": 6400
+    },
+    {
+      "epoch": 1.4849785407725322,
+      "grad_norm": 10.330431592292854,
+      "learning_rate": 2e-06,
+      "loss": 0.1908,
+      "step": 6401
+    },
+    {
+      "epoch": 1.485210532420833,
+      "grad_norm": 12.167172292182945,
+      "learning_rate": 2e-06,
+      "loss": 0.2551,
+      "step": 6402
+    },
+    {
+      "epoch": 1.4854425240691336,
+      "grad_norm": 9.095098976169192,
+      "learning_rate": 2e-06,
+      "loss": 0.203,
+      "step": 6403
+    },
+    {
+      "epoch": 1.4856745157174343,
+      "grad_norm": 21.738840895505344,
+      "learning_rate": 2e-06,
+      "loss": 0.3902,
+      "step": 6404
+    },
+    {
+      "epoch": 1.4859065073657347,
+      "grad_norm": 20.273545028134055,
+      "learning_rate": 2e-06,
+      "loss": 0.2845,
+      "step": 6405
+    },
+    {
+      "epoch": 1.4861384990140354,
+      "grad_norm": 15.994132167667901,
+      "learning_rate": 2e-06,
+      "loss": 0.2778,
+      "step": 6406
+    },
+    {
+      "epoch": 1.486370490662336,
+      "grad_norm": 13.177377098709755,
+      "learning_rate": 2e-06,
+      "loss": 0.2385,
+      "step": 6407
+    },
+    {
+      "epoch": 1.4866024823106367,
+      "grad_norm": 18.166082735053617,
+      "learning_rate": 2e-06,
+      "loss": 0.3841,
+      "step": 6408
+    },
+    {
+      "epoch": 1.4868344739589374,
+      "grad_norm": 19.653124970701338,
+      "learning_rate": 2e-06,
+      "loss": 0.2373,
+      "step": 6409
+    },
+    {
+      "epoch": 1.487066465607238,
+      "grad_norm": 20.508541376693575,
+      "learning_rate": 2e-06,
+      "loss": 0.3786,
+      "step": 6410
+    },
+    {
+      "epoch": 1.4872984572555388,
+      "grad_norm": 18.380647360810677,
+      "learning_rate": 2e-06,
+      "loss": 0.3239,
+      "step": 6411
+    },
+    {
+      "epoch": 1.4875304489038395,
+      "grad_norm": 32.4958970734748,
+      "learning_rate": 2e-06,
+      "loss": 0.2241,
+      "step": 6412
+    },
+    {
+      "epoch": 1.4877624405521401,
+      "grad_norm": 29.17435793636348,
+      "learning_rate": 2e-06,
+      "loss": 0.2003,
+      "step": 6413
+    },
+    {
+      "epoch": 1.4879944322004408,
+      "grad_norm": 14.065406253018462,
+      "learning_rate": 2e-06,
+      "loss": 0.3424,
+      "step": 6414
+    },
+    {
+      "epoch": 1.4882264238487415,
+      "grad_norm": 16.456177366838705,
+      "learning_rate": 2e-06,
+      "loss": 0.2163,
+      "step": 6415
+    },
+    {
+      "epoch": 1.4884584154970422,
+      "grad_norm": 18.49614960441739,
+      "learning_rate": 2e-06,
+      "loss": 0.3273,
+      "step": 6416
+    },
+    {
+      "epoch": 1.4886904071453428,
+      "grad_norm": 12.96237766808122,
+      "learning_rate": 2e-06,
+      "loss": 0.1475,
+      "step": 6417
+    },
+    {
+      "epoch": 1.4889223987936435,
+      "grad_norm": 13.498753630694996,
+      "learning_rate": 2e-06,
+      "loss": 0.2419,
+      "step": 6418
+    },
+    {
+      "epoch": 1.4891543904419442,
+      "grad_norm": 19.590928089711877,
+      "learning_rate": 2e-06,
+      "loss": 0.4618,
+      "step": 6419
+    },
+    {
+      "epoch": 1.4893863820902449,
+      "grad_norm": 51.909561474681176,
+      "learning_rate": 2e-06,
+      "loss": 0.3855,
+      "step": 6420
+    },
+    {
+      "epoch": 1.4896183737385453,
+      "grad_norm": 14.388437806863546,
+      "learning_rate": 2e-06,
+      "loss": 0.2515,
+      "step": 6421
+    },
+    {
+      "epoch": 1.489850365386846,
+      "grad_norm": 21.97510015415248,
+      "learning_rate": 2e-06,
+      "loss": 0.4647,
+      "step": 6422
+    },
+    {
+      "epoch": 1.4900823570351467,
+      "grad_norm": 10.430234245754157,
+      "learning_rate": 2e-06,
+      "loss": 0.1647,
+      "step": 6423
+    },
+    {
+      "epoch": 1.4903143486834474,
+      "grad_norm": 11.789116182208888,
+      "learning_rate": 2e-06,
+      "loss": 0.2325,
+      "step": 6424
+    },
+    {
+      "epoch": 1.490546340331748,
+      "grad_norm": 9.155597005212678,
+      "learning_rate": 2e-06,
+      "loss": 0.2378,
+      "step": 6425
+    },
+    {
+      "epoch": 1.4907783319800487,
+      "grad_norm": 11.338288541901479,
+      "learning_rate": 2e-06,
+      "loss": 0.3419,
+      "step": 6426
+    },
+    {
+      "epoch": 1.4910103236283494,
+      "grad_norm": 13.99855524710128,
+      "learning_rate": 2e-06,
+      "loss": 0.3002,
+      "step": 6427
+    },
+    {
+      "epoch": 1.49124231527665,
+      "grad_norm": 19.044828598098015,
+      "learning_rate": 2e-06,
+      "loss": 0.3295,
+      "step": 6428
+    },
+    {
+      "epoch": 1.4914743069249508,
+      "grad_norm": 10.062277970290738,
+      "learning_rate": 2e-06,
+      "loss": 0.2077,
+      "step": 6429
+    },
+    {
+      "epoch": 1.4917062985732514,
+      "grad_norm": 14.284423700285167,
+      "learning_rate": 2e-06,
+      "loss": 0.2801,
+      "step": 6430
+    },
+    {
+      "epoch": 1.4919382902215519,
+      "grad_norm": 35.50237344890928,
+      "learning_rate": 2e-06,
+      "loss": 0.2168,
+      "step": 6431
+    },
+    {
+      "epoch": 1.4921702818698526,
+      "grad_norm": 5.948831486266877,
+      "learning_rate": 2e-06,
+      "loss": 0.1457,
+      "step": 6432
+    },
+    {
+      "epoch": 1.4924022735181532,
+      "grad_norm": 15.893248474505135,
+      "learning_rate": 2e-06,
+      "loss": 0.2549,
+      "step": 6433
+    },
+    {
+      "epoch": 1.492634265166454,
+      "grad_norm": 17.785878418025135,
+      "learning_rate": 2e-06,
+      "loss": 0.3139,
+      "step": 6434
+    },
+    {
+      "epoch": 1.4928662568147546,
+      "grad_norm": 17.70381675074382,
+      "learning_rate": 2e-06,
+      "loss": 0.3474,
+      "step": 6435
+    },
+    {
+      "epoch": 1.4930982484630553,
+      "grad_norm": 12.609051573865418,
+      "learning_rate": 2e-06,
+      "loss": 0.2953,
+      "step": 6436
+    },
+    {
+      "epoch": 1.493330240111356,
+      "grad_norm": 5.811534282114288,
+      "learning_rate": 2e-06,
+      "loss": 0.2519,
+      "step": 6437
+    },
+    {
+      "epoch": 1.4935622317596566,
+      "grad_norm": 10.391498834861888,
+      "learning_rate": 2e-06,
+      "loss": 0.1582,
+      "step": 6438
+    },
+    {
+      "epoch": 1.4937942234079573,
+      "grad_norm": 10.597649958632756,
+      "learning_rate": 2e-06,
+      "loss": 0.2445,
+      "step": 6439
+    },
+    {
+      "epoch": 1.494026215056258,
+      "grad_norm": 9.533945431912754,
+      "learning_rate": 2e-06,
+      "loss": 0.1678,
+      "step": 6440
+    },
+    {
+      "epoch": 1.4942582067045587,
+      "grad_norm": 19.709720547193285,
+      "learning_rate": 2e-06,
+      "loss": 0.3057,
+      "step": 6441
+    },
+    {
+      "epoch": 1.4944901983528593,
+      "grad_norm": 17.841104378481514,
+      "learning_rate": 2e-06,
+      "loss": 0.2086,
+      "step": 6442
+    },
+    {
+      "epoch": 1.49472219000116,
+      "grad_norm": 13.695781339844167,
+      "learning_rate": 2e-06,
+      "loss": 0.2651,
+      "step": 6443
+    },
+    {
+      "epoch": 1.4949541816494607,
+      "grad_norm": 10.329248417507095,
+      "learning_rate": 2e-06,
+      "loss": 0.2072,
+      "step": 6444
+    },
+    {
+      "epoch": 1.4951861732977614,
+      "grad_norm": 11.273298957941135,
+      "learning_rate": 2e-06,
+      "loss": 0.3243,
+      "step": 6445
+    },
+    {
+      "epoch": 1.495418164946062,
+      "grad_norm": 11.123275926165398,
+      "learning_rate": 2e-06,
+      "loss": 0.2626,
+      "step": 6446
+    },
+    {
+      "epoch": 1.4956501565943627,
+      "grad_norm": 19.591634516882863,
+      "learning_rate": 2e-06,
+      "loss": 0.3643,
+      "step": 6447
+    },
+    {
+      "epoch": 1.4958821482426632,
+      "grad_norm": 14.93813363970523,
+      "learning_rate": 2e-06,
+      "loss": 0.2281,
+      "step": 6448
+    },
+    {
+      "epoch": 1.4961141398909639,
+      "grad_norm": 14.059895540139681,
+      "learning_rate": 2e-06,
+      "loss": 0.2931,
+      "step": 6449
+    },
+    {
+      "epoch": 1.4963461315392645,
+      "grad_norm": 13.686524407500817,
+      "learning_rate": 2e-06,
+      "loss": 0.2201,
+      "step": 6450
+    },
+    {
+      "epoch": 1.4965781231875652,
+      "grad_norm": 12.69422427186578,
+      "learning_rate": 2e-06,
+      "loss": 0.2263,
+      "step": 6451
+    },
+    {
+      "epoch": 1.496810114835866,
+      "grad_norm": 6.64847972204828,
+      "learning_rate": 2e-06,
+      "loss": 0.1778,
+      "step": 6452
+    },
+    {
+      "epoch": 1.4970421064841666,
+      "grad_norm": 10.611507182667797,
+      "learning_rate": 2e-06,
+      "loss": 0.2653,
+      "step": 6453
+    },
+    {
+      "epoch": 1.4972740981324673,
+      "grad_norm": 8.206805145949547,
+      "learning_rate": 2e-06,
+      "loss": 0.1589,
+      "step": 6454
+    },
+    {
+      "epoch": 1.497506089780768,
+      "grad_norm": 12.934838866814447,
+      "learning_rate": 2e-06,
+      "loss": 0.2327,
+      "step": 6455
+    },
+    {
+      "epoch": 1.4977380814290686,
+      "grad_norm": 8.605576047699119,
+      "learning_rate": 2e-06,
+      "loss": 0.1946,
+      "step": 6456
+    },
+    {
+      "epoch": 1.4979700730773693,
+      "grad_norm": 13.740337789467292,
+      "learning_rate": 2e-06,
+      "loss": 0.1989,
+      "step": 6457
+    },
+    {
+      "epoch": 1.4982020647256697,
+      "grad_norm": 12.719383705892868,
+      "learning_rate": 2e-06,
+      "loss": 0.3019,
+      "step": 6458
+    },
+    {
+      "epoch": 1.4984340563739704,
+      "grad_norm": 9.414119565019343,
+      "learning_rate": 2e-06,
+      "loss": 0.1875,
+      "step": 6459
+    },
+    {
+      "epoch": 1.498666048022271,
+      "grad_norm": 7.286243777258862,
+      "learning_rate": 2e-06,
+      "loss": 0.2029,
+      "step": 6460
+    },
+    {
+      "epoch": 1.4988980396705718,
+      "grad_norm": 25.087415884774984,
+      "learning_rate": 2e-06,
+      "loss": 0.3493,
+      "step": 6461
+    },
+    {
+      "epoch": 1.4991300313188725,
+      "grad_norm": 19.529758328612694,
+      "learning_rate": 2e-06,
+      "loss": 0.2511,
+      "step": 6462
+    },
+    {
+      "epoch": 1.4993620229671731,
+      "grad_norm": 10.558360675132494,
+      "learning_rate": 2e-06,
+      "loss": 0.3066,
+      "step": 6463
+    },
+    {
+      "epoch": 1.4995940146154738,
+      "grad_norm": 12.929556331585013,
+      "learning_rate": 2e-06,
+      "loss": 0.2349,
+      "step": 6464
+    },
+    {
+      "epoch": 1.4998260062637745,
+      "grad_norm": 13.317649088780495,
+      "learning_rate": 2e-06,
+      "loss": 0.2356,
+      "step": 6465
+    },
+    {
+      "epoch": 1.5000579979120752,
+      "grad_norm": 14.80770359106743,
+      "learning_rate": 2e-06,
+      "loss": 0.2201,
+      "step": 6466
+    },
+    {
+      "epoch": 1.5002899895603758,
+      "grad_norm": 10.099958255530332,
+      "learning_rate": 2e-06,
+      "loss": 0.2774,
+      "step": 6467
+    },
+    {
+      "epoch": 1.5005219812086765,
+      "grad_norm": 10.157809884548199,
+      "learning_rate": 2e-06,
+      "loss": 0.2198,
+      "step": 6468
+    },
+    {
+      "epoch": 1.5007539728569772,
+      "grad_norm": 12.343167575809934,
+      "learning_rate": 2e-06,
+      "loss": 0.26,
+      "step": 6469
+    },
+    {
+      "epoch": 1.5009859645052779,
+      "grad_norm": 16.69967960591968,
+      "learning_rate": 2e-06,
+      "loss": 0.3119,
+      "step": 6470
+    },
+    {
+      "epoch": 1.5012179561535786,
+      "grad_norm": 11.728765200927777,
+      "learning_rate": 2e-06,
+      "loss": 0.2118,
+      "step": 6471
+    },
+    {
+      "epoch": 1.5014499478018792,
+      "grad_norm": 11.541890012505004,
+      "learning_rate": 2e-06,
+      "loss": 0.1384,
+      "step": 6472
+    },
+    {
+      "epoch": 1.50168193945018,
+      "grad_norm": 9.732218961699658,
+      "learning_rate": 2e-06,
+      "loss": 0.1718,
+      "step": 6473
+    },
+    {
+      "epoch": 1.5019139310984806,
+      "grad_norm": 17.88222137496914,
+      "learning_rate": 2e-06,
+      "loss": 0.309,
+      "step": 6474
+    },
+    {
+      "epoch": 1.5021459227467813,
+      "grad_norm": 16.773658318362514,
+      "learning_rate": 2e-06,
+      "loss": 0.2618,
+      "step": 6475
+    },
+    {
+      "epoch": 1.5023779143950817,
+      "grad_norm": 12.08839410124754,
+      "learning_rate": 2e-06,
+      "loss": 0.2378,
+      "step": 6476
+    },
+    {
+      "epoch": 1.5026099060433824,
+      "grad_norm": 14.656155009066419,
+      "learning_rate": 2e-06,
+      "loss": 0.2386,
+      "step": 6477
+    },
+    {
+      "epoch": 1.502841897691683,
+      "grad_norm": 9.973123753284185,
+      "learning_rate": 2e-06,
+      "loss": 0.308,
+      "step": 6478
+    },
+    {
+      "epoch": 1.5030738893399838,
+      "grad_norm": 41.89955411808061,
+      "learning_rate": 2e-06,
+      "loss": 0.1906,
+      "step": 6479
+    },
+    {
+      "epoch": 1.5033058809882844,
+      "grad_norm": 10.945620291360028,
+      "learning_rate": 2e-06,
+      "loss": 0.204,
+      "step": 6480
+    },
+    {
+      "epoch": 1.5035378726365851,
+      "grad_norm": 10.877027582989168,
+      "learning_rate": 2e-06,
+      "loss": 0.2483,
+      "step": 6481
+    },
+    {
+      "epoch": 1.5037698642848858,
+      "grad_norm": 13.071631401641872,
+      "learning_rate": 2e-06,
+      "loss": 0.3342,
+      "step": 6482
+    },
+    {
+      "epoch": 1.5040018559331862,
+      "grad_norm": 13.171544565902238,
+      "learning_rate": 2e-06,
+      "loss": 0.3112,
+      "step": 6483
+    },
+    {
+      "epoch": 1.504233847581487,
+      "grad_norm": 12.170105872800002,
+      "learning_rate": 2e-06,
+      "loss": 0.2779,
+      "step": 6484
+    },
+    {
+      "epoch": 1.5044658392297876,
+      "grad_norm": 13.122499592129397,
+      "learning_rate": 2e-06,
+      "loss": 0.1801,
+      "step": 6485
+    },
+    {
+      "epoch": 1.5046978308780883,
+      "grad_norm": 15.47394430721565,
+      "learning_rate": 2e-06,
+      "loss": 0.2778,
+      "step": 6486
+    },
+    {
+      "epoch": 1.504929822526389,
+      "grad_norm": 11.18579845578175,
+      "learning_rate": 2e-06,
+      "loss": 0.3338,
+      "step": 6487
+    },
+    {
+      "epoch": 1.5051618141746896,
+      "grad_norm": 19.618419994055362,
+      "learning_rate": 2e-06,
+      "loss": 0.3429,
+      "step": 6488
+    },
+    {
+      "epoch": 1.5053938058229903,
+      "grad_norm": 10.515346377508802,
+      "learning_rate": 2e-06,
+      "loss": 0.2214,
+      "step": 6489
+    },
+    {
+      "epoch": 1.505625797471291,
+      "grad_norm": 16.76037884514449,
+      "learning_rate": 2e-06,
+      "loss": 0.3002,
+      "step": 6490
+    },
+    {
+      "epoch": 1.5058577891195917,
+      "grad_norm": 18.413510022031442,
+      "learning_rate": 2e-06,
+      "loss": 0.2109,
+      "step": 6491
+    },
+    {
+      "epoch": 1.5060897807678923,
+      "grad_norm": 45.908922289749256,
+      "learning_rate": 2e-06,
+      "loss": 0.2336,
+      "step": 6492
+    },
+    {
+      "epoch": 1.506321772416193,
+      "grad_norm": 15.561067182649099,
+      "learning_rate": 2e-06,
+      "loss": 0.2545,
+      "step": 6493
+    },
+    {
+      "epoch": 1.5065537640644937,
+      "grad_norm": 16.969709333468,
+      "learning_rate": 2e-06,
+      "loss": 0.3423,
+      "step": 6494
+    },
+    {
+      "epoch": 1.5067857557127944,
+      "grad_norm": 10.3194205977704,
+      "learning_rate": 2e-06,
+      "loss": 0.1431,
+      "step": 6495
+    },
+    {
+      "epoch": 1.507017747361095,
+      "grad_norm": 17.808742052277843,
+      "learning_rate": 2e-06,
+      "loss": 0.3041,
+      "step": 6496
+    },
+    {
+      "epoch": 1.5072497390093957,
+      "grad_norm": 9.31469480252244,
+      "learning_rate": 2e-06,
+      "loss": 0.2087,
+      "step": 6497
+    },
+    {
+      "epoch": 1.5074817306576964,
+      "grad_norm": 17.60457924421259,
+      "learning_rate": 2e-06,
+      "loss": 0.2772,
+      "step": 6498
+    },
+    {
+      "epoch": 1.507713722305997,
+      "grad_norm": 11.324853479445679,
+      "learning_rate": 2e-06,
+      "loss": 0.2332,
+      "step": 6499
+    },
+    {
+      "epoch": 1.5079457139542978,
+      "grad_norm": 5.4358127032744035,
+      "learning_rate": 2e-06,
+      "loss": 0.1556,
+      "step": 6500
+    },
+    {
+      "epoch": 1.5081777056025985,
+      "grad_norm": 10.692291572338746,
+      "learning_rate": 2e-06,
+      "loss": 0.264,
+      "step": 6501
+    },
+    {
+      "epoch": 1.5084096972508991,
+      "grad_norm": 23.722388037868864,
+      "learning_rate": 2e-06,
+      "loss": 0.3714,
+      "step": 6502
+    },
+    {
+      "epoch": 1.5086416888991996,
+      "grad_norm": 14.672151917575201,
+      "learning_rate": 2e-06,
+      "loss": 0.2105,
+      "step": 6503
+    },
+    {
+      "epoch": 1.5088736805475003,
+      "grad_norm": 12.031371405646743,
+      "learning_rate": 2e-06,
+      "loss": 0.2954,
+      "step": 6504
+    },
+    {
+      "epoch": 1.509105672195801,
+      "grad_norm": 3.1784302456047224,
+      "learning_rate": 2e-06,
+      "loss": 0.1228,
+      "step": 6505
+    },
+    {
+      "epoch": 1.5093376638441016,
+      "grad_norm": 14.213306200760728,
+      "learning_rate": 2e-06,
+      "loss": 0.2708,
+      "step": 6506
+    },
+    {
+      "epoch": 1.5095696554924023,
+      "grad_norm": 15.232030296744599,
+      "learning_rate": 2e-06,
+      "loss": 0.2593,
+      "step": 6507
+    },
+    {
+      "epoch": 1.509801647140703,
+      "grad_norm": 14.984208384309996,
+      "learning_rate": 2e-06,
+      "loss": 0.2175,
+      "step": 6508
+    },
+    {
+      "epoch": 1.5100336387890034,
+      "grad_norm": 15.054506568863584,
+      "learning_rate": 2e-06,
+      "loss": 0.291,
+      "step": 6509
+    },
+    {
+      "epoch": 1.510265630437304,
+      "grad_norm": 18.958083066770698,
+      "learning_rate": 2e-06,
+      "loss": 0.3165,
+      "step": 6510
+    },
+    {
+      "epoch": 1.5104976220856048,
+      "grad_norm": 15.20084589197303,
+      "learning_rate": 2e-06,
+      "loss": 0.2437,
+      "step": 6511
+    },
+    {
+      "epoch": 1.5107296137339055,
+      "grad_norm": 10.595426838724423,
+      "learning_rate": 2e-06,
+      "loss": 0.1996,
+      "step": 6512
+    },
+    {
+      "epoch": 1.5109616053822061,
+      "grad_norm": 13.323499389916211,
+      "learning_rate": 2e-06,
+      "loss": 0.2852,
+      "step": 6513
+    },
+    {
+      "epoch": 1.5111935970305068,
+      "grad_norm": 12.349992816548118,
+      "learning_rate": 2e-06,
+      "loss": 0.159,
+      "step": 6514
+    },
+    {
+      "epoch": 1.5114255886788075,
+      "grad_norm": 8.347530066491494,
+      "learning_rate": 2e-06,
+      "loss": 0.1881,
+      "step": 6515
+    },
+    {
+      "epoch": 1.5116575803271082,
+      "grad_norm": 14.539653588005612,
+      "learning_rate": 2e-06,
+      "loss": 0.233,
+      "step": 6516
+    },
+    {
+      "epoch": 1.5118895719754089,
+      "grad_norm": 14.335559702883744,
+      "learning_rate": 2e-06,
+      "loss": 0.2032,
+      "step": 6517
+    },
+    {
+      "epoch": 1.5121215636237095,
+      "grad_norm": 10.128049266081188,
+      "learning_rate": 2e-06,
+      "loss": 0.1667,
+      "step": 6518
+    },
+    {
+      "epoch": 1.5123535552720102,
+      "grad_norm": 8.385680504688892,
+      "learning_rate": 2e-06,
+      "loss": 0.2341,
+      "step": 6519
+    },
+    {
+      "epoch": 1.5125855469203109,
+      "grad_norm": 17.06497822611483,
+      "learning_rate": 2e-06,
+      "loss": 0.1913,
+      "step": 6520
+    },
+    {
+      "epoch": 1.5128175385686116,
+      "grad_norm": 16.332199808682926,
+      "learning_rate": 2e-06,
+      "loss": 0.1997,
+      "step": 6521
+    },
+    {
+      "epoch": 1.5130495302169122,
+      "grad_norm": 19.854169165602478,
+      "learning_rate": 2e-06,
+      "loss": 0.3257,
+      "step": 6522
+    },
+    {
+      "epoch": 1.513281521865213,
+      "grad_norm": 18.77711344263193,
+      "learning_rate": 2e-06,
+      "loss": 0.2624,
+      "step": 6523
+    },
+    {
+      "epoch": 1.5135135135135136,
+      "grad_norm": 13.115808526811055,
+      "learning_rate": 2e-06,
+      "loss": 0.2487,
+      "step": 6524
+    },
+    {
+      "epoch": 1.5137455051618143,
+      "grad_norm": 6.266253493130478,
+      "learning_rate": 2e-06,
+      "loss": 0.1387,
+      "step": 6525
+    },
+    {
+      "epoch": 1.513977496810115,
+      "grad_norm": 30.669033494279304,
+      "learning_rate": 2e-06,
+      "loss": 0.3013,
+      "step": 6526
+    },
+    {
+      "epoch": 1.5142094884584156,
+      "grad_norm": 17.383425126374238,
+      "learning_rate": 2e-06,
+      "loss": 0.3154,
+      "step": 6527
+    },
+    {
+      "epoch": 1.5144414801067163,
+      "grad_norm": 9.12719618391748,
+      "learning_rate": 2e-06,
+      "loss": 0.2042,
+      "step": 6528
+    },
+    {
+      "epoch": 1.5146734717550168,
+      "grad_norm": 16.01351599490197,
+      "learning_rate": 2e-06,
+      "loss": 0.2471,
+      "step": 6529
+    },
+    {
+      "epoch": 1.5149054634033174,
+      "grad_norm": 7.5545651946593155,
+      "learning_rate": 2e-06,
+      "loss": 0.1569,
+      "step": 6530
+    },
+    {
+      "epoch": 1.5151374550516181,
+      "grad_norm": 10.252329088153038,
+      "learning_rate": 2e-06,
+      "loss": 0.1647,
+      "step": 6531
+    },
+    {
+      "epoch": 1.5153694466999188,
+      "grad_norm": 12.471980786928466,
+      "learning_rate": 2e-06,
+      "loss": 0.2009,
+      "step": 6532
+    },
+    {
+      "epoch": 1.5156014383482195,
+      "grad_norm": 11.677249814799206,
+      "learning_rate": 2e-06,
+      "loss": 0.273,
+      "step": 6533
+    },
+    {
+      "epoch": 1.5158334299965202,
+      "grad_norm": 16.203498552868982,
+      "learning_rate": 2e-06,
+      "loss": 0.2866,
+      "step": 6534
+    },
+    {
+      "epoch": 1.5160654216448208,
+      "grad_norm": 16.749259296710072,
+      "learning_rate": 2e-06,
+      "loss": 0.2742,
+      "step": 6535
+    },
+    {
+      "epoch": 1.5162974132931213,
+      "grad_norm": 11.47405388582526,
+      "learning_rate": 2e-06,
+      "loss": 0.2707,
+      "step": 6536
+    },
+    {
+      "epoch": 1.516529404941422,
+      "grad_norm": 25.593684432781764,
+      "learning_rate": 2e-06,
+      "loss": 0.2175,
+      "step": 6537
+    },
+    {
+      "epoch": 1.5167613965897226,
+      "grad_norm": 20.441951801303862,
+      "learning_rate": 2e-06,
+      "loss": 0.2418,
+      "step": 6538
+    },
+    {
+      "epoch": 1.5169933882380233,
+      "grad_norm": 14.458608496737757,
+      "learning_rate": 2e-06,
+      "loss": 0.2365,
+      "step": 6539
+    },
+    {
+      "epoch": 1.517225379886324,
+      "grad_norm": 33.16233465034753,
+      "learning_rate": 2e-06,
+      "loss": 0.3357,
+      "step": 6540
+    },
+    {
+      "epoch": 1.5174573715346247,
+      "grad_norm": 14.1321734243137,
+      "learning_rate": 2e-06,
+      "loss": 0.2245,
+      "step": 6541
+    },
+    {
+      "epoch": 1.5176893631829254,
+      "grad_norm": 14.33484209040663,
+      "learning_rate": 2e-06,
+      "loss": 0.2339,
+      "step": 6542
+    },
+    {
+      "epoch": 1.517921354831226,
+      "grad_norm": 15.328236886901246,
+      "learning_rate": 2e-06,
+      "loss": 0.2048,
+      "step": 6543
+    },
+    {
+      "epoch": 1.5181533464795267,
+      "grad_norm": 10.470249220106636,
+      "learning_rate": 2e-06,
+      "loss": 0.2095,
+      "step": 6544
+    },
+    {
+      "epoch": 1.5183853381278274,
+      "grad_norm": 19.99383171597015,
+      "learning_rate": 2e-06,
+      "loss": 0.2835,
+      "step": 6545
+    },
+    {
+      "epoch": 1.518617329776128,
+      "grad_norm": 19.34033305766493,
+      "learning_rate": 2e-06,
+      "loss": 0.2156,
+      "step": 6546
+    },
+    {
+      "epoch": 1.5188493214244287,
+      "grad_norm": 20.841866636345877,
+      "learning_rate": 2e-06,
+      "loss": 0.3316,
+      "step": 6547
+    },
+    {
+      "epoch": 1.5190813130727294,
+      "grad_norm": 10.383445951544882,
+      "learning_rate": 2e-06,
+      "loss": 0.131,
+      "step": 6548
+    },
+    {
+      "epoch": 1.51931330472103,
+      "grad_norm": 14.532499443762863,
+      "learning_rate": 2e-06,
+      "loss": 0.3055,
+      "step": 6549
+    },
+    {
+      "epoch": 1.5195452963693308,
+      "grad_norm": 15.630033556591604,
+      "learning_rate": 2e-06,
+      "loss": 0.4152,
+      "step": 6550
+    },
+    {
+      "epoch": 1.5197772880176315,
+      "grad_norm": 12.030975197512266,
+      "learning_rate": 2e-06,
+      "loss": 0.2482,
+      "step": 6551
+    },
+    {
+      "epoch": 1.5200092796659321,
+      "grad_norm": 13.340481336968507,
+      "learning_rate": 2e-06,
+      "loss": 0.2646,
+      "step": 6552
+    },
+    {
+      "epoch": 1.5202412713142328,
+      "grad_norm": 14.488802750841987,
+      "learning_rate": 2e-06,
+      "loss": 0.3086,
+      "step": 6553
+    },
+    {
+      "epoch": 1.5204732629625335,
+      "grad_norm": 8.115303601704829,
+      "learning_rate": 2e-06,
+      "loss": 0.1744,
+      "step": 6554
+    },
+    {
+      "epoch": 1.5207052546108342,
+      "grad_norm": 26.135474002000866,
+      "learning_rate": 2e-06,
+      "loss": 0.4021,
+      "step": 6555
+    },
+    {
+      "epoch": 1.5209372462591346,
+      "grad_norm": 25.226920890660292,
+      "learning_rate": 2e-06,
+      "loss": 0.3901,
+      "step": 6556
+    },
+    {
+      "epoch": 1.5211692379074353,
+      "grad_norm": 16.670183683407572,
+      "learning_rate": 2e-06,
+      "loss": 0.2408,
+      "step": 6557
+    },
+    {
+      "epoch": 1.521401229555736,
+      "grad_norm": 15.625375444010585,
+      "learning_rate": 2e-06,
+      "loss": 0.2326,
+      "step": 6558
+    },
+    {
+      "epoch": 1.5216332212040367,
+      "grad_norm": 14.917535152680767,
+      "learning_rate": 2e-06,
+      "loss": 0.2647,
+      "step": 6559
+    },
+    {
+      "epoch": 1.5218652128523373,
+      "grad_norm": 13.332426191345766,
+      "learning_rate": 2e-06,
+      "loss": 0.2578,
+      "step": 6560
+    },
+    {
+      "epoch": 1.522097204500638,
+      "grad_norm": 18.671462189485453,
+      "learning_rate": 2e-06,
+      "loss": 0.3071,
+      "step": 6561
+    },
+    {
+      "epoch": 1.5223291961489387,
+      "grad_norm": 11.246730512085772,
+      "learning_rate": 2e-06,
+      "loss": 0.1709,
+      "step": 6562
+    },
+    {
+      "epoch": 1.5225611877972391,
+      "grad_norm": 19.622214769945217,
+      "learning_rate": 2e-06,
+      "loss": 0.2527,
+      "step": 6563
+    },
+    {
+      "epoch": 1.5227931794455398,
+      "grad_norm": 12.284109364278002,
+      "learning_rate": 2e-06,
+      "loss": 0.2795,
+      "step": 6564
+    },
+    {
+      "epoch": 1.5230251710938405,
+      "grad_norm": 6.678563214595849,
+      "learning_rate": 2e-06,
+      "loss": 0.1916,
+      "step": 6565
+    },
+    {
+      "epoch": 1.5232571627421412,
+      "grad_norm": 10.69967395464174,
+      "learning_rate": 2e-06,
+      "loss": 0.183,
+      "step": 6566
+    },
+    {
+      "epoch": 1.5234891543904419,
+      "grad_norm": 18.99299029345869,
+      "learning_rate": 2e-06,
+      "loss": 0.2846,
+      "step": 6567
+    },
+    {
+      "epoch": 1.5237211460387425,
+      "grad_norm": 9.86938009551779,
+      "learning_rate": 2e-06,
+      "loss": 0.2259,
+      "step": 6568
+    },
+    {
+      "epoch": 1.5239531376870432,
+      "grad_norm": 12.17829700571588,
+      "learning_rate": 2e-06,
+      "loss": 0.221,
+      "step": 6569
+    },
+    {
+      "epoch": 1.524185129335344,
+      "grad_norm": 11.699644060447417,
+      "learning_rate": 2e-06,
+      "loss": 0.2575,
+      "step": 6570
+    },
+    {
+      "epoch": 1.5244171209836446,
+      "grad_norm": 9.703258680204174,
+      "learning_rate": 2e-06,
+      "loss": 0.2,
+      "step": 6571
+    },
+    {
+      "epoch": 1.5246491126319452,
+      "grad_norm": 14.95404804361216,
+      "learning_rate": 2e-06,
+      "loss": 0.2595,
+      "step": 6572
+    },
+    {
+      "epoch": 1.524881104280246,
+      "grad_norm": 13.152934576076827,
+      "learning_rate": 2e-06,
+      "loss": 0.1685,
+      "step": 6573
+    },
+    {
+      "epoch": 1.5251130959285466,
+      "grad_norm": 15.716773941763252,
+      "learning_rate": 2e-06,
+      "loss": 0.2364,
+      "step": 6574
+    },
+    {
+      "epoch": 1.5253450875768473,
+      "grad_norm": 10.98282431021787,
+      "learning_rate": 2e-06,
+      "loss": 0.2123,
+      "step": 6575
+    },
+    {
+      "epoch": 1.525577079225148,
+      "grad_norm": 10.298974653319465,
+      "learning_rate": 2e-06,
+      "loss": 0.3139,
+      "step": 6576
+    },
+    {
+      "epoch": 1.5258090708734486,
+      "grad_norm": 15.407719180498699,
+      "learning_rate": 2e-06,
+      "loss": 0.2313,
+      "step": 6577
+    },
+    {
+      "epoch": 1.5260410625217493,
+      "grad_norm": 13.536063143083295,
+      "learning_rate": 2e-06,
+      "loss": 0.2669,
+      "step": 6578
+    },
+    {
+      "epoch": 1.52627305417005,
+      "grad_norm": 17.303253418761933,
+      "learning_rate": 2e-06,
+      "loss": 0.2706,
+      "step": 6579
+    },
+    {
+      "epoch": 1.5265050458183507,
+      "grad_norm": 8.999912205572087,
+      "learning_rate": 2e-06,
+      "loss": 0.1724,
+      "step": 6580
+    },
+    {
+      "epoch": 1.5267370374666513,
+      "grad_norm": 13.214016604293978,
+      "learning_rate": 2e-06,
+      "loss": 0.2278,
+      "step": 6581
+    },
+    {
+      "epoch": 1.526969029114952,
+      "grad_norm": 19.689311179023186,
+      "learning_rate": 2e-06,
+      "loss": 0.4611,
+      "step": 6582
+    },
+    {
+      "epoch": 1.5272010207632525,
+      "grad_norm": 17.523858510746575,
+      "learning_rate": 2e-06,
+      "loss": 0.3864,
+      "step": 6583
+    },
+    {
+      "epoch": 1.5274330124115532,
+      "grad_norm": 8.710260312691528,
+      "learning_rate": 2e-06,
+      "loss": 0.1756,
+      "step": 6584
+    },
+    {
+      "epoch": 1.5276650040598538,
+      "grad_norm": 10.848455252659218,
+      "learning_rate": 2e-06,
+      "loss": 0.224,
+      "step": 6585
+    },
+    {
+      "epoch": 1.5278969957081545,
+      "grad_norm": 16.842313323134718,
+      "learning_rate": 2e-06,
+      "loss": 0.2283,
+      "step": 6586
+    },
+    {
+      "epoch": 1.5281289873564552,
+      "grad_norm": 14.082538162407813,
+      "learning_rate": 2e-06,
+      "loss": 0.2786,
+      "step": 6587
+    },
+    {
+      "epoch": 1.5283609790047559,
+      "grad_norm": 13.690119420871433,
+      "learning_rate": 2e-06,
+      "loss": 0.213,
+      "step": 6588
+    },
+    {
+      "epoch": 1.5285929706530563,
+      "grad_norm": 8.366710908032633,
+      "learning_rate": 2e-06,
+      "loss": 0.2031,
+      "step": 6589
+    },
+    {
+      "epoch": 1.528824962301357,
+      "grad_norm": 11.625826226867636,
+      "learning_rate": 2e-06,
+      "loss": 0.2905,
+      "step": 6590
+    },
+    {
+      "epoch": 1.5290569539496577,
+      "grad_norm": 9.64366760293919,
+      "learning_rate": 2e-06,
+      "loss": 0.1721,
+      "step": 6591
+    },
+    {
+      "epoch": 1.5292889455979584,
+      "grad_norm": 11.064945785341019,
+      "learning_rate": 2e-06,
+      "loss": 0.2017,
+      "step": 6592
+    },
+    {
+      "epoch": 1.529520937246259,
+      "grad_norm": 12.392594488420682,
+      "learning_rate": 2e-06,
+      "loss": 0.3388,
+      "step": 6593
+    },
+    {
+      "epoch": 1.5297529288945597,
+      "grad_norm": 10.038529036525198,
+      "learning_rate": 2e-06,
+      "loss": 0.1767,
+      "step": 6594
+    },
+    {
+      "epoch": 1.5299849205428604,
+      "grad_norm": 18.155270257757028,
+      "learning_rate": 2e-06,
+      "loss": 0.2795,
+      "step": 6595
+    },
+    {
+      "epoch": 1.530216912191161,
+      "grad_norm": 18.631747785591013,
+      "learning_rate": 2e-06,
+      "loss": 0.4423,
+      "step": 6596
+    },
+    {
+      "epoch": 1.5304489038394617,
+      "grad_norm": 11.650253566150127,
+      "learning_rate": 2e-06,
+      "loss": 0.1601,
+      "step": 6597
+    },
+    {
+      "epoch": 1.5306808954877624,
+      "grad_norm": 11.892720717803435,
+      "learning_rate": 2e-06,
+      "loss": 0.2567,
+      "step": 6598
+    },
+    {
+      "epoch": 1.530912887136063,
+      "grad_norm": 12.76447937851006,
+      "learning_rate": 2e-06,
+      "loss": 0.2272,
+      "step": 6599
+    },
+    {
+      "epoch": 1.5311448787843638,
+      "grad_norm": 19.670009456521562,
+      "learning_rate": 2e-06,
+      "loss": 0.191,
+      "step": 6600
+    },
+    {
+      "epoch": 1.5313768704326645,
+      "grad_norm": 17.911080788040636,
+      "learning_rate": 2e-06,
+      "loss": 0.2934,
+      "step": 6601
+    },
+    {
+      "epoch": 1.5316088620809651,
+      "grad_norm": 12.02701092188653,
+      "learning_rate": 2e-06,
+      "loss": 0.2006,
+      "step": 6602
+    },
+    {
+      "epoch": 1.5318408537292658,
+      "grad_norm": 14.254489915142843,
+      "learning_rate": 2e-06,
+      "loss": 0.1326,
+      "step": 6603
+    },
+    {
+      "epoch": 1.5320728453775665,
+      "grad_norm": 16.942217221710525,
+      "learning_rate": 2e-06,
+      "loss": 0.2467,
+      "step": 6604
+    },
+    {
+      "epoch": 1.5323048370258672,
+      "grad_norm": 12.251395907968385,
+      "learning_rate": 2e-06,
+      "loss": 0.1818,
+      "step": 6605
+    },
+    {
+      "epoch": 1.5325368286741679,
+      "grad_norm": 18.873433824205463,
+      "learning_rate": 2e-06,
+      "loss": 0.304,
+      "step": 6606
+    },
+    {
+      "epoch": 1.5327688203224685,
+      "grad_norm": 19.97499742532436,
+      "learning_rate": 2e-06,
+      "loss": 0.3184,
+      "step": 6607
+    },
+    {
+      "epoch": 1.5330008119707692,
+      "grad_norm": 8.638731047589733,
+      "learning_rate": 2e-06,
+      "loss": 0.2164,
+      "step": 6608
+    },
+    {
+      "epoch": 1.5332328036190697,
+      "grad_norm": 16.66699551716635,
+      "learning_rate": 2e-06,
+      "loss": 0.2551,
+      "step": 6609
+    },
+    {
+      "epoch": 1.5334647952673703,
+      "grad_norm": 5.295961005754983,
+      "learning_rate": 2e-06,
+      "loss": 0.1288,
+      "step": 6610
+    },
+    {
+      "epoch": 1.533696786915671,
+      "grad_norm": 21.756919902176616,
+      "learning_rate": 2e-06,
+      "loss": 0.2657,
+      "step": 6611
+    },
+    {
+      "epoch": 1.5339287785639717,
+      "grad_norm": 16.63173357913042,
+      "learning_rate": 2e-06,
+      "loss": 0.3024,
+      "step": 6612
+    },
+    {
+      "epoch": 1.5341607702122724,
+      "grad_norm": 15.377279500427504,
+      "learning_rate": 2e-06,
+      "loss": 0.1583,
+      "step": 6613
+    },
+    {
+      "epoch": 1.534392761860573,
+      "grad_norm": 13.542730249995518,
+      "learning_rate": 2e-06,
+      "loss": 0.2867,
+      "step": 6614
+    },
+    {
+      "epoch": 1.5346247535088737,
+      "grad_norm": 13.45081937791314,
+      "learning_rate": 2e-06,
+      "loss": 0.3402,
+      "step": 6615
+    },
+    {
+      "epoch": 1.5348567451571742,
+      "grad_norm": 18.73627537803717,
+      "learning_rate": 2e-06,
+      "loss": 0.3025,
+      "step": 6616
+    },
+    {
+      "epoch": 1.5350887368054749,
+      "grad_norm": 17.96909572583018,
+      "learning_rate": 2e-06,
+      "loss": 0.25,
+      "step": 6617
+    },
+    {
+      "epoch": 1.5353207284537755,
+      "grad_norm": 7.296570244559349,
+      "learning_rate": 2e-06,
+      "loss": 0.2265,
+      "step": 6618
+    },
+    {
+      "epoch": 1.5355527201020762,
+      "grad_norm": 11.84219243712135,
+      "learning_rate": 2e-06,
+      "loss": 0.2714,
+      "step": 6619
+    },
+    {
+      "epoch": 1.535784711750377,
+      "grad_norm": 11.616406825047289,
+      "learning_rate": 2e-06,
+      "loss": 0.2136,
+      "step": 6620
+    },
+    {
+      "epoch": 1.5360167033986776,
+      "grad_norm": 8.454661127361028,
+      "learning_rate": 2e-06,
+      "loss": 0.2052,
+      "step": 6621
+    },
+    {
+      "epoch": 1.5362486950469783,
+      "grad_norm": 17.398773472415783,
+      "learning_rate": 2e-06,
+      "loss": 0.3175,
+      "step": 6622
+    },
+    {
+      "epoch": 1.536480686695279,
+      "grad_norm": 14.004436701132663,
+      "learning_rate": 2e-06,
+      "loss": 0.1801,
+      "step": 6623
+    },
+    {
+      "epoch": 1.5367126783435796,
+      "grad_norm": 15.782936325674424,
+      "learning_rate": 2e-06,
+      "loss": 0.1985,
+      "step": 6624
+    },
+    {
+      "epoch": 1.5369446699918803,
+      "grad_norm": 8.437744584794697,
+      "learning_rate": 2e-06,
+      "loss": 0.1646,
+      "step": 6625
+    },
+    {
+      "epoch": 1.537176661640181,
+      "grad_norm": 10.262631524625183,
+      "learning_rate": 2e-06,
+      "loss": 0.178,
+      "step": 6626
+    },
+    {
+      "epoch": 1.5374086532884816,
+      "grad_norm": 6.3501856308833755,
+      "learning_rate": 2e-06,
+      "loss": 0.228,
+      "step": 6627
+    },
+    {
+      "epoch": 1.5376406449367823,
+      "grad_norm": 10.01570986797597,
+      "learning_rate": 2e-06,
+      "loss": 0.3387,
+      "step": 6628
+    },
+    {
+      "epoch": 1.537872636585083,
+      "grad_norm": 30.458849058891083,
+      "learning_rate": 2e-06,
+      "loss": 0.4754,
+      "step": 6629
+    },
+    {
+      "epoch": 1.5381046282333837,
+      "grad_norm": 11.716062302207616,
+      "learning_rate": 2e-06,
+      "loss": 0.1712,
+      "step": 6630
+    },
+    {
+      "epoch": 1.5383366198816844,
+      "grad_norm": 18.175307041258502,
+      "learning_rate": 2e-06,
+      "loss": 0.2043,
+      "step": 6631
+    },
+    {
+      "epoch": 1.538568611529985,
+      "grad_norm": 12.24045972591269,
+      "learning_rate": 2e-06,
+      "loss": 0.179,
+      "step": 6632
+    },
+    {
+      "epoch": 1.5388006031782857,
+      "grad_norm": 11.07667745607954,
+      "learning_rate": 2e-06,
+      "loss": 0.273,
+      "step": 6633
+    },
+    {
+      "epoch": 1.5390325948265864,
+      "grad_norm": 12.272831750483638,
+      "learning_rate": 2e-06,
+      "loss": 0.179,
+      "step": 6634
+    },
+    {
+      "epoch": 1.539264586474887,
+      "grad_norm": 12.611848727210983,
+      "learning_rate": 2e-06,
+      "loss": 0.411,
+      "step": 6635
+    },
+    {
+      "epoch": 1.5394965781231875,
+      "grad_norm": 18.854779245821923,
+      "learning_rate": 2e-06,
+      "loss": 0.3932,
+      "step": 6636
+    },
+    {
+      "epoch": 1.5397285697714882,
+      "grad_norm": 7.172384964319514,
+      "learning_rate": 2e-06,
+      "loss": 0.1187,
+      "step": 6637
+    },
+    {
+      "epoch": 1.5399605614197889,
+      "grad_norm": 6.576975458207418,
+      "learning_rate": 2e-06,
+      "loss": 0.1599,
+      "step": 6638
+    },
+    {
+      "epoch": 1.5401925530680896,
+      "grad_norm": 12.24297995933426,
+      "learning_rate": 2e-06,
+      "loss": 0.2164,
+      "step": 6639
+    },
+    {
+      "epoch": 1.5404245447163902,
+      "grad_norm": 10.647747894323764,
+      "learning_rate": 2e-06,
+      "loss": 0.2583,
+      "step": 6640
+    },
+    {
+      "epoch": 1.540656536364691,
+      "grad_norm": 19.60146067514771,
+      "learning_rate": 2e-06,
+      "loss": 0.1466,
+      "step": 6641
+    },
+    {
+      "epoch": 1.5408885280129916,
+      "grad_norm": 14.684098261137065,
+      "learning_rate": 2e-06,
+      "loss": 0.2603,
+      "step": 6642
+    },
+    {
+      "epoch": 1.541120519661292,
+      "grad_norm": 16.59216010203391,
+      "learning_rate": 2e-06,
+      "loss": 0.2821,
+      "step": 6643
+    },
+    {
+      "epoch": 1.5413525113095927,
+      "grad_norm": 14.913905233436939,
+      "learning_rate": 2e-06,
+      "loss": 0.3483,
+      "step": 6644
+    },
+    {
+      "epoch": 1.5415845029578934,
+      "grad_norm": 7.166202900774562,
+      "learning_rate": 2e-06,
+      "loss": 0.1455,
+      "step": 6645
+    },
+    {
+      "epoch": 1.541816494606194,
+      "grad_norm": 21.821133413069827,
+      "learning_rate": 2e-06,
+      "loss": 0.3514,
+      "step": 6646
+    },
+    {
+      "epoch": 1.5420484862544948,
+      "grad_norm": 12.236283152701525,
+      "learning_rate": 2e-06,
+      "loss": 0.2124,
+      "step": 6647
+    },
+    {
+      "epoch": 1.5422804779027954,
+      "grad_norm": 17.098002615388406,
+      "learning_rate": 2e-06,
+      "loss": 0.2716,
+      "step": 6648
+    },
+    {
+      "epoch": 1.542512469551096,
+      "grad_norm": 14.267657162351332,
+      "learning_rate": 2e-06,
+      "loss": 0.2191,
+      "step": 6649
+    },
+    {
+      "epoch": 1.5427444611993968,
+      "grad_norm": 15.523793251848586,
+      "learning_rate": 2e-06,
+      "loss": 0.1445,
+      "step": 6650
+    },
+    {
+      "epoch": 1.5429764528476975,
+      "grad_norm": 14.559855721210122,
+      "learning_rate": 2e-06,
+      "loss": 0.2379,
+      "step": 6651
+    },
+    {
+      "epoch": 1.5432084444959981,
+      "grad_norm": 23.65012451053669,
+      "learning_rate": 2e-06,
+      "loss": 0.4326,
+      "step": 6652
+    },
+    {
+      "epoch": 1.5434404361442988,
+      "grad_norm": 8.303100619952016,
+      "learning_rate": 2e-06,
+      "loss": 0.198,
+      "step": 6653
+    },
+    {
+      "epoch": 1.5436724277925995,
+      "grad_norm": 15.935548264212265,
+      "learning_rate": 2e-06,
+      "loss": 0.1968,
+      "step": 6654
+    },
+    {
+      "epoch": 1.5439044194409002,
+      "grad_norm": 8.883382058562992,
+      "learning_rate": 2e-06,
+      "loss": 0.181,
+      "step": 6655
+    },
+    {
+      "epoch": 1.5441364110892009,
+      "grad_norm": 13.870904459442297,
+      "learning_rate": 2e-06,
+      "loss": 0.2708,
+      "step": 6656
+    },
+    {
+      "epoch": 1.5443684027375015,
+      "grad_norm": 18.77365934368007,
+      "learning_rate": 2e-06,
+      "loss": 0.3606,
+      "step": 6657
+    },
+    {
+      "epoch": 1.5446003943858022,
+      "grad_norm": 11.231032727883631,
+      "learning_rate": 2e-06,
+      "loss": 0.1979,
+      "step": 6658
+    },
+    {
+      "epoch": 1.544832386034103,
+      "grad_norm": 13.809682217206333,
+      "learning_rate": 2e-06,
+      "loss": 0.212,
+      "step": 6659
+    },
+    {
+      "epoch": 1.5450643776824036,
+      "grad_norm": 9.238173167632462,
+      "learning_rate": 2e-06,
+      "loss": 0.2164,
+      "step": 6660
+    },
+    {
+      "epoch": 1.5452963693307042,
+      "grad_norm": 13.771198076915777,
+      "learning_rate": 2e-06,
+      "loss": 0.196,
+      "step": 6661
+    },
+    {
+      "epoch": 1.5455283609790047,
+      "grad_norm": 15.557560680677668,
+      "learning_rate": 2e-06,
+      "loss": 0.1748,
+      "step": 6662
+    },
+    {
+      "epoch": 1.5457603526273054,
+      "grad_norm": 19.399216680140757,
+      "learning_rate": 2e-06,
+      "loss": 0.3062,
+      "step": 6663
+    },
+    {
+      "epoch": 1.545992344275606,
+      "grad_norm": 12.735134100589814,
+      "learning_rate": 2e-06,
+      "loss": 0.1553,
+      "step": 6664
+    },
+    {
+      "epoch": 1.5462243359239067,
+      "grad_norm": 11.721525953735028,
+      "learning_rate": 2e-06,
+      "loss": 0.1754,
+      "step": 6665
+    },
+    {
+      "epoch": 1.5464563275722074,
+      "grad_norm": 25.85267662972695,
+      "learning_rate": 2e-06,
+      "loss": 0.3786,
+      "step": 6666
+    },
+    {
+      "epoch": 1.546688319220508,
+      "grad_norm": 9.82288963464247,
+      "learning_rate": 2e-06,
+      "loss": 0.22,
+      "step": 6667
+    },
+    {
+      "epoch": 1.5469203108688088,
+      "grad_norm": 14.562619633467646,
+      "learning_rate": 2e-06,
+      "loss": 0.1804,
+      "step": 6668
+    },
+    {
+      "epoch": 1.5471523025171092,
+      "grad_norm": 12.559383643951522,
+      "learning_rate": 2e-06,
+      "loss": 0.2596,
+      "step": 6669
+    },
+    {
+      "epoch": 1.54738429416541,
+      "grad_norm": 23.46206957851582,
+      "learning_rate": 2e-06,
+      "loss": 0.2093,
+      "step": 6670
+    },
+    {
+      "epoch": 1.5476162858137106,
+      "grad_norm": 15.949501067315126,
+      "learning_rate": 2e-06,
+      "loss": 0.2572,
+      "step": 6671
+    },
+    {
+      "epoch": 1.5478482774620113,
+      "grad_norm": 24.868620049788493,
+      "learning_rate": 2e-06,
+      "loss": 0.4118,
+      "step": 6672
+    },
+    {
+      "epoch": 1.548080269110312,
+      "grad_norm": 15.287923367273436,
+      "learning_rate": 2e-06,
+      "loss": 0.2223,
+      "step": 6673
+    },
+    {
+      "epoch": 1.5483122607586126,
+      "grad_norm": 6.52400379250765,
+      "learning_rate": 2e-06,
+      "loss": 0.1518,
+      "step": 6674
+    },
+    {
+      "epoch": 1.5485442524069133,
+      "grad_norm": 14.473581077854373,
+      "learning_rate": 2e-06,
+      "loss": 0.2041,
+      "step": 6675
+    },
+    {
+      "epoch": 1.548776244055214,
+      "grad_norm": 13.152523903148891,
+      "learning_rate": 2e-06,
+      "loss": 0.2297,
+      "step": 6676
+    },
+    {
+      "epoch": 1.5490082357035146,
+      "grad_norm": 15.555084560097228,
+      "learning_rate": 2e-06,
+      "loss": 0.2117,
+      "step": 6677
+    },
+    {
+      "epoch": 1.5492402273518153,
+      "grad_norm": 18.95724262071258,
+      "learning_rate": 2e-06,
+      "loss": 0.2694,
+      "step": 6678
+    },
+    {
+      "epoch": 1.549472219000116,
+      "grad_norm": 23.805287017623378,
+      "learning_rate": 2e-06,
+      "loss": 0.2882,
+      "step": 6679
+    },
+    {
+      "epoch": 1.5497042106484167,
+      "grad_norm": 9.063455098713165,
+      "learning_rate": 2e-06,
+      "loss": 0.1562,
+      "step": 6680
+    },
+    {
+      "epoch": 1.5499362022967174,
+      "grad_norm": 16.767249600071942,
+      "learning_rate": 2e-06,
+      "loss": 0.2859,
+      "step": 6681
+    },
+    {
+      "epoch": 1.550168193945018,
+      "grad_norm": 11.262662463905274,
+      "learning_rate": 2e-06,
+      "loss": 0.2708,
+      "step": 6682
+    },
+    {
+      "epoch": 1.5504001855933187,
+      "grad_norm": 6.92155707769241,
+      "learning_rate": 2e-06,
+      "loss": 0.1165,
+      "step": 6683
+    },
+    {
+      "epoch": 1.5506321772416194,
+      "grad_norm": 13.00332862318548,
+      "learning_rate": 2e-06,
+      "loss": 0.225,
+      "step": 6684
+    },
+    {
+      "epoch": 1.55086416888992,
+      "grad_norm": 16.8149521444566,
+      "learning_rate": 2e-06,
+      "loss": 0.2741,
+      "step": 6685
+    },
+    {
+      "epoch": 1.5510961605382207,
+      "grad_norm": 19.949779449084364,
+      "learning_rate": 2e-06,
+      "loss": 0.4181,
+      "step": 6686
+    },
+    {
+      "epoch": 1.5513281521865214,
+      "grad_norm": 15.470405223572625,
+      "learning_rate": 2e-06,
+      "loss": 0.2668,
+      "step": 6687
+    },
+    {
+      "epoch": 1.551560143834822,
+      "grad_norm": 9.524958830303015,
+      "learning_rate": 2e-06,
+      "loss": 0.1999,
+      "step": 6688
+    },
+    {
+      "epoch": 1.5517921354831226,
+      "grad_norm": 10.401920612753036,
+      "learning_rate": 2e-06,
+      "loss": 0.1673,
+      "step": 6689
+    },
+    {
+      "epoch": 1.5520241271314232,
+      "grad_norm": 14.408570492378052,
+      "learning_rate": 2e-06,
+      "loss": 0.2244,
+      "step": 6690
+    },
+    {
+      "epoch": 1.552256118779724,
+      "grad_norm": 21.054179752792802,
+      "learning_rate": 2e-06,
+      "loss": 0.3153,
+      "step": 6691
+    },
+    {
+      "epoch": 1.5524881104280246,
+      "grad_norm": 12.234752323220984,
+      "learning_rate": 2e-06,
+      "loss": 0.233,
+      "step": 6692
+    },
+    {
+      "epoch": 1.5527201020763253,
+      "grad_norm": 16.21777025907242,
+      "learning_rate": 2e-06,
+      "loss": 0.2629,
+      "step": 6693
+    },
+    {
+      "epoch": 1.552952093724626,
+      "grad_norm": 12.365330556133785,
+      "learning_rate": 2e-06,
+      "loss": 0.225,
+      "step": 6694
+    },
+    {
+      "epoch": 1.5531840853729266,
+      "grad_norm": 13.604646105301338,
+      "learning_rate": 2e-06,
+      "loss": 0.1856,
+      "step": 6695
+    },
+    {
+      "epoch": 1.553416077021227,
+      "grad_norm": 11.565050637758615,
+      "learning_rate": 2e-06,
+      "loss": 0.1909,
+      "step": 6696
+    },
+    {
+      "epoch": 1.5536480686695278,
+      "grad_norm": 10.095837283548882,
+      "learning_rate": 2e-06,
+      "loss": 0.144,
+      "step": 6697
+    },
+    {
+      "epoch": 1.5538800603178284,
+      "grad_norm": 26.34064370188464,
+      "learning_rate": 2e-06,
+      "loss": 0.2701,
+      "step": 6698
+    },
+    {
+      "epoch": 1.5541120519661291,
+      "grad_norm": 15.714451308894228,
+      "learning_rate": 2e-06,
+      "loss": 0.2907,
+      "step": 6699
+    },
+    {
+      "epoch": 1.5543440436144298,
+      "grad_norm": 23.67048393742744,
+      "learning_rate": 2e-06,
+      "loss": 0.28,
+      "step": 6700
+    },
+    {
+      "epoch": 1.5545760352627305,
+      "grad_norm": 19.769761573650925,
+      "learning_rate": 2e-06,
+      "loss": 0.4424,
+      "step": 6701
+    },
+    {
+      "epoch": 1.5548080269110311,
+      "grad_norm": 13.912627564672938,
+      "learning_rate": 2e-06,
+      "loss": 0.2099,
+      "step": 6702
+    },
+    {
+      "epoch": 1.5550400185593318,
+      "grad_norm": 17.84318080318395,
+      "learning_rate": 2e-06,
+      "loss": 0.2191,
+      "step": 6703
+    },
+    {
+      "epoch": 1.5552720102076325,
+      "grad_norm": 17.882379196787138,
+      "learning_rate": 2e-06,
+      "loss": 0.2854,
+      "step": 6704
+    },
+    {
+      "epoch": 1.5555040018559332,
+      "grad_norm": 19.754686577344806,
+      "learning_rate": 2e-06,
+      "loss": 0.2216,
+      "step": 6705
+    },
+    {
+      "epoch": 1.5557359935042339,
+      "grad_norm": 12.30229564320053,
+      "learning_rate": 2e-06,
+      "loss": 0.2358,
+      "step": 6706
+    },
+    {
+      "epoch": 1.5559679851525345,
+      "grad_norm": 11.072988829052486,
+      "learning_rate": 2e-06,
+      "loss": 0.3163,
+      "step": 6707
+    },
+    {
+      "epoch": 1.5561999768008352,
+      "grad_norm": 11.920322064117377,
+      "learning_rate": 2e-06,
+      "loss": 0.184,
+      "step": 6708
+    },
+    {
+      "epoch": 1.556431968449136,
+      "grad_norm": 23.620264848489672,
+      "learning_rate": 2e-06,
+      "loss": 0.5399,
+      "step": 6709
+    },
+    {
+      "epoch": 1.5566639600974366,
+      "grad_norm": 18.340827062399427,
+      "learning_rate": 2e-06,
+      "loss": 0.2605,
+      "step": 6710
+    },
+    {
+      "epoch": 1.5568959517457372,
+      "grad_norm": 15.804678291711879,
+      "learning_rate": 2e-06,
+      "loss": 0.2082,
+      "step": 6711
+    },
+    {
+      "epoch": 1.557127943394038,
+      "grad_norm": 9.287038285372942,
+      "learning_rate": 2e-06,
+      "loss": 0.1843,
+      "step": 6712
+    },
+    {
+      "epoch": 1.5573599350423386,
+      "grad_norm": 13.862382590699733,
+      "learning_rate": 2e-06,
+      "loss": 0.2673,
+      "step": 6713
+    },
+    {
+      "epoch": 1.5575919266906393,
+      "grad_norm": 22.047374843399634,
+      "learning_rate": 2e-06,
+      "loss": 0.2695,
+      "step": 6714
+    },
+    {
+      "epoch": 1.55782391833894,
+      "grad_norm": 8.856948070107391,
+      "learning_rate": 2e-06,
+      "loss": 0.1535,
+      "step": 6715
+    },
+    {
+      "epoch": 1.5580559099872404,
+      "grad_norm": 12.628650705847866,
+      "learning_rate": 2e-06,
+      "loss": 0.223,
+      "step": 6716
+    },
+    {
+      "epoch": 1.558287901635541,
+      "grad_norm": 23.188665968133286,
+      "learning_rate": 2e-06,
+      "loss": 0.2675,
+      "step": 6717
+    },
+    {
+      "epoch": 1.5585198932838418,
+      "grad_norm": 10.686962550001526,
+      "learning_rate": 2e-06,
+      "loss": 0.3214,
+      "step": 6718
+    },
+    {
+      "epoch": 1.5587518849321425,
+      "grad_norm": 17.718995290129744,
+      "learning_rate": 2e-06,
+      "loss": 0.2649,
+      "step": 6719
+    },
+    {
+      "epoch": 1.5589838765804431,
+      "grad_norm": 13.700206729941808,
+      "learning_rate": 2e-06,
+      "loss": 0.2116,
+      "step": 6720
+    },
+    {
+      "epoch": 1.5592158682287438,
+      "grad_norm": 13.688686894063217,
+      "learning_rate": 2e-06,
+      "loss": 0.2587,
+      "step": 6721
+    },
+    {
+      "epoch": 1.5594478598770443,
+      "grad_norm": 12.49552285498753,
+      "learning_rate": 2e-06,
+      "loss": 0.2303,
+      "step": 6722
+    },
+    {
+      "epoch": 1.559679851525345,
+      "grad_norm": 13.62322311967421,
+      "learning_rate": 2e-06,
+      "loss": 0.1915,
+      "step": 6723
+    },
+    {
+      "epoch": 1.5599118431736456,
+      "grad_norm": 7.434462107481903,
+      "learning_rate": 2e-06,
+      "loss": 0.1737,
+      "step": 6724
+    },
+    {
+      "epoch": 1.5601438348219463,
+      "grad_norm": 10.92478845203364,
+      "learning_rate": 2e-06,
+      "loss": 0.2139,
+      "step": 6725
+    },
+    {
+      "epoch": 1.560375826470247,
+      "grad_norm": 12.50872192630918,
+      "learning_rate": 2e-06,
+      "loss": 0.2493,
+      "step": 6726
+    },
+    {
+      "epoch": 1.5606078181185477,
+      "grad_norm": 14.669812825175278,
+      "learning_rate": 2e-06,
+      "loss": 0.2171,
+      "step": 6727
+    },
+    {
+      "epoch": 1.5608398097668483,
+      "grad_norm": 16.867602158489653,
+      "learning_rate": 2e-06,
+      "loss": 0.1857,
+      "step": 6728
+    },
+    {
+      "epoch": 1.561071801415149,
+      "grad_norm": 24.13353406032791,
+      "learning_rate": 2e-06,
+      "loss": 0.2582,
+      "step": 6729
+    },
+    {
+      "epoch": 1.5613037930634497,
+      "grad_norm": 11.22810564138021,
+      "learning_rate": 2e-06,
+      "loss": 0.1928,
+      "step": 6730
+    },
+    {
+      "epoch": 1.5615357847117504,
+      "grad_norm": 5.337300728209669,
+      "learning_rate": 2e-06,
+      "loss": 0.1356,
+      "step": 6731
+    },
+    {
+      "epoch": 1.561767776360051,
+      "grad_norm": 14.689968164162265,
+      "learning_rate": 2e-06,
+      "loss": 0.3578,
+      "step": 6732
+    },
+    {
+      "epoch": 1.5619997680083517,
+      "grad_norm": 11.167047902657862,
+      "learning_rate": 2e-06,
+      "loss": 0.1839,
+      "step": 6733
+    },
+    {
+      "epoch": 1.5622317596566524,
+      "grad_norm": 8.165761931642459,
+      "learning_rate": 2e-06,
+      "loss": 0.1836,
+      "step": 6734
+    },
+    {
+      "epoch": 1.562463751304953,
+      "grad_norm": 10.54991480371774,
+      "learning_rate": 2e-06,
+      "loss": 0.1719,
+      "step": 6735
+    },
+    {
+      "epoch": 1.5626957429532538,
+      "grad_norm": 11.442688510228646,
+      "learning_rate": 2e-06,
+      "loss": 0.2653,
+      "step": 6736
+    },
+    {
+      "epoch": 1.5629277346015544,
+      "grad_norm": 17.070029585225544,
+      "learning_rate": 2e-06,
+      "loss": 0.2521,
+      "step": 6737
+    },
+    {
+      "epoch": 1.563159726249855,
+      "grad_norm": 19.686280230299616,
+      "learning_rate": 2e-06,
+      "loss": 0.244,
+      "step": 6738
+    },
+    {
+      "epoch": 1.5633917178981558,
+      "grad_norm": 9.948167478079785,
+      "learning_rate": 2e-06,
+      "loss": 0.2483,
+      "step": 6739
+    },
+    {
+      "epoch": 1.5636237095464565,
+      "grad_norm": 24.922215627756376,
+      "learning_rate": 2e-06,
+      "loss": 0.2743,
+      "step": 6740
+    },
+    {
+      "epoch": 1.5638557011947571,
+      "grad_norm": 16.741152692172665,
+      "learning_rate": 2e-06,
+      "loss": 0.2465,
+      "step": 6741
+    },
+    {
+      "epoch": 1.5640876928430576,
+      "grad_norm": 10.47916098313398,
+      "learning_rate": 2e-06,
+      "loss": 0.2147,
+      "step": 6742
+    },
+    {
+      "epoch": 1.5643196844913583,
+      "grad_norm": 9.617663992594,
+      "learning_rate": 2e-06,
+      "loss": 0.1909,
+      "step": 6743
+    },
+    {
+      "epoch": 1.564551676139659,
+      "grad_norm": 11.671052205411605,
+      "learning_rate": 2e-06,
+      "loss": 0.2597,
+      "step": 6744
+    },
+    {
+      "epoch": 1.5647836677879596,
+      "grad_norm": 12.058931258856934,
+      "learning_rate": 2e-06,
+      "loss": 0.2216,
+      "step": 6745
+    },
+    {
+      "epoch": 1.5650156594362603,
+      "grad_norm": 21.409795579059104,
+      "learning_rate": 2e-06,
+      "loss": 0.3596,
+      "step": 6746
+    },
+    {
+      "epoch": 1.565247651084561,
+      "grad_norm": 11.764103802660122,
+      "learning_rate": 2e-06,
+      "loss": 0.1823,
+      "step": 6747
+    },
+    {
+      "epoch": 1.5654796427328617,
+      "grad_norm": 9.706993728012272,
+      "learning_rate": 2e-06,
+      "loss": 0.169,
+      "step": 6748
+    },
+    {
+      "epoch": 1.5657116343811621,
+      "grad_norm": 15.599357074918032,
+      "learning_rate": 2e-06,
+      "loss": 0.342,
+      "step": 6749
+    },
+    {
+      "epoch": 1.5659436260294628,
+      "grad_norm": 20.41002666037572,
+      "learning_rate": 2e-06,
+      "loss": 0.2836,
+      "step": 6750
+    },
+    {
+      "epoch": 1.5661756176777635,
+      "grad_norm": 15.460943544867607,
+      "learning_rate": 2e-06,
+      "loss": 0.3194,
+      "step": 6751
+    },
+    {
+      "epoch": 1.5664076093260642,
+      "grad_norm": 18.23070983470633,
+      "learning_rate": 2e-06,
+      "loss": 0.2515,
+      "step": 6752
+    },
+    {
+      "epoch": 1.5666396009743648,
+      "grad_norm": 24.89491573709899,
+      "learning_rate": 2e-06,
+      "loss": 0.2149,
+      "step": 6753
+    },
+    {
+      "epoch": 1.5668715926226655,
+      "grad_norm": 10.42731457408122,
+      "learning_rate": 2e-06,
+      "loss": 0.1346,
+      "step": 6754
+    },
+    {
+      "epoch": 1.5671035842709662,
+      "grad_norm": 26.145121796832107,
+      "learning_rate": 2e-06,
+      "loss": 0.3824,
+      "step": 6755
+    },
+    {
+      "epoch": 1.5673355759192669,
+      "grad_norm": 19.222209564955683,
+      "learning_rate": 2e-06,
+      "loss": 0.2588,
+      "step": 6756
+    },
+    {
+      "epoch": 1.5675675675675675,
+      "grad_norm": 10.470769858587463,
+      "learning_rate": 2e-06,
+      "loss": 0.2401,
+      "step": 6757
+    },
+    {
+      "epoch": 1.5677995592158682,
+      "grad_norm": 11.749844689977595,
+      "learning_rate": 2e-06,
+      "loss": 0.1865,
+      "step": 6758
+    },
+    {
+      "epoch": 1.568031550864169,
+      "grad_norm": 24.061754616113973,
+      "learning_rate": 2e-06,
+      "loss": 0.3668,
+      "step": 6759
+    },
+    {
+      "epoch": 1.5682635425124696,
+      "grad_norm": 14.476040215882161,
+      "learning_rate": 2e-06,
+      "loss": 0.2639,
+      "step": 6760
+    },
+    {
+      "epoch": 1.5684955341607703,
+      "grad_norm": 19.156412552986147,
+      "learning_rate": 2e-06,
+      "loss": 0.2453,
+      "step": 6761
+    },
+    {
+      "epoch": 1.568727525809071,
+      "grad_norm": 9.392538890429808,
+      "learning_rate": 2e-06,
+      "loss": 0.207,
+      "step": 6762
+    },
+    {
+      "epoch": 1.5689595174573716,
+      "grad_norm": 6.850823669102928,
+      "learning_rate": 2e-06,
+      "loss": 0.145,
+      "step": 6763
+    },
+    {
+      "epoch": 1.5691915091056723,
+      "grad_norm": 13.466344294898505,
+      "learning_rate": 2e-06,
+      "loss": 0.2537,
+      "step": 6764
+    },
+    {
+      "epoch": 1.569423500753973,
+      "grad_norm": 15.669892855674025,
+      "learning_rate": 2e-06,
+      "loss": 0.2211,
+      "step": 6765
+    },
+    {
+      "epoch": 1.5696554924022736,
+      "grad_norm": 16.398892981894086,
+      "learning_rate": 2e-06,
+      "loss": 0.2775,
+      "step": 6766
+    },
+    {
+      "epoch": 1.5698874840505743,
+      "grad_norm": 15.11369594798255,
+      "learning_rate": 2e-06,
+      "loss": 0.2086,
+      "step": 6767
+    },
+    {
+      "epoch": 1.570119475698875,
+      "grad_norm": 16.09216950836462,
+      "learning_rate": 2e-06,
+      "loss": 0.2734,
+      "step": 6768
+    },
+    {
+      "epoch": 1.5703514673471755,
+      "grad_norm": 11.456504129507978,
+      "learning_rate": 2e-06,
+      "loss": 0.2696,
+      "step": 6769
+    },
+    {
+      "epoch": 1.5705834589954761,
+      "grad_norm": 31.295646221623393,
+      "learning_rate": 2e-06,
+      "loss": 0.2382,
+      "step": 6770
+    },
+    {
+      "epoch": 1.5708154506437768,
+      "grad_norm": 14.91478972462344,
+      "learning_rate": 2e-06,
+      "loss": 0.2604,
+      "step": 6771
+    },
+    {
+      "epoch": 1.5710474422920775,
+      "grad_norm": 14.084946879586926,
+      "learning_rate": 2e-06,
+      "loss": 0.2746,
+      "step": 6772
+    },
+    {
+      "epoch": 1.5712794339403782,
+      "grad_norm": 7.254192390967452,
+      "learning_rate": 2e-06,
+      "loss": 0.1817,
+      "step": 6773
+    },
+    {
+      "epoch": 1.5715114255886788,
+      "grad_norm": 21.57021793936699,
+      "learning_rate": 2e-06,
+      "loss": 0.287,
+      "step": 6774
+    },
+    {
+      "epoch": 1.5717434172369795,
+      "grad_norm": 21.625281378712074,
+      "learning_rate": 2e-06,
+      "loss": 0.2452,
+      "step": 6775
+    },
+    {
+      "epoch": 1.57197540888528,
+      "grad_norm": 14.467275186899649,
+      "learning_rate": 2e-06,
+      "loss": 0.2274,
+      "step": 6776
+    },
+    {
+      "epoch": 1.5722074005335807,
+      "grad_norm": 8.852228895085334,
+      "learning_rate": 2e-06,
+      "loss": 0.1996,
+      "step": 6777
+    },
+    {
+      "epoch": 1.5724393921818813,
+      "grad_norm": 8.286774206475272,
+      "learning_rate": 2e-06,
+      "loss": 0.187,
+      "step": 6778
+    },
+    {
+      "epoch": 1.572671383830182,
+      "grad_norm": 14.316350630272812,
+      "learning_rate": 2e-06,
+      "loss": 0.2814,
+      "step": 6779
+    },
+    {
+      "epoch": 1.5729033754784827,
+      "grad_norm": 12.459751760967073,
+      "learning_rate": 2e-06,
+      "loss": 0.2404,
+      "step": 6780
+    },
+    {
+      "epoch": 1.5731353671267834,
+      "grad_norm": 7.701203511096362,
+      "learning_rate": 2e-06,
+      "loss": 0.1813,
+      "step": 6781
+    },
+    {
+      "epoch": 1.573367358775084,
+      "grad_norm": 15.65163803258445,
+      "learning_rate": 2e-06,
+      "loss": 0.3066,
+      "step": 6782
+    },
+    {
+      "epoch": 1.5735993504233847,
+      "grad_norm": 11.415272637022873,
+      "learning_rate": 2e-06,
+      "loss": 0.1425,
+      "step": 6783
+    },
+    {
+      "epoch": 1.5738313420716854,
+      "grad_norm": 19.358539152020647,
+      "learning_rate": 2e-06,
+      "loss": 0.3298,
+      "step": 6784
+    },
+    {
+      "epoch": 1.574063333719986,
+      "grad_norm": 12.680294495610841,
+      "learning_rate": 2e-06,
+      "loss": 0.2396,
+      "step": 6785
+    },
+    {
+      "epoch": 1.5742953253682868,
+      "grad_norm": 19.26119166215534,
+      "learning_rate": 2e-06,
+      "loss": 0.2845,
+      "step": 6786
+    },
+    {
+      "epoch": 1.5745273170165874,
+      "grad_norm": 43.35671069330195,
+      "learning_rate": 2e-06,
+      "loss": 0.3221,
+      "step": 6787
+    },
+    {
+      "epoch": 1.5747593086648881,
+      "grad_norm": 25.079014100953824,
+      "learning_rate": 2e-06,
+      "loss": 0.3231,
+      "step": 6788
+    },
+    {
+      "epoch": 1.5749913003131888,
+      "grad_norm": 25.518585287352856,
+      "learning_rate": 2e-06,
+      "loss": 0.335,
+      "step": 6789
+    },
+    {
+      "epoch": 1.5752232919614895,
+      "grad_norm": 8.677770134465073,
+      "learning_rate": 2e-06,
+      "loss": 0.2009,
+      "step": 6790
+    },
+    {
+      "epoch": 1.5754552836097901,
+      "grad_norm": 16.41414521651827,
+      "learning_rate": 2e-06,
+      "loss": 0.341,
+      "step": 6791
+    },
+    {
+      "epoch": 1.5756872752580908,
+      "grad_norm": 14.16484596317382,
+      "learning_rate": 2e-06,
+      "loss": 0.1712,
+      "step": 6792
+    },
+    {
+      "epoch": 1.5759192669063915,
+      "grad_norm": 31.22819956446191,
+      "learning_rate": 2e-06,
+      "loss": 0.3924,
+      "step": 6793
+    },
+    {
+      "epoch": 1.5761512585546922,
+      "grad_norm": 10.63781472409317,
+      "learning_rate": 2e-06,
+      "loss": 0.1474,
+      "step": 6794
+    },
+    {
+      "epoch": 1.5763832502029929,
+      "grad_norm": 15.8712959128847,
+      "learning_rate": 2e-06,
+      "loss": 0.2747,
+      "step": 6795
+    },
+    {
+      "epoch": 1.5766152418512933,
+      "grad_norm": 16.81316543278334,
+      "learning_rate": 2e-06,
+      "loss": 0.2064,
+      "step": 6796
+    },
+    {
+      "epoch": 1.576847233499594,
+      "grad_norm": 10.956852812678623,
+      "learning_rate": 2e-06,
+      "loss": 0.2405,
+      "step": 6797
+    },
+    {
+      "epoch": 1.5770792251478947,
+      "grad_norm": 10.47422890616429,
+      "learning_rate": 2e-06,
+      "loss": 0.2584,
+      "step": 6798
+    },
+    {
+      "epoch": 1.5773112167961953,
+      "grad_norm": 11.247698042368608,
+      "learning_rate": 2e-06,
+      "loss": 0.2368,
+      "step": 6799
+    },
+    {
+      "epoch": 1.577543208444496,
+      "grad_norm": 11.458253221758767,
+      "learning_rate": 2e-06,
+      "loss": 0.2327,
+      "step": 6800
+    },
+    {
+      "epoch": 1.5777752000927967,
+      "grad_norm": 8.372281579949735,
+      "learning_rate": 2e-06,
+      "loss": 0.2589,
+      "step": 6801
+    },
+    {
+      "epoch": 1.5780071917410972,
+      "grad_norm": 23.002950702220993,
+      "learning_rate": 2e-06,
+      "loss": 0.2453,
+      "step": 6802
+    },
+    {
+      "epoch": 1.5782391833893978,
+      "grad_norm": 8.10206962011327,
+      "learning_rate": 2e-06,
+      "loss": 0.1706,
+      "step": 6803
+    },
+    {
+      "epoch": 1.5784711750376985,
+      "grad_norm": 5.549188344327498,
+      "learning_rate": 2e-06,
+      "loss": 0.1771,
+      "step": 6804
+    },
+    {
+      "epoch": 1.5787031666859992,
+      "grad_norm": 14.578356313178427,
+      "learning_rate": 2e-06,
+      "loss": 0.2663,
+      "step": 6805
+    },
+    {
+      "epoch": 1.5789351583342999,
+      "grad_norm": 10.42207019577819,
+      "learning_rate": 2e-06,
+      "loss": 0.2595,
+      "step": 6806
+    },
+    {
+      "epoch": 1.5791671499826005,
+      "grad_norm": 6.864112094691692,
+      "learning_rate": 2e-06,
+      "loss": 0.1364,
+      "step": 6807
+    },
+    {
+      "epoch": 1.5793991416309012,
+      "grad_norm": 4.728977241944988,
+      "learning_rate": 2e-06,
+      "loss": 0.1352,
+      "step": 6808
+    },
+    {
+      "epoch": 1.579631133279202,
+      "grad_norm": 13.672761713026164,
+      "learning_rate": 2e-06,
+      "loss": 0.2204,
+      "step": 6809
+    },
+    {
+      "epoch": 1.5798631249275026,
+      "grad_norm": 17.40182794118163,
+      "learning_rate": 2e-06,
+      "loss": 0.3978,
+      "step": 6810
+    },
+    {
+      "epoch": 1.5800951165758033,
+      "grad_norm": 15.13729096710601,
+      "learning_rate": 2e-06,
+      "loss": 0.1737,
+      "step": 6811
+    },
+    {
+      "epoch": 1.580327108224104,
+      "grad_norm": 21.127192170719233,
+      "learning_rate": 2e-06,
+      "loss": 0.5297,
+      "step": 6812
+    },
+    {
+      "epoch": 1.5805590998724046,
+      "grad_norm": 16.860482121408943,
+      "learning_rate": 2e-06,
+      "loss": 0.1868,
+      "step": 6813
+    },
+    {
+      "epoch": 1.5807910915207053,
+      "grad_norm": 10.21009934750211,
+      "learning_rate": 2e-06,
+      "loss": 0.2236,
+      "step": 6814
+    },
+    {
+      "epoch": 1.581023083169006,
+      "grad_norm": 15.657316076566811,
+      "learning_rate": 2e-06,
+      "loss": 0.2979,
+      "step": 6815
+    },
+    {
+      "epoch": 1.5812550748173066,
+      "grad_norm": 15.242342987174924,
+      "learning_rate": 2e-06,
+      "loss": 0.3382,
+      "step": 6816
+    },
+    {
+      "epoch": 1.5814870664656073,
+      "grad_norm": 8.972983513729012,
+      "learning_rate": 2e-06,
+      "loss": 0.1847,
+      "step": 6817
+    },
+    {
+      "epoch": 1.581719058113908,
+      "grad_norm": 8.861874703597897,
+      "learning_rate": 2e-06,
+      "loss": 0.2629,
+      "step": 6818
+    },
+    {
+      "epoch": 1.5819510497622087,
+      "grad_norm": 11.282644462145537,
+      "learning_rate": 2e-06,
+      "loss": 0.2448,
+      "step": 6819
+    },
+    {
+      "epoch": 1.5821830414105094,
+      "grad_norm": 16.643703301291083,
+      "learning_rate": 2e-06,
+      "loss": 0.218,
+      "step": 6820
+    },
+    {
+      "epoch": 1.58241503305881,
+      "grad_norm": 21.338207571831255,
+      "learning_rate": 2e-06,
+      "loss": 0.3162,
+      "step": 6821
+    },
+    {
+      "epoch": 1.5826470247071105,
+      "grad_norm": 10.456117076657767,
+      "learning_rate": 2e-06,
+      "loss": 0.1931,
+      "step": 6822
+    },
+    {
+      "epoch": 1.5828790163554112,
+      "grad_norm": 20.253880071565263,
+      "learning_rate": 2e-06,
+      "loss": 0.3495,
+      "step": 6823
+    },
+    {
+      "epoch": 1.5831110080037119,
+      "grad_norm": 17.609409195262224,
+      "learning_rate": 2e-06,
+      "loss": 0.3101,
+      "step": 6824
+    },
+    {
+      "epoch": 1.5833429996520125,
+      "grad_norm": 14.320650363686498,
+      "learning_rate": 2e-06,
+      "loss": 0.2719,
+      "step": 6825
+    },
+    {
+      "epoch": 1.5835749913003132,
+      "grad_norm": 18.748152428299427,
+      "learning_rate": 2e-06,
+      "loss": 0.2974,
+      "step": 6826
+    },
+    {
+      "epoch": 1.5838069829486139,
+      "grad_norm": 9.52741203825633,
+      "learning_rate": 2e-06,
+      "loss": 0.1331,
+      "step": 6827
+    },
+    {
+      "epoch": 1.5840389745969146,
+      "grad_norm": 18.27852816483889,
+      "learning_rate": 2e-06,
+      "loss": 0.3419,
+      "step": 6828
+    },
+    {
+      "epoch": 1.584270966245215,
+      "grad_norm": 9.530308694977986,
+      "learning_rate": 2e-06,
+      "loss": 0.1755,
+      "step": 6829
+    },
+    {
+      "epoch": 1.5845029578935157,
+      "grad_norm": 16.07479575909688,
+      "learning_rate": 2e-06,
+      "loss": 0.2276,
+      "step": 6830
+    },
+    {
+      "epoch": 1.5847349495418164,
+      "grad_norm": 5.195872495498322,
+      "learning_rate": 2e-06,
+      "loss": 0.1703,
+      "step": 6831
+    },
+    {
+      "epoch": 1.584966941190117,
+      "grad_norm": 9.3011970800669,
+      "learning_rate": 2e-06,
+      "loss": 0.2328,
+      "step": 6832
+    },
+    {
+      "epoch": 1.5851989328384177,
+      "grad_norm": 11.992368937834383,
+      "learning_rate": 2e-06,
+      "loss": 0.2196,
+      "step": 6833
+    },
+    {
+      "epoch": 1.5854309244867184,
+      "grad_norm": 8.603011392573826,
+      "learning_rate": 2e-06,
+      "loss": 0.2124,
+      "step": 6834
+    },
+    {
+      "epoch": 1.585662916135019,
+      "grad_norm": 12.27925572845166,
+      "learning_rate": 2e-06,
+      "loss": 0.243,
+      "step": 6835
+    },
+    {
+      "epoch": 1.5858949077833198,
+      "grad_norm": 17.79861330852824,
+      "learning_rate": 2e-06,
+      "loss": 0.2526,
+      "step": 6836
+    },
+    {
+      "epoch": 1.5861268994316204,
+      "grad_norm": 8.642526433758599,
+      "learning_rate": 2e-06,
+      "loss": 0.1874,
+      "step": 6837
+    },
+    {
+      "epoch": 1.5863588910799211,
+      "grad_norm": 7.819458952164534,
+      "learning_rate": 2e-06,
+      "loss": 0.172,
+      "step": 6838
+    },
+    {
+      "epoch": 1.5865908827282218,
+      "grad_norm": 14.04549995988459,
+      "learning_rate": 2e-06,
+      "loss": 0.2023,
+      "step": 6839
+    },
+    {
+      "epoch": 1.5868228743765225,
+      "grad_norm": 9.58691437742038,
+      "learning_rate": 2e-06,
+      "loss": 0.1658,
+      "step": 6840
+    },
+    {
+      "epoch": 1.5870548660248232,
+      "grad_norm": 7.415634246482783,
+      "learning_rate": 2e-06,
+      "loss": 0.1984,
+      "step": 6841
+    },
+    {
+      "epoch": 1.5872868576731238,
+      "grad_norm": 7.101726327086242,
+      "learning_rate": 2e-06,
+      "loss": 0.1825,
+      "step": 6842
+    },
+    {
+      "epoch": 1.5875188493214245,
+      "grad_norm": 7.409833020702544,
+      "learning_rate": 2e-06,
+      "loss": 0.173,
+      "step": 6843
+    },
+    {
+      "epoch": 1.5877508409697252,
+      "grad_norm": 5.643831335022831,
+      "learning_rate": 2e-06,
+      "loss": 0.1592,
+      "step": 6844
+    },
+    {
+      "epoch": 1.5879828326180259,
+      "grad_norm": 12.266503536628118,
+      "learning_rate": 2e-06,
+      "loss": 0.266,
+      "step": 6845
+    },
+    {
+      "epoch": 1.5882148242663265,
+      "grad_norm": 17.78772424839602,
+      "learning_rate": 2e-06,
+      "loss": 0.3874,
+      "step": 6846
+    },
+    {
+      "epoch": 1.5884468159146272,
+      "grad_norm": 15.623174768233348,
+      "learning_rate": 2e-06,
+      "loss": 0.2541,
+      "step": 6847
+    },
+    {
+      "epoch": 1.588678807562928,
+      "grad_norm": 14.267785946840403,
+      "learning_rate": 2e-06,
+      "loss": 0.3315,
+      "step": 6848
+    },
+    {
+      "epoch": 1.5889107992112284,
+      "grad_norm": 25.090609601792437,
+      "learning_rate": 2e-06,
+      "loss": 0.1686,
+      "step": 6849
+    },
+    {
+      "epoch": 1.589142790859529,
+      "grad_norm": 10.182624698844203,
+      "learning_rate": 2e-06,
+      "loss": 0.1973,
+      "step": 6850
+    },
+    {
+      "epoch": 1.5893747825078297,
+      "grad_norm": 16.37330122299087,
+      "learning_rate": 2e-06,
+      "loss": 0.2241,
+      "step": 6851
+    },
+    {
+      "epoch": 1.5896067741561304,
+      "grad_norm": 13.459030559524912,
+      "learning_rate": 2e-06,
+      "loss": 0.2656,
+      "step": 6852
+    },
+    {
+      "epoch": 1.589838765804431,
+      "grad_norm": 18.64639791565411,
+      "learning_rate": 2e-06,
+      "loss": 0.2362,
+      "step": 6853
+    },
+    {
+      "epoch": 1.5900707574527317,
+      "grad_norm": 8.167161131211438,
+      "learning_rate": 2e-06,
+      "loss": 0.2205,
+      "step": 6854
+    },
+    {
+      "epoch": 1.5903027491010322,
+      "grad_norm": 12.041969699634368,
+      "learning_rate": 2e-06,
+      "loss": 0.2306,
+      "step": 6855
+    },
+    {
+      "epoch": 1.5905347407493329,
+      "grad_norm": 14.858497088147926,
+      "learning_rate": 2e-06,
+      "loss": 0.2603,
+      "step": 6856
+    },
+    {
+      "epoch": 1.5907667323976336,
+      "grad_norm": 12.226650589555664,
+      "learning_rate": 2e-06,
+      "loss": 0.2172,
+      "step": 6857
+    },
+    {
+      "epoch": 1.5909987240459342,
+      "grad_norm": 9.08808966600926,
+      "learning_rate": 2e-06,
+      "loss": 0.1842,
+      "step": 6858
+    },
+    {
+      "epoch": 1.591230715694235,
+      "grad_norm": 13.717394027158436,
+      "learning_rate": 2e-06,
+      "loss": 0.2377,
+      "step": 6859
+    },
+    {
+      "epoch": 1.5914627073425356,
+      "grad_norm": 13.078249717134886,
+      "learning_rate": 2e-06,
+      "loss": 0.2417,
+      "step": 6860
+    },
+    {
+      "epoch": 1.5916946989908363,
+      "grad_norm": 28.056304082891295,
+      "learning_rate": 2e-06,
+      "loss": 0.2097,
+      "step": 6861
+    },
+    {
+      "epoch": 1.591926690639137,
+      "grad_norm": 14.645870930917718,
+      "learning_rate": 2e-06,
+      "loss": 0.1665,
+      "step": 6862
+    },
+    {
+      "epoch": 1.5921586822874376,
+      "grad_norm": 15.310789998713801,
+      "learning_rate": 2e-06,
+      "loss": 0.3082,
+      "step": 6863
+    },
+    {
+      "epoch": 1.5923906739357383,
+      "grad_norm": 20.451507985504687,
+      "learning_rate": 2e-06,
+      "loss": 0.2574,
+      "step": 6864
+    },
+    {
+      "epoch": 1.592622665584039,
+      "grad_norm": 21.807670305901297,
+      "learning_rate": 2e-06,
+      "loss": 0.3035,
+      "step": 6865
+    },
+    {
+      "epoch": 1.5928546572323397,
+      "grad_norm": 19.895841208070944,
+      "learning_rate": 2e-06,
+      "loss": 0.2902,
+      "step": 6866
+    },
+    {
+      "epoch": 1.5930866488806403,
+      "grad_norm": 8.457046599193541,
+      "learning_rate": 2e-06,
+      "loss": 0.2041,
+      "step": 6867
+    },
+    {
+      "epoch": 1.593318640528941,
+      "grad_norm": 22.47335248001497,
+      "learning_rate": 2e-06,
+      "loss": 0.4859,
+      "step": 6868
+    },
+    {
+      "epoch": 1.5935506321772417,
+      "grad_norm": 6.685828572757369,
+      "learning_rate": 2e-06,
+      "loss": 0.1665,
+      "step": 6869
+    },
+    {
+      "epoch": 1.5937826238255424,
+      "grad_norm": 12.58042735255098,
+      "learning_rate": 2e-06,
+      "loss": 0.1601,
+      "step": 6870
+    },
+    {
+      "epoch": 1.594014615473843,
+      "grad_norm": 22.317484002899555,
+      "learning_rate": 2e-06,
+      "loss": 0.242,
+      "step": 6871
+    },
+    {
+      "epoch": 1.5942466071221437,
+      "grad_norm": 18.39607935176384,
+      "learning_rate": 2e-06,
+      "loss": 0.3981,
+      "step": 6872
+    },
+    {
+      "epoch": 1.5944785987704444,
+      "grad_norm": 10.360346744792217,
+      "learning_rate": 2e-06,
+      "loss": 0.1472,
+      "step": 6873
+    },
+    {
+      "epoch": 1.594710590418745,
+      "grad_norm": 11.671580147838984,
+      "learning_rate": 2e-06,
+      "loss": 0.2312,
+      "step": 6874
+    },
+    {
+      "epoch": 1.5949425820670455,
+      "grad_norm": 8.029919131678305,
+      "learning_rate": 2e-06,
+      "loss": 0.234,
+      "step": 6875
+    },
+    {
+      "epoch": 1.5951745737153462,
+      "grad_norm": 14.454042603280314,
+      "learning_rate": 2e-06,
+      "loss": 0.2537,
+      "step": 6876
+    },
+    {
+      "epoch": 1.595406565363647,
+      "grad_norm": 14.486299136318895,
+      "learning_rate": 2e-06,
+      "loss": 0.2426,
+      "step": 6877
+    },
+    {
+      "epoch": 1.5956385570119476,
+      "grad_norm": 14.517852619599177,
+      "learning_rate": 2e-06,
+      "loss": 0.2604,
+      "step": 6878
+    },
+    {
+      "epoch": 1.5958705486602482,
+      "grad_norm": 13.24440098634251,
+      "learning_rate": 2e-06,
+      "loss": 0.2915,
+      "step": 6879
+    },
+    {
+      "epoch": 1.596102540308549,
+      "grad_norm": 12.914902414937012,
+      "learning_rate": 2e-06,
+      "loss": 0.3206,
+      "step": 6880
+    },
+    {
+      "epoch": 1.5963345319568496,
+      "grad_norm": 12.885798355302601,
+      "learning_rate": 2e-06,
+      "loss": 0.1833,
+      "step": 6881
+    },
+    {
+      "epoch": 1.59656652360515,
+      "grad_norm": 15.829499481850467,
+      "learning_rate": 2e-06,
+      "loss": 0.2901,
+      "step": 6882
+    },
+    {
+      "epoch": 1.5967985152534507,
+      "grad_norm": 11.329045094579913,
+      "learning_rate": 2e-06,
+      "loss": 0.241,
+      "step": 6883
+    },
+    {
+      "epoch": 1.5970305069017514,
+      "grad_norm": 14.853956669259542,
+      "learning_rate": 2e-06,
+      "loss": 0.2227,
+      "step": 6884
+    },
+    {
+      "epoch": 1.597262498550052,
+      "grad_norm": 9.383651932545247,
+      "learning_rate": 2e-06,
+      "loss": 0.1166,
+      "step": 6885
+    },
+    {
+      "epoch": 1.5974944901983528,
+      "grad_norm": 29.26111858454847,
+      "learning_rate": 2e-06,
+      "loss": 0.3973,
+      "step": 6886
+    },
+    {
+      "epoch": 1.5977264818466534,
+      "grad_norm": 6.7484712455436116,
+      "learning_rate": 2e-06,
+      "loss": 0.1273,
+      "step": 6887
+    },
+    {
+      "epoch": 1.5979584734949541,
+      "grad_norm": 10.15763033087749,
+      "learning_rate": 2e-06,
+      "loss": 0.271,
+      "step": 6888
+    },
+    {
+      "epoch": 1.5981904651432548,
+      "grad_norm": 17.35175331054969,
+      "learning_rate": 2e-06,
+      "loss": 0.4555,
+      "step": 6889
+    },
+    {
+      "epoch": 1.5984224567915555,
+      "grad_norm": 12.735975895446408,
+      "learning_rate": 2e-06,
+      "loss": 0.258,
+      "step": 6890
+    },
+    {
+      "epoch": 1.5986544484398562,
+      "grad_norm": 12.899543658492155,
+      "learning_rate": 2e-06,
+      "loss": 0.1864,
+      "step": 6891
+    },
+    {
+      "epoch": 1.5988864400881568,
+      "grad_norm": 12.698385568969032,
+      "learning_rate": 2e-06,
+      "loss": 0.1718,
+      "step": 6892
+    },
+    {
+      "epoch": 1.5991184317364575,
+      "grad_norm": 4.28535250927259,
+      "learning_rate": 2e-06,
+      "loss": 0.1266,
+      "step": 6893
+    },
+    {
+      "epoch": 1.5993504233847582,
+      "grad_norm": 15.392604216922926,
+      "learning_rate": 2e-06,
+      "loss": 0.2754,
+      "step": 6894
+    },
+    {
+      "epoch": 1.5995824150330589,
+      "grad_norm": 15.137874358989361,
+      "learning_rate": 2e-06,
+      "loss": 0.2634,
+      "step": 6895
+    },
+    {
+      "epoch": 1.5998144066813595,
+      "grad_norm": 15.254360087513062,
+      "learning_rate": 2e-06,
+      "loss": 0.3004,
+      "step": 6896
+    },
+    {
+      "epoch": 1.6000463983296602,
+      "grad_norm": 18.28465820969029,
+      "learning_rate": 2e-06,
+      "loss": 0.4048,
+      "step": 6897
+    },
+    {
+      "epoch": 1.600278389977961,
+      "grad_norm": 7.541889046152906,
+      "learning_rate": 2e-06,
+      "loss": 0.1286,
+      "step": 6898
+    },
+    {
+      "epoch": 1.6005103816262616,
+      "grad_norm": 14.525111372393546,
+      "learning_rate": 2e-06,
+      "loss": 0.1813,
+      "step": 6899
+    },
+    {
+      "epoch": 1.6007423732745623,
+      "grad_norm": 10.764190597901676,
+      "learning_rate": 2e-06,
+      "loss": 0.2426,
+      "step": 6900
+    },
+    {
+      "epoch": 1.600974364922863,
+      "grad_norm": 16.376203207837985,
+      "learning_rate": 2e-06,
+      "loss": 0.3211,
+      "step": 6901
+    },
+    {
+      "epoch": 1.6012063565711634,
+      "grad_norm": 8.72620445463533,
+      "learning_rate": 2e-06,
+      "loss": 0.1751,
+      "step": 6902
+    },
+    {
+      "epoch": 1.601438348219464,
+      "grad_norm": 15.942772043804196,
+      "learning_rate": 2e-06,
+      "loss": 0.2968,
+      "step": 6903
+    },
+    {
+      "epoch": 1.6016703398677647,
+      "grad_norm": 18.402605174063517,
+      "learning_rate": 2e-06,
+      "loss": 0.2,
+      "step": 6904
+    },
+    {
+      "epoch": 1.6019023315160654,
+      "grad_norm": 12.426769789870608,
+      "learning_rate": 2e-06,
+      "loss": 0.2583,
+      "step": 6905
+    },
+    {
+      "epoch": 1.602134323164366,
+      "grad_norm": 10.17124275196209,
+      "learning_rate": 2e-06,
+      "loss": 0.1767,
+      "step": 6906
+    },
+    {
+      "epoch": 1.6023663148126668,
+      "grad_norm": 20.138982762387695,
+      "learning_rate": 2e-06,
+      "loss": 0.301,
+      "step": 6907
+    },
+    {
+      "epoch": 1.6025983064609675,
+      "grad_norm": 14.742653023011412,
+      "learning_rate": 2e-06,
+      "loss": 0.2199,
+      "step": 6908
+    },
+    {
+      "epoch": 1.602830298109268,
+      "grad_norm": 25.703435637577172,
+      "learning_rate": 2e-06,
+      "loss": 0.3056,
+      "step": 6909
+    },
+    {
+      "epoch": 1.6030622897575686,
+      "grad_norm": 12.570448997393886,
+      "learning_rate": 2e-06,
+      "loss": 0.2229,
+      "step": 6910
+    },
+    {
+      "epoch": 1.6032942814058693,
+      "grad_norm": 7.55682126082453,
+      "learning_rate": 2e-06,
+      "loss": 0.1689,
+      "step": 6911
+    },
+    {
+      "epoch": 1.60352627305417,
+      "grad_norm": 13.806272086112495,
+      "learning_rate": 2e-06,
+      "loss": 0.2732,
+      "step": 6912
+    },
+    {
+      "epoch": 1.6037582647024706,
+      "grad_norm": 22.446697414205737,
+      "learning_rate": 2e-06,
+      "loss": 0.4215,
+      "step": 6913
+    },
+    {
+      "epoch": 1.6039902563507713,
+      "grad_norm": 17.32438262165776,
+      "learning_rate": 2e-06,
+      "loss": 0.2283,
+      "step": 6914
+    },
+    {
+      "epoch": 1.604222247999072,
+      "grad_norm": 23.79328231241564,
+      "learning_rate": 2e-06,
+      "loss": 0.3247,
+      "step": 6915
+    },
+    {
+      "epoch": 1.6044542396473727,
+      "grad_norm": 11.871426031777244,
+      "learning_rate": 2e-06,
+      "loss": 0.2369,
+      "step": 6916
+    },
+    {
+      "epoch": 1.6046862312956733,
+      "grad_norm": 12.60168201994963,
+      "learning_rate": 2e-06,
+      "loss": 0.2332,
+      "step": 6917
+    },
+    {
+      "epoch": 1.604918222943974,
+      "grad_norm": 14.391829705544197,
+      "learning_rate": 2e-06,
+      "loss": 0.2458,
+      "step": 6918
+    },
+    {
+      "epoch": 1.6051502145922747,
+      "grad_norm": 22.365046280164332,
+      "learning_rate": 2e-06,
+      "loss": 0.3598,
+      "step": 6919
+    },
+    {
+      "epoch": 1.6053822062405754,
+      "grad_norm": 9.21861265143294,
+      "learning_rate": 2e-06,
+      "loss": 0.1547,
+      "step": 6920
+    },
+    {
+      "epoch": 1.605614197888876,
+      "grad_norm": 17.501152537868485,
+      "learning_rate": 2e-06,
+      "loss": 0.241,
+      "step": 6921
+    },
+    {
+      "epoch": 1.6058461895371767,
+      "grad_norm": 13.363051800078775,
+      "learning_rate": 2e-06,
+      "loss": 0.2323,
+      "step": 6922
+    },
+    {
+      "epoch": 1.6060781811854774,
+      "grad_norm": 10.522078319379862,
+      "learning_rate": 2e-06,
+      "loss": 0.2682,
+      "step": 6923
+    },
+    {
+      "epoch": 1.606310172833778,
+      "grad_norm": 10.73490836853733,
+      "learning_rate": 2e-06,
+      "loss": 0.1414,
+      "step": 6924
+    },
+    {
+      "epoch": 1.6065421644820788,
+      "grad_norm": 8.652664717579718,
+      "learning_rate": 2e-06,
+      "loss": 0.2046,
+      "step": 6925
+    },
+    {
+      "epoch": 1.6067741561303794,
+      "grad_norm": 10.409928706645774,
+      "learning_rate": 2e-06,
+      "loss": 0.1589,
+      "step": 6926
+    },
+    {
+      "epoch": 1.6070061477786801,
+      "grad_norm": 27.671244480261826,
+      "learning_rate": 2e-06,
+      "loss": 0.3348,
+      "step": 6927
+    },
+    {
+      "epoch": 1.6072381394269808,
+      "grad_norm": 17.329101528036233,
+      "learning_rate": 2e-06,
+      "loss": 0.3062,
+      "step": 6928
+    },
+    {
+      "epoch": 1.6074701310752812,
+      "grad_norm": 14.260022102662832,
+      "learning_rate": 2e-06,
+      "loss": 0.2847,
+      "step": 6929
+    },
+    {
+      "epoch": 1.607702122723582,
+      "grad_norm": 34.96318462390881,
+      "learning_rate": 2e-06,
+      "loss": 0.4542,
+      "step": 6930
+    },
+    {
+      "epoch": 1.6079341143718826,
+      "grad_norm": 10.267362931793937,
+      "learning_rate": 2e-06,
+      "loss": 0.2643,
+      "step": 6931
+    },
+    {
+      "epoch": 1.6081661060201833,
+      "grad_norm": 17.83854286219526,
+      "learning_rate": 2e-06,
+      "loss": 0.2844,
+      "step": 6932
+    },
+    {
+      "epoch": 1.608398097668484,
+      "grad_norm": 15.649285578386333,
+      "learning_rate": 2e-06,
+      "loss": 0.2164,
+      "step": 6933
+    },
+    {
+      "epoch": 1.6086300893167846,
+      "grad_norm": 11.59450320536855,
+      "learning_rate": 2e-06,
+      "loss": 0.2121,
+      "step": 6934
+    },
+    {
+      "epoch": 1.608862080965085,
+      "grad_norm": 14.804743000417668,
+      "learning_rate": 2e-06,
+      "loss": 0.2726,
+      "step": 6935
+    },
+    {
+      "epoch": 1.6090940726133858,
+      "grad_norm": 27.013454990211674,
+      "learning_rate": 2e-06,
+      "loss": 0.2249,
+      "step": 6936
+    },
+    {
+      "epoch": 1.6093260642616865,
+      "grad_norm": 13.469542011658062,
+      "learning_rate": 2e-06,
+      "loss": 0.2222,
+      "step": 6937
+    },
+    {
+      "epoch": 1.6095580559099871,
+      "grad_norm": 6.0198227896653975,
+      "learning_rate": 2e-06,
+      "loss": 0.1241,
+      "step": 6938
+    },
+    {
+      "epoch": 1.6097900475582878,
+      "grad_norm": 15.411189902368585,
+      "learning_rate": 2e-06,
+      "loss": 0.2981,
+      "step": 6939
+    },
+    {
+      "epoch": 1.6100220392065885,
+      "grad_norm": 11.78771720357112,
+      "learning_rate": 2e-06,
+      "loss": 0.2158,
+      "step": 6940
+    },
+    {
+      "epoch": 1.6102540308548892,
+      "grad_norm": 22.74730256036679,
+      "learning_rate": 2e-06,
+      "loss": 0.3225,
+      "step": 6941
+    },
+    {
+      "epoch": 1.6104860225031898,
+      "grad_norm": 10.730879983272139,
+      "learning_rate": 2e-06,
+      "loss": 0.248,
+      "step": 6942
+    },
+    {
+      "epoch": 1.6107180141514905,
+      "grad_norm": 11.91366434944726,
+      "learning_rate": 2e-06,
+      "loss": 0.1315,
+      "step": 6943
+    },
+    {
+      "epoch": 1.6109500057997912,
+      "grad_norm": 13.894807052318972,
+      "learning_rate": 2e-06,
+      "loss": 0.2077,
+      "step": 6944
+    },
+    {
+      "epoch": 1.6111819974480919,
+      "grad_norm": 25.650238814998406,
+      "learning_rate": 2e-06,
+      "loss": 0.3684,
+      "step": 6945
+    },
+    {
+      "epoch": 1.6114139890963926,
+      "grad_norm": 20.577245785155775,
+      "learning_rate": 2e-06,
+      "loss": 0.3512,
+      "step": 6946
+    },
+    {
+      "epoch": 1.6116459807446932,
+      "grad_norm": 15.803228741453447,
+      "learning_rate": 2e-06,
+      "loss": 0.2321,
+      "step": 6947
+    },
+    {
+      "epoch": 1.611877972392994,
+      "grad_norm": 9.749549037933088,
+      "learning_rate": 2e-06,
+      "loss": 0.2645,
+      "step": 6948
+    },
+    {
+      "epoch": 1.6121099640412946,
+      "grad_norm": 19.299638527391338,
+      "learning_rate": 2e-06,
+      "loss": 0.2504,
+      "step": 6949
+    },
+    {
+      "epoch": 1.6123419556895953,
+      "grad_norm": 14.484103820826514,
+      "learning_rate": 2e-06,
+      "loss": 0.2341,
+      "step": 6950
+    },
+    {
+      "epoch": 1.612573947337896,
+      "grad_norm": 18.78291403294643,
+      "learning_rate": 2e-06,
+      "loss": 0.2847,
+      "step": 6951
+    },
+    {
+      "epoch": 1.6128059389861966,
+      "grad_norm": 11.665599286408527,
+      "learning_rate": 2e-06,
+      "loss": 0.1852,
+      "step": 6952
+    },
+    {
+      "epoch": 1.6130379306344973,
+      "grad_norm": 9.265971566495484,
+      "learning_rate": 2e-06,
+      "loss": 0.2277,
+      "step": 6953
+    },
+    {
+      "epoch": 1.613269922282798,
+      "grad_norm": 8.44029894218763,
+      "learning_rate": 2e-06,
+      "loss": 0.1743,
+      "step": 6954
+    },
+    {
+      "epoch": 1.6135019139310984,
+      "grad_norm": 13.715280174043034,
+      "learning_rate": 2e-06,
+      "loss": 0.2562,
+      "step": 6955
+    },
+    {
+      "epoch": 1.613733905579399,
+      "grad_norm": 9.560914543568476,
+      "learning_rate": 2e-06,
+      "loss": 0.1785,
+      "step": 6956
+    },
+    {
+      "epoch": 1.6139658972276998,
+      "grad_norm": 4.536975004892732,
+      "learning_rate": 2e-06,
+      "loss": 0.1289,
+      "step": 6957
+    },
+    {
+      "epoch": 1.6141978888760005,
+      "grad_norm": 9.108618213148045,
+      "learning_rate": 2e-06,
+      "loss": 0.2254,
+      "step": 6958
+    },
+    {
+      "epoch": 1.6144298805243011,
+      "grad_norm": 15.068677329144581,
+      "learning_rate": 2e-06,
+      "loss": 0.344,
+      "step": 6959
+    },
+    {
+      "epoch": 1.6146618721726018,
+      "grad_norm": 10.657346641654241,
+      "learning_rate": 2e-06,
+      "loss": 0.2706,
+      "step": 6960
+    },
+    {
+      "epoch": 1.6148938638209025,
+      "grad_norm": 10.836273054781074,
+      "learning_rate": 2e-06,
+      "loss": 0.2135,
+      "step": 6961
+    },
+    {
+      "epoch": 1.615125855469203,
+      "grad_norm": 15.3784691856773,
+      "learning_rate": 2e-06,
+      "loss": 0.2182,
+      "step": 6962
+    },
+    {
+      "epoch": 1.6153578471175036,
+      "grad_norm": 12.18357080868277,
+      "learning_rate": 2e-06,
+      "loss": 0.1148,
+      "step": 6963
+    },
+    {
+      "epoch": 1.6155898387658043,
+      "grad_norm": 16.523784611503427,
+      "learning_rate": 2e-06,
+      "loss": 0.2526,
+      "step": 6964
+    },
+    {
+      "epoch": 1.615821830414105,
+      "grad_norm": 13.457293895959346,
+      "learning_rate": 2e-06,
+      "loss": 0.2961,
+      "step": 6965
+    },
+    {
+      "epoch": 1.6160538220624057,
+      "grad_norm": 9.172409383148903,
+      "learning_rate": 2e-06,
+      "loss": 0.2673,
+      "step": 6966
+    },
+    {
+      "epoch": 1.6162858137107063,
+      "grad_norm": 10.592448387883387,
+      "learning_rate": 2e-06,
+      "loss": 0.2896,
+      "step": 6967
+    },
+    {
+      "epoch": 1.616517805359007,
+      "grad_norm": 8.043362460788911,
+      "learning_rate": 2e-06,
+      "loss": 0.2057,
+      "step": 6968
+    },
+    {
+      "epoch": 1.6167497970073077,
+      "grad_norm": 13.452763397776307,
+      "learning_rate": 2e-06,
+      "loss": 0.2028,
+      "step": 6969
+    },
+    {
+      "epoch": 1.6169817886556084,
+      "grad_norm": 11.889829935690594,
+      "learning_rate": 2e-06,
+      "loss": 0.3731,
+      "step": 6970
+    },
+    {
+      "epoch": 1.617213780303909,
+      "grad_norm": 15.55403517134958,
+      "learning_rate": 2e-06,
+      "loss": 0.2903,
+      "step": 6971
+    },
+    {
+      "epoch": 1.6174457719522097,
+      "grad_norm": 10.142859304319238,
+      "learning_rate": 2e-06,
+      "loss": 0.3078,
+      "step": 6972
+    },
+    {
+      "epoch": 1.6176777636005104,
+      "grad_norm": 22.015360469237347,
+      "learning_rate": 2e-06,
+      "loss": 0.3501,
+      "step": 6973
+    },
+    {
+      "epoch": 1.617909755248811,
+      "grad_norm": 16.099404708706622,
+      "learning_rate": 2e-06,
+      "loss": 0.2518,
+      "step": 6974
+    },
+    {
+      "epoch": 1.6181417468971118,
+      "grad_norm": 16.155557913431416,
+      "learning_rate": 2e-06,
+      "loss": 0.3623,
+      "step": 6975
+    },
+    {
+      "epoch": 1.6183737385454124,
+      "grad_norm": 12.094711432250437,
+      "learning_rate": 2e-06,
+      "loss": 0.228,
+      "step": 6976
+    },
+    {
+      "epoch": 1.6186057301937131,
+      "grad_norm": 9.624041117699276,
+      "learning_rate": 2e-06,
+      "loss": 0.2488,
+      "step": 6977
+    },
+    {
+      "epoch": 1.6188377218420138,
+      "grad_norm": 11.654084378361608,
+      "learning_rate": 2e-06,
+      "loss": 0.1419,
+      "step": 6978
+    },
+    {
+      "epoch": 1.6190697134903145,
+      "grad_norm": 10.785825490097595,
+      "learning_rate": 2e-06,
+      "loss": 0.2813,
+      "step": 6979
+    },
+    {
+      "epoch": 1.6193017051386152,
+      "grad_norm": 15.670027219809535,
+      "learning_rate": 2e-06,
+      "loss": 0.2212,
+      "step": 6980
+    },
+    {
+      "epoch": 1.6195336967869158,
+      "grad_norm": 10.37649965458519,
+      "learning_rate": 2e-06,
+      "loss": 0.3193,
+      "step": 6981
+    },
+    {
+      "epoch": 1.6197656884352163,
+      "grad_norm": 12.447543343424678,
+      "learning_rate": 2e-06,
+      "loss": 0.351,
+      "step": 6982
+    },
+    {
+      "epoch": 1.619997680083517,
+      "grad_norm": 20.356869900294978,
+      "learning_rate": 2e-06,
+      "loss": 0.3829,
+      "step": 6983
+    },
+    {
+      "epoch": 1.6202296717318176,
+      "grad_norm": 6.864750951398252,
+      "learning_rate": 2e-06,
+      "loss": 0.1783,
+      "step": 6984
+    },
+    {
+      "epoch": 1.6204616633801183,
+      "grad_norm": 12.418089662331026,
+      "learning_rate": 2e-06,
+      "loss": 0.3723,
+      "step": 6985
+    },
+    {
+      "epoch": 1.620693655028419,
+      "grad_norm": 12.540617596943656,
+      "learning_rate": 2e-06,
+      "loss": 0.3026,
+      "step": 6986
+    },
+    {
+      "epoch": 1.6209256466767197,
+      "grad_norm": 19.082797466653933,
+      "learning_rate": 2e-06,
+      "loss": 0.4161,
+      "step": 6987
+    },
+    {
+      "epoch": 1.6211576383250201,
+      "grad_norm": 7.163325569917312,
+      "learning_rate": 2e-06,
+      "loss": 0.1496,
+      "step": 6988
+    },
+    {
+      "epoch": 1.6213896299733208,
+      "grad_norm": 7.111859392755355,
+      "learning_rate": 2e-06,
+      "loss": 0.1597,
+      "step": 6989
+    },
+    {
+      "epoch": 1.6216216216216215,
+      "grad_norm": 19.57450811713846,
+      "learning_rate": 2e-06,
+      "loss": 0.2889,
+      "step": 6990
+    },
+    {
+      "epoch": 1.6218536132699222,
+      "grad_norm": 15.036470054451497,
+      "learning_rate": 2e-06,
+      "loss": 0.2912,
+      "step": 6991
+    },
+    {
+      "epoch": 1.6220856049182228,
+      "grad_norm": 15.753907717110106,
+      "learning_rate": 2e-06,
+      "loss": 0.1838,
+      "step": 6992
+    },
+    {
+      "epoch": 1.6223175965665235,
+      "grad_norm": 13.527695512482605,
+      "learning_rate": 2e-06,
+      "loss": 0.2222,
+      "step": 6993
+    },
+    {
+      "epoch": 1.6225495882148242,
+      "grad_norm": 14.951485530049725,
+      "learning_rate": 2e-06,
+      "loss": 0.4262,
+      "step": 6994
+    },
+    {
+      "epoch": 1.6227815798631249,
+      "grad_norm": 10.88914474498736,
+      "learning_rate": 2e-06,
+      "loss": 0.2445,
+      "step": 6995
+    },
+    {
+      "epoch": 1.6230135715114256,
+      "grad_norm": 7.7800678868379265,
+      "learning_rate": 2e-06,
+      "loss": 0.1926,
+      "step": 6996
+    },
+    {
+      "epoch": 1.6232455631597262,
+      "grad_norm": 8.47913440262279,
+      "learning_rate": 2e-06,
+      "loss": 0.1662,
+      "step": 6997
+    },
+    {
+      "epoch": 1.623477554808027,
+      "grad_norm": 17.86686445889657,
+      "learning_rate": 2e-06,
+      "loss": 0.3313,
+      "step": 6998
+    },
+    {
+      "epoch": 1.6237095464563276,
+      "grad_norm": 7.766073380740134,
+      "learning_rate": 2e-06,
+      "loss": 0.1742,
+      "step": 6999
+    },
+    {
+      "epoch": 1.6239415381046283,
+      "grad_norm": 15.802998754377214,
+      "learning_rate": 2e-06,
+      "loss": 0.3292,
+      "step": 7000
+    },
+    {
+      "epoch": 1.624173529752929,
+      "grad_norm": 12.543459151492442,
+      "learning_rate": 2e-06,
+      "loss": 0.254,
+      "step": 7001
+    },
+    {
+      "epoch": 1.6244055214012296,
+      "grad_norm": 13.798966596523735,
+      "learning_rate": 2e-06,
+      "loss": 0.2806,
+      "step": 7002
+    },
+    {
+      "epoch": 1.6246375130495303,
+      "grad_norm": 14.669498102811584,
+      "learning_rate": 2e-06,
+      "loss": 0.1382,
+      "step": 7003
+    },
+    {
+      "epoch": 1.624869504697831,
+      "grad_norm": 18.572808451157982,
+      "learning_rate": 2e-06,
+      "loss": 0.2693,
+      "step": 7004
+    },
+    {
+      "epoch": 1.6251014963461317,
+      "grad_norm": 16.05801422771798,
+      "learning_rate": 2e-06,
+      "loss": 0.2689,
+      "step": 7005
+    },
+    {
+      "epoch": 1.6253334879944323,
+      "grad_norm": 6.454360588304225,
+      "learning_rate": 2e-06,
+      "loss": 0.1544,
+      "step": 7006
+    },
+    {
+      "epoch": 1.625565479642733,
+      "grad_norm": 9.589180560376818,
+      "learning_rate": 2e-06,
+      "loss": 0.2748,
+      "step": 7007
+    },
+    {
+      "epoch": 1.6257974712910335,
+      "grad_norm": 14.671821741588948,
+      "learning_rate": 2e-06,
+      "loss": 0.2478,
+      "step": 7008
+    },
+    {
+      "epoch": 1.6260294629393341,
+      "grad_norm": 22.87668683869851,
+      "learning_rate": 2e-06,
+      "loss": 0.2541,
+      "step": 7009
+    },
+    {
+      "epoch": 1.6262614545876348,
+      "grad_norm": 13.872086745553338,
+      "learning_rate": 2e-06,
+      "loss": 0.3123,
+      "step": 7010
+    },
+    {
+      "epoch": 1.6264934462359355,
+      "grad_norm": 22.842902471747887,
+      "learning_rate": 2e-06,
+      "loss": 0.2549,
+      "step": 7011
+    },
+    {
+      "epoch": 1.6267254378842362,
+      "grad_norm": 15.858015016072699,
+      "learning_rate": 2e-06,
+      "loss": 0.3301,
+      "step": 7012
+    },
+    {
+      "epoch": 1.6269574295325369,
+      "grad_norm": 17.20531113729456,
+      "learning_rate": 2e-06,
+      "loss": 0.3237,
+      "step": 7013
+    },
+    {
+      "epoch": 1.6271894211808375,
+      "grad_norm": 16.205524427264955,
+      "learning_rate": 2e-06,
+      "loss": 0.2433,
+      "step": 7014
+    },
+    {
+      "epoch": 1.627421412829138,
+      "grad_norm": 15.422387385622043,
+      "learning_rate": 2e-06,
+      "loss": 0.4239,
+      "step": 7015
+    },
+    {
+      "epoch": 1.6276534044774387,
+      "grad_norm": 16.60742013346518,
+      "learning_rate": 2e-06,
+      "loss": 0.3804,
+      "step": 7016
+    },
+    {
+      "epoch": 1.6278853961257393,
+      "grad_norm": 13.290142696386331,
+      "learning_rate": 2e-06,
+      "loss": 0.2607,
+      "step": 7017
+    },
+    {
+      "epoch": 1.62811738777404,
+      "grad_norm": 14.580299839319386,
+      "learning_rate": 2e-06,
+      "loss": 0.3681,
+      "step": 7018
+    },
+    {
+      "epoch": 1.6283493794223407,
+      "grad_norm": 11.730455808829811,
+      "learning_rate": 2e-06,
+      "loss": 0.2022,
+      "step": 7019
+    },
+    {
+      "epoch": 1.6285813710706414,
+      "grad_norm": 12.826274722124499,
+      "learning_rate": 2e-06,
+      "loss": 0.2255,
+      "step": 7020
+    },
+    {
+      "epoch": 1.628813362718942,
+      "grad_norm": 12.43214970101332,
+      "learning_rate": 2e-06,
+      "loss": 0.2796,
+      "step": 7021
+    },
+    {
+      "epoch": 1.6290453543672427,
+      "grad_norm": 14.440187541395591,
+      "learning_rate": 2e-06,
+      "loss": 0.4325,
+      "step": 7022
+    },
+    {
+      "epoch": 1.6292773460155434,
+      "grad_norm": 9.727986359501164,
+      "learning_rate": 2e-06,
+      "loss": 0.2083,
+      "step": 7023
+    },
+    {
+      "epoch": 1.629509337663844,
+      "grad_norm": 15.501284501084388,
+      "learning_rate": 2e-06,
+      "loss": 0.2339,
+      "step": 7024
+    },
+    {
+      "epoch": 1.6297413293121448,
+      "grad_norm": 13.762739971342741,
+      "learning_rate": 2e-06,
+      "loss": 0.2757,
+      "step": 7025
+    },
+    {
+      "epoch": 1.6299733209604454,
+      "grad_norm": 12.265472728748575,
+      "learning_rate": 2e-06,
+      "loss": 0.3016,
+      "step": 7026
+    },
+    {
+      "epoch": 1.6302053126087461,
+      "grad_norm": 15.96070196913894,
+      "learning_rate": 2e-06,
+      "loss": 0.2472,
+      "step": 7027
+    },
+    {
+      "epoch": 1.6304373042570468,
+      "grad_norm": 14.095882691910807,
+      "learning_rate": 2e-06,
+      "loss": 0.2394,
+      "step": 7028
+    },
+    {
+      "epoch": 1.6306692959053475,
+      "grad_norm": 16.54000688770342,
+      "learning_rate": 2e-06,
+      "loss": 0.2738,
+      "step": 7029
+    },
+    {
+      "epoch": 1.6309012875536482,
+      "grad_norm": 11.581800441529591,
+      "learning_rate": 2e-06,
+      "loss": 0.2248,
+      "step": 7030
+    },
+    {
+      "epoch": 1.6311332792019488,
+      "grad_norm": 10.634740006896134,
+      "learning_rate": 2e-06,
+      "loss": 0.1975,
+      "step": 7031
+    },
+    {
+      "epoch": 1.6313652708502495,
+      "grad_norm": 5.71974278126696,
+      "learning_rate": 2e-06,
+      "loss": 0.1228,
+      "step": 7032
+    },
+    {
+      "epoch": 1.6315972624985502,
+      "grad_norm": 11.52002459240929,
+      "learning_rate": 2e-06,
+      "loss": 0.1986,
+      "step": 7033
+    },
+    {
+      "epoch": 1.6318292541468509,
+      "grad_norm": 12.300589751455139,
+      "learning_rate": 2e-06,
+      "loss": 0.2178,
+      "step": 7034
+    },
+    {
+      "epoch": 1.6320612457951513,
+      "grad_norm": 11.600142147918511,
+      "learning_rate": 2e-06,
+      "loss": 0.2138,
+      "step": 7035
+    },
+    {
+      "epoch": 1.632293237443452,
+      "grad_norm": 10.474970068593255,
+      "learning_rate": 2e-06,
+      "loss": 0.2222,
+      "step": 7036
+    },
+    {
+      "epoch": 1.6325252290917527,
+      "grad_norm": 19.9060475508325,
+      "learning_rate": 2e-06,
+      "loss": 0.2173,
+      "step": 7037
+    },
+    {
+      "epoch": 1.6327572207400534,
+      "grad_norm": 11.184774219572693,
+      "learning_rate": 2e-06,
+      "loss": 0.1366,
+      "step": 7038
+    },
+    {
+      "epoch": 1.632989212388354,
+      "grad_norm": 11.245052817363465,
+      "learning_rate": 2e-06,
+      "loss": 0.1932,
+      "step": 7039
+    },
+    {
+      "epoch": 1.6332212040366547,
+      "grad_norm": 13.971578345151503,
+      "learning_rate": 2e-06,
+      "loss": 0.3935,
+      "step": 7040
+    },
+    {
+      "epoch": 1.6334531956849554,
+      "grad_norm": 10.115989514407806,
+      "learning_rate": 2e-06,
+      "loss": 0.2755,
+      "step": 7041
+    },
+    {
+      "epoch": 1.6336851873332559,
+      "grad_norm": 9.736157752300581,
+      "learning_rate": 2e-06,
+      "loss": 0.1958,
+      "step": 7042
+    },
+    {
+      "epoch": 1.6339171789815565,
+      "grad_norm": 12.293691083260265,
+      "learning_rate": 2e-06,
+      "loss": 0.2329,
+      "step": 7043
+    },
+    {
+      "epoch": 1.6341491706298572,
+      "grad_norm": 10.391037532897096,
+      "learning_rate": 2e-06,
+      "loss": 0.1948,
+      "step": 7044
+    },
+    {
+      "epoch": 1.6343811622781579,
+      "grad_norm": 17.20398074920175,
+      "learning_rate": 2e-06,
+      "loss": 0.3274,
+      "step": 7045
+    },
+    {
+      "epoch": 1.6346131539264586,
+      "grad_norm": 26.074947278089915,
+      "learning_rate": 2e-06,
+      "loss": 0.3111,
+      "step": 7046
+    },
+    {
+      "epoch": 1.6348451455747592,
+      "grad_norm": 12.340585075621851,
+      "learning_rate": 2e-06,
+      "loss": 0.337,
+      "step": 7047
+    },
+    {
+      "epoch": 1.63507713722306,
+      "grad_norm": 11.670255160442114,
+      "learning_rate": 2e-06,
+      "loss": 0.2644,
+      "step": 7048
+    },
+    {
+      "epoch": 1.6353091288713606,
+      "grad_norm": 10.342650552953074,
+      "learning_rate": 2e-06,
+      "loss": 0.2224,
+      "step": 7049
+    },
+    {
+      "epoch": 1.6355411205196613,
+      "grad_norm": 9.819092085974871,
+      "learning_rate": 2e-06,
+      "loss": 0.2415,
+      "step": 7050
+    },
+    {
+      "epoch": 1.635773112167962,
+      "grad_norm": 11.559798052526743,
+      "learning_rate": 2e-06,
+      "loss": 0.2196,
+      "step": 7051
+    },
+    {
+      "epoch": 1.6360051038162626,
+      "grad_norm": 16.251447881217928,
+      "learning_rate": 2e-06,
+      "loss": 0.3706,
+      "step": 7052
+    },
+    {
+      "epoch": 1.6362370954645633,
+      "grad_norm": 18.046939192717193,
+      "learning_rate": 2e-06,
+      "loss": 0.3217,
+      "step": 7053
+    },
+    {
+      "epoch": 1.636469087112864,
+      "grad_norm": 5.2898743743472645,
+      "learning_rate": 2e-06,
+      "loss": 0.1283,
+      "step": 7054
+    },
+    {
+      "epoch": 1.6367010787611647,
+      "grad_norm": 14.568451000162517,
+      "learning_rate": 2e-06,
+      "loss": 0.3354,
+      "step": 7055
+    },
+    {
+      "epoch": 1.6369330704094653,
+      "grad_norm": 8.994228712771722,
+      "learning_rate": 2e-06,
+      "loss": 0.2345,
+      "step": 7056
+    },
+    {
+      "epoch": 1.637165062057766,
+      "grad_norm": 22.65822871755929,
+      "learning_rate": 2e-06,
+      "loss": 0.3173,
+      "step": 7057
+    },
+    {
+      "epoch": 1.6373970537060667,
+      "grad_norm": 18.544582583755645,
+      "learning_rate": 2e-06,
+      "loss": 0.2662,
+      "step": 7058
+    },
+    {
+      "epoch": 1.6376290453543674,
+      "grad_norm": 12.768144951770473,
+      "learning_rate": 2e-06,
+      "loss": 0.3208,
+      "step": 7059
+    },
+    {
+      "epoch": 1.637861037002668,
+      "grad_norm": 16.23576130431629,
+      "learning_rate": 2e-06,
+      "loss": 0.2045,
+      "step": 7060
+    },
+    {
+      "epoch": 1.6380930286509687,
+      "grad_norm": 10.590738727915014,
+      "learning_rate": 2e-06,
+      "loss": 0.278,
+      "step": 7061
+    },
+    {
+      "epoch": 1.6383250202992692,
+      "grad_norm": 25.543926792585626,
+      "learning_rate": 2e-06,
+      "loss": 0.3695,
+      "step": 7062
+    },
+    {
+      "epoch": 1.6385570119475699,
+      "grad_norm": 13.621056889878547,
+      "learning_rate": 2e-06,
+      "loss": 0.3308,
+      "step": 7063
+    },
+    {
+      "epoch": 1.6387890035958705,
+      "grad_norm": 11.16766633431158,
+      "learning_rate": 2e-06,
+      "loss": 0.3386,
+      "step": 7064
+    },
+    {
+      "epoch": 1.6390209952441712,
+      "grad_norm": 10.051390960951437,
+      "learning_rate": 2e-06,
+      "loss": 0.1408,
+      "step": 7065
+    },
+    {
+      "epoch": 1.639252986892472,
+      "grad_norm": 16.319182344726816,
+      "learning_rate": 2e-06,
+      "loss": 0.2849,
+      "step": 7066
+    },
+    {
+      "epoch": 1.6394849785407726,
+      "grad_norm": 8.323363441312226,
+      "learning_rate": 2e-06,
+      "loss": 0.17,
+      "step": 7067
+    },
+    {
+      "epoch": 1.639716970189073,
+      "grad_norm": 18.082884377601907,
+      "learning_rate": 2e-06,
+      "loss": 0.2853,
+      "step": 7068
+    },
+    {
+      "epoch": 1.6399489618373737,
+      "grad_norm": 17.96858123227917,
+      "learning_rate": 2e-06,
+      "loss": 0.3246,
+      "step": 7069
+    },
+    {
+      "epoch": 1.6401809534856744,
+      "grad_norm": 8.348504342658279,
+      "learning_rate": 2e-06,
+      "loss": 0.2122,
+      "step": 7070
+    },
+    {
+      "epoch": 1.640412945133975,
+      "grad_norm": 14.195859872328015,
+      "learning_rate": 2e-06,
+      "loss": 0.2579,
+      "step": 7071
+    },
+    {
+      "epoch": 1.6406449367822757,
+      "grad_norm": 8.747800547063882,
+      "learning_rate": 2e-06,
+      "loss": 0.1936,
+      "step": 7072
+    },
+    {
+      "epoch": 1.6408769284305764,
+      "grad_norm": 13.625462706249374,
+      "learning_rate": 2e-06,
+      "loss": 0.1596,
+      "step": 7073
+    },
+    {
+      "epoch": 1.641108920078877,
+      "grad_norm": 9.66076283450918,
+      "learning_rate": 2e-06,
+      "loss": 0.1452,
+      "step": 7074
+    },
+    {
+      "epoch": 1.6413409117271778,
+      "grad_norm": 15.258436926230441,
+      "learning_rate": 2e-06,
+      "loss": 0.1653,
+      "step": 7075
+    },
+    {
+      "epoch": 1.6415729033754785,
+      "grad_norm": 16.421172139281012,
+      "learning_rate": 2e-06,
+      "loss": 0.2951,
+      "step": 7076
+    },
+    {
+      "epoch": 1.6418048950237791,
+      "grad_norm": 13.484416764834592,
+      "learning_rate": 2e-06,
+      "loss": 0.2514,
+      "step": 7077
+    },
+    {
+      "epoch": 1.6420368866720798,
+      "grad_norm": 18.197461234111085,
+      "learning_rate": 2e-06,
+      "loss": 0.2852,
+      "step": 7078
+    },
+    {
+      "epoch": 1.6422688783203805,
+      "grad_norm": 21.10162923051286,
+      "learning_rate": 2e-06,
+      "loss": 0.2829,
+      "step": 7079
+    },
+    {
+      "epoch": 1.6425008699686812,
+      "grad_norm": 12.427903176075388,
+      "learning_rate": 2e-06,
+      "loss": 0.2173,
+      "step": 7080
+    },
+    {
+      "epoch": 1.6427328616169818,
+      "grad_norm": 18.285960244290226,
+      "learning_rate": 2e-06,
+      "loss": 0.3021,
+      "step": 7081
+    },
+    {
+      "epoch": 1.6429648532652825,
+      "grad_norm": 8.15178393543725,
+      "learning_rate": 2e-06,
+      "loss": 0.1949,
+      "step": 7082
+    },
+    {
+      "epoch": 1.6431968449135832,
+      "grad_norm": 10.810702543448354,
+      "learning_rate": 2e-06,
+      "loss": 0.2177,
+      "step": 7083
+    },
+    {
+      "epoch": 1.6434288365618839,
+      "grad_norm": 10.53137318843868,
+      "learning_rate": 2e-06,
+      "loss": 0.1946,
+      "step": 7084
+    },
+    {
+      "epoch": 1.6436608282101846,
+      "grad_norm": 8.979276521118402,
+      "learning_rate": 2e-06,
+      "loss": 0.2341,
+      "step": 7085
+    },
+    {
+      "epoch": 1.6438928198584852,
+      "grad_norm": 7.373128298424443,
+      "learning_rate": 2e-06,
+      "loss": 0.1703,
+      "step": 7086
+    },
+    {
+      "epoch": 1.644124811506786,
+      "grad_norm": 9.086247902153618,
+      "learning_rate": 2e-06,
+      "loss": 0.1985,
+      "step": 7087
+    },
+    {
+      "epoch": 1.6443568031550864,
+      "grad_norm": 18.829921655392535,
+      "learning_rate": 2e-06,
+      "loss": 0.2776,
+      "step": 7088
+    },
+    {
+      "epoch": 1.644588794803387,
+      "grad_norm": 16.727235065578363,
+      "learning_rate": 2e-06,
+      "loss": 0.2458,
+      "step": 7089
+    },
+    {
+      "epoch": 1.6448207864516877,
+      "grad_norm": 11.826408708970803,
+      "learning_rate": 2e-06,
+      "loss": 0.2563,
+      "step": 7090
+    },
+    {
+      "epoch": 1.6450527780999884,
+      "grad_norm": 12.812812098521128,
+      "learning_rate": 2e-06,
+      "loss": 0.2057,
+      "step": 7091
+    },
+    {
+      "epoch": 1.645284769748289,
+      "grad_norm": 18.747553139697516,
+      "learning_rate": 2e-06,
+      "loss": 0.2797,
+      "step": 7092
+    },
+    {
+      "epoch": 1.6455167613965898,
+      "grad_norm": 21.37070998648583,
+      "learning_rate": 2e-06,
+      "loss": 0.2842,
+      "step": 7093
+    },
+    {
+      "epoch": 1.6457487530448904,
+      "grad_norm": 16.031144141161292,
+      "learning_rate": 2e-06,
+      "loss": 0.2853,
+      "step": 7094
+    },
+    {
+      "epoch": 1.645980744693191,
+      "grad_norm": 11.355397426862138,
+      "learning_rate": 2e-06,
+      "loss": 0.2688,
+      "step": 7095
+    },
+    {
+      "epoch": 1.6462127363414916,
+      "grad_norm": 10.83029224429112,
+      "learning_rate": 2e-06,
+      "loss": 0.211,
+      "step": 7096
+    },
+    {
+      "epoch": 1.6464447279897922,
+      "grad_norm": 9.158647967175433,
+      "learning_rate": 2e-06,
+      "loss": 0.2457,
+      "step": 7097
+    },
+    {
+      "epoch": 1.646676719638093,
+      "grad_norm": 12.659950260456156,
+      "learning_rate": 2e-06,
+      "loss": 0.2327,
+      "step": 7098
+    },
+    {
+      "epoch": 1.6469087112863936,
+      "grad_norm": 11.799338439161634,
+      "learning_rate": 2e-06,
+      "loss": 0.2592,
+      "step": 7099
+    },
+    {
+      "epoch": 1.6471407029346943,
+      "grad_norm": 13.881708515680108,
+      "learning_rate": 2e-06,
+      "loss": 0.2502,
+      "step": 7100
+    },
+    {
+      "epoch": 1.647372694582995,
+      "grad_norm": 15.506415761028464,
+      "learning_rate": 2e-06,
+      "loss": 0.2147,
+      "step": 7101
+    },
+    {
+      "epoch": 1.6476046862312956,
+      "grad_norm": 14.239460943501205,
+      "learning_rate": 2e-06,
+      "loss": 0.2004,
+      "step": 7102
+    },
+    {
+      "epoch": 1.6478366778795963,
+      "grad_norm": 12.17574573411829,
+      "learning_rate": 2e-06,
+      "loss": 0.1786,
+      "step": 7103
+    },
+    {
+      "epoch": 1.648068669527897,
+      "grad_norm": 17.152879482952468,
+      "learning_rate": 2e-06,
+      "loss": 0.1815,
+      "step": 7104
+    },
+    {
+      "epoch": 1.6483006611761977,
+      "grad_norm": 7.326448041591375,
+      "learning_rate": 2e-06,
+      "loss": 0.1246,
+      "step": 7105
+    },
+    {
+      "epoch": 1.6485326528244983,
+      "grad_norm": 16.79864805443018,
+      "learning_rate": 2e-06,
+      "loss": 0.3143,
+      "step": 7106
+    },
+    {
+      "epoch": 1.648764644472799,
+      "grad_norm": 13.370767304403438,
+      "learning_rate": 2e-06,
+      "loss": 0.2865,
+      "step": 7107
+    },
+    {
+      "epoch": 1.6489966361210997,
+      "grad_norm": 21.51092895572337,
+      "learning_rate": 2e-06,
+      "loss": 0.295,
+      "step": 7108
+    },
+    {
+      "epoch": 1.6492286277694004,
+      "grad_norm": 18.007298325256,
+      "learning_rate": 2e-06,
+      "loss": 0.3231,
+      "step": 7109
+    },
+    {
+      "epoch": 1.649460619417701,
+      "grad_norm": 27.794812087978574,
+      "learning_rate": 2e-06,
+      "loss": 0.3555,
+      "step": 7110
+    },
+    {
+      "epoch": 1.6496926110660017,
+      "grad_norm": 9.828940646642232,
+      "learning_rate": 2e-06,
+      "loss": 0.1938,
+      "step": 7111
+    },
+    {
+      "epoch": 1.6499246027143024,
+      "grad_norm": 19.735335010849287,
+      "learning_rate": 2e-06,
+      "loss": 0.2739,
+      "step": 7112
+    },
+    {
+      "epoch": 1.650156594362603,
+      "grad_norm": 11.14559502802546,
+      "learning_rate": 2e-06,
+      "loss": 0.2884,
+      "step": 7113
+    },
+    {
+      "epoch": 1.6503885860109038,
+      "grad_norm": 12.179689951370753,
+      "learning_rate": 2e-06,
+      "loss": 0.194,
+      "step": 7114
+    },
+    {
+      "epoch": 1.6506205776592042,
+      "grad_norm": 16.684675765332578,
+      "learning_rate": 2e-06,
+      "loss": 0.2507,
+      "step": 7115
+    },
+    {
+      "epoch": 1.650852569307505,
+      "grad_norm": 6.811563807023977,
+      "learning_rate": 2e-06,
+      "loss": 0.1291,
+      "step": 7116
+    },
+    {
+      "epoch": 1.6510845609558056,
+      "grad_norm": 10.669302710321867,
+      "learning_rate": 2e-06,
+      "loss": 0.2399,
+      "step": 7117
+    },
+    {
+      "epoch": 1.6513165526041063,
+      "grad_norm": 11.669049742452437,
+      "learning_rate": 2e-06,
+      "loss": 0.2609,
+      "step": 7118
+    },
+    {
+      "epoch": 1.651548544252407,
+      "grad_norm": 18.476138275697288,
+      "learning_rate": 2e-06,
+      "loss": 0.3279,
+      "step": 7119
+    },
+    {
+      "epoch": 1.6517805359007076,
+      "grad_norm": 16.27903554685498,
+      "learning_rate": 2e-06,
+      "loss": 0.2781,
+      "step": 7120
+    },
+    {
+      "epoch": 1.652012527549008,
+      "grad_norm": 17.359246051089706,
+      "learning_rate": 2e-06,
+      "loss": 0.3157,
+      "step": 7121
+    },
+    {
+      "epoch": 1.6522445191973087,
+      "grad_norm": 12.366129458336568,
+      "learning_rate": 2e-06,
+      "loss": 0.2502,
+      "step": 7122
+    },
+    {
+      "epoch": 1.6524765108456094,
+      "grad_norm": 11.376221833141713,
+      "learning_rate": 2e-06,
+      "loss": 0.2164,
+      "step": 7123
+    },
+    {
+      "epoch": 1.65270850249391,
+      "grad_norm": 23.552326593798647,
+      "learning_rate": 2e-06,
+      "loss": 0.2096,
+      "step": 7124
+    },
+    {
+      "epoch": 1.6529404941422108,
+      "grad_norm": 21.398110960726864,
+      "learning_rate": 2e-06,
+      "loss": 0.3372,
+      "step": 7125
+    },
+    {
+      "epoch": 1.6531724857905115,
+      "grad_norm": 10.320835052153663,
+      "learning_rate": 2e-06,
+      "loss": 0.1588,
+      "step": 7126
+    },
+    {
+      "epoch": 1.6534044774388121,
+      "grad_norm": 43.51247348322184,
+      "learning_rate": 2e-06,
+      "loss": 0.3616,
+      "step": 7127
+    },
+    {
+      "epoch": 1.6536364690871128,
+      "grad_norm": 16.571909321923275,
+      "learning_rate": 2e-06,
+      "loss": 0.3267,
+      "step": 7128
+    },
+    {
+      "epoch": 1.6538684607354135,
+      "grad_norm": 12.319199115679023,
+      "learning_rate": 2e-06,
+      "loss": 0.3166,
+      "step": 7129
+    },
+    {
+      "epoch": 1.6541004523837142,
+      "grad_norm": 17.12975006935626,
+      "learning_rate": 2e-06,
+      "loss": 0.2008,
+      "step": 7130
+    },
+    {
+      "epoch": 1.6543324440320148,
+      "grad_norm": 14.052613143049053,
+      "learning_rate": 2e-06,
+      "loss": 0.2825,
+      "step": 7131
+    },
+    {
+      "epoch": 1.6545644356803155,
+      "grad_norm": 17.40443541616568,
+      "learning_rate": 2e-06,
+      "loss": 0.3052,
+      "step": 7132
+    },
+    {
+      "epoch": 1.6547964273286162,
+      "grad_norm": 27.391161447742995,
+      "learning_rate": 2e-06,
+      "loss": 0.3455,
+      "step": 7133
+    },
+    {
+      "epoch": 1.6550284189769169,
+      "grad_norm": 16.172270537691748,
+      "learning_rate": 2e-06,
+      "loss": 0.1898,
+      "step": 7134
+    },
+    {
+      "epoch": 1.6552604106252176,
+      "grad_norm": 7.745639522105556,
+      "learning_rate": 2e-06,
+      "loss": 0.1708,
+      "step": 7135
+    },
+    {
+      "epoch": 1.6554924022735182,
+      "grad_norm": 15.698002630882499,
+      "learning_rate": 2e-06,
+      "loss": 0.2375,
+      "step": 7136
+    },
+    {
+      "epoch": 1.655724393921819,
+      "grad_norm": 18.27497738405015,
+      "learning_rate": 2e-06,
+      "loss": 0.2097,
+      "step": 7137
+    },
+    {
+      "epoch": 1.6559563855701196,
+      "grad_norm": 19.399303027089903,
+      "learning_rate": 2e-06,
+      "loss": 0.2969,
+      "step": 7138
+    },
+    {
+      "epoch": 1.6561883772184203,
+      "grad_norm": 15.486733537769139,
+      "learning_rate": 2e-06,
+      "loss": 0.3188,
+      "step": 7139
+    },
+    {
+      "epoch": 1.656420368866721,
+      "grad_norm": 9.362904869460838,
+      "learning_rate": 2e-06,
+      "loss": 0.2204,
+      "step": 7140
+    },
+    {
+      "epoch": 1.6566523605150214,
+      "grad_norm": 12.366872072316477,
+      "learning_rate": 2e-06,
+      "loss": 0.2335,
+      "step": 7141
+    },
+    {
+      "epoch": 1.656884352163322,
+      "grad_norm": 12.496802501137957,
+      "learning_rate": 2e-06,
+      "loss": 0.236,
+      "step": 7142
+    },
+    {
+      "epoch": 1.6571163438116228,
+      "grad_norm": 12.254734708977596,
+      "learning_rate": 2e-06,
+      "loss": 0.1864,
+      "step": 7143
+    },
+    {
+      "epoch": 1.6573483354599234,
+      "grad_norm": 9.660063763494001,
+      "learning_rate": 2e-06,
+      "loss": 0.2146,
+      "step": 7144
+    },
+    {
+      "epoch": 1.6575803271082241,
+      "grad_norm": 10.427701070630379,
+      "learning_rate": 2e-06,
+      "loss": 0.1668,
+      "step": 7145
+    },
+    {
+      "epoch": 1.6578123187565248,
+      "grad_norm": 12.60420478831905,
+      "learning_rate": 2e-06,
+      "loss": 0.2078,
+      "step": 7146
+    },
+    {
+      "epoch": 1.6580443104048255,
+      "grad_norm": 7.42629483367214,
+      "learning_rate": 2e-06,
+      "loss": 0.1457,
+      "step": 7147
+    },
+    {
+      "epoch": 1.658276302053126,
+      "grad_norm": 18.688867676781133,
+      "learning_rate": 2e-06,
+      "loss": 0.3135,
+      "step": 7148
+    },
+    {
+      "epoch": 1.6585082937014266,
+      "grad_norm": 18.540251604423794,
+      "learning_rate": 2e-06,
+      "loss": 0.3148,
+      "step": 7149
+    },
+    {
+      "epoch": 1.6587402853497273,
+      "grad_norm": 15.21926828603947,
+      "learning_rate": 2e-06,
+      "loss": 0.3285,
+      "step": 7150
+    },
+    {
+      "epoch": 1.658972276998028,
+      "grad_norm": 17.34134795151519,
+      "learning_rate": 2e-06,
+      "loss": 0.3138,
+      "step": 7151
+    },
+    {
+      "epoch": 1.6592042686463286,
+      "grad_norm": 15.21227762992704,
+      "learning_rate": 2e-06,
+      "loss": 0.242,
+      "step": 7152
+    },
+    {
+      "epoch": 1.6594362602946293,
+      "grad_norm": 12.275563953766433,
+      "learning_rate": 2e-06,
+      "loss": 0.3142,
+      "step": 7153
+    },
+    {
+      "epoch": 1.65966825194293,
+      "grad_norm": 8.203457602806631,
+      "learning_rate": 2e-06,
+      "loss": 0.1712,
+      "step": 7154
+    },
+    {
+      "epoch": 1.6599002435912307,
+      "grad_norm": 12.874196812733794,
+      "learning_rate": 2e-06,
+      "loss": 0.2242,
+      "step": 7155
+    },
+    {
+      "epoch": 1.6601322352395314,
+      "grad_norm": 9.838459330430709,
+      "learning_rate": 2e-06,
+      "loss": 0.2479,
+      "step": 7156
+    },
+    {
+      "epoch": 1.660364226887832,
+      "grad_norm": 12.464409721834336,
+      "learning_rate": 2e-06,
+      "loss": 0.2776,
+      "step": 7157
+    },
+    {
+      "epoch": 1.6605962185361327,
+      "grad_norm": 11.060179146903785,
+      "learning_rate": 2e-06,
+      "loss": 0.2777,
+      "step": 7158
+    },
+    {
+      "epoch": 1.6608282101844334,
+      "grad_norm": 19.965716682293383,
+      "learning_rate": 2e-06,
+      "loss": 0.2935,
+      "step": 7159
+    },
+    {
+      "epoch": 1.661060201832734,
+      "grad_norm": 14.087061956800868,
+      "learning_rate": 2e-06,
+      "loss": 0.246,
+      "step": 7160
+    },
+    {
+      "epoch": 1.6612921934810347,
+      "grad_norm": 17.92066191994181,
+      "learning_rate": 2e-06,
+      "loss": 0.2955,
+      "step": 7161
+    },
+    {
+      "epoch": 1.6615241851293354,
+      "grad_norm": 24.209134033683643,
+      "learning_rate": 2e-06,
+      "loss": 0.2748,
+      "step": 7162
+    },
+    {
+      "epoch": 1.661756176777636,
+      "grad_norm": 11.876774521762682,
+      "learning_rate": 2e-06,
+      "loss": 0.249,
+      "step": 7163
+    },
+    {
+      "epoch": 1.6619881684259368,
+      "grad_norm": 11.534212929724452,
+      "learning_rate": 2e-06,
+      "loss": 0.1807,
+      "step": 7164
+    },
+    {
+      "epoch": 1.6622201600742375,
+      "grad_norm": 15.413596450223995,
+      "learning_rate": 2e-06,
+      "loss": 0.2057,
+      "step": 7165
+    },
+    {
+      "epoch": 1.6624521517225381,
+      "grad_norm": 14.949284064707648,
+      "learning_rate": 2e-06,
+      "loss": 0.1989,
+      "step": 7166
+    },
+    {
+      "epoch": 1.6626841433708388,
+      "grad_norm": 7.451658989662105,
+      "learning_rate": 2e-06,
+      "loss": 0.2982,
+      "step": 7167
+    },
+    {
+      "epoch": 1.6629161350191393,
+      "grad_norm": 17.876488522008227,
+      "learning_rate": 2e-06,
+      "loss": 0.3764,
+      "step": 7168
+    },
+    {
+      "epoch": 1.66314812666744,
+      "grad_norm": 9.601497369310469,
+      "learning_rate": 2e-06,
+      "loss": 0.1856,
+      "step": 7169
+    },
+    {
+      "epoch": 1.6633801183157406,
+      "grad_norm": 26.29643437167827,
+      "learning_rate": 2e-06,
+      "loss": 0.2852,
+      "step": 7170
+    },
+    {
+      "epoch": 1.6636121099640413,
+      "grad_norm": 26.008823115734014,
+      "learning_rate": 2e-06,
+      "loss": 0.1835,
+      "step": 7171
+    },
+    {
+      "epoch": 1.663844101612342,
+      "grad_norm": 17.775698521683385,
+      "learning_rate": 2e-06,
+      "loss": 0.3432,
+      "step": 7172
+    },
+    {
+      "epoch": 1.6640760932606427,
+      "grad_norm": 16.226258624963403,
+      "learning_rate": 2e-06,
+      "loss": 0.2569,
+      "step": 7173
+    },
+    {
+      "epoch": 1.6643080849089433,
+      "grad_norm": 11.083269886897604,
+      "learning_rate": 2e-06,
+      "loss": 0.2663,
+      "step": 7174
+    },
+    {
+      "epoch": 1.6645400765572438,
+      "grad_norm": 16.496353744352007,
+      "learning_rate": 2e-06,
+      "loss": 0.2829,
+      "step": 7175
+    },
+    {
+      "epoch": 1.6647720682055445,
+      "grad_norm": 20.16568187382624,
+      "learning_rate": 2e-06,
+      "loss": 0.2756,
+      "step": 7176
+    },
+    {
+      "epoch": 1.6650040598538451,
+      "grad_norm": 5.926606012468249,
+      "learning_rate": 2e-06,
+      "loss": 0.1369,
+      "step": 7177
+    },
+    {
+      "epoch": 1.6652360515021458,
+      "grad_norm": 12.910944857308234,
+      "learning_rate": 2e-06,
+      "loss": 0.2397,
+      "step": 7178
+    },
+    {
+      "epoch": 1.6654680431504465,
+      "grad_norm": 12.738385571388708,
+      "learning_rate": 2e-06,
+      "loss": 0.2917,
+      "step": 7179
+    },
+    {
+      "epoch": 1.6657000347987472,
+      "grad_norm": 17.232627774810663,
+      "learning_rate": 2e-06,
+      "loss": 0.2987,
+      "step": 7180
+    },
+    {
+      "epoch": 1.6659320264470479,
+      "grad_norm": 10.560890514090739,
+      "learning_rate": 2e-06,
+      "loss": 0.2315,
+      "step": 7181
+    },
+    {
+      "epoch": 1.6661640180953485,
+      "grad_norm": 5.4585419224197445,
+      "learning_rate": 2e-06,
+      "loss": 0.145,
+      "step": 7182
+    },
+    {
+      "epoch": 1.6663960097436492,
+      "grad_norm": 20.192116461853846,
+      "learning_rate": 2e-06,
+      "loss": 0.2229,
+      "step": 7183
+    },
+    {
+      "epoch": 1.6666280013919499,
+      "grad_norm": 18.001526552299513,
+      "learning_rate": 2e-06,
+      "loss": 0.2593,
+      "step": 7184
+    },
+    {
+      "epoch": 1.6668599930402506,
+      "grad_norm": 17.958796196132493,
+      "learning_rate": 2e-06,
+      "loss": 0.2876,
+      "step": 7185
+    },
+    {
+      "epoch": 1.6670919846885512,
+      "grad_norm": 13.977270787710811,
+      "learning_rate": 2e-06,
+      "loss": 0.2066,
+      "step": 7186
+    },
+    {
+      "epoch": 1.667323976336852,
+      "grad_norm": 6.077093067527741,
+      "learning_rate": 2e-06,
+      "loss": 0.1958,
+      "step": 7187
+    },
+    {
+      "epoch": 1.6675559679851526,
+      "grad_norm": 14.162961183279469,
+      "learning_rate": 2e-06,
+      "loss": 0.1998,
+      "step": 7188
+    },
+    {
+      "epoch": 1.6677879596334533,
+      "grad_norm": 15.410769413943282,
+      "learning_rate": 2e-06,
+      "loss": 0.2707,
+      "step": 7189
+    },
+    {
+      "epoch": 1.668019951281754,
+      "grad_norm": 14.002372039832355,
+      "learning_rate": 2e-06,
+      "loss": 0.249,
+      "step": 7190
+    },
+    {
+      "epoch": 1.6682519429300546,
+      "grad_norm": 11.425963241642725,
+      "learning_rate": 2e-06,
+      "loss": 0.249,
+      "step": 7191
+    },
+    {
+      "epoch": 1.6684839345783553,
+      "grad_norm": 10.424330561363904,
+      "learning_rate": 2e-06,
+      "loss": 0.1938,
+      "step": 7192
+    },
+    {
+      "epoch": 1.668715926226656,
+      "grad_norm": 7.77959555923727,
+      "learning_rate": 2e-06,
+      "loss": 0.2323,
+      "step": 7193
+    },
+    {
+      "epoch": 1.6689479178749567,
+      "grad_norm": 12.563718470613516,
+      "learning_rate": 2e-06,
+      "loss": 0.2741,
+      "step": 7194
+    },
+    {
+      "epoch": 1.6691799095232571,
+      "grad_norm": 12.248035271839441,
+      "learning_rate": 2e-06,
+      "loss": 0.2678,
+      "step": 7195
+    },
+    {
+      "epoch": 1.6694119011715578,
+      "grad_norm": 9.725214437497984,
+      "learning_rate": 2e-06,
+      "loss": 0.1983,
+      "step": 7196
+    },
+    {
+      "epoch": 1.6696438928198585,
+      "grad_norm": 83.29799388121215,
+      "learning_rate": 2e-06,
+      "loss": 0.3167,
+      "step": 7197
+    },
+    {
+      "epoch": 1.6698758844681592,
+      "grad_norm": 17.328766071210865,
+      "learning_rate": 2e-06,
+      "loss": 0.2559,
+      "step": 7198
+    },
+    {
+      "epoch": 1.6701078761164598,
+      "grad_norm": 11.736583987785513,
+      "learning_rate": 2e-06,
+      "loss": 0.1709,
+      "step": 7199
+    },
+    {
+      "epoch": 1.6703398677647605,
+      "grad_norm": 5.073238294383731,
+      "learning_rate": 2e-06,
+      "loss": 0.1439,
+      "step": 7200
+    },
+    {
+      "epoch": 1.670571859413061,
+      "grad_norm": 23.945785671567652,
+      "learning_rate": 2e-06,
+      "loss": 0.4377,
+      "step": 7201
+    },
+    {
+      "epoch": 1.6708038510613616,
+      "grad_norm": 17.812827348025948,
+      "learning_rate": 2e-06,
+      "loss": 0.2269,
+      "step": 7202
+    },
+    {
+      "epoch": 1.6710358427096623,
+      "grad_norm": 13.814090695774514,
+      "learning_rate": 2e-06,
+      "loss": 0.2267,
+      "step": 7203
+    },
+    {
+      "epoch": 1.671267834357963,
+      "grad_norm": 11.800236103771432,
+      "learning_rate": 2e-06,
+      "loss": 0.1501,
+      "step": 7204
+    },
+    {
+      "epoch": 1.6714998260062637,
+      "grad_norm": 14.133330969888693,
+      "learning_rate": 2e-06,
+      "loss": 0.2594,
+      "step": 7205
+    },
+    {
+      "epoch": 1.6717318176545644,
+      "grad_norm": 12.403132619329371,
+      "learning_rate": 2e-06,
+      "loss": 0.2171,
+      "step": 7206
+    },
+    {
+      "epoch": 1.671963809302865,
+      "grad_norm": 11.387512003348485,
+      "learning_rate": 2e-06,
+      "loss": 0.243,
+      "step": 7207
+    },
+    {
+      "epoch": 1.6721958009511657,
+      "grad_norm": 29.850020206835577,
+      "learning_rate": 2e-06,
+      "loss": 0.3552,
+      "step": 7208
+    },
+    {
+      "epoch": 1.6724277925994664,
+      "grad_norm": 14.538118662992103,
+      "learning_rate": 2e-06,
+      "loss": 0.2018,
+      "step": 7209
+    },
+    {
+      "epoch": 1.672659784247767,
+      "grad_norm": 17.847018805166996,
+      "learning_rate": 2e-06,
+      "loss": 0.3098,
+      "step": 7210
+    },
+    {
+      "epoch": 1.6728917758960677,
+      "grad_norm": 11.821794128153151,
+      "learning_rate": 2e-06,
+      "loss": 0.3643,
+      "step": 7211
+    },
+    {
+      "epoch": 1.6731237675443684,
+      "grad_norm": 20.07149739617927,
+      "learning_rate": 2e-06,
+      "loss": 0.3223,
+      "step": 7212
+    },
+    {
+      "epoch": 1.673355759192669,
+      "grad_norm": 9.991688195420922,
+      "learning_rate": 2e-06,
+      "loss": 0.2121,
+      "step": 7213
+    },
+    {
+      "epoch": 1.6735877508409698,
+      "grad_norm": 13.829416802908472,
+      "learning_rate": 2e-06,
+      "loss": 0.2223,
+      "step": 7214
+    },
+    {
+      "epoch": 1.6738197424892705,
+      "grad_norm": 5.792930798689378,
+      "learning_rate": 2e-06,
+      "loss": 0.1783,
+      "step": 7215
+    },
+    {
+      "epoch": 1.6740517341375711,
+      "grad_norm": 21.858001437576718,
+      "learning_rate": 2e-06,
+      "loss": 0.3587,
+      "step": 7216
+    },
+    {
+      "epoch": 1.6742837257858718,
+      "grad_norm": 14.953758049208334,
+      "learning_rate": 2e-06,
+      "loss": 0.261,
+      "step": 7217
+    },
+    {
+      "epoch": 1.6745157174341725,
+      "grad_norm": 15.32929982573524,
+      "learning_rate": 2e-06,
+      "loss": 0.2438,
+      "step": 7218
+    },
+    {
+      "epoch": 1.6747477090824732,
+      "grad_norm": 12.461868779926432,
+      "learning_rate": 2e-06,
+      "loss": 0.2068,
+      "step": 7219
+    },
+    {
+      "epoch": 1.6749797007307738,
+      "grad_norm": 15.767175377082367,
+      "learning_rate": 2e-06,
+      "loss": 0.2911,
+      "step": 7220
+    },
+    {
+      "epoch": 1.6752116923790743,
+      "grad_norm": 13.8931709287092,
+      "learning_rate": 2e-06,
+      "loss": 0.2982,
+      "step": 7221
+    },
+    {
+      "epoch": 1.675443684027375,
+      "grad_norm": 14.649828134211061,
+      "learning_rate": 2e-06,
+      "loss": 0.2224,
+      "step": 7222
+    },
+    {
+      "epoch": 1.6756756756756757,
+      "grad_norm": 21.45201926425029,
+      "learning_rate": 2e-06,
+      "loss": 0.2641,
+      "step": 7223
+    },
+    {
+      "epoch": 1.6759076673239763,
+      "grad_norm": 13.8367213243625,
+      "learning_rate": 2e-06,
+      "loss": 0.2521,
+      "step": 7224
+    },
+    {
+      "epoch": 1.676139658972277,
+      "grad_norm": 10.023145113017152,
+      "learning_rate": 2e-06,
+      "loss": 0.22,
+      "step": 7225
+    },
+    {
+      "epoch": 1.6763716506205777,
+      "grad_norm": 29.17788046000198,
+      "learning_rate": 2e-06,
+      "loss": 0.2037,
+      "step": 7226
+    },
+    {
+      "epoch": 1.6766036422688784,
+      "grad_norm": 9.137259677515337,
+      "learning_rate": 2e-06,
+      "loss": 0.276,
+      "step": 7227
+    },
+    {
+      "epoch": 1.6768356339171788,
+      "grad_norm": 13.453485559735219,
+      "learning_rate": 2e-06,
+      "loss": 0.2402,
+      "step": 7228
+    },
+    {
+      "epoch": 1.6770676255654795,
+      "grad_norm": 14.936132272085098,
+      "learning_rate": 2e-06,
+      "loss": 0.2982,
+      "step": 7229
+    },
+    {
+      "epoch": 1.6772996172137802,
+      "grad_norm": 11.989505567263949,
+      "learning_rate": 2e-06,
+      "loss": 0.3163,
+      "step": 7230
+    },
+    {
+      "epoch": 1.6775316088620809,
+      "grad_norm": 16.932666813685692,
+      "learning_rate": 2e-06,
+      "loss": 0.3932,
+      "step": 7231
+    },
+    {
+      "epoch": 1.6777636005103815,
+      "grad_norm": 10.189344381942949,
+      "learning_rate": 2e-06,
+      "loss": 0.2055,
+      "step": 7232
+    },
+    {
+      "epoch": 1.6779955921586822,
+      "grad_norm": 13.877659783625909,
+      "learning_rate": 2e-06,
+      "loss": 0.2089,
+      "step": 7233
+    },
+    {
+      "epoch": 1.678227583806983,
+      "grad_norm": 11.88046757505946,
+      "learning_rate": 2e-06,
+      "loss": 0.2989,
+      "step": 7234
+    },
+    {
+      "epoch": 1.6784595754552836,
+      "grad_norm": 12.790805012245048,
+      "learning_rate": 2e-06,
+      "loss": 0.2614,
+      "step": 7235
+    },
+    {
+      "epoch": 1.6786915671035842,
+      "grad_norm": 13.851701743001856,
+      "learning_rate": 2e-06,
+      "loss": 0.1686,
+      "step": 7236
+    },
+    {
+      "epoch": 1.678923558751885,
+      "grad_norm": 17.59761637474589,
+      "learning_rate": 2e-06,
+      "loss": 0.2935,
+      "step": 7237
+    },
+    {
+      "epoch": 1.6791555504001856,
+      "grad_norm": 13.48792494767899,
+      "learning_rate": 2e-06,
+      "loss": 0.3288,
+      "step": 7238
+    },
+    {
+      "epoch": 1.6793875420484863,
+      "grad_norm": 20.19153728570158,
+      "learning_rate": 2e-06,
+      "loss": 0.2486,
+      "step": 7239
+    },
+    {
+      "epoch": 1.679619533696787,
+      "grad_norm": 8.553678141416055,
+      "learning_rate": 2e-06,
+      "loss": 0.2466,
+      "step": 7240
+    },
+    {
+      "epoch": 1.6798515253450876,
+      "grad_norm": 9.832203832754326,
+      "learning_rate": 2e-06,
+      "loss": 0.2081,
+      "step": 7241
+    },
+    {
+      "epoch": 1.6800835169933883,
+      "grad_norm": 16.72323345948946,
+      "learning_rate": 2e-06,
+      "loss": 0.2816,
+      "step": 7242
+    },
+    {
+      "epoch": 1.680315508641689,
+      "grad_norm": 9.139400513114731,
+      "learning_rate": 2e-06,
+      "loss": 0.1794,
+      "step": 7243
+    },
+    {
+      "epoch": 1.6805475002899897,
+      "grad_norm": 11.173560663571159,
+      "learning_rate": 2e-06,
+      "loss": 0.21,
+      "step": 7244
+    },
+    {
+      "epoch": 1.6807794919382903,
+      "grad_norm": 13.41759061043099,
+      "learning_rate": 2e-06,
+      "loss": 0.2704,
+      "step": 7245
+    },
+    {
+      "epoch": 1.681011483586591,
+      "grad_norm": 11.530722433665698,
+      "learning_rate": 2e-06,
+      "loss": 0.2328,
+      "step": 7246
+    },
+    {
+      "epoch": 1.6812434752348917,
+      "grad_norm": 48.37803041976953,
+      "learning_rate": 2e-06,
+      "loss": 0.1837,
+      "step": 7247
+    },
+    {
+      "epoch": 1.6814754668831922,
+      "grad_norm": 9.805662828423877,
+      "learning_rate": 2e-06,
+      "loss": 0.3065,
+      "step": 7248
+    },
+    {
+      "epoch": 1.6817074585314928,
+      "grad_norm": 14.922040681149298,
+      "learning_rate": 2e-06,
+      "loss": 0.2789,
+      "step": 7249
+    },
+    {
+      "epoch": 1.6819394501797935,
+      "grad_norm": 16.518766137978755,
+      "learning_rate": 2e-06,
+      "loss": 0.197,
+      "step": 7250
+    },
+    {
+      "epoch": 1.6821714418280942,
+      "grad_norm": 9.19056226837705,
+      "learning_rate": 2e-06,
+      "loss": 0.2237,
+      "step": 7251
+    },
+    {
+      "epoch": 1.6824034334763949,
+      "grad_norm": 13.196367330577875,
+      "learning_rate": 2e-06,
+      "loss": 0.2455,
+      "step": 7252
+    },
+    {
+      "epoch": 1.6826354251246955,
+      "grad_norm": 14.036007189990867,
+      "learning_rate": 2e-06,
+      "loss": 0.2197,
+      "step": 7253
+    },
+    {
+      "epoch": 1.6828674167729962,
+      "grad_norm": 14.91431701000174,
+      "learning_rate": 2e-06,
+      "loss": 0.2718,
+      "step": 7254
+    },
+    {
+      "epoch": 1.6830994084212967,
+      "grad_norm": 16.16148852325024,
+      "learning_rate": 2e-06,
+      "loss": 0.2549,
+      "step": 7255
+    },
+    {
+      "epoch": 1.6833314000695974,
+      "grad_norm": 15.585917882713956,
+      "learning_rate": 2e-06,
+      "loss": 0.1808,
+      "step": 7256
+    },
+    {
+      "epoch": 1.683563391717898,
+      "grad_norm": 20.987773075527787,
+      "learning_rate": 2e-06,
+      "loss": 0.2464,
+      "step": 7257
+    },
+    {
+      "epoch": 1.6837953833661987,
+      "grad_norm": 9.67357503129696,
+      "learning_rate": 2e-06,
+      "loss": 0.2684,
+      "step": 7258
+    },
+    {
+      "epoch": 1.6840273750144994,
+      "grad_norm": 12.767117257443124,
+      "learning_rate": 2e-06,
+      "loss": 0.2111,
+      "step": 7259
+    },
+    {
+      "epoch": 1.6842593666628,
+      "grad_norm": 11.685125969676664,
+      "learning_rate": 2e-06,
+      "loss": 0.2216,
+      "step": 7260
+    },
+    {
+      "epoch": 1.6844913583111008,
+      "grad_norm": 15.651428326415914,
+      "learning_rate": 2e-06,
+      "loss": 0.2254,
+      "step": 7261
+    },
+    {
+      "epoch": 1.6847233499594014,
+      "grad_norm": 10.54086061454288,
+      "learning_rate": 2e-06,
+      "loss": 0.2496,
+      "step": 7262
+    },
+    {
+      "epoch": 1.684955341607702,
+      "grad_norm": 15.971847719178118,
+      "learning_rate": 2e-06,
+      "loss": 0.189,
+      "step": 7263
+    },
+    {
+      "epoch": 1.6851873332560028,
+      "grad_norm": 8.066546502110779,
+      "learning_rate": 2e-06,
+      "loss": 0.2025,
+      "step": 7264
+    },
+    {
+      "epoch": 1.6854193249043035,
+      "grad_norm": 8.468046938996284,
+      "learning_rate": 2e-06,
+      "loss": 0.2017,
+      "step": 7265
+    },
+    {
+      "epoch": 1.6856513165526041,
+      "grad_norm": 9.632765936355588,
+      "learning_rate": 2e-06,
+      "loss": 0.2203,
+      "step": 7266
+    },
+    {
+      "epoch": 1.6858833082009048,
+      "grad_norm": 14.954336739425484,
+      "learning_rate": 2e-06,
+      "loss": 0.3368,
+      "step": 7267
+    },
+    {
+      "epoch": 1.6861152998492055,
+      "grad_norm": 11.151590329091023,
+      "learning_rate": 2e-06,
+      "loss": 0.1318,
+      "step": 7268
+    },
+    {
+      "epoch": 1.6863472914975062,
+      "grad_norm": 18.76079227324024,
+      "learning_rate": 2e-06,
+      "loss": 0.2761,
+      "step": 7269
+    },
+    {
+      "epoch": 1.6865792831458069,
+      "grad_norm": 17.605320132885076,
+      "learning_rate": 2e-06,
+      "loss": 0.2647,
+      "step": 7270
+    },
+    {
+      "epoch": 1.6868112747941075,
+      "grad_norm": 12.944711743448268,
+      "learning_rate": 2e-06,
+      "loss": 0.2539,
+      "step": 7271
+    },
+    {
+      "epoch": 1.6870432664424082,
+      "grad_norm": 123.24890746611254,
+      "learning_rate": 2e-06,
+      "loss": 0.2955,
+      "step": 7272
+    },
+    {
+      "epoch": 1.6872752580907089,
+      "grad_norm": 8.159596136022413,
+      "learning_rate": 2e-06,
+      "loss": 0.1544,
+      "step": 7273
+    },
+    {
+      "epoch": 1.6875072497390093,
+      "grad_norm": 6.68692543458059,
+      "learning_rate": 2e-06,
+      "loss": 0.2265,
+      "step": 7274
+    },
+    {
+      "epoch": 1.68773924138731,
+      "grad_norm": 13.505784951239923,
+      "learning_rate": 2e-06,
+      "loss": 0.2127,
+      "step": 7275
+    },
+    {
+      "epoch": 1.6879712330356107,
+      "grad_norm": 12.820996468185626,
+      "learning_rate": 2e-06,
+      "loss": 0.322,
+      "step": 7276
+    },
+    {
+      "epoch": 1.6882032246839114,
+      "grad_norm": 16.991615458156502,
+      "learning_rate": 2e-06,
+      "loss": 0.2802,
+      "step": 7277
+    },
+    {
+      "epoch": 1.688435216332212,
+      "grad_norm": 10.463780024145759,
+      "learning_rate": 2e-06,
+      "loss": 0.1726,
+      "step": 7278
+    },
+    {
+      "epoch": 1.6886672079805127,
+      "grad_norm": 16.589872313730154,
+      "learning_rate": 2e-06,
+      "loss": 0.288,
+      "step": 7279
+    },
+    {
+      "epoch": 1.6888991996288134,
+      "grad_norm": 9.827582196248413,
+      "learning_rate": 2e-06,
+      "loss": 0.1659,
+      "step": 7280
+    },
+    {
+      "epoch": 1.6891311912771139,
+      "grad_norm": 15.061570411083014,
+      "learning_rate": 2e-06,
+      "loss": 0.2588,
+      "step": 7281
+    },
+    {
+      "epoch": 1.6893631829254145,
+      "grad_norm": 14.836146002644297,
+      "learning_rate": 2e-06,
+      "loss": 0.1934,
+      "step": 7282
+    },
+    {
+      "epoch": 1.6895951745737152,
+      "grad_norm": 10.968410038414373,
+      "learning_rate": 2e-06,
+      "loss": 0.1331,
+      "step": 7283
+    },
+    {
+      "epoch": 1.689827166222016,
+      "grad_norm": 9.697378039295229,
+      "learning_rate": 2e-06,
+      "loss": 0.1898,
+      "step": 7284
+    },
+    {
+      "epoch": 1.6900591578703166,
+      "grad_norm": 9.26353684174463,
+      "learning_rate": 2e-06,
+      "loss": 0.2501,
+      "step": 7285
+    },
+    {
+      "epoch": 1.6902911495186173,
+      "grad_norm": 16.511560222732655,
+      "learning_rate": 2e-06,
+      "loss": 0.2632,
+      "step": 7286
+    },
+    {
+      "epoch": 1.690523141166918,
+      "grad_norm": 15.538008963116829,
+      "learning_rate": 2e-06,
+      "loss": 0.2322,
+      "step": 7287
+    },
+    {
+      "epoch": 1.6907551328152186,
+      "grad_norm": 6.702030928521217,
+      "learning_rate": 2e-06,
+      "loss": 0.1563,
+      "step": 7288
+    },
+    {
+      "epoch": 1.6909871244635193,
+      "grad_norm": 18.95424278469588,
+      "learning_rate": 2e-06,
+      "loss": 0.412,
+      "step": 7289
+    },
+    {
+      "epoch": 1.69121911611182,
+      "grad_norm": 19.402775836233555,
+      "learning_rate": 2e-06,
+      "loss": 0.3471,
+      "step": 7290
+    },
+    {
+      "epoch": 1.6914511077601206,
+      "grad_norm": 18.586344150422477,
+      "learning_rate": 2e-06,
+      "loss": 0.1725,
+      "step": 7291
+    },
+    {
+      "epoch": 1.6916830994084213,
+      "grad_norm": 13.803017788235675,
+      "learning_rate": 2e-06,
+      "loss": 0.2308,
+      "step": 7292
+    },
+    {
+      "epoch": 1.691915091056722,
+      "grad_norm": 12.289919019565339,
+      "learning_rate": 2e-06,
+      "loss": 0.2321,
+      "step": 7293
+    },
+    {
+      "epoch": 1.6921470827050227,
+      "grad_norm": 14.677095371271689,
+      "learning_rate": 2e-06,
+      "loss": 0.2651,
+      "step": 7294
+    },
+    {
+      "epoch": 1.6923790743533234,
+      "grad_norm": 15.267058990164466,
+      "learning_rate": 2e-06,
+      "loss": 0.2599,
+      "step": 7295
+    },
+    {
+      "epoch": 1.692611066001624,
+      "grad_norm": 10.819449552891342,
+      "learning_rate": 2e-06,
+      "loss": 0.2296,
+      "step": 7296
+    },
+    {
+      "epoch": 1.6928430576499247,
+      "grad_norm": 7.79293963930981,
+      "learning_rate": 2e-06,
+      "loss": 0.1925,
+      "step": 7297
+    },
+    {
+      "epoch": 1.6930750492982254,
+      "grad_norm": 12.106463229541392,
+      "learning_rate": 2e-06,
+      "loss": 0.2504,
+      "step": 7298
+    },
+    {
+      "epoch": 1.693307040946526,
+      "grad_norm": 20.23700952736958,
+      "learning_rate": 2e-06,
+      "loss": 0.3133,
+      "step": 7299
+    },
+    {
+      "epoch": 1.6935390325948267,
+      "grad_norm": 13.082483003263226,
+      "learning_rate": 2e-06,
+      "loss": 0.3206,
+      "step": 7300
+    },
+    {
+      "epoch": 1.6937710242431272,
+      "grad_norm": 5.561796986154681,
+      "learning_rate": 2e-06,
+      "loss": 0.1812,
+      "step": 7301
+    },
+    {
+      "epoch": 1.6940030158914279,
+      "grad_norm": 18.485603555962598,
+      "learning_rate": 2e-06,
+      "loss": 0.2279,
+      "step": 7302
+    },
+    {
+      "epoch": 1.6942350075397286,
+      "grad_norm": 19.852227953612164,
+      "learning_rate": 2e-06,
+      "loss": 0.4063,
+      "step": 7303
+    },
+    {
+      "epoch": 1.6944669991880292,
+      "grad_norm": 19.68949775248541,
+      "learning_rate": 2e-06,
+      "loss": 0.3253,
+      "step": 7304
+    },
+    {
+      "epoch": 1.69469899083633,
+      "grad_norm": 33.77052968630431,
+      "learning_rate": 2e-06,
+      "loss": 0.3797,
+      "step": 7305
+    },
+    {
+      "epoch": 1.6949309824846306,
+      "grad_norm": 11.94857037264836,
+      "learning_rate": 2e-06,
+      "loss": 0.209,
+      "step": 7306
+    },
+    {
+      "epoch": 1.6951629741329313,
+      "grad_norm": 13.804592380035896,
+      "learning_rate": 2e-06,
+      "loss": 0.2286,
+      "step": 7307
+    },
+    {
+      "epoch": 1.6953949657812317,
+      "grad_norm": 8.742072666215575,
+      "learning_rate": 2e-06,
+      "loss": 0.1553,
+      "step": 7308
+    },
+    {
+      "epoch": 1.6956269574295324,
+      "grad_norm": 11.872405377077234,
+      "learning_rate": 2e-06,
+      "loss": 0.2513,
+      "step": 7309
+    },
+    {
+      "epoch": 1.695858949077833,
+      "grad_norm": 8.341582411825478,
+      "learning_rate": 2e-06,
+      "loss": 0.1319,
+      "step": 7310
+    },
+    {
+      "epoch": 1.6960909407261338,
+      "grad_norm": 10.44053334267485,
+      "learning_rate": 2e-06,
+      "loss": 0.2116,
+      "step": 7311
+    },
+    {
+      "epoch": 1.6963229323744344,
+      "grad_norm": 17.881643539271504,
+      "learning_rate": 2e-06,
+      "loss": 0.2348,
+      "step": 7312
+    },
+    {
+      "epoch": 1.6965549240227351,
+      "grad_norm": 19.973000179858243,
+      "learning_rate": 2e-06,
+      "loss": 0.3227,
+      "step": 7313
+    },
+    {
+      "epoch": 1.6967869156710358,
+      "grad_norm": 25.58664111962267,
+      "learning_rate": 2e-06,
+      "loss": 0.3412,
+      "step": 7314
+    },
+    {
+      "epoch": 1.6970189073193365,
+      "grad_norm": 16.170402376477018,
+      "learning_rate": 2e-06,
+      "loss": 0.2891,
+      "step": 7315
+    },
+    {
+      "epoch": 1.6972508989676371,
+      "grad_norm": 25.391042103179647,
+      "learning_rate": 2e-06,
+      "loss": 0.4034,
+      "step": 7316
+    },
+    {
+      "epoch": 1.6974828906159378,
+      "grad_norm": 15.631652628735425,
+      "learning_rate": 2e-06,
+      "loss": 0.2334,
+      "step": 7317
+    },
+    {
+      "epoch": 1.6977148822642385,
+      "grad_norm": 15.931573085004318,
+      "learning_rate": 2e-06,
+      "loss": 0.3106,
+      "step": 7318
+    },
+    {
+      "epoch": 1.6979468739125392,
+      "grad_norm": 19.12347367874556,
+      "learning_rate": 2e-06,
+      "loss": 0.1906,
+      "step": 7319
+    },
+    {
+      "epoch": 1.6981788655608399,
+      "grad_norm": 15.569423556959888,
+      "learning_rate": 2e-06,
+      "loss": 0.264,
+      "step": 7320
+    },
+    {
+      "epoch": 1.6984108572091405,
+      "grad_norm": 14.068857339247876,
+      "learning_rate": 2e-06,
+      "loss": 0.237,
+      "step": 7321
+    },
+    {
+      "epoch": 1.6986428488574412,
+      "grad_norm": 7.4346690593116,
+      "learning_rate": 2e-06,
+      "loss": 0.2656,
+      "step": 7322
+    },
+    {
+      "epoch": 1.698874840505742,
+      "grad_norm": 13.320835738404085,
+      "learning_rate": 2e-06,
+      "loss": 0.2022,
+      "step": 7323
+    },
+    {
+      "epoch": 1.6991068321540426,
+      "grad_norm": 16.618587408159758,
+      "learning_rate": 2e-06,
+      "loss": 0.3168,
+      "step": 7324
+    },
+    {
+      "epoch": 1.6993388238023432,
+      "grad_norm": 14.77136574060571,
+      "learning_rate": 2e-06,
+      "loss": 0.1647,
+      "step": 7325
+    },
+    {
+      "epoch": 1.699570815450644,
+      "grad_norm": 14.808907885630699,
+      "learning_rate": 2e-06,
+      "loss": 0.2321,
+      "step": 7326
+    },
+    {
+      "epoch": 1.6998028070989446,
+      "grad_norm": 11.36970811881377,
+      "learning_rate": 2e-06,
+      "loss": 0.2535,
+      "step": 7327
+    },
+    {
+      "epoch": 1.700034798747245,
+      "grad_norm": 11.569639321117595,
+      "learning_rate": 2e-06,
+      "loss": 0.2157,
+      "step": 7328
+    },
+    {
+      "epoch": 1.7002667903955457,
+      "grad_norm": 13.531970644424788,
+      "learning_rate": 2e-06,
+      "loss": 0.154,
+      "step": 7329
+    },
+    {
+      "epoch": 1.7004987820438464,
+      "grad_norm": 8.423003530239141,
+      "learning_rate": 2e-06,
+      "loss": 0.2078,
+      "step": 7330
+    },
+    {
+      "epoch": 1.700730773692147,
+      "grad_norm": 16.232057245838032,
+      "learning_rate": 2e-06,
+      "loss": 0.1807,
+      "step": 7331
+    },
+    {
+      "epoch": 1.7009627653404478,
+      "grad_norm": 12.855703913990952,
+      "learning_rate": 2e-06,
+      "loss": 0.2202,
+      "step": 7332
+    },
+    {
+      "epoch": 1.7011947569887484,
+      "grad_norm": 12.417491377191617,
+      "learning_rate": 2e-06,
+      "loss": 0.2327,
+      "step": 7333
+    },
+    {
+      "epoch": 1.701426748637049,
+      "grad_norm": 14.984586298988479,
+      "learning_rate": 2e-06,
+      "loss": 0.2119,
+      "step": 7334
+    },
+    {
+      "epoch": 1.7016587402853496,
+      "grad_norm": 18.66081159209174,
+      "learning_rate": 2e-06,
+      "loss": 0.3525,
+      "step": 7335
+    },
+    {
+      "epoch": 1.7018907319336503,
+      "grad_norm": 9.688077039211272,
+      "learning_rate": 2e-06,
+      "loss": 0.1497,
+      "step": 7336
+    },
+    {
+      "epoch": 1.702122723581951,
+      "grad_norm": 7.287356424854079,
+      "learning_rate": 2e-06,
+      "loss": 0.1793,
+      "step": 7337
+    },
+    {
+      "epoch": 1.7023547152302516,
+      "grad_norm": 8.409087290247287,
+      "learning_rate": 2e-06,
+      "loss": 0.2413,
+      "step": 7338
+    },
+    {
+      "epoch": 1.7025867068785523,
+      "grad_norm": 13.801898488863271,
+      "learning_rate": 2e-06,
+      "loss": 0.2279,
+      "step": 7339
+    },
+    {
+      "epoch": 1.702818698526853,
+      "grad_norm": 10.625966492141034,
+      "learning_rate": 2e-06,
+      "loss": 0.1772,
+      "step": 7340
+    },
+    {
+      "epoch": 1.7030506901751536,
+      "grad_norm": 8.734020582081474,
+      "learning_rate": 2e-06,
+      "loss": 0.2377,
+      "step": 7341
+    },
+    {
+      "epoch": 1.7032826818234543,
+      "grad_norm": 15.367119881801106,
+      "learning_rate": 2e-06,
+      "loss": 0.2827,
+      "step": 7342
+    },
+    {
+      "epoch": 1.703514673471755,
+      "grad_norm": 9.132522203727172,
+      "learning_rate": 2e-06,
+      "loss": 0.1569,
+      "step": 7343
+    },
+    {
+      "epoch": 1.7037466651200557,
+      "grad_norm": 26.578806454786072,
+      "learning_rate": 2e-06,
+      "loss": 0.3061,
+      "step": 7344
+    },
+    {
+      "epoch": 1.7039786567683564,
+      "grad_norm": 13.623761355111489,
+      "learning_rate": 2e-06,
+      "loss": 0.2544,
+      "step": 7345
+    },
+    {
+      "epoch": 1.704210648416657,
+      "grad_norm": 10.402184063489008,
+      "learning_rate": 2e-06,
+      "loss": 0.1433,
+      "step": 7346
+    },
+    {
+      "epoch": 1.7044426400649577,
+      "grad_norm": 14.848757688205776,
+      "learning_rate": 2e-06,
+      "loss": 0.2865,
+      "step": 7347
+    },
+    {
+      "epoch": 1.7046746317132584,
+      "grad_norm": 10.388087104158064,
+      "learning_rate": 2e-06,
+      "loss": 0.2016,
+      "step": 7348
+    },
+    {
+      "epoch": 1.704906623361559,
+      "grad_norm": 10.690069785947735,
+      "learning_rate": 2e-06,
+      "loss": 0.251,
+      "step": 7349
+    },
+    {
+      "epoch": 1.7051386150098597,
+      "grad_norm": 12.286173589369547,
+      "learning_rate": 2e-06,
+      "loss": 0.193,
+      "step": 7350
+    },
+    {
+      "epoch": 1.7053706066581604,
+      "grad_norm": 18.118306711239853,
+      "learning_rate": 2e-06,
+      "loss": 0.2965,
+      "step": 7351
+    },
+    {
+      "epoch": 1.705602598306461,
+      "grad_norm": 17.74489505862405,
+      "learning_rate": 2e-06,
+      "loss": 0.4117,
+      "step": 7352
+    },
+    {
+      "epoch": 1.7058345899547618,
+      "grad_norm": 14.440055676789756,
+      "learning_rate": 2e-06,
+      "loss": 0.1778,
+      "step": 7353
+    },
+    {
+      "epoch": 1.7060665816030622,
+      "grad_norm": 21.136004075754848,
+      "learning_rate": 2e-06,
+      "loss": 0.2744,
+      "step": 7354
+    },
+    {
+      "epoch": 1.706298573251363,
+      "grad_norm": 15.881275091214922,
+      "learning_rate": 2e-06,
+      "loss": 0.2271,
+      "step": 7355
+    },
+    {
+      "epoch": 1.7065305648996636,
+      "grad_norm": 7.7350361481897725,
+      "learning_rate": 2e-06,
+      "loss": 0.1523,
+      "step": 7356
+    },
+    {
+      "epoch": 1.7067625565479643,
+      "grad_norm": 10.3562729510965,
+      "learning_rate": 2e-06,
+      "loss": 0.3039,
+      "step": 7357
+    },
+    {
+      "epoch": 1.706994548196265,
+      "grad_norm": 15.919554632350538,
+      "learning_rate": 2e-06,
+      "loss": 0.2601,
+      "step": 7358
+    },
+    {
+      "epoch": 1.7072265398445656,
+      "grad_norm": 6.907861083753691,
+      "learning_rate": 2e-06,
+      "loss": 0.1562,
+      "step": 7359
+    },
+    {
+      "epoch": 1.7074585314928663,
+      "grad_norm": 13.332375865894466,
+      "learning_rate": 2e-06,
+      "loss": 0.1999,
+      "step": 7360
+    },
+    {
+      "epoch": 1.7076905231411668,
+      "grad_norm": 9.13285577347624,
+      "learning_rate": 2e-06,
+      "loss": 0.2235,
+      "step": 7361
+    },
+    {
+      "epoch": 1.7079225147894674,
+      "grad_norm": 14.410625044515026,
+      "learning_rate": 2e-06,
+      "loss": 0.3811,
+      "step": 7362
+    },
+    {
+      "epoch": 1.7081545064377681,
+      "grad_norm": 15.474633868476383,
+      "learning_rate": 2e-06,
+      "loss": 0.399,
+      "step": 7363
+    },
+    {
+      "epoch": 1.7083864980860688,
+      "grad_norm": 8.161414052092125,
+      "learning_rate": 2e-06,
+      "loss": 0.1485,
+      "step": 7364
+    },
+    {
+      "epoch": 1.7086184897343695,
+      "grad_norm": 7.173334004364632,
+      "learning_rate": 2e-06,
+      "loss": 0.2258,
+      "step": 7365
+    },
+    {
+      "epoch": 1.7088504813826702,
+      "grad_norm": 15.427045346489873,
+      "learning_rate": 2e-06,
+      "loss": 0.2198,
+      "step": 7366
+    },
+    {
+      "epoch": 1.7090824730309708,
+      "grad_norm": 12.765297230763718,
+      "learning_rate": 2e-06,
+      "loss": 0.2092,
+      "step": 7367
+    },
+    {
+      "epoch": 1.7093144646792715,
+      "grad_norm": 10.360717945103982,
+      "learning_rate": 2e-06,
+      "loss": 0.2371,
+      "step": 7368
+    },
+    {
+      "epoch": 1.7095464563275722,
+      "grad_norm": 11.163692771886856,
+      "learning_rate": 2e-06,
+      "loss": 0.2144,
+      "step": 7369
+    },
+    {
+      "epoch": 1.7097784479758729,
+      "grad_norm": 9.414971675169285,
+      "learning_rate": 2e-06,
+      "loss": 0.2496,
+      "step": 7370
+    },
+    {
+      "epoch": 1.7100104396241735,
+      "grad_norm": 12.427623155406073,
+      "learning_rate": 2e-06,
+      "loss": 0.2593,
+      "step": 7371
+    },
+    {
+      "epoch": 1.7102424312724742,
+      "grad_norm": 9.993736030250428,
+      "learning_rate": 2e-06,
+      "loss": 0.2196,
+      "step": 7372
+    },
+    {
+      "epoch": 1.710474422920775,
+      "grad_norm": 12.357927783611151,
+      "learning_rate": 2e-06,
+      "loss": 0.2762,
+      "step": 7373
+    },
+    {
+      "epoch": 1.7107064145690756,
+      "grad_norm": 17.258664349758774,
+      "learning_rate": 2e-06,
+      "loss": 0.3045,
+      "step": 7374
+    },
+    {
+      "epoch": 1.7109384062173763,
+      "grad_norm": 7.514361257025172,
+      "learning_rate": 2e-06,
+      "loss": 0.1494,
+      "step": 7375
+    },
+    {
+      "epoch": 1.711170397865677,
+      "grad_norm": 21.462401133891056,
+      "learning_rate": 2e-06,
+      "loss": 0.3879,
+      "step": 7376
+    },
+    {
+      "epoch": 1.7114023895139776,
+      "grad_norm": 14.017574262579139,
+      "learning_rate": 2e-06,
+      "loss": 0.2501,
+      "step": 7377
+    },
+    {
+      "epoch": 1.7116343811622783,
+      "grad_norm": 59.43227788764268,
+      "learning_rate": 2e-06,
+      "loss": 0.1805,
+      "step": 7378
+    },
+    {
+      "epoch": 1.711866372810579,
+      "grad_norm": 13.789051712709842,
+      "learning_rate": 2e-06,
+      "loss": 0.2946,
+      "step": 7379
+    },
+    {
+      "epoch": 1.7120983644588796,
+      "grad_norm": 11.672587587978251,
+      "learning_rate": 2e-06,
+      "loss": 0.2156,
+      "step": 7380
+    },
+    {
+      "epoch": 1.71233035610718,
+      "grad_norm": 21.34434420615057,
+      "learning_rate": 2e-06,
+      "loss": 0.325,
+      "step": 7381
+    },
+    {
+      "epoch": 1.7125623477554808,
+      "grad_norm": 16.615661247788037,
+      "learning_rate": 2e-06,
+      "loss": 0.2408,
+      "step": 7382
+    },
+    {
+      "epoch": 1.7127943394037815,
+      "grad_norm": 20.75326511437421,
+      "learning_rate": 2e-06,
+      "loss": 0.3114,
+      "step": 7383
+    },
+    {
+      "epoch": 1.7130263310520821,
+      "grad_norm": 12.286557692400283,
+      "learning_rate": 2e-06,
+      "loss": 0.2356,
+      "step": 7384
+    },
+    {
+      "epoch": 1.7132583227003828,
+      "grad_norm": 12.174885841243048,
+      "learning_rate": 2e-06,
+      "loss": 0.223,
+      "step": 7385
+    },
+    {
+      "epoch": 1.7134903143486835,
+      "grad_norm": 14.877687974756407,
+      "learning_rate": 2e-06,
+      "loss": 0.3443,
+      "step": 7386
+    },
+    {
+      "epoch": 1.7137223059969842,
+      "grad_norm": 17.07574828854014,
+      "learning_rate": 2e-06,
+      "loss": 0.3792,
+      "step": 7387
+    },
+    {
+      "epoch": 1.7139542976452846,
+      "grad_norm": 13.713405100948842,
+      "learning_rate": 2e-06,
+      "loss": 0.2943,
+      "step": 7388
+    },
+    {
+      "epoch": 1.7141862892935853,
+      "grad_norm": 10.06477622586005,
+      "learning_rate": 2e-06,
+      "loss": 0.2264,
+      "step": 7389
+    },
+    {
+      "epoch": 1.714418280941886,
+      "grad_norm": 11.263272476317919,
+      "learning_rate": 2e-06,
+      "loss": 0.2381,
+      "step": 7390
+    },
+    {
+      "epoch": 1.7146502725901867,
+      "grad_norm": 19.508196318478443,
+      "learning_rate": 2e-06,
+      "loss": 0.3064,
+      "step": 7391
+    },
+    {
+      "epoch": 1.7148822642384873,
+      "grad_norm": 11.261598830547625,
+      "learning_rate": 2e-06,
+      "loss": 0.2087,
+      "step": 7392
+    },
+    {
+      "epoch": 1.715114255886788,
+      "grad_norm": 12.405569223219322,
+      "learning_rate": 2e-06,
+      "loss": 0.3194,
+      "step": 7393
+    },
+    {
+      "epoch": 1.7153462475350887,
+      "grad_norm": 16.233821077133854,
+      "learning_rate": 2e-06,
+      "loss": 0.2638,
+      "step": 7394
+    },
+    {
+      "epoch": 1.7155782391833894,
+      "grad_norm": 18.979216214316168,
+      "learning_rate": 2e-06,
+      "loss": 0.2379,
+      "step": 7395
+    },
+    {
+      "epoch": 1.71581023083169,
+      "grad_norm": 12.002055700439051,
+      "learning_rate": 2e-06,
+      "loss": 0.2695,
+      "step": 7396
+    },
+    {
+      "epoch": 1.7160422224799907,
+      "grad_norm": 7.190049386433045,
+      "learning_rate": 2e-06,
+      "loss": 0.1919,
+      "step": 7397
+    },
+    {
+      "epoch": 1.7162742141282914,
+      "grad_norm": 9.293922544406554,
+      "learning_rate": 2e-06,
+      "loss": 0.1739,
+      "step": 7398
+    },
+    {
+      "epoch": 1.716506205776592,
+      "grad_norm": 19.113098476282847,
+      "learning_rate": 2e-06,
+      "loss": 0.2361,
+      "step": 7399
+    },
+    {
+      "epoch": 1.7167381974248928,
+      "grad_norm": 21.822492041683304,
+      "learning_rate": 2e-06,
+      "loss": 0.3029,
+      "step": 7400
+    },
+    {
+      "epoch": 1.7169701890731934,
+      "grad_norm": 13.930659725168429,
+      "learning_rate": 2e-06,
+      "loss": 0.263,
+      "step": 7401
+    },
+    {
+      "epoch": 1.717202180721494,
+      "grad_norm": 14.905559993741406,
+      "learning_rate": 2e-06,
+      "loss": 0.2522,
+      "step": 7402
+    },
+    {
+      "epoch": 1.7174341723697948,
+      "grad_norm": 11.164819303552916,
+      "learning_rate": 2e-06,
+      "loss": 0.2697,
+      "step": 7403
+    },
+    {
+      "epoch": 1.7176661640180955,
+      "grad_norm": 8.91142886247981,
+      "learning_rate": 2e-06,
+      "loss": 0.115,
+      "step": 7404
+    },
+    {
+      "epoch": 1.7178981556663961,
+      "grad_norm": 7.949290065036741,
+      "learning_rate": 2e-06,
+      "loss": 0.2455,
+      "step": 7405
+    },
+    {
+      "epoch": 1.7181301473146968,
+      "grad_norm": 10.785661327154305,
+      "learning_rate": 2e-06,
+      "loss": 0.1592,
+      "step": 7406
+    },
+    {
+      "epoch": 1.7183621389629975,
+      "grad_norm": 16.361965541625413,
+      "learning_rate": 2e-06,
+      "loss": 0.2323,
+      "step": 7407
+    },
+    {
+      "epoch": 1.718594130611298,
+      "grad_norm": 19.92901371079172,
+      "learning_rate": 2e-06,
+      "loss": 0.3203,
+      "step": 7408
+    },
+    {
+      "epoch": 1.7188261222595986,
+      "grad_norm": 9.312777707144642,
+      "learning_rate": 2e-06,
+      "loss": 0.1555,
+      "step": 7409
+    },
+    {
+      "epoch": 1.7190581139078993,
+      "grad_norm": 6.883164003465984,
+      "learning_rate": 2e-06,
+      "loss": 0.1613,
+      "step": 7410
+    },
+    {
+      "epoch": 1.7192901055562,
+      "grad_norm": 7.625503436868352,
+      "learning_rate": 2e-06,
+      "loss": 0.233,
+      "step": 7411
+    },
+    {
+      "epoch": 1.7195220972045007,
+      "grad_norm": 8.824659120759403,
+      "learning_rate": 2e-06,
+      "loss": 0.2014,
+      "step": 7412
+    },
+    {
+      "epoch": 1.7197540888528013,
+      "grad_norm": 6.026256965073181,
+      "learning_rate": 2e-06,
+      "loss": 0.187,
+      "step": 7413
+    },
+    {
+      "epoch": 1.7199860805011018,
+      "grad_norm": 13.585052566414436,
+      "learning_rate": 2e-06,
+      "loss": 0.2389,
+      "step": 7414
+    },
+    {
+      "epoch": 1.7202180721494025,
+      "grad_norm": 20.721352563891614,
+      "learning_rate": 2e-06,
+      "loss": 0.3369,
+      "step": 7415
+    },
+    {
+      "epoch": 1.7204500637977032,
+      "grad_norm": 17.802194293608064,
+      "learning_rate": 2e-06,
+      "loss": 0.291,
+      "step": 7416
+    },
+    {
+      "epoch": 1.7206820554460038,
+      "grad_norm": 12.616286867089714,
+      "learning_rate": 2e-06,
+      "loss": 0.1882,
+      "step": 7417
+    },
+    {
+      "epoch": 1.7209140470943045,
+      "grad_norm": 9.911712897105605,
+      "learning_rate": 2e-06,
+      "loss": 0.2234,
+      "step": 7418
+    },
+    {
+      "epoch": 1.7211460387426052,
+      "grad_norm": 16.06779518312088,
+      "learning_rate": 2e-06,
+      "loss": 0.2163,
+      "step": 7419
+    },
+    {
+      "epoch": 1.7213780303909059,
+      "grad_norm": 14.026915838395407,
+      "learning_rate": 2e-06,
+      "loss": 0.2335,
+      "step": 7420
+    },
+    {
+      "epoch": 1.7216100220392065,
+      "grad_norm": 16.03652589750512,
+      "learning_rate": 2e-06,
+      "loss": 0.2006,
+      "step": 7421
+    },
+    {
+      "epoch": 1.7218420136875072,
+      "grad_norm": 14.123340674487144,
+      "learning_rate": 2e-06,
+      "loss": 0.2975,
+      "step": 7422
+    },
+    {
+      "epoch": 1.722074005335808,
+      "grad_norm": 9.61097211651846,
+      "learning_rate": 2e-06,
+      "loss": 0.1643,
+      "step": 7423
+    },
+    {
+      "epoch": 1.7223059969841086,
+      "grad_norm": 6.260522196319805,
+      "learning_rate": 2e-06,
+      "loss": 0.1685,
+      "step": 7424
+    },
+    {
+      "epoch": 1.7225379886324093,
+      "grad_norm": 12.096547938872812,
+      "learning_rate": 2e-06,
+      "loss": 0.2311,
+      "step": 7425
+    },
+    {
+      "epoch": 1.72276998028071,
+      "grad_norm": 10.742362818573552,
+      "learning_rate": 2e-06,
+      "loss": 0.1912,
+      "step": 7426
+    },
+    {
+      "epoch": 1.7230019719290106,
+      "grad_norm": 9.979446739855348,
+      "learning_rate": 2e-06,
+      "loss": 0.2561,
+      "step": 7427
+    },
+    {
+      "epoch": 1.7232339635773113,
+      "grad_norm": 12.982352110270831,
+      "learning_rate": 2e-06,
+      "loss": 0.246,
+      "step": 7428
+    },
+    {
+      "epoch": 1.723465955225612,
+      "grad_norm": 10.51719727420884,
+      "learning_rate": 2e-06,
+      "loss": 0.2248,
+      "step": 7429
+    },
+    {
+      "epoch": 1.7236979468739126,
+      "grad_norm": 19.54035939653835,
+      "learning_rate": 2e-06,
+      "loss": 0.2248,
+      "step": 7430
+    },
+    {
+      "epoch": 1.7239299385222133,
+      "grad_norm": 10.282834840940072,
+      "learning_rate": 2e-06,
+      "loss": 0.1643,
+      "step": 7431
+    },
+    {
+      "epoch": 1.724161930170514,
+      "grad_norm": 15.72589351752963,
+      "learning_rate": 2e-06,
+      "loss": 0.3086,
+      "step": 7432
+    },
+    {
+      "epoch": 1.7243939218188147,
+      "grad_norm": 11.020603461535057,
+      "learning_rate": 2e-06,
+      "loss": 0.1795,
+      "step": 7433
+    },
+    {
+      "epoch": 1.7246259134671151,
+      "grad_norm": 9.55280925000355,
+      "learning_rate": 2e-06,
+      "loss": 0.2475,
+      "step": 7434
+    },
+    {
+      "epoch": 1.7248579051154158,
+      "grad_norm": 12.311849057297254,
+      "learning_rate": 2e-06,
+      "loss": 0.1943,
+      "step": 7435
+    },
+    {
+      "epoch": 1.7250898967637165,
+      "grad_norm": 8.677933835714159,
+      "learning_rate": 2e-06,
+      "loss": 0.1278,
+      "step": 7436
+    },
+    {
+      "epoch": 1.7253218884120172,
+      "grad_norm": 9.135809044634728,
+      "learning_rate": 2e-06,
+      "loss": 0.1905,
+      "step": 7437
+    },
+    {
+      "epoch": 1.7255538800603178,
+      "grad_norm": 12.128952282925505,
+      "learning_rate": 2e-06,
+      "loss": 0.2191,
+      "step": 7438
+    },
+    {
+      "epoch": 1.7257858717086185,
+      "grad_norm": 11.620391971320371,
+      "learning_rate": 2e-06,
+      "loss": 0.3023,
+      "step": 7439
+    },
+    {
+      "epoch": 1.7260178633569192,
+      "grad_norm": 7.742562300737355,
+      "learning_rate": 2e-06,
+      "loss": 0.2002,
+      "step": 7440
+    },
+    {
+      "epoch": 1.7262498550052197,
+      "grad_norm": 11.483373680485649,
+      "learning_rate": 2e-06,
+      "loss": 0.3349,
+      "step": 7441
+    },
+    {
+      "epoch": 1.7264818466535203,
+      "grad_norm": 14.756476462309388,
+      "learning_rate": 2e-06,
+      "loss": 0.2754,
+      "step": 7442
+    },
+    {
+      "epoch": 1.726713838301821,
+      "grad_norm": 24.504072386189904,
+      "learning_rate": 2e-06,
+      "loss": 0.3742,
+      "step": 7443
+    },
+    {
+      "epoch": 1.7269458299501217,
+      "grad_norm": 12.707665192584258,
+      "learning_rate": 2e-06,
+      "loss": 0.225,
+      "step": 7444
+    },
+    {
+      "epoch": 1.7271778215984224,
+      "grad_norm": 25.36059367282093,
+      "learning_rate": 2e-06,
+      "loss": 0.2895,
+      "step": 7445
+    },
+    {
+      "epoch": 1.727409813246723,
+      "grad_norm": 18.671781461183837,
+      "learning_rate": 2e-06,
+      "loss": 0.3407,
+      "step": 7446
+    },
+    {
+      "epoch": 1.7276418048950237,
+      "grad_norm": 15.746357771325753,
+      "learning_rate": 2e-06,
+      "loss": 0.2162,
+      "step": 7447
+    },
+    {
+      "epoch": 1.7278737965433244,
+      "grad_norm": 8.535603643460004,
+      "learning_rate": 2e-06,
+      "loss": 0.2024,
+      "step": 7448
+    },
+    {
+      "epoch": 1.728105788191625,
+      "grad_norm": 16.43628019510447,
+      "learning_rate": 2e-06,
+      "loss": 0.351,
+      "step": 7449
+    },
+    {
+      "epoch": 1.7283377798399258,
+      "grad_norm": 10.001938065434366,
+      "learning_rate": 2e-06,
+      "loss": 0.2177,
+      "step": 7450
+    },
+    {
+      "epoch": 1.7285697714882264,
+      "grad_norm": 12.779681705176559,
+      "learning_rate": 2e-06,
+      "loss": 0.1771,
+      "step": 7451
+    },
+    {
+      "epoch": 1.7288017631365271,
+      "grad_norm": 13.328753843002087,
+      "learning_rate": 2e-06,
+      "loss": 0.1639,
+      "step": 7452
+    },
+    {
+      "epoch": 1.7290337547848278,
+      "grad_norm": 13.858957106187253,
+      "learning_rate": 2e-06,
+      "loss": 0.2973,
+      "step": 7453
+    },
+    {
+      "epoch": 1.7292657464331285,
+      "grad_norm": 14.723632245449247,
+      "learning_rate": 2e-06,
+      "loss": 0.2721,
+      "step": 7454
+    },
+    {
+      "epoch": 1.7294977380814291,
+      "grad_norm": 11.906681071889423,
+      "learning_rate": 2e-06,
+      "loss": 0.2087,
+      "step": 7455
+    },
+    {
+      "epoch": 1.7297297297297298,
+      "grad_norm": 7.628253623661381,
+      "learning_rate": 2e-06,
+      "loss": 0.1461,
+      "step": 7456
+    },
+    {
+      "epoch": 1.7299617213780305,
+      "grad_norm": 13.74681104057977,
+      "learning_rate": 2e-06,
+      "loss": 0.2663,
+      "step": 7457
+    },
+    {
+      "epoch": 1.7301937130263312,
+      "grad_norm": 8.654851282645897,
+      "learning_rate": 2e-06,
+      "loss": 0.2446,
+      "step": 7458
+    },
+    {
+      "epoch": 1.7304257046746319,
+      "grad_norm": 11.135571891568727,
+      "learning_rate": 2e-06,
+      "loss": 0.2514,
+      "step": 7459
+    },
+    {
+      "epoch": 1.7306576963229325,
+      "grad_norm": 23.09441910446493,
+      "learning_rate": 2e-06,
+      "loss": 0.3025,
+      "step": 7460
+    },
+    {
+      "epoch": 1.730889687971233,
+      "grad_norm": 7.408623355561432,
+      "learning_rate": 2e-06,
+      "loss": 0.2236,
+      "step": 7461
+    },
+    {
+      "epoch": 1.7311216796195337,
+      "grad_norm": 14.021068585561615,
+      "learning_rate": 2e-06,
+      "loss": 0.2338,
+      "step": 7462
+    },
+    {
+      "epoch": 1.7313536712678343,
+      "grad_norm": 13.492300579390124,
+      "learning_rate": 2e-06,
+      "loss": 0.2432,
+      "step": 7463
+    },
+    {
+      "epoch": 1.731585662916135,
+      "grad_norm": 9.888204097854688,
+      "learning_rate": 2e-06,
+      "loss": 0.264,
+      "step": 7464
+    },
+    {
+      "epoch": 1.7318176545644357,
+      "grad_norm": 15.78820383154411,
+      "learning_rate": 2e-06,
+      "loss": 0.2384,
+      "step": 7465
+    },
+    {
+      "epoch": 1.7320496462127364,
+      "grad_norm": 9.480461865651494,
+      "learning_rate": 2e-06,
+      "loss": 0.2361,
+      "step": 7466
+    },
+    {
+      "epoch": 1.7322816378610368,
+      "grad_norm": 10.41053618967658,
+      "learning_rate": 2e-06,
+      "loss": 0.2749,
+      "step": 7467
+    },
+    {
+      "epoch": 1.7325136295093375,
+      "grad_norm": 12.479664664280824,
+      "learning_rate": 2e-06,
+      "loss": 0.2908,
+      "step": 7468
+    },
+    {
+      "epoch": 1.7327456211576382,
+      "grad_norm": 15.79738817329508,
+      "learning_rate": 2e-06,
+      "loss": 0.2914,
+      "step": 7469
+    },
+    {
+      "epoch": 1.7329776128059389,
+      "grad_norm": 12.33172355967031,
+      "learning_rate": 2e-06,
+      "loss": 0.2325,
+      "step": 7470
+    },
+    {
+      "epoch": 1.7332096044542395,
+      "grad_norm": 24.294313092336242,
+      "learning_rate": 2e-06,
+      "loss": 0.319,
+      "step": 7471
+    },
+    {
+      "epoch": 1.7334415961025402,
+      "grad_norm": 25.220141346783684,
+      "learning_rate": 2e-06,
+      "loss": 0.3333,
+      "step": 7472
+    },
+    {
+      "epoch": 1.733673587750841,
+      "grad_norm": 18.63395457662261,
+      "learning_rate": 2e-06,
+      "loss": 0.2178,
+      "step": 7473
+    },
+    {
+      "epoch": 1.7339055793991416,
+      "grad_norm": 15.239283398729846,
+      "learning_rate": 2e-06,
+      "loss": 0.2145,
+      "step": 7474
+    },
+    {
+      "epoch": 1.7341375710474423,
+      "grad_norm": 11.652366727377803,
+      "learning_rate": 2e-06,
+      "loss": 0.2277,
+      "step": 7475
+    },
+    {
+      "epoch": 1.734369562695743,
+      "grad_norm": 11.646906456604956,
+      "learning_rate": 2e-06,
+      "loss": 0.2013,
+      "step": 7476
+    },
+    {
+      "epoch": 1.7346015543440436,
+      "grad_norm": 8.637249927622804,
+      "learning_rate": 2e-06,
+      "loss": 0.1697,
+      "step": 7477
+    },
+    {
+      "epoch": 1.7348335459923443,
+      "grad_norm": 11.990032300591745,
+      "learning_rate": 2e-06,
+      "loss": 0.1589,
+      "step": 7478
+    },
+    {
+      "epoch": 1.735065537640645,
+      "grad_norm": 10.945905877991185,
+      "learning_rate": 2e-06,
+      "loss": 0.2319,
+      "step": 7479
+    },
+    {
+      "epoch": 1.7352975292889457,
+      "grad_norm": 7.920506695800604,
+      "learning_rate": 2e-06,
+      "loss": 0.2008,
+      "step": 7480
+    },
+    {
+      "epoch": 1.7355295209372463,
+      "grad_norm": 8.339878093891347,
+      "learning_rate": 2e-06,
+      "loss": 0.1503,
+      "step": 7481
+    },
+    {
+      "epoch": 1.735761512585547,
+      "grad_norm": 15.023425770672432,
+      "learning_rate": 2e-06,
+      "loss": 0.3408,
+      "step": 7482
+    },
+    {
+      "epoch": 1.7359935042338477,
+      "grad_norm": 9.377761857893933,
+      "learning_rate": 2e-06,
+      "loss": 0.2254,
+      "step": 7483
+    },
+    {
+      "epoch": 1.7362254958821484,
+      "grad_norm": 10.184995366885822,
+      "learning_rate": 2e-06,
+      "loss": 0.1921,
+      "step": 7484
+    },
+    {
+      "epoch": 1.736457487530449,
+      "grad_norm": 13.578407661309907,
+      "learning_rate": 2e-06,
+      "loss": 0.2303,
+      "step": 7485
+    },
+    {
+      "epoch": 1.7366894791787497,
+      "grad_norm": 7.356389683470666,
+      "learning_rate": 2e-06,
+      "loss": 0.2637,
+      "step": 7486
+    },
+    {
+      "epoch": 1.7369214708270502,
+      "grad_norm": 24.566412216006192,
+      "learning_rate": 2e-06,
+      "loss": 0.2539,
+      "step": 7487
+    },
+    {
+      "epoch": 1.7371534624753509,
+      "grad_norm": 11.865767307057718,
+      "learning_rate": 2e-06,
+      "loss": 0.2375,
+      "step": 7488
+    },
+    {
+      "epoch": 1.7373854541236515,
+      "grad_norm": 9.212238196866018,
+      "learning_rate": 2e-06,
+      "loss": 0.1615,
+      "step": 7489
+    },
+    {
+      "epoch": 1.7376174457719522,
+      "grad_norm": 11.28651869495539,
+      "learning_rate": 2e-06,
+      "loss": 0.2427,
+      "step": 7490
+    },
+    {
+      "epoch": 1.7378494374202529,
+      "grad_norm": 16.932290711659782,
+      "learning_rate": 2e-06,
+      "loss": 0.2682,
+      "step": 7491
+    },
+    {
+      "epoch": 1.7380814290685536,
+      "grad_norm": 7.522671677144873,
+      "learning_rate": 2e-06,
+      "loss": 0.1405,
+      "step": 7492
+    },
+    {
+      "epoch": 1.7383134207168542,
+      "grad_norm": 10.36334726849435,
+      "learning_rate": 2e-06,
+      "loss": 0.2022,
+      "step": 7493
+    },
+    {
+      "epoch": 1.7385454123651547,
+      "grad_norm": 10.670762139411545,
+      "learning_rate": 2e-06,
+      "loss": 0.1941,
+      "step": 7494
+    },
+    {
+      "epoch": 1.7387774040134554,
+      "grad_norm": 13.315131202322673,
+      "learning_rate": 2e-06,
+      "loss": 0.1852,
+      "step": 7495
+    },
+    {
+      "epoch": 1.739009395661756,
+      "grad_norm": 13.907336275113536,
+      "learning_rate": 2e-06,
+      "loss": 0.216,
+      "step": 7496
+    },
+    {
+      "epoch": 1.7392413873100567,
+      "grad_norm": 13.021695949675474,
+      "learning_rate": 2e-06,
+      "loss": 0.3078,
+      "step": 7497
+    },
+    {
+      "epoch": 1.7394733789583574,
+      "grad_norm": 13.732802454687654,
+      "learning_rate": 2e-06,
+      "loss": 0.1771,
+      "step": 7498
+    },
+    {
+      "epoch": 1.739705370606658,
+      "grad_norm": 11.879483859690765,
+      "learning_rate": 2e-06,
+      "loss": 0.4304,
+      "step": 7499
+    },
+    {
+      "epoch": 1.7399373622549588,
+      "grad_norm": 16.22319785403237,
+      "learning_rate": 2e-06,
+      "loss": 0.2655,
+      "step": 7500
+    },
+    {
+      "epoch": 1.7401693539032594,
+      "grad_norm": 15.214716242928388,
+      "learning_rate": 2e-06,
+      "loss": 0.2577,
+      "step": 7501
+    },
+    {
+      "epoch": 1.7404013455515601,
+      "grad_norm": 14.592929299836015,
+      "learning_rate": 2e-06,
+      "loss": 0.2414,
+      "step": 7502
+    },
+    {
+      "epoch": 1.7406333371998608,
+      "grad_norm": 20.926869969586114,
+      "learning_rate": 2e-06,
+      "loss": 0.3618,
+      "step": 7503
+    },
+    {
+      "epoch": 1.7408653288481615,
+      "grad_norm": 11.524230240811791,
+      "learning_rate": 2e-06,
+      "loss": 0.3186,
+      "step": 7504
+    },
+    {
+      "epoch": 1.7410973204964622,
+      "grad_norm": 8.307629840052147,
+      "learning_rate": 2e-06,
+      "loss": 0.256,
+      "step": 7505
+    },
+    {
+      "epoch": 1.7413293121447628,
+      "grad_norm": 14.612464879032128,
+      "learning_rate": 2e-06,
+      "loss": 0.3739,
+      "step": 7506
+    },
+    {
+      "epoch": 1.7415613037930635,
+      "grad_norm": 9.623353081994557,
+      "learning_rate": 2e-06,
+      "loss": 0.1218,
+      "step": 7507
+    },
+    {
+      "epoch": 1.7417932954413642,
+      "grad_norm": 18.58891226199371,
+      "learning_rate": 2e-06,
+      "loss": 0.208,
+      "step": 7508
+    },
+    {
+      "epoch": 1.7420252870896649,
+      "grad_norm": 10.85916321252991,
+      "learning_rate": 2e-06,
+      "loss": 0.2402,
+      "step": 7509
+    },
+    {
+      "epoch": 1.7422572787379655,
+      "grad_norm": 9.232168351677451,
+      "learning_rate": 2e-06,
+      "loss": 0.2054,
+      "step": 7510
+    },
+    {
+      "epoch": 1.7424892703862662,
+      "grad_norm": 11.909383974387465,
+      "learning_rate": 2e-06,
+      "loss": 0.1525,
+      "step": 7511
+    },
+    {
+      "epoch": 1.742721262034567,
+      "grad_norm": 10.25152784065819,
+      "learning_rate": 2e-06,
+      "loss": 0.2434,
+      "step": 7512
+    },
+    {
+      "epoch": 1.7429532536828676,
+      "grad_norm": 9.52767133429135,
+      "learning_rate": 2e-06,
+      "loss": 0.1615,
+      "step": 7513
+    },
+    {
+      "epoch": 1.743185245331168,
+      "grad_norm": 7.889990749074475,
+      "learning_rate": 2e-06,
+      "loss": 0.1444,
+      "step": 7514
+    },
+    {
+      "epoch": 1.7434172369794687,
+      "grad_norm": 11.299170921030075,
+      "learning_rate": 2e-06,
+      "loss": 0.1949,
+      "step": 7515
+    },
+    {
+      "epoch": 1.7436492286277694,
+      "grad_norm": 16.034637698952388,
+      "learning_rate": 2e-06,
+      "loss": 0.2648,
+      "step": 7516
+    },
+    {
+      "epoch": 1.74388122027607,
+      "grad_norm": 16.068494140149642,
+      "learning_rate": 2e-06,
+      "loss": 0.2912,
+      "step": 7517
+    },
+    {
+      "epoch": 1.7441132119243707,
+      "grad_norm": 11.633519815812562,
+      "learning_rate": 2e-06,
+      "loss": 0.2694,
+      "step": 7518
+    },
+    {
+      "epoch": 1.7443452035726714,
+      "grad_norm": 14.184939207792523,
+      "learning_rate": 2e-06,
+      "loss": 0.2376,
+      "step": 7519
+    },
+    {
+      "epoch": 1.744577195220972,
+      "grad_norm": 15.154376906265702,
+      "learning_rate": 2e-06,
+      "loss": 0.2754,
+      "step": 7520
+    },
+    {
+      "epoch": 1.7448091868692726,
+      "grad_norm": 16.188615959115307,
+      "learning_rate": 2e-06,
+      "loss": 0.1729,
+      "step": 7521
+    },
+    {
+      "epoch": 1.7450411785175732,
+      "grad_norm": 12.385480875263635,
+      "learning_rate": 2e-06,
+      "loss": 0.2123,
+      "step": 7522
+    },
+    {
+      "epoch": 1.745273170165874,
+      "grad_norm": 20.89350954693197,
+      "learning_rate": 2e-06,
+      "loss": 0.3894,
+      "step": 7523
+    },
+    {
+      "epoch": 1.7455051618141746,
+      "grad_norm": 12.754250138450933,
+      "learning_rate": 2e-06,
+      "loss": 0.3069,
+      "step": 7524
+    },
+    {
+      "epoch": 1.7457371534624753,
+      "grad_norm": 16.636950981836346,
+      "learning_rate": 2e-06,
+      "loss": 0.3309,
+      "step": 7525
+    },
+    {
+      "epoch": 1.745969145110776,
+      "grad_norm": 9.15472274877276,
+      "learning_rate": 2e-06,
+      "loss": 0.2223,
+      "step": 7526
+    },
+    {
+      "epoch": 1.7462011367590766,
+      "grad_norm": 12.287297688521102,
+      "learning_rate": 2e-06,
+      "loss": 0.1625,
+      "step": 7527
+    },
+    {
+      "epoch": 1.7464331284073773,
+      "grad_norm": 16.256618957006086,
+      "learning_rate": 2e-06,
+      "loss": 0.1969,
+      "step": 7528
+    },
+    {
+      "epoch": 1.746665120055678,
+      "grad_norm": 11.264582784625329,
+      "learning_rate": 2e-06,
+      "loss": 0.2765,
+      "step": 7529
+    },
+    {
+      "epoch": 1.7468971117039787,
+      "grad_norm": 7.349783736438385,
+      "learning_rate": 2e-06,
+      "loss": 0.1936,
+      "step": 7530
+    },
+    {
+      "epoch": 1.7471291033522793,
+      "grad_norm": 13.180748443177167,
+      "learning_rate": 2e-06,
+      "loss": 0.3109,
+      "step": 7531
+    },
+    {
+      "epoch": 1.74736109500058,
+      "grad_norm": 12.550800164137836,
+      "learning_rate": 2e-06,
+      "loss": 0.2776,
+      "step": 7532
+    },
+    {
+      "epoch": 1.7475930866488807,
+      "grad_norm": 18.065154984445197,
+      "learning_rate": 2e-06,
+      "loss": 0.3497,
+      "step": 7533
+    },
+    {
+      "epoch": 1.7478250782971814,
+      "grad_norm": 12.56891058573405,
+      "learning_rate": 2e-06,
+      "loss": 0.2562,
+      "step": 7534
+    },
+    {
+      "epoch": 1.748057069945482,
+      "grad_norm": 13.878529337942904,
+      "learning_rate": 2e-06,
+      "loss": 0.286,
+      "step": 7535
+    },
+    {
+      "epoch": 1.7482890615937827,
+      "grad_norm": 19.637820912174707,
+      "learning_rate": 2e-06,
+      "loss": 0.2508,
+      "step": 7536
+    },
+    {
+      "epoch": 1.7485210532420834,
+      "grad_norm": 11.758718917948533,
+      "learning_rate": 2e-06,
+      "loss": 0.2464,
+      "step": 7537
+    },
+    {
+      "epoch": 1.748753044890384,
+      "grad_norm": 7.703897971156512,
+      "learning_rate": 2e-06,
+      "loss": 0.1873,
+      "step": 7538
+    },
+    {
+      "epoch": 1.7489850365386848,
+      "grad_norm": 22.799078363783156,
+      "learning_rate": 2e-06,
+      "loss": 0.3683,
+      "step": 7539
+    },
+    {
+      "epoch": 1.7492170281869854,
+      "grad_norm": 8.022522116780017,
+      "learning_rate": 2e-06,
+      "loss": 0.2211,
+      "step": 7540
+    },
+    {
+      "epoch": 1.749449019835286,
+      "grad_norm": 9.720216247242394,
+      "learning_rate": 2e-06,
+      "loss": 0.1509,
+      "step": 7541
+    },
+    {
+      "epoch": 1.7496810114835866,
+      "grad_norm": 12.831999318303131,
+      "learning_rate": 2e-06,
+      "loss": 0.3326,
+      "step": 7542
+    },
+    {
+      "epoch": 1.7499130031318872,
+      "grad_norm": 13.350805612457814,
+      "learning_rate": 2e-06,
+      "loss": 0.2451,
+      "step": 7543
+    },
+    {
+      "epoch": 1.750144994780188,
+      "grad_norm": 29.70812030671747,
+      "learning_rate": 2e-06,
+      "loss": 0.3109,
+      "step": 7544
+    },
+    {
+      "epoch": 1.7503769864284886,
+      "grad_norm": 14.927095134968875,
+      "learning_rate": 2e-06,
+      "loss": 0.2826,
+      "step": 7545
+    },
+    {
+      "epoch": 1.7506089780767893,
+      "grad_norm": 15.271269633188103,
+      "learning_rate": 2e-06,
+      "loss": 0.2656,
+      "step": 7546
+    },
+    {
+      "epoch": 1.7508409697250897,
+      "grad_norm": 18.91591537542004,
+      "learning_rate": 2e-06,
+      "loss": 0.2331,
+      "step": 7547
+    },
+    {
+      "epoch": 1.7510729613733904,
+      "grad_norm": 15.809697419634523,
+      "learning_rate": 2e-06,
+      "loss": 0.2874,
+      "step": 7548
+    },
+    {
+      "epoch": 1.751304953021691,
+      "grad_norm": 10.111798022135611,
+      "learning_rate": 2e-06,
+      "loss": 0.2326,
+      "step": 7549
+    },
+    {
+      "epoch": 1.7515369446699918,
+      "grad_norm": 12.66921707938741,
+      "learning_rate": 2e-06,
+      "loss": 0.2217,
+      "step": 7550
+    },
+    {
+      "epoch": 1.7517689363182924,
+      "grad_norm": 7.11185655840169,
+      "learning_rate": 2e-06,
+      "loss": 0.1963,
+      "step": 7551
+    },
+    {
+      "epoch": 1.7520009279665931,
+      "grad_norm": 12.80342476480035,
+      "learning_rate": 2e-06,
+      "loss": 0.229,
+      "step": 7552
+    },
+    {
+      "epoch": 1.7522329196148938,
+      "grad_norm": 10.654623307789162,
+      "learning_rate": 2e-06,
+      "loss": 0.2072,
+      "step": 7553
+    },
+    {
+      "epoch": 1.7524649112631945,
+      "grad_norm": 31.130164040219547,
+      "learning_rate": 2e-06,
+      "loss": 0.2552,
+      "step": 7554
+    },
+    {
+      "epoch": 1.7526969029114952,
+      "grad_norm": 9.521927343087164,
+      "learning_rate": 2e-06,
+      "loss": 0.2096,
+      "step": 7555
+    },
+    {
+      "epoch": 1.7529288945597958,
+      "grad_norm": 17.004948780199133,
+      "learning_rate": 2e-06,
+      "loss": 0.2874,
+      "step": 7556
+    },
+    {
+      "epoch": 1.7531608862080965,
+      "grad_norm": 24.340497532566054,
+      "learning_rate": 2e-06,
+      "loss": 0.3665,
+      "step": 7557
+    },
+    {
+      "epoch": 1.7533928778563972,
+      "grad_norm": 9.246138535580242,
+      "learning_rate": 2e-06,
+      "loss": 0.1686,
+      "step": 7558
+    },
+    {
+      "epoch": 1.7536248695046979,
+      "grad_norm": 10.679323908025648,
+      "learning_rate": 2e-06,
+      "loss": 0.1937,
+      "step": 7559
+    },
+    {
+      "epoch": 1.7538568611529985,
+      "grad_norm": 7.459226203385063,
+      "learning_rate": 2e-06,
+      "loss": 0.2052,
+      "step": 7560
+    },
+    {
+      "epoch": 1.7540888528012992,
+      "grad_norm": 14.642865169587246,
+      "learning_rate": 2e-06,
+      "loss": 0.2596,
+      "step": 7561
+    },
+    {
+      "epoch": 1.7543208444496,
+      "grad_norm": 10.384026009480467,
+      "learning_rate": 2e-06,
+      "loss": 0.1944,
+      "step": 7562
+    },
+    {
+      "epoch": 1.7545528360979006,
+      "grad_norm": 17.544916763886047,
+      "learning_rate": 2e-06,
+      "loss": 0.3075,
+      "step": 7563
+    },
+    {
+      "epoch": 1.7547848277462013,
+      "grad_norm": 20.255564597301053,
+      "learning_rate": 2e-06,
+      "loss": 0.2725,
+      "step": 7564
+    },
+    {
+      "epoch": 1.755016819394502,
+      "grad_norm": 15.313597934383315,
+      "learning_rate": 2e-06,
+      "loss": 0.2702,
+      "step": 7565
+    },
+    {
+      "epoch": 1.7552488110428026,
+      "grad_norm": 11.868105774967077,
+      "learning_rate": 2e-06,
+      "loss": 0.1997,
+      "step": 7566
+    },
+    {
+      "epoch": 1.755480802691103,
+      "grad_norm": 8.132450737749771,
+      "learning_rate": 2e-06,
+      "loss": 0.1891,
+      "step": 7567
+    },
+    {
+      "epoch": 1.7557127943394037,
+      "grad_norm": 9.64320808739205,
+      "learning_rate": 2e-06,
+      "loss": 0.237,
+      "step": 7568
+    },
+    {
+      "epoch": 1.7559447859877044,
+      "grad_norm": 14.945149857562205,
+      "learning_rate": 2e-06,
+      "loss": 0.32,
+      "step": 7569
+    },
+    {
+      "epoch": 1.756176777636005,
+      "grad_norm": 5.6092356114404796,
+      "learning_rate": 2e-06,
+      "loss": 0.1472,
+      "step": 7570
+    },
+    {
+      "epoch": 1.7564087692843058,
+      "grad_norm": 15.171945822518653,
+      "learning_rate": 2e-06,
+      "loss": 0.2661,
+      "step": 7571
+    },
+    {
+      "epoch": 1.7566407609326065,
+      "grad_norm": 12.379076468659813,
+      "learning_rate": 2e-06,
+      "loss": 0.2324,
+      "step": 7572
+    },
+    {
+      "epoch": 1.7568727525809071,
+      "grad_norm": 10.976070574759373,
+      "learning_rate": 2e-06,
+      "loss": 0.2262,
+      "step": 7573
+    },
+    {
+      "epoch": 1.7571047442292076,
+      "grad_norm": 17.711656018486853,
+      "learning_rate": 2e-06,
+      "loss": 0.2753,
+      "step": 7574
+    },
+    {
+      "epoch": 1.7573367358775083,
+      "grad_norm": 9.414455829200904,
+      "learning_rate": 2e-06,
+      "loss": 0.2846,
+      "step": 7575
+    },
+    {
+      "epoch": 1.757568727525809,
+      "grad_norm": 10.143710465167523,
+      "learning_rate": 2e-06,
+      "loss": 0.2424,
+      "step": 7576
+    },
+    {
+      "epoch": 1.7578007191741096,
+      "grad_norm": 14.533736277795931,
+      "learning_rate": 2e-06,
+      "loss": 0.2556,
+      "step": 7577
+    },
+    {
+      "epoch": 1.7580327108224103,
+      "grad_norm": 15.913258956484471,
+      "learning_rate": 2e-06,
+      "loss": 0.2865,
+      "step": 7578
+    },
+    {
+      "epoch": 1.758264702470711,
+      "grad_norm": 12.767660833167866,
+      "learning_rate": 2e-06,
+      "loss": 0.271,
+      "step": 7579
+    },
+    {
+      "epoch": 1.7584966941190117,
+      "grad_norm": 11.294717862094545,
+      "learning_rate": 2e-06,
+      "loss": 0.2234,
+      "step": 7580
+    },
+    {
+      "epoch": 1.7587286857673123,
+      "grad_norm": 11.306623689139046,
+      "learning_rate": 2e-06,
+      "loss": 0.2039,
+      "step": 7581
+    },
+    {
+      "epoch": 1.758960677415613,
+      "grad_norm": 8.535703463216377,
+      "learning_rate": 2e-06,
+      "loss": 0.1754,
+      "step": 7582
+    },
+    {
+      "epoch": 1.7591926690639137,
+      "grad_norm": 15.506198698340077,
+      "learning_rate": 2e-06,
+      "loss": 0.1989,
+      "step": 7583
+    },
+    {
+      "epoch": 1.7594246607122144,
+      "grad_norm": 13.271148442659763,
+      "learning_rate": 2e-06,
+      "loss": 0.2016,
+      "step": 7584
+    },
+    {
+      "epoch": 1.759656652360515,
+      "grad_norm": 12.936342351595748,
+      "learning_rate": 2e-06,
+      "loss": 0.2456,
+      "step": 7585
+    },
+    {
+      "epoch": 1.7598886440088157,
+      "grad_norm": 4.865820394943794,
+      "learning_rate": 2e-06,
+      "loss": 0.1095,
+      "step": 7586
+    },
+    {
+      "epoch": 1.7601206356571164,
+      "grad_norm": 24.73588202755812,
+      "learning_rate": 2e-06,
+      "loss": 0.4227,
+      "step": 7587
+    },
+    {
+      "epoch": 1.760352627305417,
+      "grad_norm": 20.67162051161943,
+      "learning_rate": 2e-06,
+      "loss": 0.2526,
+      "step": 7588
+    },
+    {
+      "epoch": 1.7605846189537178,
+      "grad_norm": 19.6711134518649,
+      "learning_rate": 2e-06,
+      "loss": 0.3269,
+      "step": 7589
+    },
+    {
+      "epoch": 1.7608166106020184,
+      "grad_norm": 14.748934619827882,
+      "learning_rate": 2e-06,
+      "loss": 0.2186,
+      "step": 7590
+    },
+    {
+      "epoch": 1.7610486022503191,
+      "grad_norm": 11.75288363328804,
+      "learning_rate": 2e-06,
+      "loss": 0.1792,
+      "step": 7591
+    },
+    {
+      "epoch": 1.7612805938986198,
+      "grad_norm": 13.161600894829599,
+      "learning_rate": 2e-06,
+      "loss": 0.3328,
+      "step": 7592
+    },
+    {
+      "epoch": 1.7615125855469205,
+      "grad_norm": 24.39174578419128,
+      "learning_rate": 2e-06,
+      "loss": 0.276,
+      "step": 7593
+    },
+    {
+      "epoch": 1.761744577195221,
+      "grad_norm": 12.371452232880623,
+      "learning_rate": 2e-06,
+      "loss": 0.2194,
+      "step": 7594
+    },
+    {
+      "epoch": 1.7619765688435216,
+      "grad_norm": 14.749222617313963,
+      "learning_rate": 2e-06,
+      "loss": 0.2265,
+      "step": 7595
+    },
+    {
+      "epoch": 1.7622085604918223,
+      "grad_norm": 14.394177818624705,
+      "learning_rate": 2e-06,
+      "loss": 0.2763,
+      "step": 7596
+    },
+    {
+      "epoch": 1.762440552140123,
+      "grad_norm": 11.736377858916772,
+      "learning_rate": 2e-06,
+      "loss": 0.2413,
+      "step": 7597
+    },
+    {
+      "epoch": 1.7626725437884236,
+      "grad_norm": 9.37857496441924,
+      "learning_rate": 2e-06,
+      "loss": 0.2431,
+      "step": 7598
+    },
+    {
+      "epoch": 1.7629045354367243,
+      "grad_norm": 7.755353360563734,
+      "learning_rate": 2e-06,
+      "loss": 0.1467,
+      "step": 7599
+    },
+    {
+      "epoch": 1.7631365270850248,
+      "grad_norm": 17.392051182713978,
+      "learning_rate": 2e-06,
+      "loss": 0.3205,
+      "step": 7600
+    },
+    {
+      "epoch": 1.7633685187333255,
+      "grad_norm": 6.479599459078176,
+      "learning_rate": 2e-06,
+      "loss": 0.1768,
+      "step": 7601
+    },
+    {
+      "epoch": 1.7636005103816261,
+      "grad_norm": 12.443300925816132,
+      "learning_rate": 2e-06,
+      "loss": 0.2188,
+      "step": 7602
+    },
+    {
+      "epoch": 1.7638325020299268,
+      "grad_norm": 14.02188400560854,
+      "learning_rate": 2e-06,
+      "loss": 0.1731,
+      "step": 7603
+    },
+    {
+      "epoch": 1.7640644936782275,
+      "grad_norm": 13.579452802841217,
+      "learning_rate": 2e-06,
+      "loss": 0.188,
+      "step": 7604
+    },
+    {
+      "epoch": 1.7642964853265282,
+      "grad_norm": 11.894099852846338,
+      "learning_rate": 2e-06,
+      "loss": 0.2135,
+      "step": 7605
+    },
+    {
+      "epoch": 1.7645284769748288,
+      "grad_norm": 15.44436809606793,
+      "learning_rate": 2e-06,
+      "loss": 0.2343,
+      "step": 7606
+    },
+    {
+      "epoch": 1.7647604686231295,
+      "grad_norm": 25.01790094541462,
+      "learning_rate": 2e-06,
+      "loss": 0.2829,
+      "step": 7607
+    },
+    {
+      "epoch": 1.7649924602714302,
+      "grad_norm": 17.30869963511119,
+      "learning_rate": 2e-06,
+      "loss": 0.3645,
+      "step": 7608
+    },
+    {
+      "epoch": 1.7652244519197309,
+      "grad_norm": 7.544423889631266,
+      "learning_rate": 2e-06,
+      "loss": 0.1884,
+      "step": 7609
+    },
+    {
+      "epoch": 1.7654564435680316,
+      "grad_norm": 16.77983044601465,
+      "learning_rate": 2e-06,
+      "loss": 0.3064,
+      "step": 7610
+    },
+    {
+      "epoch": 1.7656884352163322,
+      "grad_norm": 13.915765880853346,
+      "learning_rate": 2e-06,
+      "loss": 0.3585,
+      "step": 7611
+    },
+    {
+      "epoch": 1.765920426864633,
+      "grad_norm": 14.45783459966466,
+      "learning_rate": 2e-06,
+      "loss": 0.234,
+      "step": 7612
+    },
+    {
+      "epoch": 1.7661524185129336,
+      "grad_norm": 13.257478868195667,
+      "learning_rate": 2e-06,
+      "loss": 0.1836,
+      "step": 7613
+    },
+    {
+      "epoch": 1.7663844101612343,
+      "grad_norm": 24.416811462153376,
+      "learning_rate": 2e-06,
+      "loss": 0.3739,
+      "step": 7614
+    },
+    {
+      "epoch": 1.766616401809535,
+      "grad_norm": 16.41812661132975,
+      "learning_rate": 2e-06,
+      "loss": 0.3323,
+      "step": 7615
+    },
+    {
+      "epoch": 1.7668483934578356,
+      "grad_norm": 13.841333663391419,
+      "learning_rate": 2e-06,
+      "loss": 0.2876,
+      "step": 7616
+    },
+    {
+      "epoch": 1.7670803851061363,
+      "grad_norm": 18.261926281948277,
+      "learning_rate": 2e-06,
+      "loss": 0.2349,
+      "step": 7617
+    },
+    {
+      "epoch": 1.767312376754437,
+      "grad_norm": 12.052038914795572,
+      "learning_rate": 2e-06,
+      "loss": 0.267,
+      "step": 7618
+    },
+    {
+      "epoch": 1.7675443684027377,
+      "grad_norm": 13.682131575458424,
+      "learning_rate": 2e-06,
+      "loss": 0.2445,
+      "step": 7619
+    },
+    {
+      "epoch": 1.767776360051038,
+      "grad_norm": 17.787255720543858,
+      "learning_rate": 2e-06,
+      "loss": 0.2742,
+      "step": 7620
+    },
+    {
+      "epoch": 1.7680083516993388,
+      "grad_norm": 15.57759468719208,
+      "learning_rate": 2e-06,
+      "loss": 0.2104,
+      "step": 7621
+    },
+    {
+      "epoch": 1.7682403433476395,
+      "grad_norm": 10.939546101600278,
+      "learning_rate": 2e-06,
+      "loss": 0.2552,
+      "step": 7622
+    },
+    {
+      "epoch": 1.7684723349959401,
+      "grad_norm": 17.21174522628649,
+      "learning_rate": 2e-06,
+      "loss": 0.3564,
+      "step": 7623
+    },
+    {
+      "epoch": 1.7687043266442408,
+      "grad_norm": 10.751055553207847,
+      "learning_rate": 2e-06,
+      "loss": 0.1995,
+      "step": 7624
+    },
+    {
+      "epoch": 1.7689363182925415,
+      "grad_norm": 39.846053247413224,
+      "learning_rate": 2e-06,
+      "loss": 0.3384,
+      "step": 7625
+    },
+    {
+      "epoch": 1.7691683099408422,
+      "grad_norm": 12.08905876035218,
+      "learning_rate": 2e-06,
+      "loss": 0.1989,
+      "step": 7626
+    },
+    {
+      "epoch": 1.7694003015891426,
+      "grad_norm": 12.793073296425936,
+      "learning_rate": 2e-06,
+      "loss": 0.2565,
+      "step": 7627
+    },
+    {
+      "epoch": 1.7696322932374433,
+      "grad_norm": 15.993395786481747,
+      "learning_rate": 2e-06,
+      "loss": 0.3398,
+      "step": 7628
+    },
+    {
+      "epoch": 1.769864284885744,
+      "grad_norm": 13.706742234967228,
+      "learning_rate": 2e-06,
+      "loss": 0.1252,
+      "step": 7629
+    },
+    {
+      "epoch": 1.7700962765340447,
+      "grad_norm": 11.953171049385569,
+      "learning_rate": 2e-06,
+      "loss": 0.2428,
+      "step": 7630
+    },
+    {
+      "epoch": 1.7703282681823453,
+      "grad_norm": 14.74479617381812,
+      "learning_rate": 2e-06,
+      "loss": 0.2547,
+      "step": 7631
+    },
+    {
+      "epoch": 1.770560259830646,
+      "grad_norm": 14.219514617403146,
+      "learning_rate": 2e-06,
+      "loss": 0.2018,
+      "step": 7632
+    },
+    {
+      "epoch": 1.7707922514789467,
+      "grad_norm": 13.951297658416964,
+      "learning_rate": 2e-06,
+      "loss": 0.2766,
+      "step": 7633
+    },
+    {
+      "epoch": 1.7710242431272474,
+      "grad_norm": 9.491752262170884,
+      "learning_rate": 2e-06,
+      "loss": 0.1672,
+      "step": 7634
+    },
+    {
+      "epoch": 1.771256234775548,
+      "grad_norm": 17.46797158449257,
+      "learning_rate": 2e-06,
+      "loss": 0.3137,
+      "step": 7635
+    },
+    {
+      "epoch": 1.7714882264238487,
+      "grad_norm": 19.06944157828703,
+      "learning_rate": 2e-06,
+      "loss": 0.2695,
+      "step": 7636
+    },
+    {
+      "epoch": 1.7717202180721494,
+      "grad_norm": 15.879264186736425,
+      "learning_rate": 2e-06,
+      "loss": 0.2672,
+      "step": 7637
+    },
+    {
+      "epoch": 1.77195220972045,
+      "grad_norm": 9.32608523384946,
+      "learning_rate": 2e-06,
+      "loss": 0.2372,
+      "step": 7638
+    },
+    {
+      "epoch": 1.7721842013687508,
+      "grad_norm": 18.03702817443358,
+      "learning_rate": 2e-06,
+      "loss": 0.2554,
+      "step": 7639
+    },
+    {
+      "epoch": 1.7724161930170514,
+      "grad_norm": 13.736295330033094,
+      "learning_rate": 2e-06,
+      "loss": 0.2727,
+      "step": 7640
+    },
+    {
+      "epoch": 1.7726481846653521,
+      "grad_norm": 8.515303235720744,
+      "learning_rate": 2e-06,
+      "loss": 0.1856,
+      "step": 7641
+    },
+    {
+      "epoch": 1.7728801763136528,
+      "grad_norm": 23.630669813146312,
+      "learning_rate": 2e-06,
+      "loss": 0.2172,
+      "step": 7642
+    },
+    {
+      "epoch": 1.7731121679619535,
+      "grad_norm": 12.887268646237283,
+      "learning_rate": 2e-06,
+      "loss": 0.3188,
+      "step": 7643
+    },
+    {
+      "epoch": 1.7733441596102542,
+      "grad_norm": 12.540798137644023,
+      "learning_rate": 2e-06,
+      "loss": 0.2652,
+      "step": 7644
+    },
+    {
+      "epoch": 1.7735761512585548,
+      "grad_norm": 14.105494099925721,
+      "learning_rate": 2e-06,
+      "loss": 0.2969,
+      "step": 7645
+    },
+    {
+      "epoch": 1.7738081429068555,
+      "grad_norm": 13.725792982960401,
+      "learning_rate": 2e-06,
+      "loss": 0.227,
+      "step": 7646
+    },
+    {
+      "epoch": 1.774040134555156,
+      "grad_norm": 21.37437022895052,
+      "learning_rate": 2e-06,
+      "loss": 0.3201,
+      "step": 7647
+    },
+    {
+      "epoch": 1.7742721262034566,
+      "grad_norm": 10.309915600828834,
+      "learning_rate": 2e-06,
+      "loss": 0.2181,
+      "step": 7648
+    },
+    {
+      "epoch": 1.7745041178517573,
+      "grad_norm": 12.343204213291852,
+      "learning_rate": 2e-06,
+      "loss": 0.1647,
+      "step": 7649
+    },
+    {
+      "epoch": 1.774736109500058,
+      "grad_norm": 12.04423941594413,
+      "learning_rate": 2e-06,
+      "loss": 0.2637,
+      "step": 7650
+    },
+    {
+      "epoch": 1.7749681011483587,
+      "grad_norm": 18.005224230213113,
+      "learning_rate": 2e-06,
+      "loss": 0.3285,
+      "step": 7651
+    },
+    {
+      "epoch": 1.7752000927966594,
+      "grad_norm": 7.775026327525456,
+      "learning_rate": 2e-06,
+      "loss": 0.1205,
+      "step": 7652
+    },
+    {
+      "epoch": 1.77543208444496,
+      "grad_norm": 15.384233415521097,
+      "learning_rate": 2e-06,
+      "loss": 0.3409,
+      "step": 7653
+    },
+    {
+      "epoch": 1.7756640760932605,
+      "grad_norm": 4.987462215279807,
+      "learning_rate": 2e-06,
+      "loss": 0.1379,
+      "step": 7654
+    },
+    {
+      "epoch": 1.7758960677415612,
+      "grad_norm": 22.727798216734325,
+      "learning_rate": 2e-06,
+      "loss": 0.4157,
+      "step": 7655
+    },
+    {
+      "epoch": 1.7761280593898618,
+      "grad_norm": 8.363343229513351,
+      "learning_rate": 2e-06,
+      "loss": 0.167,
+      "step": 7656
+    },
+    {
+      "epoch": 1.7763600510381625,
+      "grad_norm": 7.1776307695687045,
+      "learning_rate": 2e-06,
+      "loss": 0.1454,
+      "step": 7657
+    },
+    {
+      "epoch": 1.7765920426864632,
+      "grad_norm": 10.413632590231295,
+      "learning_rate": 2e-06,
+      "loss": 0.1798,
+      "step": 7658
+    },
+    {
+      "epoch": 1.7768240343347639,
+      "grad_norm": 17.375429455673192,
+      "learning_rate": 2e-06,
+      "loss": 0.3528,
+      "step": 7659
+    },
+    {
+      "epoch": 1.7770560259830646,
+      "grad_norm": 9.339877282892838,
+      "learning_rate": 2e-06,
+      "loss": 0.3373,
+      "step": 7660
+    },
+    {
+      "epoch": 1.7772880176313652,
+      "grad_norm": 15.76690432141599,
+      "learning_rate": 2e-06,
+      "loss": 0.2209,
+      "step": 7661
+    },
+    {
+      "epoch": 1.777520009279666,
+      "grad_norm": 16.580997351774784,
+      "learning_rate": 2e-06,
+      "loss": 0.248,
+      "step": 7662
+    },
+    {
+      "epoch": 1.7777520009279666,
+      "grad_norm": 8.380956961573576,
+      "learning_rate": 2e-06,
+      "loss": 0.1943,
+      "step": 7663
+    },
+    {
+      "epoch": 1.7779839925762673,
+      "grad_norm": 18.827557808833518,
+      "learning_rate": 2e-06,
+      "loss": 0.2473,
+      "step": 7664
+    },
+    {
+      "epoch": 1.778215984224568,
+      "grad_norm": 13.388074493082893,
+      "learning_rate": 2e-06,
+      "loss": 0.2367,
+      "step": 7665
+    },
+    {
+      "epoch": 1.7784479758728686,
+      "grad_norm": 26.059732419409222,
+      "learning_rate": 2e-06,
+      "loss": 0.2773,
+      "step": 7666
+    },
+    {
+      "epoch": 1.7786799675211693,
+      "grad_norm": 11.879194017751967,
+      "learning_rate": 2e-06,
+      "loss": 0.2716,
+      "step": 7667
+    },
+    {
+      "epoch": 1.77891195916947,
+      "grad_norm": 20.71252153591305,
+      "learning_rate": 2e-06,
+      "loss": 0.2782,
+      "step": 7668
+    },
+    {
+      "epoch": 1.7791439508177707,
+      "grad_norm": 8.655225424649053,
+      "learning_rate": 2e-06,
+      "loss": 0.149,
+      "step": 7669
+    },
+    {
+      "epoch": 1.7793759424660713,
+      "grad_norm": 24.443766643758085,
+      "learning_rate": 2e-06,
+      "loss": 0.378,
+      "step": 7670
+    },
+    {
+      "epoch": 1.779607934114372,
+      "grad_norm": 13.744789244886295,
+      "learning_rate": 2e-06,
+      "loss": 0.2104,
+      "step": 7671
+    },
+    {
+      "epoch": 1.7798399257626727,
+      "grad_norm": 11.431132212646698,
+      "learning_rate": 2e-06,
+      "loss": 0.2329,
+      "step": 7672
+    },
+    {
+      "epoch": 1.7800719174109734,
+      "grad_norm": 12.346610648641189,
+      "learning_rate": 2e-06,
+      "loss": 0.3629,
+      "step": 7673
+    },
+    {
+      "epoch": 1.7803039090592738,
+      "grad_norm": 13.919607998095852,
+      "learning_rate": 2e-06,
+      "loss": 0.178,
+      "step": 7674
+    },
+    {
+      "epoch": 1.7805359007075745,
+      "grad_norm": 17.970739199315016,
+      "learning_rate": 2e-06,
+      "loss": 0.3034,
+      "step": 7675
+    },
+    {
+      "epoch": 1.7807678923558752,
+      "grad_norm": 7.195667958235038,
+      "learning_rate": 2e-06,
+      "loss": 0.224,
+      "step": 7676
+    },
+    {
+      "epoch": 1.7809998840041759,
+      "grad_norm": 8.009797752370751,
+      "learning_rate": 2e-06,
+      "loss": 0.1594,
+      "step": 7677
+    },
+    {
+      "epoch": 1.7812318756524765,
+      "grad_norm": 4.141046293959366,
+      "learning_rate": 2e-06,
+      "loss": 0.1238,
+      "step": 7678
+    },
+    {
+      "epoch": 1.7814638673007772,
+      "grad_norm": 12.84582198199472,
+      "learning_rate": 2e-06,
+      "loss": 0.2748,
+      "step": 7679
+    },
+    {
+      "epoch": 1.7816958589490777,
+      "grad_norm": 15.120203097994304,
+      "learning_rate": 2e-06,
+      "loss": 0.2866,
+      "step": 7680
+    },
+    {
+      "epoch": 1.7819278505973783,
+      "grad_norm": 14.5632055267852,
+      "learning_rate": 2e-06,
+      "loss": 0.2852,
+      "step": 7681
+    },
+    {
+      "epoch": 1.782159842245679,
+      "grad_norm": 16.375652600630268,
+      "learning_rate": 2e-06,
+      "loss": 0.3306,
+      "step": 7682
+    },
+    {
+      "epoch": 1.7823918338939797,
+      "grad_norm": 15.930067366581838,
+      "learning_rate": 2e-06,
+      "loss": 0.2648,
+      "step": 7683
+    },
+    {
+      "epoch": 1.7826238255422804,
+      "grad_norm": 13.939992885928524,
+      "learning_rate": 2e-06,
+      "loss": 0.2087,
+      "step": 7684
+    },
+    {
+      "epoch": 1.782855817190581,
+      "grad_norm": 8.928404635074617,
+      "learning_rate": 2e-06,
+      "loss": 0.2156,
+      "step": 7685
+    },
+    {
+      "epoch": 1.7830878088388817,
+      "grad_norm": 6.86939899315531,
+      "learning_rate": 2e-06,
+      "loss": 0.2018,
+      "step": 7686
+    },
+    {
+      "epoch": 1.7833198004871824,
+      "grad_norm": 14.856892617680042,
+      "learning_rate": 2e-06,
+      "loss": 0.2491,
+      "step": 7687
+    },
+    {
+      "epoch": 1.783551792135483,
+      "grad_norm": 9.050658316312433,
+      "learning_rate": 2e-06,
+      "loss": 0.2238,
+      "step": 7688
+    },
+    {
+      "epoch": 1.7837837837837838,
+      "grad_norm": 10.238929511967964,
+      "learning_rate": 2e-06,
+      "loss": 0.261,
+      "step": 7689
+    },
+    {
+      "epoch": 1.7840157754320844,
+      "grad_norm": 12.373983145047632,
+      "learning_rate": 2e-06,
+      "loss": 0.2076,
+      "step": 7690
+    },
+    {
+      "epoch": 1.7842477670803851,
+      "grad_norm": 18.484115788006253,
+      "learning_rate": 2e-06,
+      "loss": 0.3735,
+      "step": 7691
+    },
+    {
+      "epoch": 1.7844797587286858,
+      "grad_norm": 13.47239908811981,
+      "learning_rate": 2e-06,
+      "loss": 0.3014,
+      "step": 7692
+    },
+    {
+      "epoch": 1.7847117503769865,
+      "grad_norm": 13.809744947458755,
+      "learning_rate": 2e-06,
+      "loss": 0.2936,
+      "step": 7693
+    },
+    {
+      "epoch": 1.7849437420252872,
+      "grad_norm": 12.90404228506961,
+      "learning_rate": 2e-06,
+      "loss": 0.2841,
+      "step": 7694
+    },
+    {
+      "epoch": 1.7851757336735878,
+      "grad_norm": 9.70647010268845,
+      "learning_rate": 2e-06,
+      "loss": 0.194,
+      "step": 7695
+    },
+    {
+      "epoch": 1.7854077253218885,
+      "grad_norm": 10.11452466154601,
+      "learning_rate": 2e-06,
+      "loss": 0.2296,
+      "step": 7696
+    },
+    {
+      "epoch": 1.7856397169701892,
+      "grad_norm": 18.241161735120254,
+      "learning_rate": 2e-06,
+      "loss": 0.2553,
+      "step": 7697
+    },
+    {
+      "epoch": 1.7858717086184899,
+      "grad_norm": 4.363594729824572,
+      "learning_rate": 2e-06,
+      "loss": 0.1453,
+      "step": 7698
+    },
+    {
+      "epoch": 1.7861037002667906,
+      "grad_norm": 19.847002056503342,
+      "learning_rate": 2e-06,
+      "loss": 0.2892,
+      "step": 7699
+    },
+    {
+      "epoch": 1.786335691915091,
+      "grad_norm": 12.037313097036915,
+      "learning_rate": 2e-06,
+      "loss": 0.1817,
+      "step": 7700
+    },
+    {
+      "epoch": 1.7865676835633917,
+      "grad_norm": 9.665541170604797,
+      "learning_rate": 2e-06,
+      "loss": 0.2488,
+      "step": 7701
+    },
+    {
+      "epoch": 1.7867996752116924,
+      "grad_norm": 9.578282900368189,
+      "learning_rate": 2e-06,
+      "loss": 0.1973,
+      "step": 7702
+    },
+    {
+      "epoch": 1.787031666859993,
+      "grad_norm": 12.011774653107265,
+      "learning_rate": 2e-06,
+      "loss": 0.1994,
+      "step": 7703
+    },
+    {
+      "epoch": 1.7872636585082937,
+      "grad_norm": 14.059264981423192,
+      "learning_rate": 2e-06,
+      "loss": 0.3322,
+      "step": 7704
+    },
+    {
+      "epoch": 1.7874956501565944,
+      "grad_norm": 11.308037303258596,
+      "learning_rate": 2e-06,
+      "loss": 0.1671,
+      "step": 7705
+    },
+    {
+      "epoch": 1.787727641804895,
+      "grad_norm": 14.830070083622193,
+      "learning_rate": 2e-06,
+      "loss": 0.2383,
+      "step": 7706
+    },
+    {
+      "epoch": 1.7879596334531955,
+      "grad_norm": 8.970729653365368,
+      "learning_rate": 2e-06,
+      "loss": 0.1878,
+      "step": 7707
+    },
+    {
+      "epoch": 1.7881916251014962,
+      "grad_norm": 15.265954978012068,
+      "learning_rate": 2e-06,
+      "loss": 0.1944,
+      "step": 7708
+    },
+    {
+      "epoch": 1.7884236167497969,
+      "grad_norm": 9.267580147523244,
+      "learning_rate": 2e-06,
+      "loss": 0.1827,
+      "step": 7709
+    },
+    {
+      "epoch": 1.7886556083980976,
+      "grad_norm": 11.816356243554782,
+      "learning_rate": 2e-06,
+      "loss": 0.2213,
+      "step": 7710
+    },
+    {
+      "epoch": 1.7888876000463982,
+      "grad_norm": 14.13892712739281,
+      "learning_rate": 2e-06,
+      "loss": 0.3183,
+      "step": 7711
+    },
+    {
+      "epoch": 1.789119591694699,
+      "grad_norm": 24.37772003449913,
+      "learning_rate": 2e-06,
+      "loss": 0.3273,
+      "step": 7712
+    },
+    {
+      "epoch": 1.7893515833429996,
+      "grad_norm": 18.881919815002593,
+      "learning_rate": 2e-06,
+      "loss": 0.2567,
+      "step": 7713
+    },
+    {
+      "epoch": 1.7895835749913003,
+      "grad_norm": 9.161757211887082,
+      "learning_rate": 2e-06,
+      "loss": 0.1735,
+      "step": 7714
+    },
+    {
+      "epoch": 1.789815566639601,
+      "grad_norm": 17.849954977305433,
+      "learning_rate": 2e-06,
+      "loss": 0.3116,
+      "step": 7715
+    },
+    {
+      "epoch": 1.7900475582879016,
+      "grad_norm": 8.87979005272126,
+      "learning_rate": 2e-06,
+      "loss": 0.137,
+      "step": 7716
+    },
+    {
+      "epoch": 1.7902795499362023,
+      "grad_norm": 17.163684039103828,
+      "learning_rate": 2e-06,
+      "loss": 0.3231,
+      "step": 7717
+    },
+    {
+      "epoch": 1.790511541584503,
+      "grad_norm": 10.957487983245123,
+      "learning_rate": 2e-06,
+      "loss": 0.2304,
+      "step": 7718
+    },
+    {
+      "epoch": 1.7907435332328037,
+      "grad_norm": 7.331413316833315,
+      "learning_rate": 2e-06,
+      "loss": 0.2005,
+      "step": 7719
+    },
+    {
+      "epoch": 1.7909755248811043,
+      "grad_norm": 10.347401603722119,
+      "learning_rate": 2e-06,
+      "loss": 0.2613,
+      "step": 7720
+    },
+    {
+      "epoch": 1.791207516529405,
+      "grad_norm": 14.864802701315973,
+      "learning_rate": 2e-06,
+      "loss": 0.2408,
+      "step": 7721
+    },
+    {
+      "epoch": 1.7914395081777057,
+      "grad_norm": 12.12204936362209,
+      "learning_rate": 2e-06,
+      "loss": 0.2537,
+      "step": 7722
+    },
+    {
+      "epoch": 1.7916714998260064,
+      "grad_norm": 11.61910317439687,
+      "learning_rate": 2e-06,
+      "loss": 0.2107,
+      "step": 7723
+    },
+    {
+      "epoch": 1.791903491474307,
+      "grad_norm": 11.021376022092555,
+      "learning_rate": 2e-06,
+      "loss": 0.2689,
+      "step": 7724
+    },
+    {
+      "epoch": 1.7921354831226077,
+      "grad_norm": 14.823213878454235,
+      "learning_rate": 2e-06,
+      "loss": 0.2617,
+      "step": 7725
+    },
+    {
+      "epoch": 1.7923674747709084,
+      "grad_norm": 13.759950442388043,
+      "learning_rate": 2e-06,
+      "loss": 0.284,
+      "step": 7726
+    },
+    {
+      "epoch": 1.7925994664192089,
+      "grad_norm": 13.73359765140498,
+      "learning_rate": 2e-06,
+      "loss": 0.1964,
+      "step": 7727
+    },
+    {
+      "epoch": 1.7928314580675095,
+      "grad_norm": 21.592029416160536,
+      "learning_rate": 2e-06,
+      "loss": 0.2878,
+      "step": 7728
+    },
+    {
+      "epoch": 1.7930634497158102,
+      "grad_norm": 13.066431196494792,
+      "learning_rate": 2e-06,
+      "loss": 0.2148,
+      "step": 7729
+    },
+    {
+      "epoch": 1.793295441364111,
+      "grad_norm": 23.518466840599984,
+      "learning_rate": 2e-06,
+      "loss": 0.3538,
+      "step": 7730
+    },
+    {
+      "epoch": 1.7935274330124116,
+      "grad_norm": 22.280556048239266,
+      "learning_rate": 2e-06,
+      "loss": 0.3328,
+      "step": 7731
+    },
+    {
+      "epoch": 1.7937594246607123,
+      "grad_norm": 18.70270509152741,
+      "learning_rate": 2e-06,
+      "loss": 0.2993,
+      "step": 7732
+    },
+    {
+      "epoch": 1.7939914163090127,
+      "grad_norm": 14.745009418359984,
+      "learning_rate": 2e-06,
+      "loss": 0.2493,
+      "step": 7733
+    },
+    {
+      "epoch": 1.7942234079573134,
+      "grad_norm": 7.020182056098816,
+      "learning_rate": 2e-06,
+      "loss": 0.1825,
+      "step": 7734
+    },
+    {
+      "epoch": 1.794455399605614,
+      "grad_norm": 8.9041045197384,
+      "learning_rate": 2e-06,
+      "loss": 0.1951,
+      "step": 7735
+    },
+    {
+      "epoch": 1.7946873912539147,
+      "grad_norm": 16.96340376259227,
+      "learning_rate": 2e-06,
+      "loss": 0.2135,
+      "step": 7736
+    },
+    {
+      "epoch": 1.7949193829022154,
+      "grad_norm": 13.063819447973446,
+      "learning_rate": 2e-06,
+      "loss": 0.3117,
+      "step": 7737
+    },
+    {
+      "epoch": 1.795151374550516,
+      "grad_norm": 17.813456129041118,
+      "learning_rate": 2e-06,
+      "loss": 0.319,
+      "step": 7738
+    },
+    {
+      "epoch": 1.7953833661988168,
+      "grad_norm": 14.21559790292514,
+      "learning_rate": 2e-06,
+      "loss": 0.3333,
+      "step": 7739
+    },
+    {
+      "epoch": 1.7956153578471175,
+      "grad_norm": 13.970632816975339,
+      "learning_rate": 2e-06,
+      "loss": 0.3119,
+      "step": 7740
+    },
+    {
+      "epoch": 1.7958473494954181,
+      "grad_norm": 20.13254991016089,
+      "learning_rate": 2e-06,
+      "loss": 0.3724,
+      "step": 7741
+    },
+    {
+      "epoch": 1.7960793411437188,
+      "grad_norm": 12.702450368500116,
+      "learning_rate": 2e-06,
+      "loss": 0.1933,
+      "step": 7742
+    },
+    {
+      "epoch": 1.7963113327920195,
+      "grad_norm": 5.033487482828364,
+      "learning_rate": 2e-06,
+      "loss": 0.1609,
+      "step": 7743
+    },
+    {
+      "epoch": 1.7965433244403202,
+      "grad_norm": 13.226202684968266,
+      "learning_rate": 2e-06,
+      "loss": 0.2027,
+      "step": 7744
+    },
+    {
+      "epoch": 1.7967753160886208,
+      "grad_norm": 11.445493042999509,
+      "learning_rate": 2e-06,
+      "loss": 0.2252,
+      "step": 7745
+    },
+    {
+      "epoch": 1.7970073077369215,
+      "grad_norm": 9.386391237175017,
+      "learning_rate": 2e-06,
+      "loss": 0.2463,
+      "step": 7746
+    },
+    {
+      "epoch": 1.7972392993852222,
+      "grad_norm": 8.137216847310562,
+      "learning_rate": 2e-06,
+      "loss": 0.2218,
+      "step": 7747
+    },
+    {
+      "epoch": 1.7974712910335229,
+      "grad_norm": 14.29784916326407,
+      "learning_rate": 2e-06,
+      "loss": 0.2736,
+      "step": 7748
+    },
+    {
+      "epoch": 1.7977032826818236,
+      "grad_norm": 32.617210845413794,
+      "learning_rate": 2e-06,
+      "loss": 0.5112,
+      "step": 7749
+    },
+    {
+      "epoch": 1.7979352743301242,
+      "grad_norm": 11.9630554774893,
+      "learning_rate": 2e-06,
+      "loss": 0.2474,
+      "step": 7750
+    },
+    {
+      "epoch": 1.798167265978425,
+      "grad_norm": 9.624929990079778,
+      "learning_rate": 2e-06,
+      "loss": 0.2495,
+      "step": 7751
+    },
+    {
+      "epoch": 1.7983992576267256,
+      "grad_norm": 14.670156870012233,
+      "learning_rate": 2e-06,
+      "loss": 0.2639,
+      "step": 7752
+    },
+    {
+      "epoch": 1.798631249275026,
+      "grad_norm": 10.86678132759406,
+      "learning_rate": 2e-06,
+      "loss": 0.2205,
+      "step": 7753
+    },
+    {
+      "epoch": 1.7988632409233267,
+      "grad_norm": 16.638640631879152,
+      "learning_rate": 2e-06,
+      "loss": 0.3019,
+      "step": 7754
+    },
+    {
+      "epoch": 1.7990952325716274,
+      "grad_norm": 10.957011788431164,
+      "learning_rate": 2e-06,
+      "loss": 0.2409,
+      "step": 7755
+    },
+    {
+      "epoch": 1.799327224219928,
+      "grad_norm": 11.232915987644578,
+      "learning_rate": 2e-06,
+      "loss": 0.2198,
+      "step": 7756
+    },
+    {
+      "epoch": 1.7995592158682288,
+      "grad_norm": 7.602008455160166,
+      "learning_rate": 2e-06,
+      "loss": 0.1659,
+      "step": 7757
+    },
+    {
+      "epoch": 1.7997912075165294,
+      "grad_norm": 10.923209403337351,
+      "learning_rate": 2e-06,
+      "loss": 0.2615,
+      "step": 7758
+    },
+    {
+      "epoch": 1.8000231991648301,
+      "grad_norm": 7.540291491452469,
+      "learning_rate": 2e-06,
+      "loss": 0.1623,
+      "step": 7759
+    },
+    {
+      "epoch": 1.8002551908131306,
+      "grad_norm": 5.588197524067868,
+      "learning_rate": 2e-06,
+      "loss": 0.1303,
+      "step": 7760
+    },
+    {
+      "epoch": 1.8004871824614312,
+      "grad_norm": 7.92180637279653,
+      "learning_rate": 2e-06,
+      "loss": 0.1813,
+      "step": 7761
+    },
+    {
+      "epoch": 1.800719174109732,
+      "grad_norm": 10.141586008302626,
+      "learning_rate": 2e-06,
+      "loss": 0.2743,
+      "step": 7762
+    },
+    {
+      "epoch": 1.8009511657580326,
+      "grad_norm": 21.33250743923904,
+      "learning_rate": 2e-06,
+      "loss": 0.334,
+      "step": 7763
+    },
+    {
+      "epoch": 1.8011831574063333,
+      "grad_norm": 10.204261149433293,
+      "learning_rate": 2e-06,
+      "loss": 0.1503,
+      "step": 7764
+    },
+    {
+      "epoch": 1.801415149054634,
+      "grad_norm": 14.724303989684397,
+      "learning_rate": 2e-06,
+      "loss": 0.2317,
+      "step": 7765
+    },
+    {
+      "epoch": 1.8016471407029346,
+      "grad_norm": 11.189179616754403,
+      "learning_rate": 2e-06,
+      "loss": 0.2597,
+      "step": 7766
+    },
+    {
+      "epoch": 1.8018791323512353,
+      "grad_norm": 8.667717048927448,
+      "learning_rate": 2e-06,
+      "loss": 0.1805,
+      "step": 7767
+    },
+    {
+      "epoch": 1.802111123999536,
+      "grad_norm": 10.099312307553099,
+      "learning_rate": 2e-06,
+      "loss": 0.2101,
+      "step": 7768
+    },
+    {
+      "epoch": 1.8023431156478367,
+      "grad_norm": 10.156468261813107,
+      "learning_rate": 2e-06,
+      "loss": 0.2434,
+      "step": 7769
+    },
+    {
+      "epoch": 1.8025751072961373,
+      "grad_norm": 6.768385693132343,
+      "learning_rate": 2e-06,
+      "loss": 0.23,
+      "step": 7770
+    },
+    {
+      "epoch": 1.802807098944438,
+      "grad_norm": 9.443324597411717,
+      "learning_rate": 2e-06,
+      "loss": 0.2353,
+      "step": 7771
+    },
+    {
+      "epoch": 1.8030390905927387,
+      "grad_norm": 10.59913449235798,
+      "learning_rate": 2e-06,
+      "loss": 0.2387,
+      "step": 7772
+    },
+    {
+      "epoch": 1.8032710822410394,
+      "grad_norm": 16.13258187524545,
+      "learning_rate": 2e-06,
+      "loss": 0.2608,
+      "step": 7773
+    },
+    {
+      "epoch": 1.80350307388934,
+      "grad_norm": 15.139015354804348,
+      "learning_rate": 2e-06,
+      "loss": 0.376,
+      "step": 7774
+    },
+    {
+      "epoch": 1.8037350655376407,
+      "grad_norm": 10.451231399468004,
+      "learning_rate": 2e-06,
+      "loss": 0.2176,
+      "step": 7775
+    },
+    {
+      "epoch": 1.8039670571859414,
+      "grad_norm": 17.00309216109801,
+      "learning_rate": 2e-06,
+      "loss": 0.2994,
+      "step": 7776
+    },
+    {
+      "epoch": 1.804199048834242,
+      "grad_norm": 12.648387109888144,
+      "learning_rate": 2e-06,
+      "loss": 0.2219,
+      "step": 7777
+    },
+    {
+      "epoch": 1.8044310404825428,
+      "grad_norm": 17.7688065718772,
+      "learning_rate": 2e-06,
+      "loss": 0.2452,
+      "step": 7778
+    },
+    {
+      "epoch": 1.8046630321308434,
+      "grad_norm": 14.367544047444886,
+      "learning_rate": 2e-06,
+      "loss": 0.2233,
+      "step": 7779
+    },
+    {
+      "epoch": 1.804895023779144,
+      "grad_norm": 9.520236345447163,
+      "learning_rate": 2e-06,
+      "loss": 0.1703,
+      "step": 7780
+    },
+    {
+      "epoch": 1.8051270154274446,
+      "grad_norm": 14.606371152511754,
+      "learning_rate": 2e-06,
+      "loss": 0.2248,
+      "step": 7781
+    },
+    {
+      "epoch": 1.8053590070757453,
+      "grad_norm": 5.579155887762543,
+      "learning_rate": 2e-06,
+      "loss": 0.1576,
+      "step": 7782
+    },
+    {
+      "epoch": 1.805590998724046,
+      "grad_norm": 12.948532567369465,
+      "learning_rate": 2e-06,
+      "loss": 0.2833,
+      "step": 7783
+    },
+    {
+      "epoch": 1.8058229903723466,
+      "grad_norm": 16.693346240562605,
+      "learning_rate": 2e-06,
+      "loss": 0.3763,
+      "step": 7784
+    },
+    {
+      "epoch": 1.8060549820206473,
+      "grad_norm": 15.82863877420163,
+      "learning_rate": 2e-06,
+      "loss": 0.2494,
+      "step": 7785
+    },
+    {
+      "epoch": 1.806286973668948,
+      "grad_norm": 8.62009366577174,
+      "learning_rate": 2e-06,
+      "loss": 0.2333,
+      "step": 7786
+    },
+    {
+      "epoch": 1.8065189653172484,
+      "grad_norm": 15.841950814591993,
+      "learning_rate": 2e-06,
+      "loss": 0.2895,
+      "step": 7787
+    },
+    {
+      "epoch": 1.806750956965549,
+      "grad_norm": 19.67162703794575,
+      "learning_rate": 2e-06,
+      "loss": 0.2885,
+      "step": 7788
+    },
+    {
+      "epoch": 1.8069829486138498,
+      "grad_norm": 9.324527872345765,
+      "learning_rate": 2e-06,
+      "loss": 0.2164,
+      "step": 7789
+    },
+    {
+      "epoch": 1.8072149402621505,
+      "grad_norm": 8.803393444656317,
+      "learning_rate": 2e-06,
+      "loss": 0.2439,
+      "step": 7790
+    },
+    {
+      "epoch": 1.8074469319104511,
+      "grad_norm": 10.688359265610206,
+      "learning_rate": 2e-06,
+      "loss": 0.2431,
+      "step": 7791
+    },
+    {
+      "epoch": 1.8076789235587518,
+      "grad_norm": 20.370846970552368,
+      "learning_rate": 2e-06,
+      "loss": 0.2405,
+      "step": 7792
+    },
+    {
+      "epoch": 1.8079109152070525,
+      "grad_norm": 12.314345405676008,
+      "learning_rate": 2e-06,
+      "loss": 0.3558,
+      "step": 7793
+    },
+    {
+      "epoch": 1.8081429068553532,
+      "grad_norm": 9.02896450167309,
+      "learning_rate": 2e-06,
+      "loss": 0.2096,
+      "step": 7794
+    },
+    {
+      "epoch": 1.8083748985036538,
+      "grad_norm": 11.910756638938176,
+      "learning_rate": 2e-06,
+      "loss": 0.1928,
+      "step": 7795
+    },
+    {
+      "epoch": 1.8086068901519545,
+      "grad_norm": 10.854730659767704,
+      "learning_rate": 2e-06,
+      "loss": 0.2081,
+      "step": 7796
+    },
+    {
+      "epoch": 1.8088388818002552,
+      "grad_norm": 14.871704613286662,
+      "learning_rate": 2e-06,
+      "loss": 0.2384,
+      "step": 7797
+    },
+    {
+      "epoch": 1.8090708734485559,
+      "grad_norm": 10.544675807311968,
+      "learning_rate": 2e-06,
+      "loss": 0.1492,
+      "step": 7798
+    },
+    {
+      "epoch": 1.8093028650968566,
+      "grad_norm": 13.806037913485634,
+      "learning_rate": 2e-06,
+      "loss": 0.2252,
+      "step": 7799
+    },
+    {
+      "epoch": 1.8095348567451572,
+      "grad_norm": 11.072548191035832,
+      "learning_rate": 2e-06,
+      "loss": 0.2812,
+      "step": 7800
+    },
+    {
+      "epoch": 1.809766848393458,
+      "grad_norm": 11.267678733025672,
+      "learning_rate": 2e-06,
+      "loss": 0.1959,
+      "step": 7801
+    },
+    {
+      "epoch": 1.8099988400417586,
+      "grad_norm": 18.377656686975637,
+      "learning_rate": 2e-06,
+      "loss": 0.241,
+      "step": 7802
+    },
+    {
+      "epoch": 1.8102308316900593,
+      "grad_norm": 25.95688759321459,
+      "learning_rate": 2e-06,
+      "loss": 0.3239,
+      "step": 7803
+    },
+    {
+      "epoch": 1.81046282333836,
+      "grad_norm": 13.683329106749722,
+      "learning_rate": 2e-06,
+      "loss": 0.1345,
+      "step": 7804
+    },
+    {
+      "epoch": 1.8106948149866606,
+      "grad_norm": 9.097969627130546,
+      "learning_rate": 2e-06,
+      "loss": 0.1568,
+      "step": 7805
+    },
+    {
+      "epoch": 1.8109268066349613,
+      "grad_norm": 20.28212796053784,
+      "learning_rate": 2e-06,
+      "loss": 0.2317,
+      "step": 7806
+    },
+    {
+      "epoch": 1.8111587982832618,
+      "grad_norm": 13.13116391176701,
+      "learning_rate": 2e-06,
+      "loss": 0.2239,
+      "step": 7807
+    },
+    {
+      "epoch": 1.8113907899315624,
+      "grad_norm": 35.367434165456125,
+      "learning_rate": 2e-06,
+      "loss": 0.2811,
+      "step": 7808
+    },
+    {
+      "epoch": 1.8116227815798631,
+      "grad_norm": 13.284596823413086,
+      "learning_rate": 2e-06,
+      "loss": 0.218,
+      "step": 7809
+    },
+    {
+      "epoch": 1.8118547732281638,
+      "grad_norm": 25.05754023794442,
+      "learning_rate": 2e-06,
+      "loss": 0.199,
+      "step": 7810
+    },
+    {
+      "epoch": 1.8120867648764645,
+      "grad_norm": 11.14006570651281,
+      "learning_rate": 2e-06,
+      "loss": 0.1748,
+      "step": 7811
+    },
+    {
+      "epoch": 1.8123187565247652,
+      "grad_norm": 9.07722743677708,
+      "learning_rate": 2e-06,
+      "loss": 0.1929,
+      "step": 7812
+    },
+    {
+      "epoch": 1.8125507481730656,
+      "grad_norm": 22.349518160097013,
+      "learning_rate": 2e-06,
+      "loss": 0.2513,
+      "step": 7813
+    },
+    {
+      "epoch": 1.8127827398213663,
+      "grad_norm": 9.248339748264183,
+      "learning_rate": 2e-06,
+      "loss": 0.1483,
+      "step": 7814
+    },
+    {
+      "epoch": 1.813014731469667,
+      "grad_norm": 11.764457418025861,
+      "learning_rate": 2e-06,
+      "loss": 0.2851,
+      "step": 7815
+    },
+    {
+      "epoch": 1.8132467231179676,
+      "grad_norm": 12.019207291970828,
+      "learning_rate": 2e-06,
+      "loss": 0.2914,
+      "step": 7816
+    },
+    {
+      "epoch": 1.8134787147662683,
+      "grad_norm": 11.780327657611625,
+      "learning_rate": 2e-06,
+      "loss": 0.1862,
+      "step": 7817
+    },
+    {
+      "epoch": 1.813710706414569,
+      "grad_norm": 7.56300305679845,
+      "learning_rate": 2e-06,
+      "loss": 0.2186,
+      "step": 7818
+    },
+    {
+      "epoch": 1.8139426980628697,
+      "grad_norm": 18.84673033730254,
+      "learning_rate": 2e-06,
+      "loss": 0.2622,
+      "step": 7819
+    },
+    {
+      "epoch": 1.8141746897111704,
+      "grad_norm": 4.401220419426188,
+      "learning_rate": 2e-06,
+      "loss": 0.1358,
+      "step": 7820
+    },
+    {
+      "epoch": 1.814406681359471,
+      "grad_norm": 22.284767666525287,
+      "learning_rate": 2e-06,
+      "loss": 0.2526,
+      "step": 7821
+    },
+    {
+      "epoch": 1.8146386730077717,
+      "grad_norm": 18.64348610947482,
+      "learning_rate": 2e-06,
+      "loss": 0.2767,
+      "step": 7822
+    },
+    {
+      "epoch": 1.8148706646560724,
+      "grad_norm": 10.983716767173828,
+      "learning_rate": 2e-06,
+      "loss": 0.1831,
+      "step": 7823
+    },
+    {
+      "epoch": 1.815102656304373,
+      "grad_norm": 12.724066637808379,
+      "learning_rate": 2e-06,
+      "loss": 0.1808,
+      "step": 7824
+    },
+    {
+      "epoch": 1.8153346479526737,
+      "grad_norm": 14.024835862214248,
+      "learning_rate": 2e-06,
+      "loss": 0.2691,
+      "step": 7825
+    },
+    {
+      "epoch": 1.8155666396009744,
+      "grad_norm": 13.519710827991254,
+      "learning_rate": 2e-06,
+      "loss": 0.1943,
+      "step": 7826
+    },
+    {
+      "epoch": 1.815798631249275,
+      "grad_norm": 14.843148579657452,
+      "learning_rate": 2e-06,
+      "loss": 0.2213,
+      "step": 7827
+    },
+    {
+      "epoch": 1.8160306228975758,
+      "grad_norm": 9.460839761780512,
+      "learning_rate": 2e-06,
+      "loss": 0.1931,
+      "step": 7828
+    },
+    {
+      "epoch": 1.8162626145458765,
+      "grad_norm": 38.541938313911785,
+      "learning_rate": 2e-06,
+      "loss": 0.3944,
+      "step": 7829
+    },
+    {
+      "epoch": 1.8164946061941771,
+      "grad_norm": 13.17265386137172,
+      "learning_rate": 2e-06,
+      "loss": 0.2638,
+      "step": 7830
+    },
+    {
+      "epoch": 1.8167265978424778,
+      "grad_norm": 12.84058855474656,
+      "learning_rate": 2e-06,
+      "loss": 0.2178,
+      "step": 7831
+    },
+    {
+      "epoch": 1.8169585894907785,
+      "grad_norm": 10.788615152278911,
+      "learning_rate": 2e-06,
+      "loss": 0.221,
+      "step": 7832
+    },
+    {
+      "epoch": 1.817190581139079,
+      "grad_norm": 19.497970514954385,
+      "learning_rate": 2e-06,
+      "loss": 0.1933,
+      "step": 7833
+    },
+    {
+      "epoch": 1.8174225727873796,
+      "grad_norm": 11.516461157637663,
+      "learning_rate": 2e-06,
+      "loss": 0.1594,
+      "step": 7834
+    },
+    {
+      "epoch": 1.8176545644356803,
+      "grad_norm": 13.070370027025735,
+      "learning_rate": 2e-06,
+      "loss": 0.1997,
+      "step": 7835
+    },
+    {
+      "epoch": 1.817886556083981,
+      "grad_norm": 11.093947259075945,
+      "learning_rate": 2e-06,
+      "loss": 0.2507,
+      "step": 7836
+    },
+    {
+      "epoch": 1.8181185477322817,
+      "grad_norm": 10.952601265349044,
+      "learning_rate": 2e-06,
+      "loss": 0.2291,
+      "step": 7837
+    },
+    {
+      "epoch": 1.8183505393805823,
+      "grad_norm": 12.583982455505959,
+      "learning_rate": 2e-06,
+      "loss": 0.2542,
+      "step": 7838
+    },
+    {
+      "epoch": 1.818582531028883,
+      "grad_norm": 10.412530305980933,
+      "learning_rate": 2e-06,
+      "loss": 0.218,
+      "step": 7839
+    },
+    {
+      "epoch": 1.8188145226771835,
+      "grad_norm": 20.58663323966539,
+      "learning_rate": 2e-06,
+      "loss": 0.3298,
+      "step": 7840
+    },
+    {
+      "epoch": 1.8190465143254841,
+      "grad_norm": 16.069589171499164,
+      "learning_rate": 2e-06,
+      "loss": 0.2718,
+      "step": 7841
+    },
+    {
+      "epoch": 1.8192785059737848,
+      "grad_norm": 15.718424709122079,
+      "learning_rate": 2e-06,
+      "loss": 0.2185,
+      "step": 7842
+    },
+    {
+      "epoch": 1.8195104976220855,
+      "grad_norm": 13.114479648383073,
+      "learning_rate": 2e-06,
+      "loss": 0.2185,
+      "step": 7843
+    },
+    {
+      "epoch": 1.8197424892703862,
+      "grad_norm": 11.817145160415038,
+      "learning_rate": 2e-06,
+      "loss": 0.2357,
+      "step": 7844
+    },
+    {
+      "epoch": 1.8199744809186869,
+      "grad_norm": 8.646679220264772,
+      "learning_rate": 2e-06,
+      "loss": 0.1634,
+      "step": 7845
+    },
+    {
+      "epoch": 1.8202064725669875,
+      "grad_norm": 12.234848247512957,
+      "learning_rate": 2e-06,
+      "loss": 0.1666,
+      "step": 7846
+    },
+    {
+      "epoch": 1.8204384642152882,
+      "grad_norm": 22.072929455242598,
+      "learning_rate": 2e-06,
+      "loss": 0.4095,
+      "step": 7847
+    },
+    {
+      "epoch": 1.8206704558635889,
+      "grad_norm": 10.690070905760146,
+      "learning_rate": 2e-06,
+      "loss": 0.1838,
+      "step": 7848
+    },
+    {
+      "epoch": 1.8209024475118896,
+      "grad_norm": 11.534125053246434,
+      "learning_rate": 2e-06,
+      "loss": 0.2829,
+      "step": 7849
+    },
+    {
+      "epoch": 1.8211344391601902,
+      "grad_norm": 7.3850766019875955,
+      "learning_rate": 2e-06,
+      "loss": 0.1526,
+      "step": 7850
+    },
+    {
+      "epoch": 1.821366430808491,
+      "grad_norm": 10.193700861814932,
+      "learning_rate": 2e-06,
+      "loss": 0.1689,
+      "step": 7851
+    },
+    {
+      "epoch": 1.8215984224567916,
+      "grad_norm": 7.246342613215927,
+      "learning_rate": 2e-06,
+      "loss": 0.1663,
+      "step": 7852
+    },
+    {
+      "epoch": 1.8218304141050923,
+      "grad_norm": 10.911410337288949,
+      "learning_rate": 2e-06,
+      "loss": 0.2319,
+      "step": 7853
+    },
+    {
+      "epoch": 1.822062405753393,
+      "grad_norm": 28.141488972018617,
+      "learning_rate": 2e-06,
+      "loss": 0.281,
+      "step": 7854
+    },
+    {
+      "epoch": 1.8222943974016936,
+      "grad_norm": 15.119773914996467,
+      "learning_rate": 2e-06,
+      "loss": 0.2547,
+      "step": 7855
+    },
+    {
+      "epoch": 1.8225263890499943,
+      "grad_norm": 8.544803647608756,
+      "learning_rate": 2e-06,
+      "loss": 0.1434,
+      "step": 7856
+    },
+    {
+      "epoch": 1.822758380698295,
+      "grad_norm": 13.881430004783747,
+      "learning_rate": 2e-06,
+      "loss": 0.3103,
+      "step": 7857
+    },
+    {
+      "epoch": 1.8229903723465957,
+      "grad_norm": 20.911031864699904,
+      "learning_rate": 2e-06,
+      "loss": 0.3389,
+      "step": 7858
+    },
+    {
+      "epoch": 1.8232223639948963,
+      "grad_norm": 18.0973394831566,
+      "learning_rate": 2e-06,
+      "loss": 0.2147,
+      "step": 7859
+    },
+    {
+      "epoch": 1.8234543556431968,
+      "grad_norm": 8.356805904776557,
+      "learning_rate": 2e-06,
+      "loss": 0.1853,
+      "step": 7860
+    },
+    {
+      "epoch": 1.8236863472914975,
+      "grad_norm": 9.609173826516205,
+      "learning_rate": 2e-06,
+      "loss": 0.1511,
+      "step": 7861
+    },
+    {
+      "epoch": 1.8239183389397982,
+      "grad_norm": 14.17171157061349,
+      "learning_rate": 2e-06,
+      "loss": 0.3058,
+      "step": 7862
+    },
+    {
+      "epoch": 1.8241503305880988,
+      "grad_norm": 8.80752026693199,
+      "learning_rate": 2e-06,
+      "loss": 0.2101,
+      "step": 7863
+    },
+    {
+      "epoch": 1.8243823222363995,
+      "grad_norm": 7.813485615307996,
+      "learning_rate": 2e-06,
+      "loss": 0.2748,
+      "step": 7864
+    },
+    {
+      "epoch": 1.8246143138847002,
+      "grad_norm": 9.32680535990223,
+      "learning_rate": 2e-06,
+      "loss": 0.2522,
+      "step": 7865
+    },
+    {
+      "epoch": 1.8248463055330006,
+      "grad_norm": 9.896185768039878,
+      "learning_rate": 2e-06,
+      "loss": 0.1993,
+      "step": 7866
+    },
+    {
+      "epoch": 1.8250782971813013,
+      "grad_norm": 16.489932650037563,
+      "learning_rate": 2e-06,
+      "loss": 0.4062,
+      "step": 7867
+    },
+    {
+      "epoch": 1.825310288829602,
+      "grad_norm": 12.21555760793375,
+      "learning_rate": 2e-06,
+      "loss": 0.2114,
+      "step": 7868
+    },
+    {
+      "epoch": 1.8255422804779027,
+      "grad_norm": 10.668708934887045,
+      "learning_rate": 2e-06,
+      "loss": 0.2042,
+      "step": 7869
+    },
+    {
+      "epoch": 1.8257742721262034,
+      "grad_norm": 19.506231923470228,
+      "learning_rate": 2e-06,
+      "loss": 0.2417,
+      "step": 7870
+    },
+    {
+      "epoch": 1.826006263774504,
+      "grad_norm": 15.243980399678323,
+      "learning_rate": 2e-06,
+      "loss": 0.3222,
+      "step": 7871
+    },
+    {
+      "epoch": 1.8262382554228047,
+      "grad_norm": 8.266212958892414,
+      "learning_rate": 2e-06,
+      "loss": 0.1827,
+      "step": 7872
+    },
+    {
+      "epoch": 1.8264702470711054,
+      "grad_norm": 9.228484859678517,
+      "learning_rate": 2e-06,
+      "loss": 0.1577,
+      "step": 7873
+    },
+    {
+      "epoch": 1.826702238719406,
+      "grad_norm": 15.119166628188635,
+      "learning_rate": 2e-06,
+      "loss": 0.1612,
+      "step": 7874
+    },
+    {
+      "epoch": 1.8269342303677067,
+      "grad_norm": 7.614144747643755,
+      "learning_rate": 2e-06,
+      "loss": 0.1429,
+      "step": 7875
+    },
+    {
+      "epoch": 1.8271662220160074,
+      "grad_norm": 13.10715413771453,
+      "learning_rate": 2e-06,
+      "loss": 0.2377,
+      "step": 7876
+    },
+    {
+      "epoch": 1.827398213664308,
+      "grad_norm": 17.644967445732938,
+      "learning_rate": 2e-06,
+      "loss": 0.2515,
+      "step": 7877
+    },
+    {
+      "epoch": 1.8276302053126088,
+      "grad_norm": 19.092950755612446,
+      "learning_rate": 2e-06,
+      "loss": 0.3464,
+      "step": 7878
+    },
+    {
+      "epoch": 1.8278621969609095,
+      "grad_norm": 12.387344155425442,
+      "learning_rate": 2e-06,
+      "loss": 0.3138,
+      "step": 7879
+    },
+    {
+      "epoch": 1.8280941886092101,
+      "grad_norm": 14.011979558604576,
+      "learning_rate": 2e-06,
+      "loss": 0.2179,
+      "step": 7880
+    },
+    {
+      "epoch": 1.8283261802575108,
+      "grad_norm": 15.390569673859822,
+      "learning_rate": 2e-06,
+      "loss": 0.2641,
+      "step": 7881
+    },
+    {
+      "epoch": 1.8285581719058115,
+      "grad_norm": 14.684780900009232,
+      "learning_rate": 2e-06,
+      "loss": 0.1935,
+      "step": 7882
+    },
+    {
+      "epoch": 1.8287901635541122,
+      "grad_norm": 11.430777807918494,
+      "learning_rate": 2e-06,
+      "loss": 0.18,
+      "step": 7883
+    },
+    {
+      "epoch": 1.8290221552024128,
+      "grad_norm": 7.791418141941399,
+      "learning_rate": 2e-06,
+      "loss": 0.1834,
+      "step": 7884
+    },
+    {
+      "epoch": 1.8292541468507135,
+      "grad_norm": 8.64636567131907,
+      "learning_rate": 2e-06,
+      "loss": 0.1954,
+      "step": 7885
+    },
+    {
+      "epoch": 1.829486138499014,
+      "grad_norm": 15.346495917999354,
+      "learning_rate": 2e-06,
+      "loss": 0.2054,
+      "step": 7886
+    },
+    {
+      "epoch": 1.8297181301473147,
+      "grad_norm": 12.296776262277993,
+      "learning_rate": 2e-06,
+      "loss": 0.1336,
+      "step": 7887
+    },
+    {
+      "epoch": 1.8299501217956153,
+      "grad_norm": 19.494222119023746,
+      "learning_rate": 2e-06,
+      "loss": 0.2837,
+      "step": 7888
+    },
+    {
+      "epoch": 1.830182113443916,
+      "grad_norm": 13.25582425439502,
+      "learning_rate": 2e-06,
+      "loss": 0.3047,
+      "step": 7889
+    },
+    {
+      "epoch": 1.8304141050922167,
+      "grad_norm": 9.671077020296584,
+      "learning_rate": 2e-06,
+      "loss": 0.1777,
+      "step": 7890
+    },
+    {
+      "epoch": 1.8306460967405174,
+      "grad_norm": 10.511830121222385,
+      "learning_rate": 2e-06,
+      "loss": 0.1911,
+      "step": 7891
+    },
+    {
+      "epoch": 1.830878088388818,
+      "grad_norm": 12.49834553629901,
+      "learning_rate": 2e-06,
+      "loss": 0.2534,
+      "step": 7892
+    },
+    {
+      "epoch": 1.8311100800371185,
+      "grad_norm": 17.583887619416416,
+      "learning_rate": 2e-06,
+      "loss": 0.2884,
+      "step": 7893
+    },
+    {
+      "epoch": 1.8313420716854192,
+      "grad_norm": 13.841502274963958,
+      "learning_rate": 2e-06,
+      "loss": 0.1891,
+      "step": 7894
+    },
+    {
+      "epoch": 1.8315740633337199,
+      "grad_norm": 12.67870442340447,
+      "learning_rate": 2e-06,
+      "loss": 0.2628,
+      "step": 7895
+    },
+    {
+      "epoch": 1.8318060549820205,
+      "grad_norm": 16.384562089918386,
+      "learning_rate": 2e-06,
+      "loss": 0.2482,
+      "step": 7896
+    },
+    {
+      "epoch": 1.8320380466303212,
+      "grad_norm": 16.046864886905126,
+      "learning_rate": 2e-06,
+      "loss": 0.2183,
+      "step": 7897
+    },
+    {
+      "epoch": 1.832270038278622,
+      "grad_norm": 11.993529864813677,
+      "learning_rate": 2e-06,
+      "loss": 0.2744,
+      "step": 7898
+    },
+    {
+      "epoch": 1.8325020299269226,
+      "grad_norm": 16.457508380663167,
+      "learning_rate": 2e-06,
+      "loss": 0.3141,
+      "step": 7899
+    },
+    {
+      "epoch": 1.8327340215752232,
+      "grad_norm": 11.896321643614053,
+      "learning_rate": 2e-06,
+      "loss": 0.1972,
+      "step": 7900
+    },
+    {
+      "epoch": 1.832966013223524,
+      "grad_norm": 10.634033931621662,
+      "learning_rate": 2e-06,
+      "loss": 0.2399,
+      "step": 7901
+    },
+    {
+      "epoch": 1.8331980048718246,
+      "grad_norm": 17.425657647560143,
+      "learning_rate": 2e-06,
+      "loss": 0.305,
+      "step": 7902
+    },
+    {
+      "epoch": 1.8334299965201253,
+      "grad_norm": 7.202441697727795,
+      "learning_rate": 2e-06,
+      "loss": 0.1521,
+      "step": 7903
+    },
+    {
+      "epoch": 1.833661988168426,
+      "grad_norm": 17.19205030465811,
+      "learning_rate": 2e-06,
+      "loss": 0.2845,
+      "step": 7904
+    },
+    {
+      "epoch": 1.8338939798167266,
+      "grad_norm": 19.90010493299395,
+      "learning_rate": 2e-06,
+      "loss": 0.2641,
+      "step": 7905
+    },
+    {
+      "epoch": 1.8341259714650273,
+      "grad_norm": 19.52854387682745,
+      "learning_rate": 2e-06,
+      "loss": 0.3048,
+      "step": 7906
+    },
+    {
+      "epoch": 1.834357963113328,
+      "grad_norm": 10.902417230263483,
+      "learning_rate": 2e-06,
+      "loss": 0.2184,
+      "step": 7907
+    },
+    {
+      "epoch": 1.8345899547616287,
+      "grad_norm": 17.362599545004098,
+      "learning_rate": 2e-06,
+      "loss": 0.2387,
+      "step": 7908
+    },
+    {
+      "epoch": 1.8348219464099293,
+      "grad_norm": 16.18254568800934,
+      "learning_rate": 2e-06,
+      "loss": 0.2198,
+      "step": 7909
+    },
+    {
+      "epoch": 1.83505393805823,
+      "grad_norm": 27.008496402497844,
+      "learning_rate": 2e-06,
+      "loss": 0.5212,
+      "step": 7910
+    },
+    {
+      "epoch": 1.8352859297065307,
+      "grad_norm": 23.159668358951915,
+      "learning_rate": 2e-06,
+      "loss": 0.3255,
+      "step": 7911
+    },
+    {
+      "epoch": 1.8355179213548314,
+      "grad_norm": 21.345873761286768,
+      "learning_rate": 2e-06,
+      "loss": 0.2788,
+      "step": 7912
+    },
+    {
+      "epoch": 1.8357499130031318,
+      "grad_norm": 16.364693029156975,
+      "learning_rate": 2e-06,
+      "loss": 0.2494,
+      "step": 7913
+    },
+    {
+      "epoch": 1.8359819046514325,
+      "grad_norm": 21.265562874143434,
+      "learning_rate": 2e-06,
+      "loss": 0.2334,
+      "step": 7914
+    },
+    {
+      "epoch": 1.8362138962997332,
+      "grad_norm": 8.42226495039441,
+      "learning_rate": 2e-06,
+      "loss": 0.2178,
+      "step": 7915
+    },
+    {
+      "epoch": 1.8364458879480339,
+      "grad_norm": 14.120536069491699,
+      "learning_rate": 2e-06,
+      "loss": 0.2106,
+      "step": 7916
+    },
+    {
+      "epoch": 1.8366778795963346,
+      "grad_norm": 16.753426930658062,
+      "learning_rate": 2e-06,
+      "loss": 0.2631,
+      "step": 7917
+    },
+    {
+      "epoch": 1.8369098712446352,
+      "grad_norm": 9.596391419187736,
+      "learning_rate": 2e-06,
+      "loss": 0.1716,
+      "step": 7918
+    },
+    {
+      "epoch": 1.837141862892936,
+      "grad_norm": 14.833436969950283,
+      "learning_rate": 2e-06,
+      "loss": 0.2554,
+      "step": 7919
+    },
+    {
+      "epoch": 1.8373738545412364,
+      "grad_norm": 6.294776122050837,
+      "learning_rate": 2e-06,
+      "loss": 0.1314,
+      "step": 7920
+    },
+    {
+      "epoch": 1.837605846189537,
+      "grad_norm": 8.58136538816046,
+      "learning_rate": 2e-06,
+      "loss": 0.1775,
+      "step": 7921
+    },
+    {
+      "epoch": 1.8378378378378377,
+      "grad_norm": 6.839034671602942,
+      "learning_rate": 2e-06,
+      "loss": 0.147,
+      "step": 7922
+    },
+    {
+      "epoch": 1.8380698294861384,
+      "grad_norm": 11.75226477118561,
+      "learning_rate": 2e-06,
+      "loss": 0.2727,
+      "step": 7923
+    },
+    {
+      "epoch": 1.838301821134439,
+      "grad_norm": 16.951382202621378,
+      "learning_rate": 2e-06,
+      "loss": 0.387,
+      "step": 7924
+    },
+    {
+      "epoch": 1.8385338127827398,
+      "grad_norm": 9.669318688045863,
+      "learning_rate": 2e-06,
+      "loss": 0.2122,
+      "step": 7925
+    },
+    {
+      "epoch": 1.8387658044310404,
+      "grad_norm": 10.658857823286702,
+      "learning_rate": 2e-06,
+      "loss": 0.1693,
+      "step": 7926
+    },
+    {
+      "epoch": 1.838997796079341,
+      "grad_norm": 11.164269201204279,
+      "learning_rate": 2e-06,
+      "loss": 0.2675,
+      "step": 7927
+    },
+    {
+      "epoch": 1.8392297877276418,
+      "grad_norm": 11.292803059157091,
+      "learning_rate": 2e-06,
+      "loss": 0.3351,
+      "step": 7928
+    },
+    {
+      "epoch": 1.8394617793759425,
+      "grad_norm": 13.662105944550406,
+      "learning_rate": 2e-06,
+      "loss": 0.2295,
+      "step": 7929
+    },
+    {
+      "epoch": 1.8396937710242431,
+      "grad_norm": 9.984871014211448,
+      "learning_rate": 2e-06,
+      "loss": 0.1737,
+      "step": 7930
+    },
+    {
+      "epoch": 1.8399257626725438,
+      "grad_norm": 15.528006175833218,
+      "learning_rate": 2e-06,
+      "loss": 0.2634,
+      "step": 7931
+    },
+    {
+      "epoch": 1.8401577543208445,
+      "grad_norm": 11.802930981454681,
+      "learning_rate": 2e-06,
+      "loss": 0.3148,
+      "step": 7932
+    },
+    {
+      "epoch": 1.8403897459691452,
+      "grad_norm": 25.823960323458206,
+      "learning_rate": 2e-06,
+      "loss": 0.3875,
+      "step": 7933
+    },
+    {
+      "epoch": 1.8406217376174459,
+      "grad_norm": 5.4731564900955645,
+      "learning_rate": 2e-06,
+      "loss": 0.1652,
+      "step": 7934
+    },
+    {
+      "epoch": 1.8408537292657465,
+      "grad_norm": 12.465598680342927,
+      "learning_rate": 2e-06,
+      "loss": 0.2969,
+      "step": 7935
+    },
+    {
+      "epoch": 1.8410857209140472,
+      "grad_norm": 10.890761811575437,
+      "learning_rate": 2e-06,
+      "loss": 0.1851,
+      "step": 7936
+    },
+    {
+      "epoch": 1.8413177125623479,
+      "grad_norm": 11.075531771245522,
+      "learning_rate": 2e-06,
+      "loss": 0.19,
+      "step": 7937
+    },
+    {
+      "epoch": 1.8415497042106486,
+      "grad_norm": 10.76191718337455,
+      "learning_rate": 2e-06,
+      "loss": 0.2794,
+      "step": 7938
+    },
+    {
+      "epoch": 1.8417816958589492,
+      "grad_norm": 13.447884654447204,
+      "learning_rate": 2e-06,
+      "loss": 0.2633,
+      "step": 7939
+    },
+    {
+      "epoch": 1.8420136875072497,
+      "grad_norm": 8.35862362187622,
+      "learning_rate": 2e-06,
+      "loss": 0.2139,
+      "step": 7940
+    },
+    {
+      "epoch": 1.8422456791555504,
+      "grad_norm": 11.615926552765899,
+      "learning_rate": 2e-06,
+      "loss": 0.1836,
+      "step": 7941
+    },
+    {
+      "epoch": 1.842477670803851,
+      "grad_norm": 16.797906163493213,
+      "learning_rate": 2e-06,
+      "loss": 0.2207,
+      "step": 7942
+    },
+    {
+      "epoch": 1.8427096624521517,
+      "grad_norm": 11.526737998281494,
+      "learning_rate": 2e-06,
+      "loss": 0.2684,
+      "step": 7943
+    },
+    {
+      "epoch": 1.8429416541004524,
+      "grad_norm": 14.813675655160232,
+      "learning_rate": 2e-06,
+      "loss": 0.2797,
+      "step": 7944
+    },
+    {
+      "epoch": 1.843173645748753,
+      "grad_norm": 16.873756448801746,
+      "learning_rate": 2e-06,
+      "loss": 0.1525,
+      "step": 7945
+    },
+    {
+      "epoch": 1.8434056373970535,
+      "grad_norm": 8.490637239098254,
+      "learning_rate": 2e-06,
+      "loss": 0.1979,
+      "step": 7946
+    },
+    {
+      "epoch": 1.8436376290453542,
+      "grad_norm": 8.188933475422282,
+      "learning_rate": 2e-06,
+      "loss": 0.1951,
+      "step": 7947
+    },
+    {
+      "epoch": 1.843869620693655,
+      "grad_norm": 12.989538195546992,
+      "learning_rate": 2e-06,
+      "loss": 0.2963,
+      "step": 7948
+    },
+    {
+      "epoch": 1.8441016123419556,
+      "grad_norm": 25.921803203614267,
+      "learning_rate": 2e-06,
+      "loss": 0.2511,
+      "step": 7949
+    },
+    {
+      "epoch": 1.8443336039902563,
+      "grad_norm": 13.38270787112594,
+      "learning_rate": 2e-06,
+      "loss": 0.278,
+      "step": 7950
+    },
+    {
+      "epoch": 1.844565595638557,
+      "grad_norm": 15.901728624066433,
+      "learning_rate": 2e-06,
+      "loss": 0.2949,
+      "step": 7951
+    },
+    {
+      "epoch": 1.8447975872868576,
+      "grad_norm": 30.953565323139976,
+      "learning_rate": 2e-06,
+      "loss": 0.3273,
+      "step": 7952
+    },
+    {
+      "epoch": 1.8450295789351583,
+      "grad_norm": 12.810216707183498,
+      "learning_rate": 2e-06,
+      "loss": 0.1859,
+      "step": 7953
+    },
+    {
+      "epoch": 1.845261570583459,
+      "grad_norm": 11.592796853621131,
+      "learning_rate": 2e-06,
+      "loss": 0.1906,
+      "step": 7954
+    },
+    {
+      "epoch": 1.8454935622317596,
+      "grad_norm": 8.935925628117879,
+      "learning_rate": 2e-06,
+      "loss": 0.2226,
+      "step": 7955
+    },
+    {
+      "epoch": 1.8457255538800603,
+      "grad_norm": 9.500849218443497,
+      "learning_rate": 2e-06,
+      "loss": 0.215,
+      "step": 7956
+    },
+    {
+      "epoch": 1.845957545528361,
+      "grad_norm": 10.944130618833324,
+      "learning_rate": 2e-06,
+      "loss": 0.2248,
+      "step": 7957
+    },
+    {
+      "epoch": 1.8461895371766617,
+      "grad_norm": 9.18069076395506,
+      "learning_rate": 2e-06,
+      "loss": 0.1856,
+      "step": 7958
+    },
+    {
+      "epoch": 1.8464215288249624,
+      "grad_norm": 9.685371708277902,
+      "learning_rate": 2e-06,
+      "loss": 0.1926,
+      "step": 7959
+    },
+    {
+      "epoch": 1.846653520473263,
+      "grad_norm": 26.997531508708946,
+      "learning_rate": 2e-06,
+      "loss": 0.3127,
+      "step": 7960
+    },
+    {
+      "epoch": 1.8468855121215637,
+      "grad_norm": 12.585809388667009,
+      "learning_rate": 2e-06,
+      "loss": 0.1861,
+      "step": 7961
+    },
+    {
+      "epoch": 1.8471175037698644,
+      "grad_norm": 11.3527907629372,
+      "learning_rate": 2e-06,
+      "loss": 0.2552,
+      "step": 7962
+    },
+    {
+      "epoch": 1.847349495418165,
+      "grad_norm": 23.095925633732367,
+      "learning_rate": 2e-06,
+      "loss": 0.3313,
+      "step": 7963
+    },
+    {
+      "epoch": 1.8475814870664657,
+      "grad_norm": 13.084226012680856,
+      "learning_rate": 2e-06,
+      "loss": 0.2949,
+      "step": 7964
+    },
+    {
+      "epoch": 1.8478134787147664,
+      "grad_norm": 18.085654292620408,
+      "learning_rate": 2e-06,
+      "loss": 0.3053,
+      "step": 7965
+    },
+    {
+      "epoch": 1.8480454703630669,
+      "grad_norm": 14.712791411772923,
+      "learning_rate": 2e-06,
+      "loss": 0.3434,
+      "step": 7966
+    },
+    {
+      "epoch": 1.8482774620113676,
+      "grad_norm": 8.334839560631009,
+      "learning_rate": 2e-06,
+      "loss": 0.1873,
+      "step": 7967
+    },
+    {
+      "epoch": 1.8485094536596682,
+      "grad_norm": 10.518803209871267,
+      "learning_rate": 2e-06,
+      "loss": 0.3561,
+      "step": 7968
+    },
+    {
+      "epoch": 1.848741445307969,
+      "grad_norm": 22.56750349666928,
+      "learning_rate": 2e-06,
+      "loss": 0.286,
+      "step": 7969
+    },
+    {
+      "epoch": 1.8489734369562696,
+      "grad_norm": 18.491907036686882,
+      "learning_rate": 2e-06,
+      "loss": 0.321,
+      "step": 7970
+    },
+    {
+      "epoch": 1.8492054286045703,
+      "grad_norm": 10.100982824992236,
+      "learning_rate": 2e-06,
+      "loss": 0.1683,
+      "step": 7971
+    },
+    {
+      "epoch": 1.849437420252871,
+      "grad_norm": 15.074645252715337,
+      "learning_rate": 2e-06,
+      "loss": 0.2232,
+      "step": 7972
+    },
+    {
+      "epoch": 1.8496694119011714,
+      "grad_norm": 14.046821339662545,
+      "learning_rate": 2e-06,
+      "loss": 0.2664,
+      "step": 7973
+    },
+    {
+      "epoch": 1.849901403549472,
+      "grad_norm": 13.59824054448896,
+      "learning_rate": 2e-06,
+      "loss": 0.3291,
+      "step": 7974
+    },
+    {
+      "epoch": 1.8501333951977728,
+      "grad_norm": 13.790300671692236,
+      "learning_rate": 2e-06,
+      "loss": 0.2642,
+      "step": 7975
+    },
+    {
+      "epoch": 1.8503653868460734,
+      "grad_norm": 10.412428158867494,
+      "learning_rate": 2e-06,
+      "loss": 0.1941,
+      "step": 7976
+    },
+    {
+      "epoch": 1.8505973784943741,
+      "grad_norm": 9.241196274154442,
+      "learning_rate": 2e-06,
+      "loss": 0.2431,
+      "step": 7977
+    },
+    {
+      "epoch": 1.8508293701426748,
+      "grad_norm": 18.006977742518487,
+      "learning_rate": 2e-06,
+      "loss": 0.3314,
+      "step": 7978
+    },
+    {
+      "epoch": 1.8510613617909755,
+      "grad_norm": 8.735578702425071,
+      "learning_rate": 2e-06,
+      "loss": 0.1915,
+      "step": 7979
+    },
+    {
+      "epoch": 1.8512933534392761,
+      "grad_norm": 11.066825745768144,
+      "learning_rate": 2e-06,
+      "loss": 0.3201,
+      "step": 7980
+    },
+    {
+      "epoch": 1.8515253450875768,
+      "grad_norm": 12.934799799849172,
+      "learning_rate": 2e-06,
+      "loss": 0.2668,
+      "step": 7981
+    },
+    {
+      "epoch": 1.8517573367358775,
+      "grad_norm": 7.652593730052824,
+      "learning_rate": 2e-06,
+      "loss": 0.2211,
+      "step": 7982
+    },
+    {
+      "epoch": 1.8519893283841782,
+      "grad_norm": 17.360645095568245,
+      "learning_rate": 2e-06,
+      "loss": 0.2471,
+      "step": 7983
+    },
+    {
+      "epoch": 1.8522213200324789,
+      "grad_norm": 9.342320938908026,
+      "learning_rate": 2e-06,
+      "loss": 0.2092,
+      "step": 7984
+    },
+    {
+      "epoch": 1.8524533116807795,
+      "grad_norm": 28.26538822948439,
+      "learning_rate": 2e-06,
+      "loss": 0.3829,
+      "step": 7985
+    },
+    {
+      "epoch": 1.8526853033290802,
+      "grad_norm": 12.575102199523327,
+      "learning_rate": 2e-06,
+      "loss": 0.2662,
+      "step": 7986
+    },
+    {
+      "epoch": 1.852917294977381,
+      "grad_norm": 12.649445490250741,
+      "learning_rate": 2e-06,
+      "loss": 0.2228,
+      "step": 7987
+    },
+    {
+      "epoch": 1.8531492866256816,
+      "grad_norm": 10.65686234012664,
+      "learning_rate": 2e-06,
+      "loss": 0.2862,
+      "step": 7988
+    },
+    {
+      "epoch": 1.8533812782739822,
+      "grad_norm": 13.377746222563427,
+      "learning_rate": 2e-06,
+      "loss": 0.2825,
+      "step": 7989
+    },
+    {
+      "epoch": 1.853613269922283,
+      "grad_norm": 8.499970384087424,
+      "learning_rate": 2e-06,
+      "loss": 0.2493,
+      "step": 7990
+    },
+    {
+      "epoch": 1.8538452615705836,
+      "grad_norm": 8.641111636956055,
+      "learning_rate": 2e-06,
+      "loss": 0.1472,
+      "step": 7991
+    },
+    {
+      "epoch": 1.8540772532188843,
+      "grad_norm": 18.614890935703087,
+      "learning_rate": 2e-06,
+      "loss": 0.2897,
+      "step": 7992
+    },
+    {
+      "epoch": 1.8543092448671847,
+      "grad_norm": 9.830085780138903,
+      "learning_rate": 2e-06,
+      "loss": 0.2636,
+      "step": 7993
+    },
+    {
+      "epoch": 1.8545412365154854,
+      "grad_norm": 7.283981619222619,
+      "learning_rate": 2e-06,
+      "loss": 0.2193,
+      "step": 7994
+    },
+    {
+      "epoch": 1.854773228163786,
+      "grad_norm": 16.965974276924957,
+      "learning_rate": 2e-06,
+      "loss": 0.2347,
+      "step": 7995
+    },
+    {
+      "epoch": 1.8550052198120868,
+      "grad_norm": 11.141632499056673,
+      "learning_rate": 2e-06,
+      "loss": 0.2715,
+      "step": 7996
+    },
+    {
+      "epoch": 1.8552372114603874,
+      "grad_norm": 13.656532823530375,
+      "learning_rate": 2e-06,
+      "loss": 0.2832,
+      "step": 7997
+    },
+    {
+      "epoch": 1.8554692031086881,
+      "grad_norm": 8.84745272325778,
+      "learning_rate": 2e-06,
+      "loss": 0.2077,
+      "step": 7998
+    },
+    {
+      "epoch": 1.8557011947569888,
+      "grad_norm": 9.806319010310668,
+      "learning_rate": 2e-06,
+      "loss": 0.2299,
+      "step": 7999
+    },
+    {
+      "epoch": 1.8559331864052893,
+      "grad_norm": 8.63698914343859,
+      "learning_rate": 2e-06,
+      "loss": 0.2444,
+      "step": 8000
+    },
+    {
+      "epoch": 1.85616517805359,
+      "grad_norm": 19.746684480722426,
+      "learning_rate": 2e-06,
+      "loss": 0.3618,
+      "step": 8001
+    },
+    {
+      "epoch": 1.8563971697018906,
+      "grad_norm": 10.050691436850114,
+      "learning_rate": 2e-06,
+      "loss": 0.2331,
+      "step": 8002
+    },
+    {
+      "epoch": 1.8566291613501913,
+      "grad_norm": 23.133285568263815,
+      "learning_rate": 2e-06,
+      "loss": 0.2847,
+      "step": 8003
+    },
+    {
+      "epoch": 1.856861152998492,
+      "grad_norm": 9.853651040793197,
+      "learning_rate": 2e-06,
+      "loss": 0.1745,
+      "step": 8004
+    },
+    {
+      "epoch": 1.8570931446467926,
+      "grad_norm": 6.07637328157186,
+      "learning_rate": 2e-06,
+      "loss": 0.1538,
+      "step": 8005
+    },
+    {
+      "epoch": 1.8573251362950933,
+      "grad_norm": 10.665413106483905,
+      "learning_rate": 2e-06,
+      "loss": 0.303,
+      "step": 8006
+    },
+    {
+      "epoch": 1.857557127943394,
+      "grad_norm": 5.973799154259255,
+      "learning_rate": 2e-06,
+      "loss": 0.1648,
+      "step": 8007
+    },
+    {
+      "epoch": 1.8577891195916947,
+      "grad_norm": 11.260741284363519,
+      "learning_rate": 2e-06,
+      "loss": 0.2539,
+      "step": 8008
+    },
+    {
+      "epoch": 1.8580211112399954,
+      "grad_norm": 12.612140361854662,
+      "learning_rate": 2e-06,
+      "loss": 0.1837,
+      "step": 8009
+    },
+    {
+      "epoch": 1.858253102888296,
+      "grad_norm": 6.64501874593489,
+      "learning_rate": 2e-06,
+      "loss": 0.127,
+      "step": 8010
+    },
+    {
+      "epoch": 1.8584850945365967,
+      "grad_norm": 12.641913011016891,
+      "learning_rate": 2e-06,
+      "loss": 0.2472,
+      "step": 8011
+    },
+    {
+      "epoch": 1.8587170861848974,
+      "grad_norm": 16.01004189377708,
+      "learning_rate": 2e-06,
+      "loss": 0.2818,
+      "step": 8012
+    },
+    {
+      "epoch": 1.858949077833198,
+      "grad_norm": 17.650523284827663,
+      "learning_rate": 2e-06,
+      "loss": 0.3115,
+      "step": 8013
+    },
+    {
+      "epoch": 1.8591810694814987,
+      "grad_norm": 67.9849712980046,
+      "learning_rate": 2e-06,
+      "loss": 0.2104,
+      "step": 8014
+    },
+    {
+      "epoch": 1.8594130611297994,
+      "grad_norm": 18.553952981423404,
+      "learning_rate": 2e-06,
+      "loss": 0.3201,
+      "step": 8015
+    },
+    {
+      "epoch": 1.8596450527781,
+      "grad_norm": 19.8359578410062,
+      "learning_rate": 2e-06,
+      "loss": 0.2287,
+      "step": 8016
+    },
+    {
+      "epoch": 1.8598770444264008,
+      "grad_norm": 13.506166327338736,
+      "learning_rate": 2e-06,
+      "loss": 0.2029,
+      "step": 8017
+    },
+    {
+      "epoch": 1.8601090360747015,
+      "grad_norm": 13.173351169007882,
+      "learning_rate": 2e-06,
+      "loss": 0.1775,
+      "step": 8018
+    },
+    {
+      "epoch": 1.860341027723002,
+      "grad_norm": 16.9773971290991,
+      "learning_rate": 2e-06,
+      "loss": 0.2437,
+      "step": 8019
+    },
+    {
+      "epoch": 1.8605730193713026,
+      "grad_norm": 11.166665045313033,
+      "learning_rate": 2e-06,
+      "loss": 0.2486,
+      "step": 8020
+    },
+    {
+      "epoch": 1.8608050110196033,
+      "grad_norm": 17.771633832862648,
+      "learning_rate": 2e-06,
+      "loss": 0.3317,
+      "step": 8021
+    },
+    {
+      "epoch": 1.861037002667904,
+      "grad_norm": 13.650262107790912,
+      "learning_rate": 2e-06,
+      "loss": 0.2381,
+      "step": 8022
+    },
+    {
+      "epoch": 1.8612689943162046,
+      "grad_norm": 14.93861953518762,
+      "learning_rate": 2e-06,
+      "loss": 0.2917,
+      "step": 8023
+    },
+    {
+      "epoch": 1.8615009859645053,
+      "grad_norm": 50.26662400292754,
+      "learning_rate": 2e-06,
+      "loss": 0.1945,
+      "step": 8024
+    },
+    {
+      "epoch": 1.861732977612806,
+      "grad_norm": 11.603664768722052,
+      "learning_rate": 2e-06,
+      "loss": 0.3147,
+      "step": 8025
+    },
+    {
+      "epoch": 1.8619649692611064,
+      "grad_norm": 12.386479170427224,
+      "learning_rate": 2e-06,
+      "loss": 0.201,
+      "step": 8026
+    },
+    {
+      "epoch": 1.8621969609094071,
+      "grad_norm": 20.00899319309306,
+      "learning_rate": 2e-06,
+      "loss": 0.2685,
+      "step": 8027
+    },
+    {
+      "epoch": 1.8624289525577078,
+      "grad_norm": 14.307786611545218,
+      "learning_rate": 2e-06,
+      "loss": 0.3113,
+      "step": 8028
+    },
+    {
+      "epoch": 1.8626609442060085,
+      "grad_norm": 10.419544932315354,
+      "learning_rate": 2e-06,
+      "loss": 0.2153,
+      "step": 8029
+    },
+    {
+      "epoch": 1.8628929358543092,
+      "grad_norm": 14.944128238574825,
+      "learning_rate": 2e-06,
+      "loss": 0.2304,
+      "step": 8030
+    },
+    {
+      "epoch": 1.8631249275026098,
+      "grad_norm": 14.828493347874334,
+      "learning_rate": 2e-06,
+      "loss": 0.2339,
+      "step": 8031
+    },
+    {
+      "epoch": 1.8633569191509105,
+      "grad_norm": 8.43164667293264,
+      "learning_rate": 2e-06,
+      "loss": 0.1578,
+      "step": 8032
+    },
+    {
+      "epoch": 1.8635889107992112,
+      "grad_norm": 8.225180499158036,
+      "learning_rate": 2e-06,
+      "loss": 0.2174,
+      "step": 8033
+    },
+    {
+      "epoch": 1.8638209024475119,
+      "grad_norm": 8.714004111926837,
+      "learning_rate": 2e-06,
+      "loss": 0.1848,
+      "step": 8034
+    },
+    {
+      "epoch": 1.8640528940958125,
+      "grad_norm": 7.209489143883522,
+      "learning_rate": 2e-06,
+      "loss": 0.1175,
+      "step": 8035
+    },
+    {
+      "epoch": 1.8642848857441132,
+      "grad_norm": 19.47233075136568,
+      "learning_rate": 2e-06,
+      "loss": 0.3315,
+      "step": 8036
+    },
+    {
+      "epoch": 1.864516877392414,
+      "grad_norm": 11.761565148017262,
+      "learning_rate": 2e-06,
+      "loss": 0.2358,
+      "step": 8037
+    },
+    {
+      "epoch": 1.8647488690407146,
+      "grad_norm": 22.946883432516643,
+      "learning_rate": 2e-06,
+      "loss": 0.2992,
+      "step": 8038
+    },
+    {
+      "epoch": 1.8649808606890153,
+      "grad_norm": 10.099079023837012,
+      "learning_rate": 2e-06,
+      "loss": 0.2294,
+      "step": 8039
+    },
+    {
+      "epoch": 1.865212852337316,
+      "grad_norm": 11.982340199964142,
+      "learning_rate": 2e-06,
+      "loss": 0.225,
+      "step": 8040
+    },
+    {
+      "epoch": 1.8654448439856166,
+      "grad_norm": 7.414469724974377,
+      "learning_rate": 2e-06,
+      "loss": 0.1843,
+      "step": 8041
+    },
+    {
+      "epoch": 1.8656768356339173,
+      "grad_norm": 14.110442842535285,
+      "learning_rate": 2e-06,
+      "loss": 0.3108,
+      "step": 8042
+    },
+    {
+      "epoch": 1.865908827282218,
+      "grad_norm": 8.306951984448888,
+      "learning_rate": 2e-06,
+      "loss": 0.222,
+      "step": 8043
+    },
+    {
+      "epoch": 1.8661408189305186,
+      "grad_norm": 8.020722441848871,
+      "learning_rate": 2e-06,
+      "loss": 0.2296,
+      "step": 8044
+    },
+    {
+      "epoch": 1.8663728105788193,
+      "grad_norm": 14.329130788383804,
+      "learning_rate": 2e-06,
+      "loss": 0.3023,
+      "step": 8045
+    },
+    {
+      "epoch": 1.8666048022271198,
+      "grad_norm": 12.243991467568708,
+      "learning_rate": 2e-06,
+      "loss": 0.2126,
+      "step": 8046
+    },
+    {
+      "epoch": 1.8668367938754205,
+      "grad_norm": 11.49317180930521,
+      "learning_rate": 2e-06,
+      "loss": 0.1405,
+      "step": 8047
+    },
+    {
+      "epoch": 1.8670687855237211,
+      "grad_norm": 12.517361716203531,
+      "learning_rate": 2e-06,
+      "loss": 0.2399,
+      "step": 8048
+    },
+    {
+      "epoch": 1.8673007771720218,
+      "grad_norm": 14.413158198703666,
+      "learning_rate": 2e-06,
+      "loss": 0.2407,
+      "step": 8049
+    },
+    {
+      "epoch": 1.8675327688203225,
+      "grad_norm": 20.62454734785914,
+      "learning_rate": 2e-06,
+      "loss": 0.224,
+      "step": 8050
+    },
+    {
+      "epoch": 1.8677647604686232,
+      "grad_norm": 11.221516745189199,
+      "learning_rate": 2e-06,
+      "loss": 0.2663,
+      "step": 8051
+    },
+    {
+      "epoch": 1.8679967521169238,
+      "grad_norm": 14.150459004169447,
+      "learning_rate": 2e-06,
+      "loss": 0.2292,
+      "step": 8052
+    },
+    {
+      "epoch": 1.8682287437652243,
+      "grad_norm": 16.522724664226192,
+      "learning_rate": 2e-06,
+      "loss": 0.2884,
+      "step": 8053
+    },
+    {
+      "epoch": 1.868460735413525,
+      "grad_norm": 11.491441717510634,
+      "learning_rate": 2e-06,
+      "loss": 0.2876,
+      "step": 8054
+    },
+    {
+      "epoch": 1.8686927270618257,
+      "grad_norm": 19.27225227412006,
+      "learning_rate": 2e-06,
+      "loss": 0.2874,
+      "step": 8055
+    },
+    {
+      "epoch": 1.8689247187101263,
+      "grad_norm": 7.696650394731279,
+      "learning_rate": 2e-06,
+      "loss": 0.1884,
+      "step": 8056
+    },
+    {
+      "epoch": 1.869156710358427,
+      "grad_norm": 13.935186097930975,
+      "learning_rate": 2e-06,
+      "loss": 0.2719,
+      "step": 8057
+    },
+    {
+      "epoch": 1.8693887020067277,
+      "grad_norm": 7.901974875172806,
+      "learning_rate": 2e-06,
+      "loss": 0.2239,
+      "step": 8058
+    },
+    {
+      "epoch": 1.8696206936550284,
+      "grad_norm": 14.279966452089885,
+      "learning_rate": 2e-06,
+      "loss": 0.309,
+      "step": 8059
+    },
+    {
+      "epoch": 1.869852685303329,
+      "grad_norm": 21.199725575895616,
+      "learning_rate": 2e-06,
+      "loss": 0.3386,
+      "step": 8060
+    },
+    {
+      "epoch": 1.8700846769516297,
+      "grad_norm": 9.775247696139246,
+      "learning_rate": 2e-06,
+      "loss": 0.2333,
+      "step": 8061
+    },
+    {
+      "epoch": 1.8703166685999304,
+      "grad_norm": 17.735314243096184,
+      "learning_rate": 2e-06,
+      "loss": 0.274,
+      "step": 8062
+    },
+    {
+      "epoch": 1.870548660248231,
+      "grad_norm": 9.118166555172795,
+      "learning_rate": 2e-06,
+      "loss": 0.2709,
+      "step": 8063
+    },
+    {
+      "epoch": 1.8707806518965318,
+      "grad_norm": 10.560664539208704,
+      "learning_rate": 2e-06,
+      "loss": 0.1477,
+      "step": 8064
+    },
+    {
+      "epoch": 1.8710126435448324,
+      "grad_norm": 21.11132835177889,
+      "learning_rate": 2e-06,
+      "loss": 0.2022,
+      "step": 8065
+    },
+    {
+      "epoch": 1.871244635193133,
+      "grad_norm": 12.965268707465519,
+      "learning_rate": 2e-06,
+      "loss": 0.2166,
+      "step": 8066
+    },
+    {
+      "epoch": 1.8714766268414338,
+      "grad_norm": 16.22190776075008,
+      "learning_rate": 2e-06,
+      "loss": 0.3488,
+      "step": 8067
+    },
+    {
+      "epoch": 1.8717086184897345,
+      "grad_norm": 18.535484259025015,
+      "learning_rate": 2e-06,
+      "loss": 0.2677,
+      "step": 8068
+    },
+    {
+      "epoch": 1.8719406101380351,
+      "grad_norm": 12.973138242724986,
+      "learning_rate": 2e-06,
+      "loss": 0.2302,
+      "step": 8069
+    },
+    {
+      "epoch": 1.8721726017863358,
+      "grad_norm": 17.321787407067543,
+      "learning_rate": 2e-06,
+      "loss": 0.424,
+      "step": 8070
+    },
+    {
+      "epoch": 1.8724045934346365,
+      "grad_norm": 8.641817252315901,
+      "learning_rate": 2e-06,
+      "loss": 0.2172,
+      "step": 8071
+    },
+    {
+      "epoch": 1.8726365850829372,
+      "grad_norm": 13.667276124603173,
+      "learning_rate": 2e-06,
+      "loss": 0.1638,
+      "step": 8072
+    },
+    {
+      "epoch": 1.8728685767312376,
+      "grad_norm": 12.34392074525113,
+      "learning_rate": 2e-06,
+      "loss": 0.1892,
+      "step": 8073
+    },
+    {
+      "epoch": 1.8731005683795383,
+      "grad_norm": 13.175216729247062,
+      "learning_rate": 2e-06,
+      "loss": 0.2071,
+      "step": 8074
+    },
+    {
+      "epoch": 1.873332560027839,
+      "grad_norm": 10.656296949914095,
+      "learning_rate": 2e-06,
+      "loss": 0.2188,
+      "step": 8075
+    },
+    {
+      "epoch": 1.8735645516761397,
+      "grad_norm": 21.55205836915179,
+      "learning_rate": 2e-06,
+      "loss": 0.2832,
+      "step": 8076
+    },
+    {
+      "epoch": 1.8737965433244403,
+      "grad_norm": 14.915656126608042,
+      "learning_rate": 2e-06,
+      "loss": 0.2965,
+      "step": 8077
+    },
+    {
+      "epoch": 1.874028534972741,
+      "grad_norm": 15.797128126647014,
+      "learning_rate": 2e-06,
+      "loss": 0.1939,
+      "step": 8078
+    },
+    {
+      "epoch": 1.8742605266210415,
+      "grad_norm": 15.63902295953352,
+      "learning_rate": 2e-06,
+      "loss": 0.2841,
+      "step": 8079
+    },
+    {
+      "epoch": 1.8744925182693422,
+      "grad_norm": 7.098117949851355,
+      "learning_rate": 2e-06,
+      "loss": 0.1646,
+      "step": 8080
+    },
+    {
+      "epoch": 1.8747245099176428,
+      "grad_norm": 16.878481122030315,
+      "learning_rate": 2e-06,
+      "loss": 0.212,
+      "step": 8081
+    },
+    {
+      "epoch": 1.8749565015659435,
+      "grad_norm": 14.102839925698367,
+      "learning_rate": 2e-06,
+      "loss": 0.259,
+      "step": 8082
+    },
+    {
+      "epoch": 1.8751884932142442,
+      "grad_norm": 8.015270991643696,
+      "learning_rate": 2e-06,
+      "loss": 0.2123,
+      "step": 8083
+    },
+    {
+      "epoch": 1.8754204848625449,
+      "grad_norm": 10.75137113578418,
+      "learning_rate": 2e-06,
+      "loss": 0.1887,
+      "step": 8084
+    },
+    {
+      "epoch": 1.8756524765108455,
+      "grad_norm": 12.192149049115477,
+      "learning_rate": 2e-06,
+      "loss": 0.2144,
+      "step": 8085
+    },
+    {
+      "epoch": 1.8758844681591462,
+      "grad_norm": 16.200978015518178,
+      "learning_rate": 2e-06,
+      "loss": 0.1866,
+      "step": 8086
+    },
+    {
+      "epoch": 1.876116459807447,
+      "grad_norm": 13.604564928941926,
+      "learning_rate": 2e-06,
+      "loss": 0.2224,
+      "step": 8087
+    },
+    {
+      "epoch": 1.8763484514557476,
+      "grad_norm": 18.27121505748652,
+      "learning_rate": 2e-06,
+      "loss": 0.3974,
+      "step": 8088
+    },
+    {
+      "epoch": 1.8765804431040483,
+      "grad_norm": 20.230308054727587,
+      "learning_rate": 2e-06,
+      "loss": 0.2641,
+      "step": 8089
+    },
+    {
+      "epoch": 1.876812434752349,
+      "grad_norm": 19.98312531455832,
+      "learning_rate": 2e-06,
+      "loss": 0.2823,
+      "step": 8090
+    },
+    {
+      "epoch": 1.8770444264006496,
+      "grad_norm": 13.997583676506107,
+      "learning_rate": 2e-06,
+      "loss": 0.3769,
+      "step": 8091
+    },
+    {
+      "epoch": 1.8772764180489503,
+      "grad_norm": 16.245997317386614,
+      "learning_rate": 2e-06,
+      "loss": 0.265,
+      "step": 8092
+    },
+    {
+      "epoch": 1.877508409697251,
+      "grad_norm": 6.987822813233787,
+      "learning_rate": 2e-06,
+      "loss": 0.1468,
+      "step": 8093
+    },
+    {
+      "epoch": 1.8777404013455516,
+      "grad_norm": 13.866209639245193,
+      "learning_rate": 2e-06,
+      "loss": 0.2698,
+      "step": 8094
+    },
+    {
+      "epoch": 1.8779723929938523,
+      "grad_norm": 9.262533352772289,
+      "learning_rate": 2e-06,
+      "loss": 0.2204,
+      "step": 8095
+    },
+    {
+      "epoch": 1.878204384642153,
+      "grad_norm": 17.625891059497523,
+      "learning_rate": 2e-06,
+      "loss": 0.3638,
+      "step": 8096
+    },
+    {
+      "epoch": 1.8784363762904537,
+      "grad_norm": 15.109355853479892,
+      "learning_rate": 2e-06,
+      "loss": 0.1749,
+      "step": 8097
+    },
+    {
+      "epoch": 1.8786683679387544,
+      "grad_norm": 23.316790726338905,
+      "learning_rate": 2e-06,
+      "loss": 0.3017,
+      "step": 8098
+    },
+    {
+      "epoch": 1.8789003595870548,
+      "grad_norm": 11.417710489306446,
+      "learning_rate": 2e-06,
+      "loss": 0.2762,
+      "step": 8099
+    },
+    {
+      "epoch": 1.8791323512353555,
+      "grad_norm": 7.568858560857904,
+      "learning_rate": 2e-06,
+      "loss": 0.1969,
+      "step": 8100
+    },
+    {
+      "epoch": 1.8793643428836562,
+      "grad_norm": 24.255230697484876,
+      "learning_rate": 2e-06,
+      "loss": 0.3469,
+      "step": 8101
+    },
+    {
+      "epoch": 1.8795963345319568,
+      "grad_norm": 10.709730914448418,
+      "learning_rate": 2e-06,
+      "loss": 0.1396,
+      "step": 8102
+    },
+    {
+      "epoch": 1.8798283261802575,
+      "grad_norm": 15.12566667585955,
+      "learning_rate": 2e-06,
+      "loss": 0.2438,
+      "step": 8103
+    },
+    {
+      "epoch": 1.8800603178285582,
+      "grad_norm": 10.69027851351895,
+      "learning_rate": 2e-06,
+      "loss": 0.1836,
+      "step": 8104
+    },
+    {
+      "epoch": 1.8802923094768589,
+      "grad_norm": 8.275867129578193,
+      "learning_rate": 2e-06,
+      "loss": 0.1433,
+      "step": 8105
+    },
+    {
+      "epoch": 1.8805243011251593,
+      "grad_norm": 18.796644003060116,
+      "learning_rate": 2e-06,
+      "loss": 0.3486,
+      "step": 8106
+    },
+    {
+      "epoch": 1.88075629277346,
+      "grad_norm": 11.441232359659232,
+      "learning_rate": 2e-06,
+      "loss": 0.267,
+      "step": 8107
+    },
+    {
+      "epoch": 1.8809882844217607,
+      "grad_norm": 13.62519582991966,
+      "learning_rate": 2e-06,
+      "loss": 0.1969,
+      "step": 8108
+    },
+    {
+      "epoch": 1.8812202760700614,
+      "grad_norm": 12.018684733601567,
+      "learning_rate": 2e-06,
+      "loss": 0.2544,
+      "step": 8109
+    },
+    {
+      "epoch": 1.881452267718362,
+      "grad_norm": 18.41427434587355,
+      "learning_rate": 2e-06,
+      "loss": 0.2961,
+      "step": 8110
+    },
+    {
+      "epoch": 1.8816842593666627,
+      "grad_norm": 11.72345637527048,
+      "learning_rate": 2e-06,
+      "loss": 0.2071,
+      "step": 8111
+    },
+    {
+      "epoch": 1.8819162510149634,
+      "grad_norm": 9.216262154161301,
+      "learning_rate": 2e-06,
+      "loss": 0.2301,
+      "step": 8112
+    },
+    {
+      "epoch": 1.882148242663264,
+      "grad_norm": 17.853558650966367,
+      "learning_rate": 2e-06,
+      "loss": 0.2497,
+      "step": 8113
+    },
+    {
+      "epoch": 1.8823802343115648,
+      "grad_norm": 15.682849567438378,
+      "learning_rate": 2e-06,
+      "loss": 0.2568,
+      "step": 8114
+    },
+    {
+      "epoch": 1.8826122259598654,
+      "grad_norm": 23.39963649706194,
+      "learning_rate": 2e-06,
+      "loss": 0.3241,
+      "step": 8115
+    },
+    {
+      "epoch": 1.8828442176081661,
+      "grad_norm": 11.397005268434754,
+      "learning_rate": 2e-06,
+      "loss": 0.2757,
+      "step": 8116
+    },
+    {
+      "epoch": 1.8830762092564668,
+      "grad_norm": 11.623011034774825,
+      "learning_rate": 2e-06,
+      "loss": 0.2046,
+      "step": 8117
+    },
+    {
+      "epoch": 1.8833082009047675,
+      "grad_norm": 16.244386942551117,
+      "learning_rate": 2e-06,
+      "loss": 0.2464,
+      "step": 8118
+    },
+    {
+      "epoch": 1.8835401925530681,
+      "grad_norm": 6.50215631748504,
+      "learning_rate": 2e-06,
+      "loss": 0.1487,
+      "step": 8119
+    },
+    {
+      "epoch": 1.8837721842013688,
+      "grad_norm": 13.968814913063218,
+      "learning_rate": 2e-06,
+      "loss": 0.24,
+      "step": 8120
+    },
+    {
+      "epoch": 1.8840041758496695,
+      "grad_norm": 14.418952475658717,
+      "learning_rate": 2e-06,
+      "loss": 0.2565,
+      "step": 8121
+    },
+    {
+      "epoch": 1.8842361674979702,
+      "grad_norm": 12.90293443244331,
+      "learning_rate": 2e-06,
+      "loss": 0.2305,
+      "step": 8122
+    },
+    {
+      "epoch": 1.8844681591462709,
+      "grad_norm": 11.279220611971848,
+      "learning_rate": 2e-06,
+      "loss": 0.2261,
+      "step": 8123
+    },
+    {
+      "epoch": 1.8847001507945715,
+      "grad_norm": 12.477288015553475,
+      "learning_rate": 2e-06,
+      "loss": 0.2421,
+      "step": 8124
+    },
+    {
+      "epoch": 1.8849321424428722,
+      "grad_norm": 12.22566777943852,
+      "learning_rate": 2e-06,
+      "loss": 0.2004,
+      "step": 8125
+    },
+    {
+      "epoch": 1.8851641340911727,
+      "grad_norm": 10.719972656821513,
+      "learning_rate": 2e-06,
+      "loss": 0.2487,
+      "step": 8126
+    },
+    {
+      "epoch": 1.8853961257394733,
+      "grad_norm": 12.691193607855086,
+      "learning_rate": 2e-06,
+      "loss": 0.3267,
+      "step": 8127
+    },
+    {
+      "epoch": 1.885628117387774,
+      "grad_norm": 11.938062303337183,
+      "learning_rate": 2e-06,
+      "loss": 0.2,
+      "step": 8128
+    },
+    {
+      "epoch": 1.8858601090360747,
+      "grad_norm": 16.38343012082749,
+      "learning_rate": 2e-06,
+      "loss": 0.3129,
+      "step": 8129
+    },
+    {
+      "epoch": 1.8860921006843754,
+      "grad_norm": 13.360707433601444,
+      "learning_rate": 2e-06,
+      "loss": 0.297,
+      "step": 8130
+    },
+    {
+      "epoch": 1.886324092332676,
+      "grad_norm": 8.182444282018277,
+      "learning_rate": 2e-06,
+      "loss": 0.2032,
+      "step": 8131
+    },
+    {
+      "epoch": 1.8865560839809767,
+      "grad_norm": 10.785681550556108,
+      "learning_rate": 2e-06,
+      "loss": 0.198,
+      "step": 8132
+    },
+    {
+      "epoch": 1.8867880756292772,
+      "grad_norm": 32.651483206036424,
+      "learning_rate": 2e-06,
+      "loss": 0.2223,
+      "step": 8133
+    },
+    {
+      "epoch": 1.8870200672775779,
+      "grad_norm": 12.78225895745723,
+      "learning_rate": 2e-06,
+      "loss": 0.3003,
+      "step": 8134
+    },
+    {
+      "epoch": 1.8872520589258786,
+      "grad_norm": 16.118136488781317,
+      "learning_rate": 2e-06,
+      "loss": 0.3566,
+      "step": 8135
+    },
+    {
+      "epoch": 1.8874840505741792,
+      "grad_norm": 16.010772026878108,
+      "learning_rate": 2e-06,
+      "loss": 0.2491,
+      "step": 8136
+    },
+    {
+      "epoch": 1.88771604222248,
+      "grad_norm": 19.512267007684628,
+      "learning_rate": 2e-06,
+      "loss": 0.298,
+      "step": 8137
+    },
+    {
+      "epoch": 1.8879480338707806,
+      "grad_norm": 12.91385202920017,
+      "learning_rate": 2e-06,
+      "loss": 0.2058,
+      "step": 8138
+    },
+    {
+      "epoch": 1.8881800255190813,
+      "grad_norm": 13.799231253139101,
+      "learning_rate": 2e-06,
+      "loss": 0.1785,
+      "step": 8139
+    },
+    {
+      "epoch": 1.888412017167382,
+      "grad_norm": 22.71783170409705,
+      "learning_rate": 2e-06,
+      "loss": 0.3614,
+      "step": 8140
+    },
+    {
+      "epoch": 1.8886440088156826,
+      "grad_norm": 17.47576531872985,
+      "learning_rate": 2e-06,
+      "loss": 0.2828,
+      "step": 8141
+    },
+    {
+      "epoch": 1.8888760004639833,
+      "grad_norm": 17.212646341870357,
+      "learning_rate": 2e-06,
+      "loss": 0.2752,
+      "step": 8142
+    },
+    {
+      "epoch": 1.889107992112284,
+      "grad_norm": 11.878432237227706,
+      "learning_rate": 2e-06,
+      "loss": 0.216,
+      "step": 8143
+    },
+    {
+      "epoch": 1.8893399837605847,
+      "grad_norm": 14.344434202109612,
+      "learning_rate": 2e-06,
+      "loss": 0.2197,
+      "step": 8144
+    },
+    {
+      "epoch": 1.8895719754088853,
+      "grad_norm": 24.818920744905732,
+      "learning_rate": 2e-06,
+      "loss": 0.2646,
+      "step": 8145
+    },
+    {
+      "epoch": 1.889803967057186,
+      "grad_norm": 10.651778755421654,
+      "learning_rate": 2e-06,
+      "loss": 0.1695,
+      "step": 8146
+    },
+    {
+      "epoch": 1.8900359587054867,
+      "grad_norm": 14.289848463260581,
+      "learning_rate": 2e-06,
+      "loss": 0.2756,
+      "step": 8147
+    },
+    {
+      "epoch": 1.8902679503537874,
+      "grad_norm": 7.128992669850312,
+      "learning_rate": 2e-06,
+      "loss": 0.1521,
+      "step": 8148
+    },
+    {
+      "epoch": 1.890499942002088,
+      "grad_norm": 17.376190428609913,
+      "learning_rate": 2e-06,
+      "loss": 0.2186,
+      "step": 8149
+    },
+    {
+      "epoch": 1.8907319336503887,
+      "grad_norm": 14.718226807755526,
+      "learning_rate": 2e-06,
+      "loss": 0.2132,
+      "step": 8150
+    },
+    {
+      "epoch": 1.8909639252986894,
+      "grad_norm": 35.42613465223065,
+      "learning_rate": 2e-06,
+      "loss": 0.3643,
+      "step": 8151
+    },
+    {
+      "epoch": 1.89119591694699,
+      "grad_norm": 29.24116725879479,
+      "learning_rate": 2e-06,
+      "loss": 0.3953,
+      "step": 8152
+    },
+    {
+      "epoch": 1.8914279085952905,
+      "grad_norm": 11.504743827237121,
+      "learning_rate": 2e-06,
+      "loss": 0.1929,
+      "step": 8153
+    },
+    {
+      "epoch": 1.8916599002435912,
+      "grad_norm": 15.527277538736046,
+      "learning_rate": 2e-06,
+      "loss": 0.2618,
+      "step": 8154
+    },
+    {
+      "epoch": 1.8918918918918919,
+      "grad_norm": 34.41089193580936,
+      "learning_rate": 2e-06,
+      "loss": 0.3138,
+      "step": 8155
+    },
+    {
+      "epoch": 1.8921238835401926,
+      "grad_norm": 17.68286000532179,
+      "learning_rate": 2e-06,
+      "loss": 0.2378,
+      "step": 8156
+    },
+    {
+      "epoch": 1.8923558751884932,
+      "grad_norm": 13.15350512468965,
+      "learning_rate": 2e-06,
+      "loss": 0.134,
+      "step": 8157
+    },
+    {
+      "epoch": 1.892587866836794,
+      "grad_norm": 10.329227674219053,
+      "learning_rate": 2e-06,
+      "loss": 0.1932,
+      "step": 8158
+    },
+    {
+      "epoch": 1.8928198584850944,
+      "grad_norm": 13.006155127117866,
+      "learning_rate": 2e-06,
+      "loss": 0.2922,
+      "step": 8159
+    },
+    {
+      "epoch": 1.893051850133395,
+      "grad_norm": 12.329676024489999,
+      "learning_rate": 2e-06,
+      "loss": 0.2295,
+      "step": 8160
+    },
+    {
+      "epoch": 1.8932838417816957,
+      "grad_norm": 7.363387316454607,
+      "learning_rate": 2e-06,
+      "loss": 0.1269,
+      "step": 8161
+    },
+    {
+      "epoch": 1.8935158334299964,
+      "grad_norm": 17.318169665052793,
+      "learning_rate": 2e-06,
+      "loss": 0.2694,
+      "step": 8162
+    },
+    {
+      "epoch": 1.893747825078297,
+      "grad_norm": 19.869125915460828,
+      "learning_rate": 2e-06,
+      "loss": 0.2753,
+      "step": 8163
+    },
+    {
+      "epoch": 1.8939798167265978,
+      "grad_norm": 7.136062236164506,
+      "learning_rate": 2e-06,
+      "loss": 0.2068,
+      "step": 8164
+    },
+    {
+      "epoch": 1.8942118083748984,
+      "grad_norm": 12.456129038837584,
+      "learning_rate": 2e-06,
+      "loss": 0.2313,
+      "step": 8165
+    },
+    {
+      "epoch": 1.8944438000231991,
+      "grad_norm": 17.244983081129227,
+      "learning_rate": 2e-06,
+      "loss": 0.3142,
+      "step": 8166
+    },
+    {
+      "epoch": 1.8946757916714998,
+      "grad_norm": 16.69022335332699,
+      "learning_rate": 2e-06,
+      "loss": 0.2655,
+      "step": 8167
+    },
+    {
+      "epoch": 1.8949077833198005,
+      "grad_norm": 15.769969245776224,
+      "learning_rate": 2e-06,
+      "loss": 0.2282,
+      "step": 8168
+    },
+    {
+      "epoch": 1.8951397749681012,
+      "grad_norm": 8.289435804973813,
+      "learning_rate": 2e-06,
+      "loss": 0.1708,
+      "step": 8169
+    },
+    {
+      "epoch": 1.8953717666164018,
+      "grad_norm": 10.823837362743815,
+      "learning_rate": 2e-06,
+      "loss": 0.234,
+      "step": 8170
+    },
+    {
+      "epoch": 1.8956037582647025,
+      "grad_norm": 11.207424384436552,
+      "learning_rate": 2e-06,
+      "loss": 0.1962,
+      "step": 8171
+    },
+    {
+      "epoch": 1.8958357499130032,
+      "grad_norm": 9.150322406339379,
+      "learning_rate": 2e-06,
+      "loss": 0.2094,
+      "step": 8172
+    },
+    {
+      "epoch": 1.8960677415613039,
+      "grad_norm": 17.94134879246771,
+      "learning_rate": 2e-06,
+      "loss": 0.2231,
+      "step": 8173
+    },
+    {
+      "epoch": 1.8962997332096045,
+      "grad_norm": 22.93012671108854,
+      "learning_rate": 2e-06,
+      "loss": 0.2061,
+      "step": 8174
+    },
+    {
+      "epoch": 1.8965317248579052,
+      "grad_norm": 16.504892149263032,
+      "learning_rate": 2e-06,
+      "loss": 0.2255,
+      "step": 8175
+    },
+    {
+      "epoch": 1.896763716506206,
+      "grad_norm": 17.36398239932726,
+      "learning_rate": 2e-06,
+      "loss": 0.2867,
+      "step": 8176
+    },
+    {
+      "epoch": 1.8969957081545066,
+      "grad_norm": 22.820471052406855,
+      "learning_rate": 2e-06,
+      "loss": 0.1776,
+      "step": 8177
+    },
+    {
+      "epoch": 1.8972276998028073,
+      "grad_norm": 12.508140962756473,
+      "learning_rate": 2e-06,
+      "loss": 0.2118,
+      "step": 8178
+    },
+    {
+      "epoch": 1.8974596914511077,
+      "grad_norm": 15.703706473722347,
+      "learning_rate": 2e-06,
+      "loss": 0.2247,
+      "step": 8179
+    },
+    {
+      "epoch": 1.8976916830994084,
+      "grad_norm": 34.07236853610495,
+      "learning_rate": 2e-06,
+      "loss": 0.4813,
+      "step": 8180
+    },
+    {
+      "epoch": 1.897923674747709,
+      "grad_norm": 15.111982169520063,
+      "learning_rate": 2e-06,
+      "loss": 0.3347,
+      "step": 8181
+    },
+    {
+      "epoch": 1.8981556663960097,
+      "grad_norm": 18.108042990665616,
+      "learning_rate": 2e-06,
+      "loss": 0.3716,
+      "step": 8182
+    },
+    {
+      "epoch": 1.8983876580443104,
+      "grad_norm": 10.371113738441565,
+      "learning_rate": 2e-06,
+      "loss": 0.1723,
+      "step": 8183
+    },
+    {
+      "epoch": 1.898619649692611,
+      "grad_norm": 11.176901885735523,
+      "learning_rate": 2e-06,
+      "loss": 0.2535,
+      "step": 8184
+    },
+    {
+      "epoch": 1.8988516413409118,
+      "grad_norm": 11.546728228280674,
+      "learning_rate": 2e-06,
+      "loss": 0.1787,
+      "step": 8185
+    },
+    {
+      "epoch": 1.8990836329892122,
+      "grad_norm": 18.031724353604417,
+      "learning_rate": 2e-06,
+      "loss": 0.263,
+      "step": 8186
+    },
+    {
+      "epoch": 1.899315624637513,
+      "grad_norm": 7.97087998375236,
+      "learning_rate": 2e-06,
+      "loss": 0.128,
+      "step": 8187
+    },
+    {
+      "epoch": 1.8995476162858136,
+      "grad_norm": 10.91288450703131,
+      "learning_rate": 2e-06,
+      "loss": 0.1763,
+      "step": 8188
+    },
+    {
+      "epoch": 1.8997796079341143,
+      "grad_norm": 15.287091653761278,
+      "learning_rate": 2e-06,
+      "loss": 0.2395,
+      "step": 8189
+    },
+    {
+      "epoch": 1.900011599582415,
+      "grad_norm": 9.40977049520668,
+      "learning_rate": 2e-06,
+      "loss": 0.1625,
+      "step": 8190
+    },
+    {
+      "epoch": 1.9002435912307156,
+      "grad_norm": 16.861117607116114,
+      "learning_rate": 2e-06,
+      "loss": 0.2271,
+      "step": 8191
+    },
+    {
+      "epoch": 1.9004755828790163,
+      "grad_norm": 11.037214769342043,
+      "learning_rate": 2e-06,
+      "loss": 0.1528,
+      "step": 8192
+    },
+    {
+      "epoch": 1.900707574527317,
+      "grad_norm": 10.195558248624783,
+      "learning_rate": 2e-06,
+      "loss": 0.1352,
+      "step": 8193
+    },
+    {
+      "epoch": 1.9009395661756177,
+      "grad_norm": 19.357736986172732,
+      "learning_rate": 2e-06,
+      "loss": 0.2074,
+      "step": 8194
+    },
+    {
+      "epoch": 1.9011715578239183,
+      "grad_norm": 22.36942699777849,
+      "learning_rate": 2e-06,
+      "loss": 0.3187,
+      "step": 8195
+    },
+    {
+      "epoch": 1.901403549472219,
+      "grad_norm": 8.715362015576734,
+      "learning_rate": 2e-06,
+      "loss": 0.2511,
+      "step": 8196
+    },
+    {
+      "epoch": 1.9016355411205197,
+      "grad_norm": 10.25230107442266,
+      "learning_rate": 2e-06,
+      "loss": 0.1778,
+      "step": 8197
+    },
+    {
+      "epoch": 1.9018675327688204,
+      "grad_norm": 23.561928495390756,
+      "learning_rate": 2e-06,
+      "loss": 0.4416,
+      "step": 8198
+    },
+    {
+      "epoch": 1.902099524417121,
+      "grad_norm": 18.159330177747112,
+      "learning_rate": 2e-06,
+      "loss": 0.265,
+      "step": 8199
+    },
+    {
+      "epoch": 1.9023315160654217,
+      "grad_norm": 18.474659179229178,
+      "learning_rate": 2e-06,
+      "loss": 0.2264,
+      "step": 8200
+    },
+    {
+      "epoch": 1.9025635077137224,
+      "grad_norm": 11.213883376723222,
+      "learning_rate": 2e-06,
+      "loss": 0.2187,
+      "step": 8201
+    },
+    {
+      "epoch": 1.902795499362023,
+      "grad_norm": 11.325475746908815,
+      "learning_rate": 2e-06,
+      "loss": 0.1437,
+      "step": 8202
+    },
+    {
+      "epoch": 1.9030274910103238,
+      "grad_norm": 12.667406137830117,
+      "learning_rate": 2e-06,
+      "loss": 0.2505,
+      "step": 8203
+    },
+    {
+      "epoch": 1.9032594826586244,
+      "grad_norm": 14.219083128107396,
+      "learning_rate": 2e-06,
+      "loss": 0.1709,
+      "step": 8204
+    },
+    {
+      "epoch": 1.9034914743069251,
+      "grad_norm": 21.08465805561743,
+      "learning_rate": 2e-06,
+      "loss": 0.4011,
+      "step": 8205
+    },
+    {
+      "epoch": 1.9037234659552256,
+      "grad_norm": 16.782938445686284,
+      "learning_rate": 2e-06,
+      "loss": 0.2381,
+      "step": 8206
+    },
+    {
+      "epoch": 1.9039554576035262,
+      "grad_norm": 12.519800931564225,
+      "learning_rate": 2e-06,
+      "loss": 0.1999,
+      "step": 8207
+    },
+    {
+      "epoch": 1.904187449251827,
+      "grad_norm": 14.461461522859151,
+      "learning_rate": 2e-06,
+      "loss": 0.3346,
+      "step": 8208
+    },
+    {
+      "epoch": 1.9044194409001276,
+      "grad_norm": 14.345881539762717,
+      "learning_rate": 2e-06,
+      "loss": 0.2979,
+      "step": 8209
+    },
+    {
+      "epoch": 1.9046514325484283,
+      "grad_norm": 21.421155368467556,
+      "learning_rate": 2e-06,
+      "loss": 0.3399,
+      "step": 8210
+    },
+    {
+      "epoch": 1.904883424196729,
+      "grad_norm": 14.75430383047075,
+      "learning_rate": 2e-06,
+      "loss": 0.1874,
+      "step": 8211
+    },
+    {
+      "epoch": 1.9051154158450294,
+      "grad_norm": 16.452560204189812,
+      "learning_rate": 2e-06,
+      "loss": 0.2955,
+      "step": 8212
+    },
+    {
+      "epoch": 1.90534740749333,
+      "grad_norm": 12.58674547104881,
+      "learning_rate": 2e-06,
+      "loss": 0.2175,
+      "step": 8213
+    },
+    {
+      "epoch": 1.9055793991416308,
+      "grad_norm": 32.0433974432042,
+      "learning_rate": 2e-06,
+      "loss": 0.3187,
+      "step": 8214
+    },
+    {
+      "epoch": 1.9058113907899314,
+      "grad_norm": 27.844541614260155,
+      "learning_rate": 2e-06,
+      "loss": 0.3755,
+      "step": 8215
+    },
+    {
+      "epoch": 1.9060433824382321,
+      "grad_norm": 10.550625708302274,
+      "learning_rate": 2e-06,
+      "loss": 0.1614,
+      "step": 8216
+    },
+    {
+      "epoch": 1.9062753740865328,
+      "grad_norm": 20.29109359857516,
+      "learning_rate": 2e-06,
+      "loss": 0.3744,
+      "step": 8217
+    },
+    {
+      "epoch": 1.9065073657348335,
+      "grad_norm": 18.18674145651999,
+      "learning_rate": 2e-06,
+      "loss": 0.3038,
+      "step": 8218
+    },
+    {
+      "epoch": 1.9067393573831342,
+      "grad_norm": 26.554136681725186,
+      "learning_rate": 2e-06,
+      "loss": 0.3285,
+      "step": 8219
+    },
+    {
+      "epoch": 1.9069713490314348,
+      "grad_norm": 17.913901890935545,
+      "learning_rate": 2e-06,
+      "loss": 0.2838,
+      "step": 8220
+    },
+    {
+      "epoch": 1.9072033406797355,
+      "grad_norm": 19.446754092117164,
+      "learning_rate": 2e-06,
+      "loss": 0.2213,
+      "step": 8221
+    },
+    {
+      "epoch": 1.9074353323280362,
+      "grad_norm": 19.892146142585084,
+      "learning_rate": 2e-06,
+      "loss": 0.3371,
+      "step": 8222
+    },
+    {
+      "epoch": 1.9076673239763369,
+      "grad_norm": 13.182404615018127,
+      "learning_rate": 2e-06,
+      "loss": 0.2448,
+      "step": 8223
+    },
+    {
+      "epoch": 1.9078993156246375,
+      "grad_norm": 20.5755363946165,
+      "learning_rate": 2e-06,
+      "loss": 0.2963,
+      "step": 8224
+    },
+    {
+      "epoch": 1.9081313072729382,
+      "grad_norm": 12.559442015884914,
+      "learning_rate": 2e-06,
+      "loss": 0.2099,
+      "step": 8225
+    },
+    {
+      "epoch": 1.908363298921239,
+      "grad_norm": 9.005075480299361,
+      "learning_rate": 2e-06,
+      "loss": 0.1641,
+      "step": 8226
+    },
+    {
+      "epoch": 1.9085952905695396,
+      "grad_norm": 12.662288781435917,
+      "learning_rate": 2e-06,
+      "loss": 0.3134,
+      "step": 8227
+    },
+    {
+      "epoch": 1.9088272822178403,
+      "grad_norm": 12.007012421122173,
+      "learning_rate": 2e-06,
+      "loss": 0.2553,
+      "step": 8228
+    },
+    {
+      "epoch": 1.909059273866141,
+      "grad_norm": 11.816634760711612,
+      "learning_rate": 2e-06,
+      "loss": 0.2257,
+      "step": 8229
+    },
+    {
+      "epoch": 1.9092912655144416,
+      "grad_norm": 12.862882132997411,
+      "learning_rate": 2e-06,
+      "loss": 0.1843,
+      "step": 8230
+    },
+    {
+      "epoch": 1.9095232571627423,
+      "grad_norm": 26.931106829079567,
+      "learning_rate": 2e-06,
+      "loss": 0.236,
+      "step": 8231
+    },
+    {
+      "epoch": 1.9097552488110427,
+      "grad_norm": 9.259853349503151,
+      "learning_rate": 2e-06,
+      "loss": 0.2687,
+      "step": 8232
+    },
+    {
+      "epoch": 1.9099872404593434,
+      "grad_norm": 12.53754130652307,
+      "learning_rate": 2e-06,
+      "loss": 0.2398,
+      "step": 8233
+    },
+    {
+      "epoch": 1.910219232107644,
+      "grad_norm": 18.023157803986198,
+      "learning_rate": 2e-06,
+      "loss": 0.3218,
+      "step": 8234
+    },
+    {
+      "epoch": 1.9104512237559448,
+      "grad_norm": 8.078029810459396,
+      "learning_rate": 2e-06,
+      "loss": 0.1691,
+      "step": 8235
+    },
+    {
+      "epoch": 1.9106832154042455,
+      "grad_norm": 14.604188755436693,
+      "learning_rate": 2e-06,
+      "loss": 0.2846,
+      "step": 8236
+    },
+    {
+      "epoch": 1.9109152070525461,
+      "grad_norm": 9.915303648016163,
+      "learning_rate": 2e-06,
+      "loss": 0.1822,
+      "step": 8237
+    },
+    {
+      "epoch": 1.9111471987008468,
+      "grad_norm": 8.937008541033672,
+      "learning_rate": 2e-06,
+      "loss": 0.1915,
+      "step": 8238
+    },
+    {
+      "epoch": 1.9113791903491473,
+      "grad_norm": 15.495136841108435,
+      "learning_rate": 2e-06,
+      "loss": 0.1943,
+      "step": 8239
+    },
+    {
+      "epoch": 1.911611181997448,
+      "grad_norm": 18.594498427474978,
+      "learning_rate": 2e-06,
+      "loss": 0.2526,
+      "step": 8240
+    },
+    {
+      "epoch": 1.9118431736457486,
+      "grad_norm": 11.037879862277668,
+      "learning_rate": 2e-06,
+      "loss": 0.2421,
+      "step": 8241
+    },
+    {
+      "epoch": 1.9120751652940493,
+      "grad_norm": 24.98641628833053,
+      "learning_rate": 2e-06,
+      "loss": 0.2371,
+      "step": 8242
+    },
+    {
+      "epoch": 1.91230715694235,
+      "grad_norm": 16.29291873274218,
+      "learning_rate": 2e-06,
+      "loss": 0.3609,
+      "step": 8243
+    },
+    {
+      "epoch": 1.9125391485906507,
+      "grad_norm": 21.25534135384721,
+      "learning_rate": 2e-06,
+      "loss": 0.3265,
+      "step": 8244
+    },
+    {
+      "epoch": 1.9127711402389513,
+      "grad_norm": 10.772522409197187,
+      "learning_rate": 2e-06,
+      "loss": 0.2059,
+      "step": 8245
+    },
+    {
+      "epoch": 1.913003131887252,
+      "grad_norm": 16.822319519518206,
+      "learning_rate": 2e-06,
+      "loss": 0.2291,
+      "step": 8246
+    },
+    {
+      "epoch": 1.9132351235355527,
+      "grad_norm": 14.701476350588532,
+      "learning_rate": 2e-06,
+      "loss": 0.2077,
+      "step": 8247
+    },
+    {
+      "epoch": 1.9134671151838534,
+      "grad_norm": 10.151290714699144,
+      "learning_rate": 2e-06,
+      "loss": 0.1983,
+      "step": 8248
+    },
+    {
+      "epoch": 1.913699106832154,
+      "grad_norm": 11.486781285690082,
+      "learning_rate": 2e-06,
+      "loss": 0.1778,
+      "step": 8249
+    },
+    {
+      "epoch": 1.9139310984804547,
+      "grad_norm": 16.169915834829315,
+      "learning_rate": 2e-06,
+      "loss": 0.2844,
+      "step": 8250
+    },
+    {
+      "epoch": 1.9141630901287554,
+      "grad_norm": 18.397321285879713,
+      "learning_rate": 2e-06,
+      "loss": 0.2849,
+      "step": 8251
+    },
+    {
+      "epoch": 1.914395081777056,
+      "grad_norm": 11.317068990602722,
+      "learning_rate": 2e-06,
+      "loss": 0.2229,
+      "step": 8252
+    },
+    {
+      "epoch": 1.9146270734253568,
+      "grad_norm": 19.361350050762294,
+      "learning_rate": 2e-06,
+      "loss": 0.2358,
+      "step": 8253
+    },
+    {
+      "epoch": 1.9148590650736574,
+      "grad_norm": 14.06912297548446,
+      "learning_rate": 2e-06,
+      "loss": 0.3069,
+      "step": 8254
+    },
+    {
+      "epoch": 1.9150910567219581,
+      "grad_norm": 17.87175382915879,
+      "learning_rate": 2e-06,
+      "loss": 0.2287,
+      "step": 8255
+    },
+    {
+      "epoch": 1.9153230483702588,
+      "grad_norm": 15.518948701012697,
+      "learning_rate": 2e-06,
+      "loss": 0.2169,
+      "step": 8256
+    },
+    {
+      "epoch": 1.9155550400185595,
+      "grad_norm": 9.438148701509457,
+      "learning_rate": 2e-06,
+      "loss": 0.1725,
+      "step": 8257
+    },
+    {
+      "epoch": 1.9157870316668602,
+      "grad_norm": 17.408432016873743,
+      "learning_rate": 2e-06,
+      "loss": 0.2678,
+      "step": 8258
+    },
+    {
+      "epoch": 1.9160190233151606,
+      "grad_norm": 8.518150459279795,
+      "learning_rate": 2e-06,
+      "loss": 0.1942,
+      "step": 8259
+    },
+    {
+      "epoch": 1.9162510149634613,
+      "grad_norm": 11.916440509055764,
+      "learning_rate": 2e-06,
+      "loss": 0.2316,
+      "step": 8260
+    },
+    {
+      "epoch": 1.916483006611762,
+      "grad_norm": 19.22244658410548,
+      "learning_rate": 2e-06,
+      "loss": 0.2165,
+      "step": 8261
+    },
+    {
+      "epoch": 1.9167149982600626,
+      "grad_norm": 22.489005047076123,
+      "learning_rate": 2e-06,
+      "loss": 0.3285,
+      "step": 8262
+    },
+    {
+      "epoch": 1.9169469899083633,
+      "grad_norm": 13.18035388952861,
+      "learning_rate": 2e-06,
+      "loss": 0.2058,
+      "step": 8263
+    },
+    {
+      "epoch": 1.917178981556664,
+      "grad_norm": 16.614831732754023,
+      "learning_rate": 2e-06,
+      "loss": 0.2375,
+      "step": 8264
+    },
+    {
+      "epoch": 1.9174109732049647,
+      "grad_norm": 11.545535466753725,
+      "learning_rate": 2e-06,
+      "loss": 0.1232,
+      "step": 8265
+    },
+    {
+      "epoch": 1.9176429648532651,
+      "grad_norm": 12.093689927206457,
+      "learning_rate": 2e-06,
+      "loss": 0.2792,
+      "step": 8266
+    },
+    {
+      "epoch": 1.9178749565015658,
+      "grad_norm": 9.959454490978523,
+      "learning_rate": 2e-06,
+      "loss": 0.2145,
+      "step": 8267
+    },
+    {
+      "epoch": 1.9181069481498665,
+      "grad_norm": 12.2647047989879,
+      "learning_rate": 2e-06,
+      "loss": 0.2634,
+      "step": 8268
+    },
+    {
+      "epoch": 1.9183389397981672,
+      "grad_norm": 16.72304117556915,
+      "learning_rate": 2e-06,
+      "loss": 0.3142,
+      "step": 8269
+    },
+    {
+      "epoch": 1.9185709314464678,
+      "grad_norm": 19.31870618615468,
+      "learning_rate": 2e-06,
+      "loss": 0.382,
+      "step": 8270
+    },
+    {
+      "epoch": 1.9188029230947685,
+      "grad_norm": 6.752555646575119,
+      "learning_rate": 2e-06,
+      "loss": 0.1397,
+      "step": 8271
+    },
+    {
+      "epoch": 1.9190349147430692,
+      "grad_norm": 7.376211731271454,
+      "learning_rate": 2e-06,
+      "loss": 0.1096,
+      "step": 8272
+    },
+    {
+      "epoch": 1.9192669063913699,
+      "grad_norm": 14.1403769816287,
+      "learning_rate": 2e-06,
+      "loss": 0.2297,
+      "step": 8273
+    },
+    {
+      "epoch": 1.9194988980396706,
+      "grad_norm": 18.630262161070675,
+      "learning_rate": 2e-06,
+      "loss": 0.2822,
+      "step": 8274
+    },
+    {
+      "epoch": 1.9197308896879712,
+      "grad_norm": 10.637772181676997,
+      "learning_rate": 2e-06,
+      "loss": 0.2134,
+      "step": 8275
+    },
+    {
+      "epoch": 1.919962881336272,
+      "grad_norm": 10.803230879586474,
+      "learning_rate": 2e-06,
+      "loss": 0.2273,
+      "step": 8276
+    },
+    {
+      "epoch": 1.9201948729845726,
+      "grad_norm": 14.041361424353319,
+      "learning_rate": 2e-06,
+      "loss": 0.2323,
+      "step": 8277
+    },
+    {
+      "epoch": 1.9204268646328733,
+      "grad_norm": 25.059508495132192,
+      "learning_rate": 2e-06,
+      "loss": 0.3291,
+      "step": 8278
+    },
+    {
+      "epoch": 1.920658856281174,
+      "grad_norm": 17.070058774407894,
+      "learning_rate": 2e-06,
+      "loss": 0.2736,
+      "step": 8279
+    },
+    {
+      "epoch": 1.9208908479294746,
+      "grad_norm": 11.351872788066839,
+      "learning_rate": 2e-06,
+      "loss": 0.2799,
+      "step": 8280
+    },
+    {
+      "epoch": 1.9211228395777753,
+      "grad_norm": 18.216056033597024,
+      "learning_rate": 2e-06,
+      "loss": 0.2414,
+      "step": 8281
+    },
+    {
+      "epoch": 1.921354831226076,
+      "grad_norm": 13.080737756337564,
+      "learning_rate": 2e-06,
+      "loss": 0.3448,
+      "step": 8282
+    },
+    {
+      "epoch": 1.9215868228743767,
+      "grad_norm": 20.208896374040844,
+      "learning_rate": 2e-06,
+      "loss": 0.259,
+      "step": 8283
+    },
+    {
+      "epoch": 1.9218188145226773,
+      "grad_norm": 2.866532475964635,
+      "learning_rate": 2e-06,
+      "loss": 0.1151,
+      "step": 8284
+    },
+    {
+      "epoch": 1.922050806170978,
+      "grad_norm": 7.573395237380319,
+      "learning_rate": 2e-06,
+      "loss": 0.1573,
+      "step": 8285
+    },
+    {
+      "epoch": 1.9222827978192785,
+      "grad_norm": 18.308860336279682,
+      "learning_rate": 2e-06,
+      "loss": 0.2621,
+      "step": 8286
+    },
+    {
+      "epoch": 1.9225147894675791,
+      "grad_norm": 11.665796555064631,
+      "learning_rate": 2e-06,
+      "loss": 0.1945,
+      "step": 8287
+    },
+    {
+      "epoch": 1.9227467811158798,
+      "grad_norm": 15.296308819790674,
+      "learning_rate": 2e-06,
+      "loss": 0.2331,
+      "step": 8288
+    },
+    {
+      "epoch": 1.9229787727641805,
+      "grad_norm": 9.646629419972417,
+      "learning_rate": 2e-06,
+      "loss": 0.1939,
+      "step": 8289
+    },
+    {
+      "epoch": 1.9232107644124812,
+      "grad_norm": 103.93129525379511,
+      "learning_rate": 2e-06,
+      "loss": 0.2315,
+      "step": 8290
+    },
+    {
+      "epoch": 1.9234427560607819,
+      "grad_norm": 13.058749888640612,
+      "learning_rate": 2e-06,
+      "loss": 0.2627,
+      "step": 8291
+    },
+    {
+      "epoch": 1.9236747477090823,
+      "grad_norm": 17.24259009477687,
+      "learning_rate": 2e-06,
+      "loss": 0.2115,
+      "step": 8292
+    },
+    {
+      "epoch": 1.923906739357383,
+      "grad_norm": 10.338322982938008,
+      "learning_rate": 2e-06,
+      "loss": 0.2494,
+      "step": 8293
+    },
+    {
+      "epoch": 1.9241387310056837,
+      "grad_norm": 13.010569352737638,
+      "learning_rate": 2e-06,
+      "loss": 0.2818,
+      "step": 8294
+    },
+    {
+      "epoch": 1.9243707226539843,
+      "grad_norm": 10.243548363837144,
+      "learning_rate": 2e-06,
+      "loss": 0.2308,
+      "step": 8295
+    },
+    {
+      "epoch": 1.924602714302285,
+      "grad_norm": 26.38176885433789,
+      "learning_rate": 2e-06,
+      "loss": 0.2453,
+      "step": 8296
+    },
+    {
+      "epoch": 1.9248347059505857,
+      "grad_norm": 18.326231746899325,
+      "learning_rate": 2e-06,
+      "loss": 0.2372,
+      "step": 8297
+    },
+    {
+      "epoch": 1.9250666975988864,
+      "grad_norm": 15.340498799445832,
+      "learning_rate": 2e-06,
+      "loss": 0.2787,
+      "step": 8298
+    },
+    {
+      "epoch": 1.925298689247187,
+      "grad_norm": 11.107298049649948,
+      "learning_rate": 2e-06,
+      "loss": 0.2725,
+      "step": 8299
+    },
+    {
+      "epoch": 1.9255306808954877,
+      "grad_norm": 15.409151955869918,
+      "learning_rate": 2e-06,
+      "loss": 0.3218,
+      "step": 8300
+    },
+    {
+      "epoch": 1.9257626725437884,
+      "grad_norm": 13.18529269749906,
+      "learning_rate": 2e-06,
+      "loss": 0.3214,
+      "step": 8301
+    },
+    {
+      "epoch": 1.925994664192089,
+      "grad_norm": 12.997759326811828,
+      "learning_rate": 2e-06,
+      "loss": 0.1669,
+      "step": 8302
+    },
+    {
+      "epoch": 1.9262266558403898,
+      "grad_norm": 11.75141841418822,
+      "learning_rate": 2e-06,
+      "loss": 0.2567,
+      "step": 8303
+    },
+    {
+      "epoch": 1.9264586474886904,
+      "grad_norm": 13.803993613795765,
+      "learning_rate": 2e-06,
+      "loss": 0.3057,
+      "step": 8304
+    },
+    {
+      "epoch": 1.9266906391369911,
+      "grad_norm": 11.200192090314818,
+      "learning_rate": 2e-06,
+      "loss": 0.2454,
+      "step": 8305
+    },
+    {
+      "epoch": 1.9269226307852918,
+      "grad_norm": 10.956488078705902,
+      "learning_rate": 2e-06,
+      "loss": 0.2506,
+      "step": 8306
+    },
+    {
+      "epoch": 1.9271546224335925,
+      "grad_norm": 38.90968870785902,
+      "learning_rate": 2e-06,
+      "loss": 0.26,
+      "step": 8307
+    },
+    {
+      "epoch": 1.9273866140818932,
+      "grad_norm": 12.874901051121027,
+      "learning_rate": 2e-06,
+      "loss": 0.3622,
+      "step": 8308
+    },
+    {
+      "epoch": 1.9276186057301938,
+      "grad_norm": 18.50058674378736,
+      "learning_rate": 2e-06,
+      "loss": 0.2987,
+      "step": 8309
+    },
+    {
+      "epoch": 1.9278505973784945,
+      "grad_norm": 14.402281423883219,
+      "learning_rate": 2e-06,
+      "loss": 0.2354,
+      "step": 8310
+    },
+    {
+      "epoch": 1.9280825890267952,
+      "grad_norm": 14.094001216596647,
+      "learning_rate": 2e-06,
+      "loss": 0.3899,
+      "step": 8311
+    },
+    {
+      "epoch": 1.9283145806750956,
+      "grad_norm": 18.630838256596093,
+      "learning_rate": 2e-06,
+      "loss": 0.2709,
+      "step": 8312
+    },
+    {
+      "epoch": 1.9285465723233963,
+      "grad_norm": 16.355351861756276,
+      "learning_rate": 2e-06,
+      "loss": 0.3076,
+      "step": 8313
+    },
+    {
+      "epoch": 1.928778563971697,
+      "grad_norm": 13.490106187278478,
+      "learning_rate": 2e-06,
+      "loss": 0.2527,
+      "step": 8314
+    },
+    {
+      "epoch": 1.9290105556199977,
+      "grad_norm": 17.257607194236968,
+      "learning_rate": 2e-06,
+      "loss": 0.2722,
+      "step": 8315
+    },
+    {
+      "epoch": 1.9292425472682984,
+      "grad_norm": 10.354500315551464,
+      "learning_rate": 2e-06,
+      "loss": 0.2323,
+      "step": 8316
+    },
+    {
+      "epoch": 1.929474538916599,
+      "grad_norm": 16.34349615296237,
+      "learning_rate": 2e-06,
+      "loss": 0.2311,
+      "step": 8317
+    },
+    {
+      "epoch": 1.9297065305648997,
+      "grad_norm": 8.077491084678064,
+      "learning_rate": 2e-06,
+      "loss": 0.177,
+      "step": 8318
+    },
+    {
+      "epoch": 1.9299385222132002,
+      "grad_norm": 19.54466387276364,
+      "learning_rate": 2e-06,
+      "loss": 0.3311,
+      "step": 8319
+    },
+    {
+      "epoch": 1.9301705138615008,
+      "grad_norm": 11.75052050959431,
+      "learning_rate": 2e-06,
+      "loss": 0.2188,
+      "step": 8320
+    },
+    {
+      "epoch": 1.9304025055098015,
+      "grad_norm": 15.061184790407024,
+      "learning_rate": 2e-06,
+      "loss": 0.2134,
+      "step": 8321
+    },
+    {
+      "epoch": 1.9306344971581022,
+      "grad_norm": 14.553576441539253,
+      "learning_rate": 2e-06,
+      "loss": 0.2815,
+      "step": 8322
+    },
+    {
+      "epoch": 1.9308664888064029,
+      "grad_norm": 14.718620141267312,
+      "learning_rate": 2e-06,
+      "loss": 0.2289,
+      "step": 8323
+    },
+    {
+      "epoch": 1.9310984804547036,
+      "grad_norm": 20.38486878391961,
+      "learning_rate": 2e-06,
+      "loss": 0.2531,
+      "step": 8324
+    },
+    {
+      "epoch": 1.9313304721030042,
+      "grad_norm": 14.426701058009007,
+      "learning_rate": 2e-06,
+      "loss": 0.2395,
+      "step": 8325
+    },
+    {
+      "epoch": 1.931562463751305,
+      "grad_norm": 7.443348328961412,
+      "learning_rate": 2e-06,
+      "loss": 0.1931,
+      "step": 8326
+    },
+    {
+      "epoch": 1.9317944553996056,
+      "grad_norm": 13.279828797272623,
+      "learning_rate": 2e-06,
+      "loss": 0.2471,
+      "step": 8327
+    },
+    {
+      "epoch": 1.9320264470479063,
+      "grad_norm": 14.691548590843379,
+      "learning_rate": 2e-06,
+      "loss": 0.2645,
+      "step": 8328
+    },
+    {
+      "epoch": 1.932258438696207,
+      "grad_norm": 18.380291854873782,
+      "learning_rate": 2e-06,
+      "loss": 0.2756,
+      "step": 8329
+    },
+    {
+      "epoch": 1.9324904303445076,
+      "grad_norm": 16.434077330676157,
+      "learning_rate": 2e-06,
+      "loss": 0.2942,
+      "step": 8330
+    },
+    {
+      "epoch": 1.9327224219928083,
+      "grad_norm": 23.01022355060371,
+      "learning_rate": 2e-06,
+      "loss": 0.337,
+      "step": 8331
+    },
+    {
+      "epoch": 1.932954413641109,
+      "grad_norm": 15.352569187881347,
+      "learning_rate": 2e-06,
+      "loss": 0.2557,
+      "step": 8332
+    },
+    {
+      "epoch": 1.9331864052894097,
+      "grad_norm": 31.26671018444753,
+      "learning_rate": 2e-06,
+      "loss": 0.2989,
+      "step": 8333
+    },
+    {
+      "epoch": 1.9334183969377103,
+      "grad_norm": 4.16190214180648,
+      "learning_rate": 2e-06,
+      "loss": 0.1166,
+      "step": 8334
+    },
+    {
+      "epoch": 1.933650388586011,
+      "grad_norm": 19.25567276550992,
+      "learning_rate": 2e-06,
+      "loss": 0.3374,
+      "step": 8335
+    },
+    {
+      "epoch": 1.9338823802343117,
+      "grad_norm": 16.127640357750774,
+      "learning_rate": 2e-06,
+      "loss": 0.1962,
+      "step": 8336
+    },
+    {
+      "epoch": 1.9341143718826124,
+      "grad_norm": 9.010495407693012,
+      "learning_rate": 2e-06,
+      "loss": 0.1595,
+      "step": 8337
+    },
+    {
+      "epoch": 1.934346363530913,
+      "grad_norm": 13.093907921328455,
+      "learning_rate": 2e-06,
+      "loss": 0.2079,
+      "step": 8338
+    },
+    {
+      "epoch": 1.9345783551792135,
+      "grad_norm": 11.93257976455428,
+      "learning_rate": 2e-06,
+      "loss": 0.1789,
+      "step": 8339
+    },
+    {
+      "epoch": 1.9348103468275142,
+      "grad_norm": 12.455310510281812,
+      "learning_rate": 2e-06,
+      "loss": 0.1926,
+      "step": 8340
+    },
+    {
+      "epoch": 1.9350423384758149,
+      "grad_norm": 9.872377564603019,
+      "learning_rate": 2e-06,
+      "loss": 0.1529,
+      "step": 8341
+    },
+    {
+      "epoch": 1.9352743301241155,
+      "grad_norm": 16.326098772159725,
+      "learning_rate": 2e-06,
+      "loss": 0.1564,
+      "step": 8342
+    },
+    {
+      "epoch": 1.9355063217724162,
+      "grad_norm": 8.191776245582423,
+      "learning_rate": 2e-06,
+      "loss": 0.1847,
+      "step": 8343
+    },
+    {
+      "epoch": 1.935738313420717,
+      "grad_norm": 15.177847539987345,
+      "learning_rate": 2e-06,
+      "loss": 0.2256,
+      "step": 8344
+    },
+    {
+      "epoch": 1.9359703050690173,
+      "grad_norm": 15.284297069013228,
+      "learning_rate": 2e-06,
+      "loss": 0.2495,
+      "step": 8345
+    },
+    {
+      "epoch": 1.936202296717318,
+      "grad_norm": 10.819774117608612,
+      "learning_rate": 2e-06,
+      "loss": 0.1997,
+      "step": 8346
+    },
+    {
+      "epoch": 1.9364342883656187,
+      "grad_norm": 24.980519277732025,
+      "learning_rate": 2e-06,
+      "loss": 0.2839,
+      "step": 8347
+    },
+    {
+      "epoch": 1.9366662800139194,
+      "grad_norm": 16.686504111394285,
+      "learning_rate": 2e-06,
+      "loss": 0.256,
+      "step": 8348
+    },
+    {
+      "epoch": 1.93689827166222,
+      "grad_norm": 15.473655593524072,
+      "learning_rate": 2e-06,
+      "loss": 0.2852,
+      "step": 8349
+    },
+    {
+      "epoch": 1.9371302633105207,
+      "grad_norm": 13.78303381473132,
+      "learning_rate": 2e-06,
+      "loss": 0.3147,
+      "step": 8350
+    },
+    {
+      "epoch": 1.9373622549588214,
+      "grad_norm": 15.809047183386966,
+      "learning_rate": 2e-06,
+      "loss": 0.2181,
+      "step": 8351
+    },
+    {
+      "epoch": 1.937594246607122,
+      "grad_norm": 14.087075162865043,
+      "learning_rate": 2e-06,
+      "loss": 0.2784,
+      "step": 8352
+    },
+    {
+      "epoch": 1.9378262382554228,
+      "grad_norm": 6.435673804344862,
+      "learning_rate": 2e-06,
+      "loss": 0.144,
+      "step": 8353
+    },
+    {
+      "epoch": 1.9380582299037235,
+      "grad_norm": 12.885130455224658,
+      "learning_rate": 2e-06,
+      "loss": 0.3624,
+      "step": 8354
+    },
+    {
+      "epoch": 1.9382902215520241,
+      "grad_norm": 13.006022998323044,
+      "learning_rate": 2e-06,
+      "loss": 0.1781,
+      "step": 8355
+    },
+    {
+      "epoch": 1.9385222132003248,
+      "grad_norm": 12.62404139588954,
+      "learning_rate": 2e-06,
+      "loss": 0.2423,
+      "step": 8356
+    },
+    {
+      "epoch": 1.9387542048486255,
+      "grad_norm": 9.91038736403853,
+      "learning_rate": 2e-06,
+      "loss": 0.2856,
+      "step": 8357
+    },
+    {
+      "epoch": 1.9389861964969262,
+      "grad_norm": 24.05402448981315,
+      "learning_rate": 2e-06,
+      "loss": 0.3787,
+      "step": 8358
+    },
+    {
+      "epoch": 1.9392181881452268,
+      "grad_norm": 15.102452709625638,
+      "learning_rate": 2e-06,
+      "loss": 0.2343,
+      "step": 8359
+    },
+    {
+      "epoch": 1.9394501797935275,
+      "grad_norm": 13.076728471248122,
+      "learning_rate": 2e-06,
+      "loss": 0.1743,
+      "step": 8360
+    },
+    {
+      "epoch": 1.9396821714418282,
+      "grad_norm": 9.24397385494658,
+      "learning_rate": 2e-06,
+      "loss": 0.2097,
+      "step": 8361
+    },
+    {
+      "epoch": 1.9399141630901289,
+      "grad_norm": 16.266485171871672,
+      "learning_rate": 2e-06,
+      "loss": 0.2397,
+      "step": 8362
+    },
+    {
+      "epoch": 1.9401461547384296,
+      "grad_norm": 12.217629891584592,
+      "learning_rate": 2e-06,
+      "loss": 0.16,
+      "step": 8363
+    },
+    {
+      "epoch": 1.9403781463867302,
+      "grad_norm": 11.591707102973945,
+      "learning_rate": 2e-06,
+      "loss": 0.2571,
+      "step": 8364
+    },
+    {
+      "epoch": 1.9406101380350307,
+      "grad_norm": 8.643280962767939,
+      "learning_rate": 2e-06,
+      "loss": 0.1058,
+      "step": 8365
+    },
+    {
+      "epoch": 1.9408421296833314,
+      "grad_norm": 40.216672592940476,
+      "learning_rate": 2e-06,
+      "loss": 0.378,
+      "step": 8366
+    },
+    {
+      "epoch": 1.941074121331632,
+      "grad_norm": 6.157949903531771,
+      "learning_rate": 2e-06,
+      "loss": 0.1596,
+      "step": 8367
+    },
+    {
+      "epoch": 1.9413061129799327,
+      "grad_norm": 11.693409202021451,
+      "learning_rate": 2e-06,
+      "loss": 0.2576,
+      "step": 8368
+    },
+    {
+      "epoch": 1.9415381046282334,
+      "grad_norm": 10.782026224472514,
+      "learning_rate": 2e-06,
+      "loss": 0.1393,
+      "step": 8369
+    },
+    {
+      "epoch": 1.941770096276534,
+      "grad_norm": 23.328197673210656,
+      "learning_rate": 2e-06,
+      "loss": 0.3011,
+      "step": 8370
+    },
+    {
+      "epoch": 1.9420020879248348,
+      "grad_norm": 20.896594843390798,
+      "learning_rate": 2e-06,
+      "loss": 0.2877,
+      "step": 8371
+    },
+    {
+      "epoch": 1.9422340795731352,
+      "grad_norm": 15.411905183941613,
+      "learning_rate": 2e-06,
+      "loss": 0.2703,
+      "step": 8372
+    },
+    {
+      "epoch": 1.9424660712214359,
+      "grad_norm": 13.43832227814687,
+      "learning_rate": 2e-06,
+      "loss": 0.209,
+      "step": 8373
+    },
+    {
+      "epoch": 1.9426980628697366,
+      "grad_norm": 11.963382454873283,
+      "learning_rate": 2e-06,
+      "loss": 0.3074,
+      "step": 8374
+    },
+    {
+      "epoch": 1.9429300545180372,
+      "grad_norm": 11.95775311491443,
+      "learning_rate": 2e-06,
+      "loss": 0.223,
+      "step": 8375
+    },
+    {
+      "epoch": 1.943162046166338,
+      "grad_norm": 17.428277570749742,
+      "learning_rate": 2e-06,
+      "loss": 0.3112,
+      "step": 8376
+    },
+    {
+      "epoch": 1.9433940378146386,
+      "grad_norm": 15.230805680496852,
+      "learning_rate": 2e-06,
+      "loss": 0.2687,
+      "step": 8377
+    },
+    {
+      "epoch": 1.9436260294629393,
+      "grad_norm": 11.831562166444332,
+      "learning_rate": 2e-06,
+      "loss": 0.2156,
+      "step": 8378
+    },
+    {
+      "epoch": 1.94385802111124,
+      "grad_norm": 11.559861634564646,
+      "learning_rate": 2e-06,
+      "loss": 0.205,
+      "step": 8379
+    },
+    {
+      "epoch": 1.9440900127595406,
+      "grad_norm": 20.769356547787247,
+      "learning_rate": 2e-06,
+      "loss": 0.2047,
+      "step": 8380
+    },
+    {
+      "epoch": 1.9443220044078413,
+      "grad_norm": 14.79081337535743,
+      "learning_rate": 2e-06,
+      "loss": 0.3526,
+      "step": 8381
+    },
+    {
+      "epoch": 1.944553996056142,
+      "grad_norm": 26.84448752847015,
+      "learning_rate": 2e-06,
+      "loss": 0.2859,
+      "step": 8382
+    },
+    {
+      "epoch": 1.9447859877044427,
+      "grad_norm": 24.08944134225375,
+      "learning_rate": 2e-06,
+      "loss": 0.2332,
+      "step": 8383
+    },
+    {
+      "epoch": 1.9450179793527433,
+      "grad_norm": 15.492793474505326,
+      "learning_rate": 2e-06,
+      "loss": 0.2011,
+      "step": 8384
+    },
+    {
+      "epoch": 1.945249971001044,
+      "grad_norm": 13.641540417777914,
+      "learning_rate": 2e-06,
+      "loss": 0.2306,
+      "step": 8385
+    },
+    {
+      "epoch": 1.9454819626493447,
+      "grad_norm": 10.80299252573023,
+      "learning_rate": 2e-06,
+      "loss": 0.1511,
+      "step": 8386
+    },
+    {
+      "epoch": 1.9457139542976454,
+      "grad_norm": 13.673854162605489,
+      "learning_rate": 2e-06,
+      "loss": 0.2412,
+      "step": 8387
+    },
+    {
+      "epoch": 1.945945945945946,
+      "grad_norm": 10.380498854463934,
+      "learning_rate": 2e-06,
+      "loss": 0.1457,
+      "step": 8388
+    },
+    {
+      "epoch": 1.9461779375942467,
+      "grad_norm": 22.70279296127693,
+      "learning_rate": 2e-06,
+      "loss": 0.2985,
+      "step": 8389
+    },
+    {
+      "epoch": 1.9464099292425474,
+      "grad_norm": 10.8828809073517,
+      "learning_rate": 2e-06,
+      "loss": 0.2901,
+      "step": 8390
+    },
+    {
+      "epoch": 1.946641920890848,
+      "grad_norm": 12.740824574439682,
+      "learning_rate": 2e-06,
+      "loss": 0.2331,
+      "step": 8391
+    },
+    {
+      "epoch": 1.9468739125391485,
+      "grad_norm": 10.986567109093164,
+      "learning_rate": 2e-06,
+      "loss": 0.2182,
+      "step": 8392
+    },
+    {
+      "epoch": 1.9471059041874492,
+      "grad_norm": 12.355360192244024,
+      "learning_rate": 2e-06,
+      "loss": 0.3446,
+      "step": 8393
+    },
+    {
+      "epoch": 1.94733789583575,
+      "grad_norm": 21.81122795551603,
+      "learning_rate": 2e-06,
+      "loss": 0.3911,
+      "step": 8394
+    },
+    {
+      "epoch": 1.9475698874840506,
+      "grad_norm": 10.852009445706015,
+      "learning_rate": 2e-06,
+      "loss": 0.1806,
+      "step": 8395
+    },
+    {
+      "epoch": 1.9478018791323513,
+      "grad_norm": 15.858822478149348,
+      "learning_rate": 2e-06,
+      "loss": 0.307,
+      "step": 8396
+    },
+    {
+      "epoch": 1.948033870780652,
+      "grad_norm": 7.640288640596409,
+      "learning_rate": 2e-06,
+      "loss": 0.2044,
+      "step": 8397
+    },
+    {
+      "epoch": 1.9482658624289526,
+      "grad_norm": 12.081191805075523,
+      "learning_rate": 2e-06,
+      "loss": 0.2311,
+      "step": 8398
+    },
+    {
+      "epoch": 1.948497854077253,
+      "grad_norm": 14.423207635176416,
+      "learning_rate": 2e-06,
+      "loss": 0.2262,
+      "step": 8399
+    },
+    {
+      "epoch": 1.9487298457255537,
+      "grad_norm": 11.753273624897368,
+      "learning_rate": 2e-06,
+      "loss": 0.1815,
+      "step": 8400
+    },
+    {
+      "epoch": 1.9489618373738544,
+      "grad_norm": 14.569786005421452,
+      "learning_rate": 2e-06,
+      "loss": 0.1847,
+      "step": 8401
+    },
+    {
+      "epoch": 1.949193829022155,
+      "grad_norm": 12.004180040477387,
+      "learning_rate": 2e-06,
+      "loss": 0.2279,
+      "step": 8402
+    },
+    {
+      "epoch": 1.9494258206704558,
+      "grad_norm": 11.534062119713278,
+      "learning_rate": 2e-06,
+      "loss": 0.2343,
+      "step": 8403
+    },
+    {
+      "epoch": 1.9496578123187565,
+      "grad_norm": 16.491537593130033,
+      "learning_rate": 2e-06,
+      "loss": 0.4219,
+      "step": 8404
+    },
+    {
+      "epoch": 1.9498898039670571,
+      "grad_norm": 20.454712496719754,
+      "learning_rate": 2e-06,
+      "loss": 0.3419,
+      "step": 8405
+    },
+    {
+      "epoch": 1.9501217956153578,
+      "grad_norm": 14.254428634177303,
+      "learning_rate": 2e-06,
+      "loss": 0.229,
+      "step": 8406
+    },
+    {
+      "epoch": 1.9503537872636585,
+      "grad_norm": 17.131492337479326,
+      "learning_rate": 2e-06,
+      "loss": 0.2399,
+      "step": 8407
+    },
+    {
+      "epoch": 1.9505857789119592,
+      "grad_norm": 10.012126423335909,
+      "learning_rate": 2e-06,
+      "loss": 0.176,
+      "step": 8408
+    },
+    {
+      "epoch": 1.9508177705602598,
+      "grad_norm": 15.887064355909393,
+      "learning_rate": 2e-06,
+      "loss": 0.2605,
+      "step": 8409
+    },
+    {
+      "epoch": 1.9510497622085605,
+      "grad_norm": 18.013675269362956,
+      "learning_rate": 2e-06,
+      "loss": 0.2112,
+      "step": 8410
+    },
+    {
+      "epoch": 1.9512817538568612,
+      "grad_norm": 16.777442020004855,
+      "learning_rate": 2e-06,
+      "loss": 0.2361,
+      "step": 8411
+    },
+    {
+      "epoch": 1.9515137455051619,
+      "grad_norm": 14.27637380350311,
+      "learning_rate": 2e-06,
+      "loss": 0.1877,
+      "step": 8412
+    },
+    {
+      "epoch": 1.9517457371534626,
+      "grad_norm": 5.247172375194371,
+      "learning_rate": 2e-06,
+      "loss": 0.1402,
+      "step": 8413
+    },
+    {
+      "epoch": 1.9519777288017632,
+      "grad_norm": 12.012920149131409,
+      "learning_rate": 2e-06,
+      "loss": 0.2541,
+      "step": 8414
+    },
+    {
+      "epoch": 1.952209720450064,
+      "grad_norm": 8.75045047925571,
+      "learning_rate": 2e-06,
+      "loss": 0.2358,
+      "step": 8415
+    },
+    {
+      "epoch": 1.9524417120983646,
+      "grad_norm": 11.665687192552866,
+      "learning_rate": 2e-06,
+      "loss": 0.2806,
+      "step": 8416
+    },
+    {
+      "epoch": 1.9526737037466653,
+      "grad_norm": 11.848604742375645,
+      "learning_rate": 2e-06,
+      "loss": 0.2743,
+      "step": 8417
+    },
+    {
+      "epoch": 1.952905695394966,
+      "grad_norm": 7.23663640660394,
+      "learning_rate": 2e-06,
+      "loss": 0.1936,
+      "step": 8418
+    },
+    {
+      "epoch": 1.9531376870432664,
+      "grad_norm": 12.568935820789614,
+      "learning_rate": 2e-06,
+      "loss": 0.3088,
+      "step": 8419
+    },
+    {
+      "epoch": 1.953369678691567,
+      "grad_norm": 8.327932198506467,
+      "learning_rate": 2e-06,
+      "loss": 0.2023,
+      "step": 8420
+    },
+    {
+      "epoch": 1.9536016703398678,
+      "grad_norm": 7.20468914494589,
+      "learning_rate": 2e-06,
+      "loss": 0.164,
+      "step": 8421
+    },
+    {
+      "epoch": 1.9538336619881684,
+      "grad_norm": 11.390215281587516,
+      "learning_rate": 2e-06,
+      "loss": 0.1818,
+      "step": 8422
+    },
+    {
+      "epoch": 1.9540656536364691,
+      "grad_norm": 7.102489376796252,
+      "learning_rate": 2e-06,
+      "loss": 0.1231,
+      "step": 8423
+    },
+    {
+      "epoch": 1.9542976452847698,
+      "grad_norm": 15.509530220768998,
+      "learning_rate": 2e-06,
+      "loss": 0.2076,
+      "step": 8424
+    },
+    {
+      "epoch": 1.9545296369330702,
+      "grad_norm": 16.151138514300136,
+      "learning_rate": 2e-06,
+      "loss": 0.3305,
+      "step": 8425
+    },
+    {
+      "epoch": 1.954761628581371,
+      "grad_norm": 8.82614805592662,
+      "learning_rate": 2e-06,
+      "loss": 0.1806,
+      "step": 8426
+    },
+    {
+      "epoch": 1.9549936202296716,
+      "grad_norm": 20.327694912149823,
+      "learning_rate": 2e-06,
+      "loss": 0.3804,
+      "step": 8427
+    },
+    {
+      "epoch": 1.9552256118779723,
+      "grad_norm": 13.419014003838033,
+      "learning_rate": 2e-06,
+      "loss": 0.25,
+      "step": 8428
+    },
+    {
+      "epoch": 1.955457603526273,
+      "grad_norm": 24.40155706745644,
+      "learning_rate": 2e-06,
+      "loss": 0.4561,
+      "step": 8429
+    },
+    {
+      "epoch": 1.9556895951745736,
+      "grad_norm": 14.019688656480323,
+      "learning_rate": 2e-06,
+      "loss": 0.1941,
+      "step": 8430
+    },
+    {
+      "epoch": 1.9559215868228743,
+      "grad_norm": 12.53301727015699,
+      "learning_rate": 2e-06,
+      "loss": 0.2806,
+      "step": 8431
+    },
+    {
+      "epoch": 1.956153578471175,
+      "grad_norm": 14.361679193213567,
+      "learning_rate": 2e-06,
+      "loss": 0.3081,
+      "step": 8432
+    },
+    {
+      "epoch": 1.9563855701194757,
+      "grad_norm": 26.510147634656878,
+      "learning_rate": 2e-06,
+      "loss": 0.4196,
+      "step": 8433
+    },
+    {
+      "epoch": 1.9566175617677763,
+      "grad_norm": 13.513129730273098,
+      "learning_rate": 2e-06,
+      "loss": 0.2441,
+      "step": 8434
+    },
+    {
+      "epoch": 1.956849553416077,
+      "grad_norm": 14.561957210594727,
+      "learning_rate": 2e-06,
+      "loss": 0.279,
+      "step": 8435
+    },
+    {
+      "epoch": 1.9570815450643777,
+      "grad_norm": 26.445220538157606,
+      "learning_rate": 2e-06,
+      "loss": 0.2931,
+      "step": 8436
+    },
+    {
+      "epoch": 1.9573135367126784,
+      "grad_norm": 13.783969377146583,
+      "learning_rate": 2e-06,
+      "loss": 0.2249,
+      "step": 8437
+    },
+    {
+      "epoch": 1.957545528360979,
+      "grad_norm": 27.928379555788805,
+      "learning_rate": 2e-06,
+      "loss": 0.4553,
+      "step": 8438
+    },
+    {
+      "epoch": 1.9577775200092797,
+      "grad_norm": 17.986716161532808,
+      "learning_rate": 2e-06,
+      "loss": 0.2743,
+      "step": 8439
+    },
+    {
+      "epoch": 1.9580095116575804,
+      "grad_norm": 14.087517558232287,
+      "learning_rate": 2e-06,
+      "loss": 0.3162,
+      "step": 8440
+    },
+    {
+      "epoch": 1.958241503305881,
+      "grad_norm": 12.892139932310869,
+      "learning_rate": 2e-06,
+      "loss": 0.2236,
+      "step": 8441
+    },
+    {
+      "epoch": 1.9584734949541818,
+      "grad_norm": 11.508804773043272,
+      "learning_rate": 2e-06,
+      "loss": 0.2051,
+      "step": 8442
+    },
+    {
+      "epoch": 1.9587054866024824,
+      "grad_norm": 22.660170295163322,
+      "learning_rate": 2e-06,
+      "loss": 0.3131,
+      "step": 8443
+    },
+    {
+      "epoch": 1.9589374782507831,
+      "grad_norm": 19.71433988724558,
+      "learning_rate": 2e-06,
+      "loss": 0.3338,
+      "step": 8444
+    },
+    {
+      "epoch": 1.9591694698990836,
+      "grad_norm": 13.96157075577866,
+      "learning_rate": 2e-06,
+      "loss": 0.2529,
+      "step": 8445
+    },
+    {
+      "epoch": 1.9594014615473843,
+      "grad_norm": 19.057960436934465,
+      "learning_rate": 2e-06,
+      "loss": 0.2632,
+      "step": 8446
+    },
+    {
+      "epoch": 1.959633453195685,
+      "grad_norm": 11.549432793969682,
+      "learning_rate": 2e-06,
+      "loss": 0.275,
+      "step": 8447
+    },
+    {
+      "epoch": 1.9598654448439856,
+      "grad_norm": 21.336870167414542,
+      "learning_rate": 2e-06,
+      "loss": 0.2588,
+      "step": 8448
+    },
+    {
+      "epoch": 1.9600974364922863,
+      "grad_norm": 11.302086911448749,
+      "learning_rate": 2e-06,
+      "loss": 0.2778,
+      "step": 8449
+    },
+    {
+      "epoch": 1.960329428140587,
+      "grad_norm": 15.244429894793514,
+      "learning_rate": 2e-06,
+      "loss": 0.2052,
+      "step": 8450
+    },
+    {
+      "epoch": 1.9605614197888876,
+      "grad_norm": 8.502244674605915,
+      "learning_rate": 2e-06,
+      "loss": 0.2113,
+      "step": 8451
+    },
+    {
+      "epoch": 1.960793411437188,
+      "grad_norm": 26.1900071729061,
+      "learning_rate": 2e-06,
+      "loss": 0.3883,
+      "step": 8452
+    },
+    {
+      "epoch": 1.9610254030854888,
+      "grad_norm": 14.440024116328479,
+      "learning_rate": 2e-06,
+      "loss": 0.2319,
+      "step": 8453
+    },
+    {
+      "epoch": 1.9612573947337895,
+      "grad_norm": 10.80774147458881,
+      "learning_rate": 2e-06,
+      "loss": 0.1606,
+      "step": 8454
+    },
+    {
+      "epoch": 1.9614893863820901,
+      "grad_norm": 15.094895749753553,
+      "learning_rate": 2e-06,
+      "loss": 0.2338,
+      "step": 8455
+    },
+    {
+      "epoch": 1.9617213780303908,
+      "grad_norm": 15.965997277124025,
+      "learning_rate": 2e-06,
+      "loss": 0.1719,
+      "step": 8456
+    },
+    {
+      "epoch": 1.9619533696786915,
+      "grad_norm": 19.69390806864314,
+      "learning_rate": 2e-06,
+      "loss": 0.2806,
+      "step": 8457
+    },
+    {
+      "epoch": 1.9621853613269922,
+      "grad_norm": 9.577296538084363,
+      "learning_rate": 2e-06,
+      "loss": 0.2495,
+      "step": 8458
+    },
+    {
+      "epoch": 1.9624173529752929,
+      "grad_norm": 26.75682900422842,
+      "learning_rate": 2e-06,
+      "loss": 0.2845,
+      "step": 8459
+    },
+    {
+      "epoch": 1.9626493446235935,
+      "grad_norm": 9.733980937932044,
+      "learning_rate": 2e-06,
+      "loss": 0.2037,
+      "step": 8460
+    },
+    {
+      "epoch": 1.9628813362718942,
+      "grad_norm": 22.218555882854805,
+      "learning_rate": 2e-06,
+      "loss": 0.3003,
+      "step": 8461
+    },
+    {
+      "epoch": 1.9631133279201949,
+      "grad_norm": 7.914416513340461,
+      "learning_rate": 2e-06,
+      "loss": 0.137,
+      "step": 8462
+    },
+    {
+      "epoch": 1.9633453195684956,
+      "grad_norm": 10.2317656645032,
+      "learning_rate": 2e-06,
+      "loss": 0.1882,
+      "step": 8463
+    },
+    {
+      "epoch": 1.9635773112167962,
+      "grad_norm": 10.376146409636112,
+      "learning_rate": 2e-06,
+      "loss": 0.1614,
+      "step": 8464
+    },
+    {
+      "epoch": 1.963809302865097,
+      "grad_norm": 17.186088573393064,
+      "learning_rate": 2e-06,
+      "loss": 0.2096,
+      "step": 8465
+    },
+    {
+      "epoch": 1.9640412945133976,
+      "grad_norm": 7.104689584111022,
+      "learning_rate": 2e-06,
+      "loss": 0.1568,
+      "step": 8466
+    },
+    {
+      "epoch": 1.9642732861616983,
+      "grad_norm": 21.66118189834683,
+      "learning_rate": 2e-06,
+      "loss": 0.3,
+      "step": 8467
+    },
+    {
+      "epoch": 1.964505277809999,
+      "grad_norm": 19.41182700985375,
+      "learning_rate": 2e-06,
+      "loss": 0.3131,
+      "step": 8468
+    },
+    {
+      "epoch": 1.9647372694582996,
+      "grad_norm": 9.70519193662155,
+      "learning_rate": 2e-06,
+      "loss": 0.3182,
+      "step": 8469
+    },
+    {
+      "epoch": 1.9649692611066003,
+      "grad_norm": 15.098456257992314,
+      "learning_rate": 2e-06,
+      "loss": 0.2525,
+      "step": 8470
+    },
+    {
+      "epoch": 1.965201252754901,
+      "grad_norm": 8.604709981076459,
+      "learning_rate": 2e-06,
+      "loss": 0.285,
+      "step": 8471
+    },
+    {
+      "epoch": 1.9654332444032014,
+      "grad_norm": 16.637331812879484,
+      "learning_rate": 2e-06,
+      "loss": 0.3224,
+      "step": 8472
+    },
+    {
+      "epoch": 1.9656652360515021,
+      "grad_norm": 21.586022385819454,
+      "learning_rate": 2e-06,
+      "loss": 0.4029,
+      "step": 8473
+    },
+    {
+      "epoch": 1.9658972276998028,
+      "grad_norm": 7.344537697532603,
+      "learning_rate": 2e-06,
+      "loss": 0.1618,
+      "step": 8474
+    },
+    {
+      "epoch": 1.9661292193481035,
+      "grad_norm": 11.941410369839828,
+      "learning_rate": 2e-06,
+      "loss": 0.2365,
+      "step": 8475
+    },
+    {
+      "epoch": 1.9663612109964042,
+      "grad_norm": 10.13680904838394,
+      "learning_rate": 2e-06,
+      "loss": 0.1756,
+      "step": 8476
+    },
+    {
+      "epoch": 1.9665932026447048,
+      "grad_norm": 9.082398998611101,
+      "learning_rate": 2e-06,
+      "loss": 0.2709,
+      "step": 8477
+    },
+    {
+      "epoch": 1.9668251942930053,
+      "grad_norm": 18.60080880891386,
+      "learning_rate": 2e-06,
+      "loss": 0.3045,
+      "step": 8478
+    },
+    {
+      "epoch": 1.967057185941306,
+      "grad_norm": 16.36407242270446,
+      "learning_rate": 2e-06,
+      "loss": 0.2604,
+      "step": 8479
+    },
+    {
+      "epoch": 1.9672891775896066,
+      "grad_norm": 14.299477048910845,
+      "learning_rate": 2e-06,
+      "loss": 0.2473,
+      "step": 8480
+    },
+    {
+      "epoch": 1.9675211692379073,
+      "grad_norm": 16.17852866015562,
+      "learning_rate": 2e-06,
+      "loss": 0.3183,
+      "step": 8481
+    },
+    {
+      "epoch": 1.967753160886208,
+      "grad_norm": 19.005085835163797,
+      "learning_rate": 2e-06,
+      "loss": 0.2687,
+      "step": 8482
+    },
+    {
+      "epoch": 1.9679851525345087,
+      "grad_norm": 12.339758285163436,
+      "learning_rate": 2e-06,
+      "loss": 0.2092,
+      "step": 8483
+    },
+    {
+      "epoch": 1.9682171441828094,
+      "grad_norm": 17.454199640819382,
+      "learning_rate": 2e-06,
+      "loss": 0.2479,
+      "step": 8484
+    },
+    {
+      "epoch": 1.96844913583111,
+      "grad_norm": 20.698836453320574,
+      "learning_rate": 2e-06,
+      "loss": 0.405,
+      "step": 8485
+    },
+    {
+      "epoch": 1.9686811274794107,
+      "grad_norm": 6.671319954473673,
+      "learning_rate": 2e-06,
+      "loss": 0.2192,
+      "step": 8486
+    },
+    {
+      "epoch": 1.9689131191277114,
+      "grad_norm": 13.887452692343361,
+      "learning_rate": 2e-06,
+      "loss": 0.2378,
+      "step": 8487
+    },
+    {
+      "epoch": 1.969145110776012,
+      "grad_norm": 15.72369330544798,
+      "learning_rate": 2e-06,
+      "loss": 0.2271,
+      "step": 8488
+    },
+    {
+      "epoch": 1.9693771024243127,
+      "grad_norm": 22.1359744407644,
+      "learning_rate": 2e-06,
+      "loss": 0.3123,
+      "step": 8489
+    },
+    {
+      "epoch": 1.9696090940726134,
+      "grad_norm": 9.595794165045705,
+      "learning_rate": 2e-06,
+      "loss": 0.2045,
+      "step": 8490
+    },
+    {
+      "epoch": 1.969841085720914,
+      "grad_norm": 8.637735599591288,
+      "learning_rate": 2e-06,
+      "loss": 0.1908,
+      "step": 8491
+    },
+    {
+      "epoch": 1.9700730773692148,
+      "grad_norm": 18.339224502699604,
+      "learning_rate": 2e-06,
+      "loss": 0.2914,
+      "step": 8492
+    },
+    {
+      "epoch": 1.9703050690175155,
+      "grad_norm": 13.889624100787907,
+      "learning_rate": 2e-06,
+      "loss": 0.2646,
+      "step": 8493
+    },
+    {
+      "epoch": 1.9705370606658161,
+      "grad_norm": 14.503935064188033,
+      "learning_rate": 2e-06,
+      "loss": 0.2686,
+      "step": 8494
+    },
+    {
+      "epoch": 1.9707690523141168,
+      "grad_norm": 7.739563925770556,
+      "learning_rate": 2e-06,
+      "loss": 0.1703,
+      "step": 8495
+    },
+    {
+      "epoch": 1.9710010439624175,
+      "grad_norm": 10.551688065569241,
+      "learning_rate": 2e-06,
+      "loss": 0.3515,
+      "step": 8496
+    },
+    {
+      "epoch": 1.9712330356107182,
+      "grad_norm": 9.902050304949746,
+      "learning_rate": 2e-06,
+      "loss": 0.2068,
+      "step": 8497
+    },
+    {
+      "epoch": 1.9714650272590186,
+      "grad_norm": 6.983275072558769,
+      "learning_rate": 2e-06,
+      "loss": 0.1782,
+      "step": 8498
+    },
+    {
+      "epoch": 1.9716970189073193,
+      "grad_norm": 13.785828158122156,
+      "learning_rate": 2e-06,
+      "loss": 0.3464,
+      "step": 8499
+    },
+    {
+      "epoch": 1.97192901055562,
+      "grad_norm": 14.076067112016535,
+      "learning_rate": 2e-06,
+      "loss": 0.3078,
+      "step": 8500
+    },
+    {
+      "epoch": 1.9721610022039207,
+      "grad_norm": 12.483396942471327,
+      "learning_rate": 2e-06,
+      "loss": 0.2082,
+      "step": 8501
+    },
+    {
+      "epoch": 1.9723929938522213,
+      "grad_norm": 12.186631988394545,
+      "learning_rate": 2e-06,
+      "loss": 0.2424,
+      "step": 8502
+    },
+    {
+      "epoch": 1.972624985500522,
+      "grad_norm": 9.102218431247547,
+      "learning_rate": 2e-06,
+      "loss": 0.3024,
+      "step": 8503
+    },
+    {
+      "epoch": 1.9728569771488227,
+      "grad_norm": 10.886415063131288,
+      "learning_rate": 2e-06,
+      "loss": 0.2534,
+      "step": 8504
+    },
+    {
+      "epoch": 1.9730889687971231,
+      "grad_norm": 12.836982023025246,
+      "learning_rate": 2e-06,
+      "loss": 0.2293,
+      "step": 8505
+    },
+    {
+      "epoch": 1.9733209604454238,
+      "grad_norm": 7.1802546184489255,
+      "learning_rate": 2e-06,
+      "loss": 0.1422,
+      "step": 8506
+    },
+    {
+      "epoch": 1.9735529520937245,
+      "grad_norm": 10.10370150232109,
+      "learning_rate": 2e-06,
+      "loss": 0.2613,
+      "step": 8507
+    },
+    {
+      "epoch": 1.9737849437420252,
+      "grad_norm": 12.471189951292699,
+      "learning_rate": 2e-06,
+      "loss": 0.2318,
+      "step": 8508
+    },
+    {
+      "epoch": 1.9740169353903259,
+      "grad_norm": 21.52531851333419,
+      "learning_rate": 2e-06,
+      "loss": 0.3975,
+      "step": 8509
+    },
+    {
+      "epoch": 1.9742489270386265,
+      "grad_norm": 9.963227113698434,
+      "learning_rate": 2e-06,
+      "loss": 0.2103,
+      "step": 8510
+    },
+    {
+      "epoch": 1.9744809186869272,
+      "grad_norm": 11.232264002401344,
+      "learning_rate": 2e-06,
+      "loss": 0.2087,
+      "step": 8511
+    },
+    {
+      "epoch": 1.9747129103352279,
+      "grad_norm": 16.55607130186788,
+      "learning_rate": 2e-06,
+      "loss": 0.3419,
+      "step": 8512
+    },
+    {
+      "epoch": 1.9749449019835286,
+      "grad_norm": 8.849922497856467,
+      "learning_rate": 2e-06,
+      "loss": 0.224,
+      "step": 8513
+    },
+    {
+      "epoch": 1.9751768936318292,
+      "grad_norm": 21.600280522778597,
+      "learning_rate": 2e-06,
+      "loss": 0.2065,
+      "step": 8514
+    },
+    {
+      "epoch": 1.97540888528013,
+      "grad_norm": 19.806163133323412,
+      "learning_rate": 2e-06,
+      "loss": 0.3007,
+      "step": 8515
+    },
+    {
+      "epoch": 1.9756408769284306,
+      "grad_norm": 20.337850588002727,
+      "learning_rate": 2e-06,
+      "loss": 0.237,
+      "step": 8516
+    },
+    {
+      "epoch": 1.9758728685767313,
+      "grad_norm": 8.776756918468648,
+      "learning_rate": 2e-06,
+      "loss": 0.1599,
+      "step": 8517
+    },
+    {
+      "epoch": 1.976104860225032,
+      "grad_norm": 11.533070577796943,
+      "learning_rate": 2e-06,
+      "loss": 0.1877,
+      "step": 8518
+    },
+    {
+      "epoch": 1.9763368518733326,
+      "grad_norm": 12.84657385589497,
+      "learning_rate": 2e-06,
+      "loss": 0.2746,
+      "step": 8519
+    },
+    {
+      "epoch": 1.9765688435216333,
+      "grad_norm": 16.518796669554234,
+      "learning_rate": 2e-06,
+      "loss": 0.295,
+      "step": 8520
+    },
+    {
+      "epoch": 1.976800835169934,
+      "grad_norm": 16.963351981037178,
+      "learning_rate": 2e-06,
+      "loss": 0.2811,
+      "step": 8521
+    },
+    {
+      "epoch": 1.9770328268182347,
+      "grad_norm": 4.346222300206503,
+      "learning_rate": 2e-06,
+      "loss": 0.1426,
+      "step": 8522
+    },
+    {
+      "epoch": 1.9772648184665353,
+      "grad_norm": 11.419250342857609,
+      "learning_rate": 2e-06,
+      "loss": 0.2009,
+      "step": 8523
+    },
+    {
+      "epoch": 1.977496810114836,
+      "grad_norm": 18.097650604177662,
+      "learning_rate": 2e-06,
+      "loss": 0.3862,
+      "step": 8524
+    },
+    {
+      "epoch": 1.9777288017631365,
+      "grad_norm": 20.249542109447514,
+      "learning_rate": 2e-06,
+      "loss": 0.3922,
+      "step": 8525
+    },
+    {
+      "epoch": 1.9779607934114372,
+      "grad_norm": 13.871707038549962,
+      "learning_rate": 2e-06,
+      "loss": 0.1997,
+      "step": 8526
+    },
+    {
+      "epoch": 1.9781927850597378,
+      "grad_norm": 13.029721724629182,
+      "learning_rate": 2e-06,
+      "loss": 0.1359,
+      "step": 8527
+    },
+    {
+      "epoch": 1.9784247767080385,
+      "grad_norm": 8.589532756585148,
+      "learning_rate": 2e-06,
+      "loss": 0.203,
+      "step": 8528
+    },
+    {
+      "epoch": 1.9786567683563392,
+      "grad_norm": 19.65133151674858,
+      "learning_rate": 2e-06,
+      "loss": 0.3163,
+      "step": 8529
+    },
+    {
+      "epoch": 1.9788887600046399,
+      "grad_norm": 10.528372449802873,
+      "learning_rate": 2e-06,
+      "loss": 0.2023,
+      "step": 8530
+    },
+    {
+      "epoch": 1.9791207516529405,
+      "grad_norm": 13.753512180806643,
+      "learning_rate": 2e-06,
+      "loss": 0.2618,
+      "step": 8531
+    },
+    {
+      "epoch": 1.979352743301241,
+      "grad_norm": 10.418585440923946,
+      "learning_rate": 2e-06,
+      "loss": 0.2562,
+      "step": 8532
+    },
+    {
+      "epoch": 1.9795847349495417,
+      "grad_norm": 9.050020397117102,
+      "learning_rate": 2e-06,
+      "loss": 0.2337,
+      "step": 8533
+    },
+    {
+      "epoch": 1.9798167265978424,
+      "grad_norm": 10.006801325208919,
+      "learning_rate": 2e-06,
+      "loss": 0.2385,
+      "step": 8534
+    },
+    {
+      "epoch": 1.980048718246143,
+      "grad_norm": 11.336402703687698,
+      "learning_rate": 2e-06,
+      "loss": 0.205,
+      "step": 8535
+    },
+    {
+      "epoch": 1.9802807098944437,
+      "grad_norm": 12.825693375037904,
+      "learning_rate": 2e-06,
+      "loss": 0.2171,
+      "step": 8536
+    },
+    {
+      "epoch": 1.9805127015427444,
+      "grad_norm": 5.524550556954387,
+      "learning_rate": 2e-06,
+      "loss": 0.1233,
+      "step": 8537
+    },
+    {
+      "epoch": 1.980744693191045,
+      "grad_norm": 8.083841633609582,
+      "learning_rate": 2e-06,
+      "loss": 0.1239,
+      "step": 8538
+    },
+    {
+      "epoch": 1.9809766848393457,
+      "grad_norm": 11.573832770938761,
+      "learning_rate": 2e-06,
+      "loss": 0.2471,
+      "step": 8539
+    },
+    {
+      "epoch": 1.9812086764876464,
+      "grad_norm": 7.392452023260391,
+      "learning_rate": 2e-06,
+      "loss": 0.1823,
+      "step": 8540
+    },
+    {
+      "epoch": 1.981440668135947,
+      "grad_norm": 21.98841567744905,
+      "learning_rate": 2e-06,
+      "loss": 0.3394,
+      "step": 8541
+    },
+    {
+      "epoch": 1.9816726597842478,
+      "grad_norm": 6.8136512886727445,
+      "learning_rate": 2e-06,
+      "loss": 0.1468,
+      "step": 8542
+    },
+    {
+      "epoch": 1.9819046514325485,
+      "grad_norm": 19.584918378180546,
+      "learning_rate": 2e-06,
+      "loss": 0.247,
+      "step": 8543
+    },
+    {
+      "epoch": 1.9821366430808491,
+      "grad_norm": 13.017497347406255,
+      "learning_rate": 2e-06,
+      "loss": 0.2514,
+      "step": 8544
+    },
+    {
+      "epoch": 1.9823686347291498,
+      "grad_norm": 16.294108863284833,
+      "learning_rate": 2e-06,
+      "loss": 0.2779,
+      "step": 8545
+    },
+    {
+      "epoch": 1.9826006263774505,
+      "grad_norm": 8.147068821027043,
+      "learning_rate": 2e-06,
+      "loss": 0.1924,
+      "step": 8546
+    },
+    {
+      "epoch": 1.9828326180257512,
+      "grad_norm": 9.830621743145237,
+      "learning_rate": 2e-06,
+      "loss": 0.24,
+      "step": 8547
+    },
+    {
+      "epoch": 1.9830646096740518,
+      "grad_norm": 22.846163306008492,
+      "learning_rate": 2e-06,
+      "loss": 0.3014,
+      "step": 8548
+    },
+    {
+      "epoch": 1.9832966013223525,
+      "grad_norm": 18.613538314175056,
+      "learning_rate": 2e-06,
+      "loss": 0.4346,
+      "step": 8549
+    },
+    {
+      "epoch": 1.9835285929706532,
+      "grad_norm": 13.590506060031638,
+      "learning_rate": 2e-06,
+      "loss": 0.2421,
+      "step": 8550
+    },
+    {
+      "epoch": 1.9837605846189539,
+      "grad_norm": 13.04190187877181,
+      "learning_rate": 2e-06,
+      "loss": 0.2047,
+      "step": 8551
+    },
+    {
+      "epoch": 1.9839925762672543,
+      "grad_norm": 10.602516907217499,
+      "learning_rate": 2e-06,
+      "loss": 0.2152,
+      "step": 8552
+    },
+    {
+      "epoch": 1.984224567915555,
+      "grad_norm": 12.780620813300855,
+      "learning_rate": 2e-06,
+      "loss": 0.2415,
+      "step": 8553
+    },
+    {
+      "epoch": 1.9844565595638557,
+      "grad_norm": 12.627158631424301,
+      "learning_rate": 2e-06,
+      "loss": 0.1876,
+      "step": 8554
+    },
+    {
+      "epoch": 1.9846885512121564,
+      "grad_norm": 19.562289971581983,
+      "learning_rate": 2e-06,
+      "loss": 0.2719,
+      "step": 8555
+    },
+    {
+      "epoch": 1.984920542860457,
+      "grad_norm": 19.225652733982006,
+      "learning_rate": 2e-06,
+      "loss": 0.3401,
+      "step": 8556
+    },
+    {
+      "epoch": 1.9851525345087577,
+      "grad_norm": 10.428271595257316,
+      "learning_rate": 2e-06,
+      "loss": 0.2353,
+      "step": 8557
+    },
+    {
+      "epoch": 1.9853845261570582,
+      "grad_norm": 9.14187799266162,
+      "learning_rate": 2e-06,
+      "loss": 0.2572,
+      "step": 8558
+    },
+    {
+      "epoch": 1.9856165178053589,
+      "grad_norm": 12.648061748666258,
+      "learning_rate": 2e-06,
+      "loss": 0.2418,
+      "step": 8559
+    },
+    {
+      "epoch": 1.9858485094536595,
+      "grad_norm": 10.629099645791218,
+      "learning_rate": 2e-06,
+      "loss": 0.235,
+      "step": 8560
+    },
+    {
+      "epoch": 1.9860805011019602,
+      "grad_norm": 4.342336370217446,
+      "learning_rate": 2e-06,
+      "loss": 0.1194,
+      "step": 8561
+    },
+    {
+      "epoch": 1.986312492750261,
+      "grad_norm": 16.143025239136698,
+      "learning_rate": 2e-06,
+      "loss": 0.2595,
+      "step": 8562
+    },
+    {
+      "epoch": 1.9865444843985616,
+      "grad_norm": 9.754089385774627,
+      "learning_rate": 2e-06,
+      "loss": 0.2004,
+      "step": 8563
+    },
+    {
+      "epoch": 1.9867764760468622,
+      "grad_norm": 15.841521291649542,
+      "learning_rate": 2e-06,
+      "loss": 0.2853,
+      "step": 8564
+    },
+    {
+      "epoch": 1.987008467695163,
+      "grad_norm": 14.760371139310362,
+      "learning_rate": 2e-06,
+      "loss": 0.2199,
+      "step": 8565
+    },
+    {
+      "epoch": 1.9872404593434636,
+      "grad_norm": 10.32037228954311,
+      "learning_rate": 2e-06,
+      "loss": 0.2619,
+      "step": 8566
+    },
+    {
+      "epoch": 1.9874724509917643,
+      "grad_norm": 11.920341608033274,
+      "learning_rate": 2e-06,
+      "loss": 0.1909,
+      "step": 8567
+    },
+    {
+      "epoch": 1.987704442640065,
+      "grad_norm": 11.357525599468852,
+      "learning_rate": 2e-06,
+      "loss": 0.172,
+      "step": 8568
+    },
+    {
+      "epoch": 1.9879364342883656,
+      "grad_norm": 12.124003669396531,
+      "learning_rate": 2e-06,
+      "loss": 0.229,
+      "step": 8569
+    },
+    {
+      "epoch": 1.9881684259366663,
+      "grad_norm": 19.62315796808522,
+      "learning_rate": 2e-06,
+      "loss": 0.2791,
+      "step": 8570
+    },
+    {
+      "epoch": 1.988400417584967,
+      "grad_norm": 11.750587502060243,
+      "learning_rate": 2e-06,
+      "loss": 0.2483,
+      "step": 8571
+    },
+    {
+      "epoch": 1.9886324092332677,
+      "grad_norm": 15.404513427589219,
+      "learning_rate": 2e-06,
+      "loss": 0.2913,
+      "step": 8572
+    },
+    {
+      "epoch": 1.9888644008815684,
+      "grad_norm": 5.384890915201693,
+      "learning_rate": 2e-06,
+      "loss": 0.149,
+      "step": 8573
+    },
+    {
+      "epoch": 1.989096392529869,
+      "grad_norm": 11.17698439178024,
+      "learning_rate": 2e-06,
+      "loss": 0.236,
+      "step": 8574
+    },
+    {
+      "epoch": 1.9893283841781697,
+      "grad_norm": 25.955564600183294,
+      "learning_rate": 2e-06,
+      "loss": 0.3429,
+      "step": 8575
+    },
+    {
+      "epoch": 1.9895603758264704,
+      "grad_norm": 9.573922131436106,
+      "learning_rate": 2e-06,
+      "loss": 0.2218,
+      "step": 8576
+    },
+    {
+      "epoch": 1.989792367474771,
+      "grad_norm": 10.128885271798515,
+      "learning_rate": 2e-06,
+      "loss": 0.2154,
+      "step": 8577
+    },
+    {
+      "epoch": 1.9900243591230715,
+      "grad_norm": 12.827613625373983,
+      "learning_rate": 2e-06,
+      "loss": 0.2551,
+      "step": 8578
+    },
+    {
+      "epoch": 1.9902563507713722,
+      "grad_norm": 12.181597847453952,
+      "learning_rate": 2e-06,
+      "loss": 0.2036,
+      "step": 8579
+    },
+    {
+      "epoch": 1.9904883424196729,
+      "grad_norm": 11.204479881120816,
+      "learning_rate": 2e-06,
+      "loss": 0.2682,
+      "step": 8580
+    },
+    {
+      "epoch": 1.9907203340679736,
+      "grad_norm": 14.7149776713658,
+      "learning_rate": 2e-06,
+      "loss": 0.1934,
+      "step": 8581
+    },
+    {
+      "epoch": 1.9909523257162742,
+      "grad_norm": 14.559952872618316,
+      "learning_rate": 2e-06,
+      "loss": 0.3198,
+      "step": 8582
+    },
+    {
+      "epoch": 1.991184317364575,
+      "grad_norm": 17.65883394269422,
+      "learning_rate": 2e-06,
+      "loss": 0.2938,
+      "step": 8583
+    },
+    {
+      "epoch": 1.9914163090128756,
+      "grad_norm": 11.148759808439678,
+      "learning_rate": 2e-06,
+      "loss": 0.2647,
+      "step": 8584
+    },
+    {
+      "epoch": 1.991648300661176,
+      "grad_norm": 11.688280097465956,
+      "learning_rate": 2e-06,
+      "loss": 0.2514,
+      "step": 8585
+    },
+    {
+      "epoch": 1.9918802923094767,
+      "grad_norm": 12.486888404537797,
+      "learning_rate": 2e-06,
+      "loss": 0.2763,
+      "step": 8586
+    },
+    {
+      "epoch": 1.9921122839577774,
+      "grad_norm": 11.592115694074757,
+      "learning_rate": 2e-06,
+      "loss": 0.2743,
+      "step": 8587
+    },
+    {
+      "epoch": 1.992344275606078,
+      "grad_norm": 11.680842392664022,
+      "learning_rate": 2e-06,
+      "loss": 0.2208,
+      "step": 8588
+    },
+    {
+      "epoch": 1.9925762672543788,
+      "grad_norm": 18.036521977119452,
+      "learning_rate": 2e-06,
+      "loss": 0.3916,
+      "step": 8589
+    },
+    {
+      "epoch": 1.9928082589026794,
+      "grad_norm": 8.003475013553198,
+      "learning_rate": 2e-06,
+      "loss": 0.1752,
+      "step": 8590
+    },
+    {
+      "epoch": 1.99304025055098,
+      "grad_norm": 13.081533947310318,
+      "learning_rate": 2e-06,
+      "loss": 0.2328,
+      "step": 8591
+    },
+    {
+      "epoch": 1.9932722421992808,
+      "grad_norm": 12.915462699759964,
+      "learning_rate": 2e-06,
+      "loss": 0.2749,
+      "step": 8592
+    },
+    {
+      "epoch": 1.9935042338475815,
+      "grad_norm": 10.856135360597765,
+      "learning_rate": 2e-06,
+      "loss": 0.376,
+      "step": 8593
+    },
+    {
+      "epoch": 1.9937362254958821,
+      "grad_norm": 7.591018863066772,
+      "learning_rate": 2e-06,
+      "loss": 0.1334,
+      "step": 8594
+    },
+    {
+      "epoch": 1.9939682171441828,
+      "grad_norm": 16.73779734886259,
+      "learning_rate": 2e-06,
+      "loss": 0.3323,
+      "step": 8595
+    },
+    {
+      "epoch": 1.9942002087924835,
+      "grad_norm": 17.095465102448927,
+      "learning_rate": 2e-06,
+      "loss": 0.2698,
+      "step": 8596
+    },
+    {
+      "epoch": 1.9944322004407842,
+      "grad_norm": 16.201500392677026,
+      "learning_rate": 2e-06,
+      "loss": 0.3091,
+      "step": 8597
+    },
+    {
+      "epoch": 1.9946641920890849,
+      "grad_norm": 10.38865379973004,
+      "learning_rate": 2e-06,
+      "loss": 0.2829,
+      "step": 8598
+    },
+    {
+      "epoch": 1.9948961837373855,
+      "grad_norm": 9.53782022055871,
+      "learning_rate": 2e-06,
+      "loss": 0.2776,
+      "step": 8599
+    },
+    {
+      "epoch": 1.9951281753856862,
+      "grad_norm": 8.811020125502003,
+      "learning_rate": 2e-06,
+      "loss": 0.1423,
+      "step": 8600
+    },
+    {
+      "epoch": 1.9953601670339869,
+      "grad_norm": 15.270821714163787,
+      "learning_rate": 2e-06,
+      "loss": 0.2136,
+      "step": 8601
+    },
+    {
+      "epoch": 1.9955921586822876,
+      "grad_norm": 18.901918550352107,
+      "learning_rate": 2e-06,
+      "loss": 0.3078,
+      "step": 8602
+    },
+    {
+      "epoch": 1.9958241503305882,
+      "grad_norm": 9.205250195423652,
+      "learning_rate": 2e-06,
+      "loss": 0.1244,
+      "step": 8603
+    },
+    {
+      "epoch": 1.996056141978889,
+      "grad_norm": 12.846303081127727,
+      "learning_rate": 2e-06,
+      "loss": 0.2795,
+      "step": 8604
+    },
+    {
+      "epoch": 1.9962881336271894,
+      "grad_norm": 15.88210580620789,
+      "learning_rate": 2e-06,
+      "loss": 0.2166,
+      "step": 8605
+    },
+    {
+      "epoch": 1.99652012527549,
+      "grad_norm": 10.421273958167372,
+      "learning_rate": 2e-06,
+      "loss": 0.2085,
+      "step": 8606
+    },
+    {
+      "epoch": 1.9967521169237907,
+      "grad_norm": 11.151099582393117,
+      "learning_rate": 2e-06,
+      "loss": 0.3363,
+      "step": 8607
+    },
+    {
+      "epoch": 1.9969841085720914,
+      "grad_norm": 14.37049931427536,
+      "learning_rate": 2e-06,
+      "loss": 0.2388,
+      "step": 8608
+    },
+    {
+      "epoch": 1.997216100220392,
+      "grad_norm": 14.278416156594247,
+      "learning_rate": 2e-06,
+      "loss": 0.1511,
+      "step": 8609
+    },
+    {
+      "epoch": 1.9974480918686928,
+      "grad_norm": 12.665305808679461,
+      "learning_rate": 2e-06,
+      "loss": 0.1878,
+      "step": 8610
+    },
+    {
+      "epoch": 1.9976800835169934,
+      "grad_norm": 15.377701007457812,
+      "learning_rate": 2e-06,
+      "loss": 0.3327,
+      "step": 8611
+    },
+    {
+      "epoch": 1.997912075165294,
+      "grad_norm": 8.160117309260238,
+      "learning_rate": 2e-06,
+      "loss": 0.1728,
+      "step": 8612
+    },
+    {
+      "epoch": 1.9981440668135946,
+      "grad_norm": 10.501757269939331,
+      "learning_rate": 2e-06,
+      "loss": 0.218,
+      "step": 8613
+    },
+    {
+      "epoch": 1.9983760584618953,
+      "grad_norm": 12.866025684053145,
+      "learning_rate": 2e-06,
+      "loss": 0.3282,
+      "step": 8614
+    },
+    {
+      "epoch": 1.998608050110196,
+      "grad_norm": 13.778437356571668,
+      "learning_rate": 2e-06,
+      "loss": 0.2975,
+      "step": 8615
+    },
+    {
+      "epoch": 1.9988400417584966,
+      "grad_norm": 13.683219081874212,
+      "learning_rate": 2e-06,
+      "loss": 0.2543,
+      "step": 8616
+    },
+    {
+      "epoch": 1.9990720334067973,
+      "grad_norm": 7.564723850571399,
+      "learning_rate": 2e-06,
+      "loss": 0.1385,
+      "step": 8617
+    },
+    {
+      "epoch": 1.999304025055098,
+      "grad_norm": 11.625230524676542,
+      "learning_rate": 2e-06,
+      "loss": 0.3084,
+      "step": 8618
+    },
+    {
+      "epoch": 1.9995360167033986,
+      "grad_norm": 17.15768884295696,
+      "learning_rate": 2e-06,
+      "loss": 0.3042,
+      "step": 8619
+    },
+    {
+      "epoch": 1.9997680083516993,
+      "grad_norm": 9.186683071124046,
+      "learning_rate": 2e-06,
+      "loss": 0.1746,
+      "step": 8620
+    },
+    {
+      "epoch": 1.9997680083516993,
+      "step": 8620,
+      "total_flos": 2997308663627776.0,
+      "train_loss": 0.283832880054231,
+      "train_runtime": 20710.279,
+      "train_samples_per_second": 6.66,
+      "train_steps_per_second": 0.416
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 8620,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2997308663627776.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}